Sadiq Khan’s re-election as Mayor of London on 2 May 2024 was historic. No previous Mayor has won a third term – Ken Livingstone tried and failed in 2008 and 2012, and Boris Johnson had moved on to other things by 2016. Khan’s 11.1 percentage point margin was also a record for a mayoral contest between candidates for the two biggest parties, though he fell just short of the distance by which Livingstone, running as an Independent, defeated the Conservatives’ Steve Norris in 2000. His vote share was also just short of the highest yet, achieved by Khan himself in 2016, and Johnson’s in 2012.
For the third time, Khan polled over a million votes, a mandate no other British politician has enjoyed, with the exception of Johnson in 2008. The scale of his victory was similar to that projected by pollsters Savanta and Redfield & Wilton. YouGov erred in Khan’s favour while whispers just before the count that it would be a close-run result turned out to be highly inaccurate, revealing the credulity of senior BBC journalists.
The London Assembly elections conformed to the broad pattern of the mayoral outcome. Labour maintained their leading position with 11 seats, the Conservatives dropped by one to eight, the Liberal Democrats and Greens held steady at two and three respectively, and Reform UK won a seat for the first time. The Tories lost West Central constituency to Labour and the South West seat to the Lib Dems, the latter being the first time a party other than one of the big two has achieved a constituency win. However, the London-wide list element of the Assembly elections, determined by proportional representation, compensated the Tories with one additional seat.
There is less to say about Khan’s third win than there was about some earlier mayoral elections. That is partly because it conformed to expectations but mostly because so much less detail has been published about the local results. Unlike previous elections, no data has been published about how each electoral ward voted and, as yet, there is no data is available below the level of Assembly constituency. It is not possible to match individual voters’ choices between the three ballots, so inferences from the aggregate numbers about what people were doing with their votes are not rigorous. But it is likely, that we can use the large numbers to tell a story, particularly if the simplest explanation also has a politically plausible logic.
Both Khan and Conservative candidate Susan Hall ran substantially ahead of their respective parties’ vote in the Assembly elections, particularly the London-wide list component. The same thing happened in 2021, though to differing degrees: Hall ran 5.5 points ahead of the Tory list compared to 2021 Tory candidate Shaun Bailey’s 4.6 per cent, but Khan’s advantage over the Labour list more than doubled from 1.9 points in 2021 to 4.1 points in 2024.
Neither appealed more than their parties to the extent that Boris Johnson did in 2012, as reflected in a 12-point advantage over the Tory list in 2012. But it seems likely that the change in electoral system to First Past the Post encouraged tactical voting for the frontrunners among electors who would previously have given them second preferences.
The pattern of mayoral over-performance of the party baseline and its changes since 2021 add support to the argument that it reflects tactical factors, particularly on the Labour side. Khan’s best performances relative to the Labour list were in Lambeth & Southwark (+11.2 points), South West (+11.1 points) and North East (+10.6 points).
Correspondingly, the Lib Dems in South West and the Greens in the other two constituencies were down by double-digit margins in the mayoral vote compared to the list. In each case, the net switch to Khan was greater than in 2021. It is plausible that there were 10,000 voters in each of North East and Lambeth & Southwark who chose the Greens on the list and Khan for Mayor.
Hall’s best performances relative to the Conservative list were in the strongest Tory areas of outer London Bexley & Bromley (+12.6 points, see chart above), Havering & Redbridge (+11.8) and Croydon & Sutton (+9.5). These were also the constituencies with the largest gaps between Reform UK vote shares of the list and mayoral ballots. It is likely that around half of Reform UK list voters supported Hall for Mayor.
The tactical argument is the same as it was for the roughly half of the Green vote who opted for Khan. However, while the Greens and South West Lib Dems account for most of Khan’s advantage over the Labour baseline, the Reform UK factor only explains about half of Hall’s overperformance. The rest is harder to trace. Hall did relatively well in South West (+7.6) so this may reflect the choice made by right-wing Lib Dems. It is also possible that some normally Labour voters in outer London chose to support her for Mayor instead of Khan.
Voters in South West (chart below) faced a particularly complex set of calculations. The mayoral election was predominantly Conservative v Labour, while the constituency seat was a three-way contest that could credibly be represented as being mainly between the Lib Dems and Conservatives, and the list election was a free-for-all. The results were personal and tactical victories for Khan for Mayor and Lib Dem Gareth Roberts for constituency AM. Nobody got much more than a quarter of the vote in the non-tactical Assembly list vote.
Compared to 2021, the biggest pro-Labour swing in the mayoral election was in the City & East constituency, home of London’s largest concentration of Muslims. Labour’s list and constituency Assembly votes were down in the constituency, but the party does not seem to have suffered a Gaza revolt on the scale seen in elections in the West Midlands or the towns of Lancashire and Yorkshire. Given that Labour’s leads were already very large in Newham and Tower Hamlets in 2021, it may be that Barking & Dagenham, where Khan won only narrowly in 2021, was what powered the local pro-Khan swing.
While Khan’s victory was to be expected in a city that has voted Labour at every election since 1997, at a time when his party was riding high in the national polls his comfortable but not landslide margin does point to a distinctive quality of City Hall politics. Incumbent mayors in other metro areas, including even Andy Street who lost narrowly in the West Midlands, polled above their party baseline. Yet Khan’s vote was below what Labour can expect in a general election.
London’s devolved politics is mature enough to be demanding even of a candidate with the city’s favoured party label. The result also reflects a price, willingly paid, for Khan’s political courage in persevering with Ultra-Low Emission Zone expansion despite sometimes hysterical opposition and the faint-heartedness of some of his Labour colleagues after the Uxbridge & South Ruislip parliamentary by-election outcome last July. It also, less honourably, reflects the unwillingness of some – as seen in the Brent & Harrow Assembly seat results ever since he became a mayoral candidate – to vote for a Muslim, no matter how respectful he has been towards London’s complex tapestry of faith.
Because London’s electorate is unwilling to make City Hall safe for any party, the next Mayor of London could well be a Conservative. Despite everything – a terrible candidate, bad polling, campaign videos that make sense only as job applications to the US Republican Party – nearly a third of voters supported Hall. London might be safe Labour territory in general elections, but at City Hall level the electorate seems to want to keep it competitive, despite the apparent determination of the Tories to throw the result.
If the party can find a candidate for 2028 who doesn’t appear to hate London, they should be able to capitalise on a “time for a change” mood against the likely background of Labour facing the costs of governing nationally. However, if it chooses to spend its first term in national opposition embracing madness, as parties tend to when evicted from power, it could probably contrive to lose in London yet again.
X/Twitter: Lewis Baston and OnLondon. Support OnLondon.co.uk and its writers for just £5 a month or £50 a year and get things for your money too. Details HERE.