Why I’m liking the tracker

I’m getting to rather like the YouGov tracker simply it provides an opportunity to test my theories about what drives the polls.

I’ve always argued here that so much of poll findings outside election campaigns is driven by which party or politician is getting the most attention. For most people are not really that interested in politics.

So far the tracker seems to have been backing this up. The Tories have hardly been heard from for weeks, the Lib Dems are still benefiting from their week at the sea-side and it’s Labour that’s been getting the big profile in the past few days.

Anyway let’s all think towards next week and try to guess what the tracker will be saying on the day after Cameron’s speech.

Mike Smithson

CON 37% (-3) LAB 30% (+1) LD 21%(+3)

But most of fieldwork took place BEFORE the Sun’s news?

Peter Kellner has just been in touch with news of today YouGov tracker poll. He emphasises that “most fieldwork reflects impact of yesterday evening TV news coverage but not this mornings papers. Tomorrow’s figures will measure full impact of media coverage, Sun switch etc”.

It’s important to note that the numbers we are seeing each are essentially a reaction to the news of the previous evening.

So what these number tell us is that Brown’s speech and the coverage of it on TV helped Labour up a point - but provided the biggest boost for the Liberal Democrats who are up to 21%. The Tories are the big loser. Quite why the poll has moved in the way it has I do not know.

If it had not have been for the Sun overnight then Labour would have been ecstatic. A 7% deficit would have surely meant a hung parliament with Labour and the Tories very close to each other on seats.

The numbers tomorrow will start to show whether the bad news for Labour has had an impact. At the very minimum the Sun’s decision became today’s big political story over-shadowing the conference thus depriving Labour of a whole news cycle which it could have expected to be positive. My guess is that Labour will be back into the 20s tomorrow evening.

Mike Smithson

PaddyPower

Do the terms of the market make it a good bet?

Thanks to StJohn on the previous thread for spotting this new market fromPaddyPower on who, of a named list, will be the first to call on Gordon Brown to step down.

The chances are, as far as I can see, that none of them would go public and the likelihood is that Brown will leave of his own accord - probably after a general election defeat.

But under the terms of the market all bets are void “…should any of the below not call for his resignation before he leaves the position.” This means that you only lose if another “runner” on the list makes such a call.

Clearly Labour is going through a febrile period and tensions could be heightened if, next week, the Tories get a conference bounce in the polls. The more a Labour defeat looks likely the greater the chances of public calls for Brown to step aside.

The opening odds on Tony Blair were 100/1 - that’s now moved in to 20/1 - and I wondered whether John Cruddas was the best bet. Of all of them on the list he’s the most independent and most controversial. Purnell might be another possibility but his price is too tight.

Mike Smithson

What will today’s news do to YouGov’s daily tracker?

In trying to assess the political influence of the Sun’s switch Labour MP, Nick Palmer, made this astute comment on the previous thread:-

“..It occurs to me that the daily YouGovs are going to give us an excellent opportunity to test the power of the Sun endorsement to shift things. The YouGov taken today and reported tomorrow will show the impact of the speech plus a little bit of Sun. The one taken tomorrow and reported on Thursday will show the full impact of the Sun’s endorsement, probably magnified a bit by whatever effect the speech had starting to wear off…”

So to put this into context I dug out the above from an earlier poll from the firm this month where the firm’s newspaper weighting data appeared.

As I’ve pointed out here before YouGov tends to have far fewer Sun/Star readers in its samples compared to the population as a whole and generally they are weighted upwards. Thus in the poll example above 299 of the 2009 people who took are identified as Sun/Star readers but that is increased to 442 after the weighting has been applied. Thus YouGov assumes that 22-23% of the electorate are Sun/Star readers.

Given their respective circulations my guess is that just under a fifth of the weighted sample of YouGov polls are Sun readers.

Will they be influenced by today’s front page? We’ll have to see.

Mike Smithson