"Things can only get better"

Two more stonking by-election wins for Starmer; ReformUK now a real threat

This was another happy night for Labour activists, as two more Conservative seats fell to the opposition on massive swings. Labour overturned a 23 per cent majority in the Bristol suburbs to take Kingswood on a 16.6 point swing. But it will be Wellingborough that grabs the headlines, with Labour comfortably overturned a whopping 36 point majority on a massive 28.5 point swing, the second largest ever recorded in a by-election since the Second World War. The Wellingborough Conservative candidate, Helen Harrison, now has the dubious honour of securing the largest drop in Conservative support ever recorded in a by-election, as her party’s vote collapsed by over 37 percentage points.

There were also intriguing signals in the performance of the smaller parties. ReformUK secured their best two by-election results to date, and while they have not yet matched UKIP’s performances in the coalition years, they now pose a genuine electoral threat to the Conservatives. The Liberal Democrat vote fell sharply in both seats, suggesting Lib Dems remain willing to vote tactically for Labour in seats where they are best placed to oust the Conservatives. But the Green vote did not - the party was flat in Wellingborough and up sharply in Kingswood. Starmer may be paying a price for his decision to row back on Labour’s £28 billion green investment pledge, as the Green voters who care about this issue most become less willing to loan him their support.

The scores on the doors:

15th February 2024 by-election results

These results continue an exceptionally strong run of Labour performances in “big two” by-elections. Keir Starmer has now achieved swings of 20 points or more four times in the last year, equalling Tony Blair’s record in opposition. And the 28.5 point swing in Wellingborough was just a whisker away from Blair’s record for the largest swing to Labour ever achieved in a by-election, the 29.1 point swing recorded in Dudley West in 1994. Even the more modest-seeming swing of 16.6 points in Kingswood is in fact very large by historical standards - it is larger than Labour managed in any by election between the mid 1990s and 2022. A performance which would have been hailed as a smashing breakthrough just three years ago now looks like business as usual. This in itself tells the tale of an opposition on the path to victory.

Swings of 20 points or more to Labour in by-elections since 1945

Adding these latest results to the regular Swingometer performance charts confirm that Labour’s performances now put it firmlt on the path to victory. By-elections are weak signals, but we can reduce the noise by focus our attention on “big two” contests between the Conservatives and Labour through the decades. In the chart below I take a look at average swings from the government to the opposition in every parliament since 1945, and compare this with updated averages for Labour over the whole of the Parliament to date (grey bar) and in by-elections held since the “partygate” scandals broke (black bar) in the autumn of 2021.

Labour’s performance over the Parliament as a whole now matches or beats five of the seven oppositions who went on to win. The average swing to Labour over the whole Parliament equals that achieved by David Cameron, and exceeds that achieved by Margaret Thatcher, Harold Wilson or Winston Churchill. Starmer does even better if we exclude the early by-elections held during the Johnson government’s “vaccine honeymoon”, and instead look at performances since the “partygate” scandal triggered the first pronounced slump in Conservative support. Starmer’s performances in “big two” contests since December 2021 are now better than every opposition leader except one - Tony Blair. Labour’s average swing of 13.9 points under Starmer is just narrowly behind the 14.4 point average swing the party achieved in the the 1992-1997 Parliament.

Average government to opposition swings, “big two” by-elections, Parliaments since 1945

The growing strength of Starmer’s position can be made even clearer with a more basic comparison. In the graph below, I calculate the average performance in “big two” by election contests for every post-war opposition which went on to lose (grey bar) and every opposition which went on to win (black bar) and compare Labour’s track record with these. Losing oppositions achieve an average swing of four points, winning opposition manage an average swing of nine points. Labour has now managed an average swing of nearly eleven points in the Parliament overall, and nearly 14 points since the “partygate” story broke. By-election history suggests this is an opposition on course to win, and win well.

Average swings to the opposition since 1945 in Parliaments where the incumbent wins (grey) and when the opposition wins (black), and Labour’s performances since 2019 (light red) and since December 2021 (dark red)

UKIP 2.0? The rise of ReformUK

The other big headline from this week’s by-elections is this: the radical right revolt is back. ReformUK posted its two best election performances by far in these contests, winning double digit shares in both. A debate has been raging about the seriousness of the ReformUK threat, ever since the latest UKIP successor party began rising in the polls last autumn, with sceptics wondering whether the strong numbers might be a polling artefact, or whether Reform lacked the organisational strength to convert polling interest into ballot box performance. These by-elections provide a strong suggestion that Reform are real contenders, able for the first time to back up strong polling with solid election performances.

Reform’s resurgence is real, but Richard Tice’s turquoise troublemakers are no match yet electorally for Nigel Farage’s purple people. The chart below details all the double digit vote shares achieved by radical right parties in by-elections since 2010. UKIP dominate the chart - they regularly posted twenty point plus vote shares during the coalition years, culminating in two massive 2014 by-election wins for the Tory defectors Douglas Carswell and Mark Reckless in Clacton and Rochester and Strood in 2014. Reform’s immediate predecessor the Brexit party also polled close to 30 per cent in Peterborough. Reform’s recent showings are modest by comparison.

We can also see that Reform are under performing UKIP when we directly compare their performances this week to UKIP’s 2015 showing in the same seats. In Wellingborough Reform managed 13 per cent, but UKIP managed nearly 20 per cent in the 2015 general election. Reform managed 10 points in Kingswood, running five points behind UKIP’s 2015 outcome.

2024 ReformUK performance in Wellingborough and Kingswood by-elections, and 2015 UKIP performance in the same seats

While ReformUK are a pale shadow of their purple predecessor, they nonetheless pose a major threat to the Conservatives for two reasons. Firstly, much of the Conservative advance in general elections since 2015 has come through winning over UKIP voters by embracing Brexit. This leaves the Conservatives much more exposed to a challenge on the radical right flank - while UKIP were most attractive to Tory voters in the Coalition years, they won substantial numbers of voters from Labour and the Lib Dems too. ReformUK are drawing much more heavily from 2019 Conservative supporters, and the Tory voters they are winning and generally people who would be unlikely to back any other opposition party. Reform, in short, splits off votes the Tories would otherwise hope to hold and cannot afford to lose in an already harsh electoral climate.

That’s bad enough for the Conservatives, but the news gets worse because of a quirk of the 2019 general election. Nigel Farage stood his Brexit party candidates down in all Conservative held seats ahead of that contest, meaning Farage friendly voters in over 300 Tory seats could not support a local Brexit Party candidate. Most of those who voted likely followed Farage’s signal and backed the Conservatives Current ReformUK leader Richard Tice has pledged to stand candidates in every constituency, so in all of these seats, voters who previously had no option to vote radical right last time will get a chance next time. A split on the right is certain to reduce the Conservatives’ majority in most of these seats, making it easier for Labour to win. And the better Reform do, the bigger the split will be.

Turnout - yawnslide coming?

The Conservatives have sought to downplay this latest drubbing by point to a sharp decline in turnout. This is a familiar line which all struggling governments reach for after bad by-election results - turnout is always lower in these contests, so “voters didn’t turn against us, they stayed at home” is always available as a crutch for government representatives.

But there is a grain of truth here. Turnout has been unusually low in recent competitive by-elections, and the decline in turnout in the past couple of years has been substantially larger than it was in the mid 1990s, as the below chart shows. Turnout was down an average of 23 points across all by-elections at the tail end of the Major government - the average decline in the last couple of years is eight points higher. In competitive seats where Labour stood to defeat a Tory incumbent, the average decline was 16 points in 1996-7, but in the last few years it has been 12 points higher at 28 points. Faced with a choice between an exhausted and unpopular government and a cautious and uninspiring opposition, more voters than ever are opting to sit the contest out - even in seats where there is a strong chance of giving the government a bloody nose.

The last time by-election turnouts were this low, according to Sir John Curtice, was in the run-up to the 2001 general election, which saw turnout slump to a post-war record low. Getting voters to the polls may be a challenge for all parties in the general election to come.

Average turnout change in all by-elections (dark bars) and competitive seat by elections (light bars) 1996-97 and 2022-24

Nothing has changed

Labour have faced a wave of critical media coverage in the run-up to these by-elections. A protracted row over the party’s decision to water down its £28bn green investment pledge played out in public, then a scandal over leaked anti-semitic comments forced the party to drop its candidate in the coming Rochdale by-election and remove a prospective candidate for the general election. Polls showing a slight softening in Labour’s lead were seized upon as evidence of a growing crisis for the opposition, while other polls showing the lead growing were largely ignored.

We can expect more of this in the months to come - crisis and volatility are much more exciting stories than big, stable poll leads. These by-elections confirm what is also evident from the polling averages: nothing has changed. Labour remain dominant, and on course for a crushing win. The chaos likely to come in Rochdale two weeks from now will no doubt be covered widely, and drive a new range of speculation about Labour’s woes. But a by-election with so many unusual features will tell us little. Rochdale will come and go, but unless the polling starts to move, Labour will remain on course to win.

While team Starmer will be reassured that recent turbulence has not blown them off course, team Sunak will face yet more trouble from the Tory right after record ReformUK performances. Socially conservative Tory voices on the back benches and in the press will insist that only hardline action on immigration can prevent a split right and electoral oblivion. But devoting yet more time to migration could be the worst of all worlds for Sunak, encouraging the nightmare scenario rather than averting it. His “Rwanda scheme” is not going to deliver on his pledge to “stop the boats”, and is therefore unlikely to restore the trust of migration sceptical voters backing ReformUK, who, as Sam Freedman’s recent analysis has shown, have little intention of returning to the Tory fold.

More months of Parliamentary feuding over small boats and the Rwanda scheme may fail to win back ReformUK supporters, while further alienating the moderate voters the Tories are losing to Labour and the Liberal Democrats, who don’t see immigration as a priority and may conclude the government has nothing to offer on the issues that care about most.

Sunak himself created this trap by making an impossible, undeliverable pledge - to “Stop the Boats” - the centrepiece of his government. This doomed policy has helped foment the radical right revolt it was meant to see off, and the Prime Minister is now out of options. Failure on immigration means he cannot unify the right. Failure on everything else means he cannot compete in the centre ground. And losing both would mean electoral oblivion, an outcome which looks just a little more likely after this week’s results.