YouGov/Telegraph 26/07/06 38 (-1) 33 (nc) 18 (nc) +5
Ipsos-MORI/FT 24/07/06 36 (nc) 32 (-1) 24 (+3) +4
ICM/Guardian 23/07/06 39 (+3) 35 (nc) 17 (-1) +4
Populus/Times 09/07/06 36 (-1) 34 (nc) 19 (+1) +2

July’s opinion polls largely show no change. Labour’s level of support has remained almost static, despite the arrest of Lord Levy and most of the month’s polls being conducted at at time when the government was facing criticism for not condemning Israel’s action in Lebanon, which polls suggested was highly unpopular in Britain. The Conservative and Lib Dem shares of the vote were also static in most polls, the exception being MORI showing a leap in the Lib Dem that appears to be entirely out of line with other pollsters, and ICM moving from showing the lowest Tory level of support amongst the main pollsters to showing the highest.

Without much change in the headline voting intention polls, there were more interesting changes to the approval ratings of the party leaders. Tony Blair’s ratings have hit new lows, but perhaps more significantly David Cameron seems to have finally ended his honeymoon. YouGov’s poll showed his approval rating falling from +14 to +2, and MORI gave him his first negative approval rating. YouGov’s tracker poll suggested the declines dates from the “hug-a-hoodie” speech, though it is probably a combination of that and the fading away of the temporary boost he had enjoyed since the local elections.

Because ICM are now doing their monthly poll at the end of the month (presumably because they delayed it a week at Easter as bank holiday weekends tend to produce dodgy samples), we’ve reached a point where three of the four main pollsters are carrying out their polls in the final week of the month. I understand that Populus aren’t doing a poll this week, so unless any of the Sunday papers commission voting intention polls during August these may be last voting intention polls for almost a month.

With my usual caveats about the value of averaging out polls and swing calculators, if the sort of shares of the vote suggested by these polls was repeated at a general election it would produce a hung Parliament with Labour as the largest party.

(Oh, and here are the polls since the general election in pretty picture form, from Chris Lightfoot)

A golden rule in looking at polls should be to examine to the big picture, the trend in the polls as a whole, rather than trying to draw conclusions from individual polls. However, because I try to report every new political poll that’s published, this blog sometimes misses out on the bigger picture. Hence in addition to the normal reports on new polls, I’m also going to re-introduce something I used to do back in 2004 – a monthly round up of voting intention polls.

  Date CON LAB LDEM Lead
YouGov/Telegraph 28/06/06 39 (nc) 33 (+1) 18 (+1) +6
YouGov/Telegraph 23/06/06 39 (+1) 32 (nc) 17 (+1) +7
ICM/Guardian 18/06/06 37 (-1) 32 (-2) 21 (+1) +5
Ipsos-MORI/Observer 18/06/06 41 (nc) 34 (+3) 18 (nc) +7
Populus/Times 04/06/06 37 (-1) 34 (+4) 18 (-2) +3

The voting intention polls paint a fairy consistent pattern this month – the boost in Conservative support that saw them rise to the high thirties seems to have been consolidated. None of the pollsters reported any significant change in their support, and it’s fair to assume that their support has stabilized at around 38%. MORI show a slightly higher level of support for the Tories, probably because of their very harsh filter by people’s reported likelihood to vote.

Labour‘s position is less clear cut, but I suspect they are starting to recover slightly from their nadir after the local elections – ICM’s poll that showed Labour falling is from a Labour score of 34% in May that seemed anomolously high (ICM had shown Labour increasing support after the local elections when everyone else showed them slumping). The first data from YouGov’s daily political trackers did indicate a degree of Labour recovery, with the negative impressions of some ministers recovering and the Tory lead on public services like the NHS fading away. That said, Labour’s reputation for competence had not staged a recovery, and their polling figures are still noticably below those from before mid-April.

There is an unusual lack of variance in reported levels of Lib Dem support. In recent months the main contrast between the pollsters has been the level of support for the Liberal Democrats indicated by the different methodologies (at the most extreme level, after Kennedy’s resignation YouGov showed the Lib Dems diving to 13% in the same week that ICM showed them at 19%). Now three of the pollsters (YouGov, Populus and MORI) all have the Lib Dems at 18% in their latest poll. While the Liberal Democrats are obviously a long way below the share of the vote they achieved at the General Election, they have recovered from the despairs of Kennedy’s ousting and, if anything, the trend now seems to once again be in their favour, with the last three published polls all showing small increases in their support. Ratings for Ming Campbell however remain atrocious – the YouGov tracking data showed him slumping after the local elections and YouGov’s monthly poll found only 6% of people thought that he would make the best Prime Minister (though that said, like the Dunfermline by-election during their leadership crisis, the Bromley & Chistlehurst by-election shows it doesn’t necessarily stop them winning votes).

As I mentioned this morning, YouGov and MORI’s polls at least indicate a drop in the “Others” vote – YouGov has shown quite a consistent trend downwards, from a high of 15% down to 14%, to 12% to 10% in today’s poll. A large proportion of this is the BNP vote fading away having peaked at the local elections.

I am sceptical of the value of averaging out polls, and indeed on the value of using uniform national swing to predict how shares of the vote would translate into seats (see here), but if the sort of figures suggested by this month’s polls – say, CON 38%, LAB 33%, LDEM 18% – were repeated at the next election, then on a uniform swing on the new boundaries Labour would remain the largest political party, albeit short of an overall majority – CON 280 seats, LAB 298 seats, LDEM 42 seats, Others 30.