Are these a better pointer than voting intentions?

There is a theory, supported by polling for the 1992 general election, that leader approval ratings are a better indicator of electoral outcomes than the standard voting intention findings.

I sort of buy into this and have begun reporting in more detail the numbers from the two regular polls that ask this in a standard form - YouGov for the Sunday Times and the MORI monitor.

Above are the latest findings tucked away in the detailed data from the latest YouGov poll and the trends are not encouraging for Mr. Brown but provide something of a boost for Mr. Cameron and Mr. Clegg.

One well argued academic theory is that in the UK the PM needs a positive approval rating that is bigger than half of the two party vote. That does not happen in the latest numbers.

Mike Smithson

Is this a problem with having greater appeal to the young?

The table shows two sets of figures from the latest ICM poll: the proportion who voted for one of the main parties in 2005 and the proportion of their overall support now coming from the 18-24 old age-group.

This is relevant because past voters are more likely to vote next time and, amongst the young, there is a problem with the unregistered.

A couple of weeks ago the Electoral Commission published a report showing that 56% of the 18 -24 group are unregistered and 31% of people with black or minority ethnic backgrounds are not on the register.

Surprisingly this is not a question that is asked by pollsters - so a part of the vote shares being reported cannot be translated into real votes on election day.

Part of the smaller past vote proportion for Labour is down to those who did not vote on principle following the Iraq war. But an even bigger element is because of the greater appeal that Brown’s party has to the younger age groups - and if so many are not on the register what does that say about current poll numbers?

Things could be so tight in this election that we have to look at every angle.

UPDATE: Mark Park makes a valid point on the thread about the 56% unregistered figure which was how it was reported in the media. The figure, he notes, came from a series of pilot investigations which the Electoral Commission’s report explicitly says “cannot be used to report on national rates”. But the broad point remains - younger voters are less likely to be on the register than the rest of the electorate and that pollsters ought to be checking this.

Mike Smithson

Will this get the core vote out in the marginals?

On the latest Radio 4’s Any Questions the schools secretary, Ed Balls, made a revealing comment on what I took to be part of Labour’s defence strategy in the marginals.

When being pressed over UNITE’s financial support for Labour including his own seat on the outskirts of Leeds he replied: ” “Unite members make contributions from their pay packets to help the fight against the BNP…..”.

That helped him handle the issue during the programme and was, I suggest, a pointer to the general election campaign.

For traditionally Labour can do well when it can persuade its core supporters that there is a something to come out to vote against. In 1997, 2001 and 2005 it was the Tories - in 2010 in some seats it could be the BNP.

For although the Tory share and Cameron’s ratings have slipped, Labour and Brown are still some way behind, so the full frontal “beastly Tories” rhetoric might not be as potent.

Only five months ago Labour saw the BNP tactic working a treat in the Glasgow NE by-election as I commented at the time. Then a closing argument to core voters was about preventing Griffin’s party from saving its deposit.

Of course this only comes into play if the BNP is standing in a marginal. So a key element to look at in assessing LAB-CON encounters is whether there’s a BNP candidate. If there is then that might just make Labour’s defence a touch easier.

Mike Smithson