The usual pattern for opinion polling over conference season is that each party gets a shortlived bump in support stemming from their conference, but the effect vanishes quickly and within a couple of weeks of the end of the Conservative conference the parties’ poll ratings will be back to where they were before any of these events happened.

This pattern has been followed in 2013.

However, polls in mid-October have been rather variable, with Labour leads measured at three, six and 11 points in surveys published on Sunday 20 October. But, as Bob Worcester has always said, ‘watch the share, not the gap’. The Labour share seems to be reasonably consistent, within sampling variations, of 37-38 per cent, more or less where it was over the summer. The Conservative share seems to have a wider range of variation than Labour’s, with different polls in the last few days finding figures as different as 27 per cent (Opinium) and 35 per cent (Ipsos MORI). The United Kingdom Independence party share is even more inconsistent, with a high of 17 per cent and a low of eight per cent (the combined Conservative and UKIP share varies between 42 and 48 per cent). The Liberal Democrats are at 9-10 per cent except for one outlier at 14 per cent.

Labour’s lead is therefore probably somewhere around the 5-6 per cent mark – again, much like the pattern over summer 2013 although a much smaller lead than the party was enjoying at times during the year between March 2012 and March 2013. Supporting the evidence for a slight narrowing of the gap, public opinion seems to be taking a more appreciative view of the government’s record, with a net 25 per cent or so disapproving compared to net 40 per cent disapproval at the time of the ‘omnishambles’ in 2012 and 35 per cent or so at the start of the summer.

Among middle-class voters (ABC1 in social classification) it is startling to note that Labour’s performance in the polls at the moment is good by any historical standards. YouGov persistently finds Labour support among this group running at 36-38 per cent, compared to 34 per cent in the 1997 general election! The Conservatives are running a few points below their 1997 performance, suggesting a real swing to Labour among the middle classes. Why, then, is Labour not miles ahead in the polls? The answer is partly that the working class is still only a bit more Labour than ABC1s, and that social change over 16 years has increased the middle-class proportion of the electorate. The other is that, since 1997, and particularly in 2001 and 2005, the swing to the Conservatives was higher in the working-class social groups, and the drop in turnout was worse (it is, historically, quite a new phenomenon for social class to make such a difference to turnout in the UK).

Labour’s relative weakness among working-class voters is one of its more pressing problems as a party. Low turnout is a sign that people – particularly young and working-class people – lack a sense that politics can be used to effect any meaningful change. The challenge for Labour is to find a policy that is big in terms of changing people’s lives, achievable and (given the financial context) not too expensive. Sounds easy, doesn’t it? The public support and appreciate many of Labour’s smaller ‘on your side’ policies on energy pricing and payday lenders, but these perhaps do not seem big enough. Labour probably also needs to think about the centre, and the lost voters, more in terms of alienated Midlands and southern working-class voters than the classic picture of a middle-class car-driving mum.

———————————————————

Lewis Baston is senior research fellow at Democratic Audit and a contributing editor to Progress. He writes the Poll positions column, part of the Campaign for a Labour Majority. He tweets @Lewis_Baston