What really happened in London’s mayoral election in 2008?

By Declan Gaffney

In a piece for LabourList on Tuesday, Paul Richards offered some reasons for Ken Livingstone’s defeat in the 2008 Mayoral Election: “Most of all, he lost the support in outer London of white, hard-working families.” Paul cited other contributory factors as well, but it’s this assertion about where Ken lost votes in 2008 I want to consider.

Let’s deal with the obvious point first. To anyone familiar with London, this racialised contrast between inner and outer London is wholly inaccurate. Inner London is indeed by far the most ethnically diverse area in the UK, but much of outer London has levels of ethnic diversity which are unrivalled in most of the rest of the UK. 30% of working parents in those outer London boroughs are from BAME groups, compared to 5% in the rest of the UK[1].

“Hard-working families” is of course Labour code for skilled manual workers, the ‘C2′ social grade of market research jargon. Paul asserts that it was Ken’s loss of support among this group (or the white, child rearing part of this group) which had the biggest impact on the election result. But how does this compare with the actual voting patterns in 2004 and 2008? And how does Ken’s support compare to that of Labour, as evidenced in the simultaneous London Assembly elections?

The table below shows the correlation between ward level demographic characteristics and vote share for Ken and for Labour in 2004 and 2008, from the GLA’s excellent study. Where the correlation coefficient is positive, that means that the share of the vote tends to increase with the variable: vice versa for negative coefficients.

The correlation between white British ethnicity and voting for Ken was strongly negative in 2008, but it was also negative in 2004, and it was negative to roughly the same extent for Labour in both years. There is no evidence of a ‘Ken factor’ affecting his share of white British ethnic group votes more than Labour’s.

As for social grade, Paul’s story is pretty much the opposite of what happened in 2008. It is true that Ken did worse than Labour among ‘C2′ voters in both elections, but in 2008 the correlation with the C2 grade switched from very slightly negative to very slightly positive – in other words, nothing much changed. The significant changes were elsewhere. In 2004, Ken had done better among higher income groups and worse among lower income groups than Labour. In 2008 he did better among lower income demographics (D and E) than in 2004, and much worse among higher income demographics (AB and C1).

Do these data explain why Ken lost in 2008? Of course not. They just tell us something about how he lost, and what they tell us shows that Labour’s standard all-purpose explanation for poor performance – that it’s all to do with the ‘C2′s – doesn’t apply.

I don’t know how Ken did among ‘white hard-working families’, any more I suspect than Paul does, because as far as I’m aware that data hasn’t been produced. But, regardless, the question of how Labour’s mayoral candidate rebuilds support in outer London should not be framed in these terms. This is London, where ‘hard working families’, like all families, are far more likely to belong to ethnic minority groups than in other regions.

I have commented previously on Labour’s tendency to look at electoral defeat through the distorting prism of ‘C2′ votes taken in isolation. When applied to London Mayoral elections, this is even more deceptive, and more so again if assumptions about ethnicity are imported into the category of ‘hard-working families’.

Fortunately, both the current candidates for Labour’s mayoral nomination know their city far too well to be too influenced by off-the-peg accounts of why Ken lost in 2008.

[1] Author’s calculations, from figures in Spence L., Parents and Work in London (GLA 2006) Tables C16, C17