In poll position (3 April 2008)
London elections 08: So far, the opinion polls have varied widely on the winners’ margin. Why is that?
The ICM poll showing Boris Johnson narrowly ahead of Ken Livingstone in the contest to be mayor of London is welcome news for Labour, who had feared that the contest was slipping away after two YouGov polls had shown Johnson 12 and 10 points in the lead. Has there been a pro-Livingstone swing, or is one or other of the polls just wrong?
Opinion poll methodology is an abstruse field, and blundering into it unwarily is dangerous, but the striking difference between the two polls may be something to do with how the polls were taken. Getting London – a vast, diverse place where the voters move around more frequently then elsewhere – right in polling terms is a challenge. Both ICM and YouGov are respectable and professional firms that try to get the answer right – the fact that YouGov’s client is the vehemently anti-Livingstone Evening Standard does not affect the way the pollsters do their job.
Sampling for YouGov was taken over the Easter weekend, and holiday times sometimes produce funny numbers (although a YouGov poll taken the previous week also produced a similar topline result). But there are some other issues. ICM ensured that ethnic minority voters were represented in the poll in proportion to the population in London (29% is the official number), and this is important because black and minority ethnic voters are particularly pro-Livingstone.
Some of the differences in weighting are also interesting. YouGov did not adjust for likely turnout. Does anyone expect, as the first YouGov poll implies, that for every three votes cast by people aged 45-54 there will be five cast by people aged 18-24? London may have a young population, but no real election, with the possible exception of one or two of the Democratic primaries this year, shows that sort of turnout pattern. However, ICM’s sample showed Johnson further ahead (48-40) among people who claimed to be certain to vote, so Labour cannot assume that the army of young pro-Johnson voters found by YouGov is entirely illusory.
There are staggering differences in the detail of ICM and YouGov, with YouGov finding that women were more pro-Johnson than men (a 14-point lead, rather than six-point) and ICM the reverse – men giving Johnson an 11-point lead but women giving him an eight-point deficit. The last YouGov poll shows that more people who identify with the Liberal Democrats intend to vote for Johnson (40%) than their own party’s candidate Brian Paddick (31%) which although possible seems peculiar and implies a 7% swing of Lib Dems from Paddick to Johnson in a period of just over a week since their previous poll.
The past two mayoral elections have seen unusual constellations of voters. In 2004 Livingstone’s strength was in inner London, even in areas that are normally Conservative or Lib Dem. He won all but one ward in Camden, for instance, even leading in the upper middle class heights of Hampstead Town. Steve Norris polled best in the outer boroughs, many of which are ambivalent about their allegiance to “London”. Johnson seems, whichever poll one looks at, to have maintained the Conservative advantage in outer London and pulled away some non-Labour voters in inner London who supported Livingstone in 2004. Interestingly, YouGov finds Labour’s vote in the assembly election pretty much identical to Livingstone’s support, also suggesting that Livingstone’s vote this time will be a more conventional Labour party coalition of support.
The polls are also consistent in showing that Johnson is doing better in the contest for Lib Dem second preferences, in marked contrast to 2004, when Livingstone tended to benefit. This reflects Johnson’s greater crossover appeal, Livingstone’s accumulated discontents of eight years in office, and the closer relations between Lib Dems and Tories in London. The two parties run several boroughs in coalition and work together on the Greater London Assembly in opposition to an informal Labour-Green alliance.
Livingstone has had a rotten few weeks. Labour’s national popularity has sagged badly in several polls, with the Conservatives taking a strong national lead that is also reflected in Londoners’ voting intentions. Closer to home, there has been the resignation of two of his City Hall advisers, including Lee Jasper, under clouds of suspicion, something that has encouraged a “time for a change” feeling. Voters who dislike the apparent cronyism of Livingstone’s City Hall and distrust his honesty and currently intend to vote for Johnson may not feel the same after a look at Johnson’s own chequered record in precisely these matters. Livingstone also has formidable basic strength in public opinion, in that he is regarded even by people not currently intending to vote for him as having done a good job as mayor. His specific policies attract more support than Johnson’s on issues such as the congestion charge and public transport. Livingstone is also articulate and clear on detail, while Johnson is not. While Johnson has done enough to dispel the impression that the campaign is merely an exercise in vanity, he still looks flimsy on detail and competence and is being kept out of the way by his hard-right Australian campaign adviser Lynton Crosby.
Polls, whatever technical issues may arise and however they are reported, are more reliable than the only alternatives – self-interested claims by the parties, and the water-divining approach to public opinion by which a journalist sniffs the air and gets a “feel” for how it is going. If ICM had shown a Johnson lead even half as large as YouGov’s, it would be tempting to claim that the race was already over. Defeatism was starting to pervade Labour’s approach to the campaign. The new poll should dispel that, and if the party knows what it is doing, prompt a vigorous fightback that has every chance of success. Livingstone’s choice is to give up wearily but gracefully – which has looked a possibility – or hammer Johnson on policies and personal competence. Labour needs to energise its difficult to reach electorate, and the closer and more publicised the election, the better for them. Johnson, Crosby and the Evening Standard – salivating at the chance to inflict a fatal blow in its long vendetta against Livingstone – are not going to play nice either. With so much at stake, and (according to ICM at least) still all to fight for, it will be a brutal April in London politics.