Category: Polling

Latest polls…

Good to see the Lib Dems up to 24% in the latest IPSOS-MORI poll.

Looking at MORI’s long term trends (and using their unweighted figure of 23% for comparability), we seem to be up +1%pt vs. the General Election and up up +3%pts vs. the last poll before Ming became leader.

Five months after Paddy started, we were on 8%, up +1%pt vs. the last poll before Paddy became leader

Five months after Charles started, we were on 15%, up +1%pt vs. the last poll before Charles became leader.

Thirteen months after the 2001 General Election, we were on 16%, down -2%pts on MORI’s General Election poll.

There are arguments about MORI’s methodology, but with the exception of ICM (which looks like an outlier), the other polls also show no burning need to panic, with our polling up flat or slightly up since Ming took over. Generally long-term trends are by far the most useful things to look at with these polls since methodology makes short-term trends and absolute numbers unreliable bases for comparison.

Looking at the current political situation, Ming also has the non-trivial advantage of having done the important thing of being right on the major issues. Cameron’s failure to show any leadership in the face of the crisis in the Middle East will not have done him any favours in the long term. It all suggests that there is no need to panic just yet.

Old chums connection?

Stephen Tall wonders why the BBC continue to retain Frank Luntz over at ‘A Liberal Goes A Long Way’.

I wonder if he’s seen the earlier comments on the ‘David Davis for leader’ blog?

BBC management probably like Luntz because he makes entertaining TV and he predicted the Cameron victory – although several commentators have made the case that the causality actually went the other way and that his Newsnight Focus Group made a major contribution to the Cameron victory (discussed at length in this thread on politicalbetting.com).

However, anyone who’s spent any time in marketing or market research knows how easy it is to ‘throw’ qualitative research, how easy it is to bias research group stimulus, and, particularly, how easy it to selectively edit the results, particularly if you’re looking for people to endorse your product (or politician).

A couple of minutes’ research shows pretty clear evidence from the US is that Frank Luntz has a strong link to the Bush administration and the conservative right as well – this PBS interview is particularly illuminating. There’s also evidence that he has misled media companies about those links. In light of this, I’m surprised that the BBC continue to use him as an ‘independent pollster’ rather than as a ‘conservative pollster and commentator’. Perhaps entertaining TV provides enough justification to throw any attempt at neutrality out of the window.

Perhaps we should get Media Matters for America – who wrote to MSNBC to complain about their use of Frank Luntz in 2004 – to write to the BBC in 2006?