The Sun and Voodoo Polls

On Thursday, February 28, the Sun headlined:

121,764 VOTE TO SAVE £

Neil Syson's article began: 'A MASSIVE army of Sun readers voted to save the Pound yesterday in our biggest You The Jury poll response ever.' The splash continued on, and the 'Full Story' was reported on pages 8 and 9. There it reported 'THE 15-1 vote against scrapping the Pound in The Sun's historic poll [sic] could signal that Britons will NEVER agree to handing our economy over to Brussels … And it overshadowed surveys done by polling organisations like MORI and NOP … [which] rarely cover more than 1,000 voters at a time.'

As well as a cartoon showing Blair and Brown steamrollering towards the Euro through the wall of public opinion, its THE SUN SAYS leader made clear their disregard for the results of the MORI poll reported the same day in The Times, their sister paper: 'UP THE POLL: There's a baffling opinion poll in The Times today. According to MORI, three million Sun readers agree with scrapping the Pound. Three million? Can you believe that? We can't. Who are these people? And where did MORI conduct this poll … in Brussels?

'The irony of all of this was that the Sun had commissioned a proper opinion poll from its pollsters, MORI, to assess public reaction to the police following the Lawrence 'scandal'. (See p. 3 for details) After its attack on MORI in its leader on the Thursday, what did Monday's paper headline?

'BRITAIN BACKS OUR BOBBIES: Sun poll boosts under-fire cops'

The second paragraph reported 'An exclusive Sun poll reveals officers still have the support of ordinary people — white, black and Asian.'

So one 'Sun poll' — their unrepresentative phone-in 'voodoo' poll — was more accurate than MORI on the Thursday, because of its huge, if unrepresentative response, but the one we did led the same newspaper on the following Monday. No wonder Sun readers are confused, and in audience after audience I address, I am asked about the accuracy of polls. What else would I find when such inconsistencies are reported so vividly?

I have long crusaded against the reporting of such voodoo polls, but alas, the Sun, the BBC, LBC and Talk Radio continue, though not, I am happy to say, The Times, Telegraph, Guardian, Indy or Financial Times. Thus, I suspect, there will be increasing scepticism of polls and polling among readers of the redtops and audiences of the TV and radio programmes who 'give their readers/listeners/viewers a chance to express their views'. I don't object to that; I do object when they are passed off as anything other than unrepresentative, a measure of those that bothered to phone in, paid the cost of the call, and responded, as they sometimes do, to the pressure group or party who urged them to 'vote', once, twice, or more (as did one Desmond who wrote to the Evening Standard to say how pleased he was that his side 'won', having voted 157 times himself on an LBC phone-in).

Even more cynical was the fax-in poll on the future of Europe by 'Fax Polling Associates'. These enterprising entrepreneurs sent out hundreds of thousands of 'The Referendum on Europe' forms, 'Last chance to vote!' in 1997 and asked mugs to fax them back to vote to 'stay in' or 'get out' of 'Europe?' to what turned out to be a premium-rate phone line, at a cost to the sender of £1 per minute at all times. They claimed they got back some 15% of the hundreds of thousands they got back, so must have pocketed a pretty penny.

Clearly more dangerous are constituency polls carried in the local paper. On 25 March 1997, the Guardian reported 'a telephone poll by the Worcester Evening News', putting Labour on 73% support in the key marginal and the Conservatives on just 22%. This ludicrous poll was, in fact, a phone-in voodoo poll. Labour actually secured 50% of the vote and the Conservatives 36%. (By comparison, the average difference between poll findings and the final election vote share of the three main parties and SNP in the 28 constituency polls carried out across the country during the election by ICM and MORI was 2.4%.)

Naturally enough, interested parties frequently avail themselves of this opportunity. The former Labour Chief Whip Michael Cocks admitted a couple of years ago 'When I was Chief Whip, I would occasionally sit colleagues down with instructions to make repeated telephone calls to telephone poll numbers to distort the result' , and the other parties are no more innocent.

Who can forget the fiasco of the 1996 BBC Radio 4 Today programme Personality of the Year poll? Tony Blair disqualified at the nominations stage on the suspicion of Labour Party supporters rigging the voting, but John Major qualifying for the shortlist and subsequently winning (beating Lisa Potts, the primary school teacher who protected her class against a madman wielding a machete, into second place). Even the BBC had to shamefacedly admit that there had also been some multiple voting in favour of Mr. Major, but not in their opinion enough to affect the result. Similar distortions happened throughout the history of the annual Today poll — Mrs. Thatcher was a frequent winner.

None of these examples had any real impact. But voodoo polls amount to spreading propaganda for a point of view through the media, on the basis of figures which the perpetrators must known are at the best unreliable and at the worst untrue. In extreme cases this behaviour becomes a deliberate con, often by groups that dare not commission a representative poll because they dare not face the truth of public opinion. Politicians are frequent users of such tricks, especially by their quoting of canvas returns; some pressure groups, sadly, behave similarly. Used in this way, especially if used in an attempt to counter or deny the results of a reputably conducted sample survey, voodoo polls are dishonest.

Sincerely Sir Robert Worcester