EXCLUSIVE: YouGov Political Trackers


Data from the first two months of YouGov’s daily political trackers on BrandIndex shows that the week before the local elections caused lasting damage to Tony Blair’s own reputation and to the Labour party’s reputation for competence. Local election success turned around public perceptions of David Cameron, whose ratings had been beginning to flag at the end of April. Meanwhile opinions of Ming Campbell have fallen sharply.

Since the mid-April YouGov’s daily BrandIndex consumer brand surveys have included political trackers, including approval ratings for the party leaders, whether people have positive or negative opinions of high profile politicians, party image, which party is best on particular issues and various other things. BrandIndex itself is a subscription service, and how the political data will be available in the long run is still being decided - but I’ve had the opportunity to look over and analyse the daily tracking data from the first two months.

Firstly there needs to be a caveat. BrandIndex is slightly skewed towards younger and wealthier people, because it is primarily a product aimed at business and clients are more interested in young trendsetting affluent people’s opinions. This means that, at present, the absolute figures would be skewed against Labour, but the changes and trends in the figures would still be significant. However, in order to counteract this bias we did a normal YouGov political survey a week or so ago, using YouGov’s normal political weightings, asked exactly the same questions, and then adjusted the tracker data to match.

Three main findings come out of the recent data - firstly Labour have sustained lasting damage from the recent difficulties that have engulfed them. As one would expect, their figures plummeted during the week of ‘Black Wednesday’, opinions of ministers fell, government approval fell, they fell behind the Tories on major issues and so on. This is hardly surprising given that the media were constantly reporting that the government were in crisis. What is more important is what measures fell and then stayed down.

While opinions of ministers like Patricia Hewitt have gradually recovered since the local elections, Tony Blair’s own ratings have remained significantly down. Perhaps more worryingly for Labour, since Blair himself will be gone by the next election, the party’s reputation for competence dropped sharply during the week of ‘Black Wednesday’ and has not staged any sort of recovery. In April around 25% of people associated Labour with being competent, the figure has now settled at around 17%. Labour aren’t just suffering because of bad headlines - there has been a shift in people’s perception of their competence.

Secondly there is David Cameron’s approval ratings - there has been some discussion in the media about Cameron’s star beginning to fade (notably an article by Andrew Rawnsley that described Cameron’s figures as being in steep decline that provoked much discussion on Mike Smithson’s Political Betting site). Those stories were based on old and sporadic data; the YouGov trackers do indeed show that Cameron’s approval ratings were just beginning to decline prior to the local election when he was being faced with criticism about having his chauffeur follow his bike carrying his briefcase and attacked by Labour as a blue chameleon… but the local election results changed that.

Cameron’s approval ratings and positive or negative perception rating both peaked after the Conservatives made the most gains in the local elections on May 4th and have been on a steadly upwards trend since then. The difference is presumably because Cameron is now seen as a ‘winner’, putting him in a positive light. The local elections came along at just to right time to give his image boost; it seems he has once again been very lucky indeed.

Finally there is Ming Campbell. There has been a lot of questions raised over Campbell’s performance as leader by the commentariat, there has been scant evidence until now though of whether real people think he is performing badly. Looking at the YouGov BrandIndex trackers there can be little doubt that this is not just a story within the Westminister bubble. Campbell’s approval ratings prior to the local elections were broadly neutral, with those thinking he was doing a bad job balanced out by those thinking he was doing a good job. Since the local elections they have declined sharply, and his approval rating is now -23. Interestingly the decline didn’t start straight after the election, but in the middle of the week afterwards when the commentariat started questioning his PMQs performance and mutterings about his leadership began.

The last few days of data covered were after a successful PMQs performance by Campbell and positive coverage of the Lib Dems tax plans. As yet they haven’t improved matters, but obviously these things can take time. The only good news for Campbell is that the proportion of people who don’t know what they think about his leadership - while declining - is still in the mid-40s, so there are plenty of people who haven’t mind their minds up yet.

More details, and graphs showing all these trends, can be downloaded in the report below:

pdf Download full report HERE.

These icons link to social bookmarking sites where readers can share and discover new web pages.
  • Digg
  • del.icio.us
  • Netvouz
  • DZone
  • ThisNext
  • MisterWong
  • Wists
  • Reddit
  • StumbleUpon
  • Technorati
  • blogmarks
  • feedmelinks
  • Furl
  • Spurl
  • NewsVine
  • Facebook

9 Responses to “EXCLUSIVE: YouGov Political Trackers”

  1. Anthony, offtopic, but have you tried adapting your “swing calculator” to show the effects of seats of regional swings (using YouGov’s regions, and perhaps taking some sort of average over 3 months to get large enough samples)? Or would this be pointless because the regional data isn’t weighted properly?

  2. Regional data isn’t weighted separately, but when samples are drawn up IIRC they are representative within regions, so they should still be representative-ish (YouGov polls are essentially “pre-weighted” - the sample is designed so it is representative to start with. They are weighted again afterwards to correct any deviation from that caused by differential response rates).

    The problem with using YouGov data is that the normal cross-breaks produced for the final tables are for the Television regions, which cannot be easily split up into constituencies. To get actual government regions extra cross-breaks would have to be produced.

    I might well do something like it in the run up to the next election, but I think it’s a bit silly this far out.

  3. Hi Anthony , Well done on getting hold of these interesting figures . Given the slating that the daily tracker polls received at the last GE , what is the M of E in the daily figures in this poll . Could not some of the short term trends actually be no more than normal poll variations ?

  4. The survey size each day is 2,000 people, but there is a split sample, so each of the political tracker questions is answered by about 625 people a day. That’s a margin of error of 4%.

    The figures in the report though are 5-day rolling averages, rather than single days figures, so are based on samples of around 3,125 people - a margin of error of just under 2%.

    Tracker polls are good at eliminating the worry of rogue polls. If you have a single poll then there is always the chance that a figure is just an outlier or a rogue. The probability of you coincidentally getting a similar outlier the next day is pretty small. The probability of you getting get another outlier in the same direction the following day is absurdly small, after a while the trend is undeniable (for five day averages read the same figure 5 days later and 5 days after that). Bottom line is still, just like normall polls, look at the trends, not the day to day changes.

    The Populus tracker at the election is somewhat unfairly used to have a go at them. It suffered from some specific problems related to tracker polls - firstly Populus had a relatively sample size - 250 a day. Secondly phone pollsters face a problem with non-contact bias on tracker polls, normally phone pollsters will ring back people they can’t contact the next day, or the day after - they try people they can’t reach about 8 times. On tracker polls where the data relates to just one day they can’t do that - to their credit Populus sought to minimise this by starting fieldwork very early in the day. Thirdly Populus had the misfortune to have a bank holiday weekend (which mess up poll results) just before election day - if you are doing a daily poll EVERY day you can’t get round that (this one doesn’t poll on bank holidays). Finally there is the inescapable fact that if you are doing a rolling poll, if you get one day’s duff data you are stuck with it for 3 or 5 days until it drops out. Populus got stuck with a duff days data from the bank holiday weekend and were stuck with it through till the election. Just bad luck and bad timing.

    It’s worth noting they did a second tracker for Michael Ashcroft that was lucky enough to get a decent sample on the Bank Holiday, and which gave them a wholly credible result.

  5. Thanks Anthony , will you be able to publish this data on a regular basis ?

  6. Nothings decided yet

  7. I think it is. Its time for change, its time for a Conservative government!

  8. I hope not, it would be such a pity to see this golden age come to an end. The Conservatives will give us high inflation, unemployment, crime, interest rate. Please not again!

  9. LET’S HOPE THAT PEOPLE WILL SEE SENSE IN TIME!! CAMERON IS JUST ANYBODY’S LAP DOG ON ANY PARTICULAR DAY!

    IMAGINE HIM AND HIS CHUM GEORGEY OSBORNE RUNNING THE COUNTRY!! WHAT ON EARTH DO EITHER OF THEM KNOW ABOUT FINANCES APART FROM COUNTING WHAT THEY HAVE HAD GIVEN TO THEM!

    LOOK AT WHAT THEY SUPPORTED IN THE PAST - RECORD INFLATION (27% 1980), RECORD INTEREST RATES (16%), RECORD UNEMPLOYMENT, PRIVATISATION OF GAS,ELECTRICITY, WATER AND NOW THE HIGH BILLS TO GO WITH IT, DECIMATION OF THE RAILWAYS, BUS SERVICES, PUBLIC SECTOR ETC! ALL OF THESE THINGS THE TORY TWO WERE IN SUPPORT OF.

    AS FOR HUG A HOODIE - WHAT WILL IT BE NEXT - KISS A COKE ADDICT?

    NOT FOR ME THANK YOU!!

    KEEP EM OUT!!!

    WILMSLOW CHESHIRE
    (GEORGEY BOYS CONSTITUENCY!!)

Leave a Reply