Populus Perspective
July 2008
Brown & out
The Conservative lead over Labour has narrowed slightly according to a new poll by Populus for The Times. The poll, conducted last weekend, puts Labour up a little to 28% and the Conservatives down a little, on 41%, with the Lib Dems on 19%.
Labour’s average in published polls, which stood at 40% last autumn, has been at the sub-Michael Foot level of 26% for the last three months, with the last twenty published polls all showing double-digit Conservative leads.
It says much about Labour’s parlous poll position that the 13% Conservative lead in this latest poll may be regarded as a sign that Labour has bottomed out. But the Populus poll findings are unambiguously bad for Gordon Brown personally. Asked which word, from various pairs, best describes each of the three main party leaders, more than two thirds of voters view the Prime Minister as ‘weak’ rather than ‘strong’ and nearly three quarters see him as ‘a loser’, not ‘a winner’, as ‘bad for me & my family’, and as ‘not up to the job of being Prime Minister’. Further analysis of the data reveals that the Prime Minister rates worse among women than men, and especially badly among the C2 socio-economic group – skilled manual workers who have been a key part of Labour’s traditional electoral base.
The Populus poll provides further evidence that the electorate has firmly made up its mind about Mr Brown – following last month’s Times poll which found that the Prime Minister’s rating on its ‘leader index’ has now dropped below the previous record low, achieved five years ago by the hapless Iain Duncan Smith.
Click here to see the detailed poll results
The limits of Conservative cut-through
The latest Populus poll gauged which party is most trusted on a range of policy topics. There has been a marked shift on most issues since the end of Gordon Brown’s honeymoon last autumn, with the Conservatives now enjoying leads on most issues.
The poll results also enable a comparison of where the Conservatives have gained most ground and where, relatively at least, they have had limited cut-through with voters.
Since September last year the Conservatives have made a 23% net advance in vote share, relative to Labour (i.e. they have gone from a 10% deficit to a 13% lead). The chart below shows how the party’s net improvement on various policy issues compares with this increase in topline support.
The significance of the swing to the Conservatives over the last few months on the issue of economic management is immediately clear. The chart also shows, however, that the improvement in the Conservative position has been substantially lower on most policy issues than it has been in voting intention. On most issues the number of voters rating the Conservatives as best party is well below their overall level of support – and in some cases this number has barely increased at all during the last few months: on the NHS, for example, 28% preferred the Conservatives in September 2007 and this has risen to only 31% now; 31% thought the Conservative Party best on school standards last September – only 1% more think this now.
Click here to see the detailed poll results
Petrol prices & political popularity
It has become an accepted wisdom in the US that rising petrol prices have a negative – and, many claim, direct - impact on Presidential approval ratings).
In the graph below a downward blue graph line indicates rising gasoline prices, providing an easy comparison with the (inexorable) slippage in support for the President.
Prompted by our client the AA, Populus has looked at the same comparison for the UK, plotting the Labour government’s average monthly poll rating against petrol prices over the last few years.
It is interesting to note that the price of a litre of petrol is markedly higher, and the government substantially more unpopular, now than in autumn 2000, when fuel protests brought the country virtually to a standstill for a few days. The graph indicates that petrol prices are not directly related to Labour’s poll rating: early last year, for example, fuel prices dropped somewhat but Labour’s support did too. But many will feel that there is some correlation, at least, in the steep and very similar angles of the two graph lines over the past few months.
Lisbon loss
It was Irish women in particular, rather than Irish voters in general, who rejected the Lisbon Treaty in last month’s referendum, according to a post-referendum poll just published by the European Commission. Men voted in favour of the treaty by a narrow 51/49 margin, but it was defeated because 56% of Irish women voted ‘no’ and only 44% voted ‘yes’.
The European Commission may be concerned to note that the only age-group that backed the treaty was over-55s – and that disagreement was highest among young voters: nearly two thirds of 18-24s voted ‘no’.
The poll found agreement on both sides of the argument that the ‘No’ campaign had been the more convincing – even among those who voted to approve the treaty, 57% thought that the campaign to reject it had been more persuasive.
The biggest single reason given for voting against the treaty was ‘not knowing enough about it’, with ‘protecting Ireland’s identity’ the next most common reason, followed by the desire ‘to protect Irish neutrality in defence matters’.
Nearly two thirds of those who voted ‘yes’ think the referendum result will ‘weaken Ireland’s position in the EU’ – a view shared by 42% of the majority of the Irish electorate who opted not to vote at all in the referendum, and by 24% of those who voted ‘no’.
A reputation to Google for
Google has the most positive reputation of any major US company, according to the latest reputation quotient (RQ) survey of American consumers by Harris Interactive – the first time Google has topped the list.
The technology sector as a whole is regarded as having the best reputation overall, with 73% positive ratings, well ahead of the travel & tourism sector, which is ranked second (55% positive). For four sectors (airlines, financial services, pharmaceuticals and energy) barely a quarter of ratings are positive; perhaps predictably the tobacco industry ranks lowest. The research found that men tend to give higher reputation scores to technology and finance companies, women to consumer goods companies.
When it comes to America’s corporate reputation as a whole, around 30% think it is good, but 70% think it is bad – and 10% think it is ‘terrible – and there is not much it can do to improve’.
The Harris Reputation Quotient research has established a close correlation between overall corporate reputation and the likelihood of consumers recommending or purchasing a company’s products or services.
Nearly half of those surveyed think that companies have a responsibility to ‘help in the global efforts to address social problems, including hunger and disease’. 3% of those surveyed take the opposite view, believing that ‘a company’s only responsibility is to generate profits’.
American PIE
There were moments during the US Presidential primaries when the polls were so wrong as to make news; most memorably the polls before the New Hampshire primary significantly understated support for Hillary Clinton. But how did the US pollsters do overall?
The excellent American poll analysis website fivethirtyeight.com has provided the answer. Starting with the overall ‘average error’ for each company conducting surveys during the primary season – i.e. the difference between the final poll before voting and the actual result – it then subtracted the maximum statistical margin of error for each poll (defined by its sample size) to produce a figure for ‘pollster introduced error’ (PIE). This represents “unforced error: the error that results from poor methodology” (as opposed to the error arising from the limits of the number of people interviewed, which is imposed by the laws of statistics and is beyond a pollster’s control).
The table below shows the primary season PIE for some of the most prominent polling organisations in America.
| Survey USA | 1.02% |
| Rasmussen | 1.03% |
| Field Poll | 1.47% |
| Mason-Dixon | 1.73% |
| ABC/Washington Post | 2.12% |
| Zogby | 2.16% |
| CNN/Opinion Research | 2.22% |
| LA Times/Bloomberg | 2.23% |
| American Research Group | 2.32% |
| Fox/Opinion Dynamics | 2.36% |
| Gallup | 2.40% |
| CBS/New York Times | 2.64% |
| Zogby Interactive | 5.73% |
This means, for example, that the final Rasmussen poll in each primary race differed from the result by +1.03% above its statistical margin of error. The average PIE for all the pollsters analysed was +2.3%. The last time British polling companies erred to this extent – during the 1992 General Election – it prompted a wholescale methodological review of political polling across the industry which led to innovations such as past vote weighting, and adjustments for spirals of silence. These have led the main polling companies in Britain to produce eve of General Election polls that actually differed from the final parties’ vote shares by less than the maximum sample-induced margin of error margin in May 2005 – giving them a PIE of less than zero – this even though British pollsters typically work with sample sizes of 1,000 or more – generally larger than their American counterparts - giving them a smaller potential sample error to start with.
It will be interesting to see whether the relatively poor performance of US pollsters during the Primary season will prompt similar revisions in methodology in America between now and the Presidential Election in November.
Holiday cutbacks
Nearly three voters in five (58%) have cut-back on their summer holiday plans as a direct result of the credit squeeze, the rising cost of living and the strength of the Euro against the pound, according to the latest Populus poll published this week in The Times.
41% have cut the length of their holiday, a third are either going to a less expensive overseas destination or holidaying in Britain instead of abroad, and nearly one in five (19%) have cancelled their summer holiday altogether.
Scotland (which splits exactly 50:50) is the only part of the UK where there is not a majority who have made cutbacks in their holiday plans as a result of concerns about the economic situation.
Subscribe
To receive regular polling updates click here
©2009 All text copyright Populus Limited