Burnham!

The consensus in the media that’s developed over the weekend is that British prime minister Keir Starmer is ready to give up the fight to retain his job and will shortly, perhaps today, announce a plan for his departure. His successor is universally expected to be former mayor of Manchester Andy Burnham, who last week convincingly won a by-election for the seat of Makerfield to return to the House of Commons.

I’ll talk about Starmer’s departure another time; for now I want to look more closely at Makerfield, and whether the result justifies the dramatic aftermath. Although Britain’s circumstances are distinctive, there are also some lessons in it for other countries, including Australia.

Makerfield is a fairly typical working-class constituency in the west of greater Manchester; very white, reasonably well-off, and traditionally regarded as safe Labour. Its previous MP (who resigned to give Burnham his chance) won it in 2024 with 45.2%, 5,400 votes ahead of his nearest rival, the far-right Reform, on 31.8%. But that was in the context of a landslide victory; things have turned bad nationwide for Labour since then, and this was very much the sort of seat where Reform did well at last month’s local elections.

So there was no certainty, at least initially, that Burnham would win the seat. He did, however, have some big advantages: he was a popular and high-profile local mayor, and he was able to capitalise on both pro- and anti-government feeling. Those who liked the Starmer government would support him, but those who were disappointed with it could also back him as the most likely agent of change.

As the campaign progressed, and as pundits thought more about the realities of the situation, the sense that Burnham was at serious risk of losing fell away. Nonetheless, the size of his victory was unexpected: Labour’s vote was up by almost ten points and he beat Reform by more than twenty points, 54.8% to 34.5%, a margin of about 9,200 votes. It’s those numbers that have set off the collapse in Starmer’s internal support.

It’s pretty clear where the votes came from. The Liberal Democrats and the Greens, which in 2024 had 11.2% between them, fell to a combined 1.0%, almost matching the rise in the Labour vote. Conversely, the combined gain for Reform and for its new openly fascist rival, Restore (which had 6.8%), was very close to the fall in the Conservative vote, down from 10.9% to 2.2%. Turnout was 58.8%, up more than six points.

In other words, the British have got the hang of tactical voting. Faced with a case where the result was uncertain but there was no doubt about who the top two would be, left of centre voters rallied almost unanimously to Burnham to beat Reform. Conservative voters, on the other hand, seem to have rallied to the far right, although that may be deceptive: given the way the government has been travelling there has probably been some movement directly from Labour to Reform, counterbalanced by some Conservatives switching to Burnham.

Back in 2024, when it won 0.8% of the seats nationwide for its 14.3% of the vote, Reform was a strong advocate for some sort of proportional representation. But we haven’t heard much about that lately. Once it started polling in the mid-20s or higher, the prospect that first-past-the-post might deliver it a majority with around 30% of the vote became too tempting.

But that only works as long as your opponents are divided. If they combine effectively, 30% isn’t going to cut it. And as if to rub in the message, the only real vote-splitting in Makerfield came from Restore, whose voters, at the extreme anti-system end of the spectrum, showed no interest in helping Reform – just as Reform’s voters had no interest in helping the Conservatives in 2024. That wasn’t enough to make the difference between winning and losing, but in a more marginal seat it easily might have.

Many politicians, even in countries that don’t use the system, have a first-past-the-post mentality: if they’re ahead of their rivals they think that somehow means they “should” win, even if their support is well short of a majority. Bernie Sanders had that problem in 2020, when he thought he could leverage a third of the Democrat vote into the party’s presidential nomination. And we’re seeing it in Australia, where once One Nation hits 30% in the polls, people start talking about it winning a majority.

In reality, One Nation is eminently beatable even if its support should remain in the 30s. Our preferential system even removes most of the need for tactical voting – and proportional representation, which Burnham has previously supported, would do even better. But as Reform’s Nigel Farage is discovering to his cost, a sufficiently aware and interested electorate can do the job even under first-past-the-post.

How much of this Burnham can take personal credit for is unclear. Michael Jacobs at Inside Story tells us that “any other Labour candidate would undoubtedly have lost”; maybe they would, but that’s a brave call to make about a twenty-point victory. And as noted earlier, the circumstances in Makerfield were particularly favorable for Burnham: no-one really knows how well he will be received outside the north, especially when he is no longer able to simultaneously appeal to both the government’s supporters and its critics.

But there’s no doubt that Labour needs to do something, and Makerfield couldn’t realistically have been any better from Burnham’s point of view. Now it looks as if he’ll get to show his stuff on a wider stage.

(There were also two by-elections in Scotland that are worth a look, but this post is long enough already – we’ll get to them another day.)