On 30 June and 4 July two political parties achieved similar shares of votes – the Rassemblement National 33.2%. Labour 33.7% (excluding Northern Ireland, as do most opinion polls, it was 34.6%). The rest is history. Keir Starmer joined Tony Blair as the only post-war UK Prime Minister to bring in a parliamentary party of over 400 – not even Margaret Thatcher managed that. In another extraordinary outcome, the Liberal Democrats increased their MPs from 11 in 2019 to 72 with a vote share rising only from 11.5% to 12.2%. The Conservatives slipped to an abject 121 seats and 23.7%., the worst share for either of the two main parties since Labour’s 22.2% in 1918, before it even was one. Meanwhile the RN felt cheated as the two-round French system allowed their candidates to be picked off in straight fights by strategic opposition withdrawals between rounds; their second round vote was 37.1%, less than the total of far-right parties in the first round but a stark contrast to Reform’s 14.4% in the UK. All choices of electoral systems are indeed a journey into the unknown.
Meanwhile in Scotland there was a more familiar pattern as the SNP made a fatal slippage to the level where their candidates fell like ninepins to Labour. In 2017 37% of the vote yielded 35 seats for the SNP, in 2024 30% yielded nine. The Scottish Conservatives did well to hang on to five seats out of six with their lowest vote share in constituency electoral history (12.7%); Reform’s 7.0% was half their level south of the border. Liberal Democrats outdid first-past-the-post with six seats on 9.7%: victories in two huge Highland constituencies will make the electoral map a pleasingly-coloured sight for them.
Labour’s boost from one to 37 MPs was a reminder that seeing an electoral wipe-out as a disaster taking several elections or maybe a whole generation to retrieve is a misreading of how voters behave. The Labour-SNP vote gap is only 5% (35.3% to 30.0%) and sets up well the 2026 Holyrood elections when Scottish voters will be invited to complete the Labour project by returning Anas Sarwar as First Minister, with the uneasy collaboration of 1997-2007 still in memory. The SNP will surely be tempted to replace John Swinney with the increasingly prominent Kate Forbes as their standard-bearer.
Labour’s English vote level is similar to the estimated national equivalent vote in the English local elections on 2 May, when Conservative strategists correctly observed Labour’s vulnerability to pro-Palestinian independents in some constituencies. Indeed their seat losses to them (Birmingham Perry Bar, Blackburn, Dewsbury and Batley, Leicester South, and Jeremy Corbyn in Islington North) matched Reform’s MPs. They also precipitated the one Conservative loss to Labour, in Leicester East. This factor influenced Rishi Sunak’s decision on election timing, which gave up six months of a large parliamentary majority able to pass laws and manage the economy. As with Emmanuel Macron, another young man in a hurry, Sunak took a self-inflicted hit to his place in history, dragging himself down to Liz Truss territory.
The pollsters’ reputation also suffered. Regularly showing Labour at 40% of the national vote, as in 2017 they were five points out on the Labour vote but in the opposite direction. The much-discussed regression polls, which aggregated samples over days and weeks to allow segmentation by economic, social and political variables and then applied them to the profile of constituencies, were correct in alerting us to the hard-to-credit Liberal Democrat potential. YouGov was precisely right at 72 seats for them, but all these polls exaggerated the Labour-Conservative gap.
Most of the Scottish polls correctly picked up the Labour-SNP gap but unfortunately the most discrepant one was the final one, Savanta’s showing a 34-31 SNP lead. The broadcasters’ exit poll got SNP nearly right after anxieties about their Scottish sampling (10 against 9) and UK-wide were bang-on with Labour (410 against 411, or 412 counting the formerly-Labour Speaker). Coming out less well was Scottish electoral management, even with turnout down from 68% to 59%. Postal voting issues seemed to affect Scotland disproportionately, the first Scottish declaration was not until after 2.30 am, and three out of the final five results were from Scotland, Inverness humiliatingly delayed until Saturday after ballot paper discrepancies.
The 2024 election was both dramatic and boring, Labour’s dogged caution leaving space for Conservative mistakes. The voters’ sense of ‘time for a change’ showed an unerring ability to make a hit on the governing party – but in Scotland they could not turn them out of office. Nicola Sturgeon’s ‘de facto referendum’ concept, and her decision to admit Greens into government, are now retreating into history, leaving the SNP with some scope to recover the loyalty of pro-independence forces before 2026.