There is a fascinating – if not bizarre – Populus poll in today’s Times. David Davis has moved ahead of David Cameron amongst Conservative supporters, despite the fact that the same supporters think that Cameron would stand a better chance of winning, would lead a more united party and would put forward more bold and compelling policies. The same poll suggests that Cameron would cut the present 8 point Labour lead to just 2 points, while Davis would increase it to 11.

First the good news for the Davis camp – in Populus’s post-conference poll David Cameron was preferred slightly to Davis amongst voters in general (by 29% to 21%) but enjoyed a large lead amongst Tory supporters (by 45% to 15%). Today’s poll still has Cameron leading amongst the wider public (by 37% to 30%), but amongst Tory supporters Davis has caught up and overtaken Cameron and now leads him by 50% to 37%. While part of this is obviously a fall in Cameron’s support, the majority is due to a huge collapse in the number of people saying they don’t know – last month 41% said they didn’t know, this month only 13% said they didn’t know and those former don’t knows have apparently overwhelmingly broken in favour of Davis.

It is important to remember that this a poll of Conservative voters, not the Conservative members who will (or in many cases have) actually cast the votes in the leadership election. The question is whether or not this swing in support will be reflected amongst normal party members. The last YouGov poll of party members showed there had been a swing towards Davis – but that Cameron still enjoyed a large lead. While fieldwork for both polls started on Friday, Populus’s continued across the whole weekend so could have picked up a later swing. Opinions of Tory voters do not necessarily reflect those of party members, so we have no way of knowing until there is another poll of party members.

More interesting is why opinion has changed like this – public opinion doesn’t just change at random, there are almost always identifiable causes. Here is the mystery – asked who would make the Conservative party more in touch with ordinary people, Conservative voters think Cameron by 42% to 15%, they think Cameron is more likely to lead a united party by 36% to 12% and, by a considerable distance, think Cameron is more likely to win a General Election by 45% to 11%.

While the poll was taken after the Question Time debate, given the relatively low proportion of the public who actually watch Question Time, this is unlikely to be the direct cause of the whole swing. The natural assumption is that Conservative supporters have swung behind Davis in response to his well publicised policy statements on taxation, Europe and the return of Grammar schools. This would explain why Davis’s boost amongst Tory voters is not reflected amongst voters in general – he has been concentrating on policies that appeal directly to Conservative voters and those Conservative voters prefer the candidate with the ‘right’ policies to the candidate they think has the best chance of winning.

The problem with this explanation is that the poll also found that Conservative voters thought that David Cameron was the candidate most likely to “put forward bold and compelling policies on the economy, tax and public services” (by 33% to 14%). Clearly Conservative voters must have found something they like about David Davis or something they dislike about David Cameron to lead them to prefer Davis, but this poll doesn’t tell us what it is!

Finally the bad news for David Davis – the same poll included hypothetical voting intentions for a General Election where the party leaders were Gordon Brown, Charlie Kennedy and either David Davis or David Cameron. As ever, I shall add the caveat that these hypothetical voting intention questions are just that – hypothetical. Respondents know the context they are being asked in and use them to indicate preferences for leadership candidates. That said the contrasts are very large indeed. With the Conservatives led by David Cameron Labour would lead by only 2 points (CON 35%, LAB 37%, LD 20%); against David Davis Labour would enjoy an 11 point lead (CON 32%, LAB 43%, LD 18%).

UPDATE: A comment by Andrew Cooper on Mike Smithson’s website reveals that the Populus poll used a split sample for the questions meaning that the respondents who said they preferred Davis weren’t actually the same respondents who answered the questions about which candidate would be better at various things. In itself this makes no difference – once weighted both halves of the sample would be equally representative of the population. What it does mean is that the sample size for those questions was only 750 and, once you take only the people who say they will vote Tory at the election, the headline figure of 50% of Tory voters preferring Davis was based on only 122 so respondents. A sample that size has a margin of error of just under 9%, so potentially Davis’s lead could all be down to sample error (though it is more likely that he is ahead). That said, even at the extremes of the margin of error, Davis has still made huge strides in his level of support since the party conference.

Comments are closed.