« Sticking with our John Kerry Call
Labour eases to 1.4 - is now the time to bet? »
h1

Where’ve the opinion poll protestors been since 1992?

June 23rd, 2004

Kinnock Sheffield
Neil Kinnock: Sheffield Rally: 1992 General Election

    Labour bias +6.5% IGNORE; UKIP bias +4.9% PROTEST

For nearly a decade and a half the opinion polls have had, when tested against real General Election results, an average pro-Labour bias of 6.5% and there’s been hardly a whisper from Labour MPs. The bias led to political gamblers making huge errors and losses at the 1992 General Election in which Neil Kinnock, above, made what amounted to a pre-election victory speech a few days ahead of his own party going down on votes by 8%.

Now, for the first time since then, we have a pollster showing a bias to Labour that’s down to 2.1% and large numbers of Labour politicians are protesting in an ongoing campaign.

    Ostensibly they are focussing on an over-estimate of 4.9% in the Euro Elections for UKIP, but is their real concern that what has been a given of British politics for so long - the pro-Labour bias of UK opinion polls - is no more?

The essence of their position would seem to be that it’s OK for the polls to consistently over-state Labour by 6.5%, but not to make errors of 4.9% with other parties.

What pollsters report, particularly in the run-up to an election, has a huge effect on the voters and it is not always the party that’s reported to be in the lead that benefits.

  • Would Kinnock’s Sheffield speech in 1992 have happened if there had been accurate polls?
  • What part did the pollsters play in causing the turnout in the 2001 General Election to drop below 60%? In the weeks before Labour leads of upto 28% had been reported and voters could have been forgiven for thinking that the result was a foregone conclusion. In the end Labour won on votes by just 9%
  • The pollsters’ embarassment was saved by the distortions in the Commons seat distribution that left Tony Blair with a 165 majority even though his vote share was down significantly.

    • Figures that are just plain wrong help nobody and it is not just political gamblers who have got an interest.

      That is why the recent comments have made such depressing reading. They seem to be attacking some of polling firms because they are not reporting what they want to hear.

    The internet pollster YouGov has come in for stick and there is little doubt that it is not as good at predicting vote shares for the smaller parties than it is for the Conservatives and Labour. The UKIP figure was an exaggeration and YouGov seem to record the Lib Dems at much lower levels than the other polls or much higher.

    But for the first time in a decade and a half political gamblers have a proven and reliable guide to how the main two parties are doing. Let’s be thankful for this and be just a bit wary of the pollster’s figures for the Lib Dems and the smaller parties.

    US ELECTIONS UPDATE. The latest polls show big leads for John Kerry. This follows some recent polls showing a Bush recovery and has happened in spite of a huge anti-Kerry ad campaign by the Republicans. The latest prices have him at 2.1 or 2.2. For some reason the Iowa price has taken a steep drop. Our call is BACK KERRY.

    Picture - http://news.bbc.co.uk/furniture/vote2001/in_depth/election_battles/92sheffield_rally.jpg



    MessageSpace Advertising

    5 comments to “Where’ve the opinion poll protestors been since 1992?”

    1. I don\’t understand your statement \”The bias led to political gamblers making huge errors and losses at the 1992 General Election\”, since the bookies were surely influenced by the same factors that you were. Unless, of course, you are suggesting that Ladbrokes and William Hill teamed up to produce a private, and the only accurate, opinion poll of the electrion. If you lost money on a bet, that\’s becuase you weighed up the evidence less well than the bookies. Full stop.

           by Sam June 23rd, 2004 at 8:49 am

    2. All I\’m saying is that punters who backed Labour in 1992, as many did, might have made different bets if the polls had been giving a better reflection of party support. The standard answer from the pollsters is that they reported what was true at the time but there was, in each election, a late swing to the Conservatives. Who knows? But we can only judge pollsters by comparing their output to real results, which is what I\’m doing, and if the \”late swing\” is a factor then they need to develop methodlogies to deal with it.

      The current issue stems from the UKIP surge story coming first from YouGov and, in the end, it was nothing like as large as the firm had predicted. Two days before the elections I posted the following - In particular the extraordinary figures that YouGov has been putting out on UK Independence Party support will be subject to huge scrutiny. Other pollsters have showed a UKIP surge but to nothing like the same extent. For if YouGov are getting it right then we can give more weight to their predictions on the Tory share, which until the UKIP affect clicked in for the Euro Election, was substantially higher than with the conventional pollsters. We will also be more critical of the conventional pollsters. If YouGov are not as accurate then the credence that can be attached to their figures will be lowered and we would imagine more political criticism.

      My view is that the YG methodoogy involving surveys restricted to those who volunteer to go on the YouGov panel might mean that a proportion of those questioned are much more politically aware than the norm. These people are more likely to react fast to trends.

           by Mike Smithson June 23rd, 2004 at 11:33 am

    3. humidifier…

      …

           by humidifiers January 13th, 2006 at 7:28 am

    4. slot-game…

      …

           by slot-game April 18th, 2006 at 10:41 pm

    5. Clothes steamer….

      Baby clothes. Infant girls spring clothes. Kids clothes. Hollister clothes. Clothes….

           by Clothes. July 1st, 2008 at 10:01 pm