Another case of media sloppiness when reporting a survey

On Sunday morning the post-Dunfermline decline in the Gordon Brown betting price on succeeding Tony Blair was stopped and punters began betting on him again. This was prompted by widespread coverage of a YouGov poll in the Sunday Times that seemed to state that Labour would do better under the Chancellor than with Blair.

This was the critical section of the report: “On the eve of a speech by Brown on terrorism, a YouGov poll for The Sunday Times has found that Labour would be six percentage points ahead of David Cameron’s Tories, by a margin of 43% to 37%, if Brown was leader. Liberal Democrat supporters, in particular, favour the chancellor.”

But from the detailed data about the poll, now available, it is clear the the question of which party people would vote for if Brown was leader was not put.

The actual question was “If you had to choose between the two, which would you prefer to govern Britain? A Conservative government led by David Cameron: A Labour government led by Gordon Brown.” It was a straight Brown or Cameron choice with no option for respondents to indicate the Lib Dems or any other party.

The reason why, in the Sunday Times’s words, “Liberal Democrat supporters, in particular, favour the chancellor” was that those who filled in the survey were not offered the possibility of supporting their own party.

The last time that YouGov put the Brown-Cameron question in such terms the Chancellor was ahead 46-37 – so there has been a three point decline in his position.

From the new poll data Lib Dems split 57-22 for Brown. If the option of saying they could vote Lib Dem had been offered then, looking at the figures, it is probable that Brown would not have been ahead.

This latest case follows our report on Monday of how parts of the media misinterpreted the implication of the Lib Dem leadership questions in the same survey.

The problem we have is that when a poll comes out we are totally reliant on how the paper that commissioned it actually reports it. If things are presented wrongly then there is nothing we can do until the hard information from the pollster is made available.

COMMENTS ON THE SITE Can we turn the noise level down and stop insulting others who post here. PBC works because people of all allegiances feel able to come and discuss political betting and political outcomes. There are other places where you can abuse supporters of rival parties.

Mike Smithson