The Multi-Party Trap: How 31 Percent of the Vote Could Buy a Commons Majority

Published: 7 June, 2026

A system built for two parties is now being asked to referee a contest between six. Start with the arithmetic, and the verdict follows: this is not democracy recording a choice. It is a machine manufacturing one.

There is a number that ought to stop British politics in its tracks, and almost nobody is treating it with the seriousness it deserves.

On 4 January 2026, the polling firm More in Common published a constituency-by-constituency projection of the next general election. Based on a survey of more than 16,000 Britons, conducted between late November and mid-December 2025, its model produced a single, startling headline. If an election were held tomorrow, Reform UK would win 381 of the 650 seats in the House of Commons. That is a working majority of 112. It would be achieved on 31 percent of the national vote.

Let that sit for a moment. Sixty percent of the seats. Thirty-one percent of the vote. More in Common, to its credit, did not bury the implication. The firm described the projected outcome as one that would see Reform "winning 60 per cent of seats on 31 per cent of the vote, rivalling the 2024 General Election as one of the most disproportionate results in modern British history." In the same breath, the model had Labour - the party that won a landslide eighteen months earlier - collapsing to just 85 seats, a loss of 326. No major party in British history has experienced such a rapid reversal of fortune.

It would be easy, and lazy, to read this as a story about Reform, or about Nigel Farage, or about the collapse of the Labour government under Keir Starmer. Those are real stories, and they matter. But they are not the deepest story. The deepest story is the machine itself - the electoral system that takes 31 percent of a fragmented electorate and converts it into untrammelled power. The same machine that, in July 2024, took 33.7 percent of the vote and handed Keir Starmer a 172-seat majority. The same machine that, in 2015, gave the UK Independence Party one seat for 3.9 million votes.

First Past the Post was designed for a world that no longer exists: a world of two big parties trading power between them, each commanding well over 40 percent of the vote, the rest a rounding error. That world is gone. Britain now has six viable parties, and the system has not been told. What we are living through is not a realignment that the electoral system is faithfully recording. It is a distortion that the electoral system is actively manufacturing.

I want to be straight about method, because it governs everything that follows. This is a polemic, and I am not going to pretend otherwise - I think the system is indefensible, and by the end I will have said so in plain language. But the opinion is meant to be earned by the numbers, not substituted for them. So the deal I am offering the reader is this: I will do the arithmetic first, honestly, including the parts that complicate my case; I will concede the strongest arguments against my position rather than the weakest; and where I reach for a heated word, it will be because the figure underneath it has earned the heat, not because the figure was insufficient. There is a structural argument here - that First Past the Post cannot fairly represent a six-party electorate, which is true regardless of who is currently winning - and there is a crisis argument - that it is about to hand one specific party untrammelled power on a minority of the vote. The two are often run together. I will try to keep them apart, because the first is the durable one and the second depends on a number that is already moving.

Part One: The Anchor

Let us begin by establishing that the More in Common projection is not an outlier or a stunt. It is a serious model, and its conclusions are echoed - in direction if not always in magnitude - across the entire field of British psephology.

The full projected seat table is worth reproducing, because the shape of it is the whole argument:

Stare at the bottom half of that table for a while, because that is where the system reveals itself most nakedly. The Greens and the Liberal Democrats are projected to win almost identical shares of the national vote - 12 percent and 11 percent respectively. The Liberal Democrats convert their share into 35 seats. The Greens convert their slightly larger share into 9. Same votes, four times the seats. The difference is not popularity. It is geography. The Liberal Democrat vote is piled efficiently into the south-west and a string of affluent commuter seats; the Green vote is smeared thinly across hundreds of constituencies where it finishes a respectable third or fourth and elects nobody. More in Common's own director, Luke Tryl, flagged precisely this: most of the constituencies in the projection would be won on under 35 percent of the vote, and nearly half of Reform's seats would be

marginal - "broad but shallow, and potentially fragile."

That fragility cuts both ways, and we will return to it. But first, the cross-checks, because a single poll proves nothing.

The June 2025 MRP from YouGov, the first of the parliament, projected Reform on 271 seats - already the largest party, ahead of Labour on 178 and the Conservatives reduced to fourth place behind the Liberal Democrats. By September 2025, YouGov's updated model, built on a sample of 13,000, had Reform on 311 seats, just 15 short of an outright majority. In October, Electoral Calculus, modelling a Find Out Now poll, put Reform on 367 seats and a majority of 84. More in Common's January figure of 381 sits at the upper end of this range but is not alone there. Even Electoral Calculus's own January 2026 model, on a slightly different set of inputs, produced 335 Reform seats and a majority of 20.

The models do not agree on the precise number. They range from the high 200s to the high 300s. What every single one of them agrees on is the shape: Reform as the largest party, winning a quantity of seats wildly out of proportion to its vote share, while a vote share in the high 20s or low 30s is enough to threaten or achieve an absolute Commons majority. The disagreement is about whether Reform clears the 326-seat winning line. There is no disagreement that a party supported by fewer than one in three voters is in striking distance of total control of the legislature.

And the polling context behind these models is itself remarkable. By the spring of 2026, the combined vote share of Labour and the Conservatives - the two parties that have alternated in government for a century - had fallen to around 37 to 38 percent. At the 2024 election it had been 57.4 percent, itself a historic low. The Electoral Reform Society noted that just over 14 points separated the first-placed party from the fifth-placed party in the polling averages, the narrowest spread, and the most fragmented field, in modern British history. The PollCheck moving average in early June 2026 had Reform around 27 percent, Labour around 18, the Conservatives around 18, the Greens around 15, and the Liberal Democrats around 12. Five parties bunched within fifteen points of each other. This is a multi-party system in everything but its electoral plumbing.

Part Two: The Arithmetic of Distortion

Here is where we have to slow down and do the work that the news cycle never quite gets around to. What would these vote shares actually produce under a proportional system? Not as a utopian thought experiment, but as a simple sum.

There is a temptation here that I am going to resist, because resisting it is the difference between an argument and a stunt. The temptation is to allocate all 650 seats in direct proportion to the national vote - pure, unthresholded list PR - and report the result as the "true" distortion. Do that and you get the most dramatic possible number: Reform's 31 percent yields about 201 seats, against the 381 First Past the Post awards it, a distortion of +180. It is a striking figure and I have seen it deployed many times. But it is not honest, because nobody is proposing pure national list PR. No serious British reformer advocates it, and no established democracy uses an unthresholded national list. To compare the system we have against a system nobody wants is to win the argument against an opponent who isn't in the room. The +180 is a ceiling, not a fair measure, and I will treat it as such.

The honest comparison is against the systems reformers actually propose and that comparable democracies actually use. Take regional list PR allocated by the d'Hondt method - the system Britain itself used for European Parliament elections until 2020, applied within regions rather than as one national pool, which is roughly how Scotland and Wales already top up their parliaments. Regional pools and the d'Hondt formula both modestly favour larger parties and impose an effective threshold, so the leading party keeps a real bonus and the smallest, most thinly spread parties still lose out. Run the More in Common vote shares through that machinery and the picture looks like this:

(These regional figures are my own estimates and should be read as approximate; the precise totals depend on how regional boundaries are drawn and how high any threshold is set. They are deliberately conservative - they hand the leading party more than pure proportionality would. The SNP keeps a small bonus even here, because d'Hondt applied within regions reduces but does not eliminate the advantage that geographic concentration confers on a nationalist party whose vote is already stacked into Scottish seats; the same regional pooling that trims Reform's national bonus leaves a party that only stands in one part of the country relatively favoured.)

The single most important figure in this article is now in that table, and it is not the +180 ceiling. It is +136. Even under a realistic, larger-party-friendly proportional system of the kind Britain has actually operated, First Past the Post would still hand Reform around 136 seats more than its vote share warrants. Note what that bonus does: it carries Reform from roughly 245 seats - the largest party by a street, but 80 short of a majority and forced to govern by negotiation - to 381 seats and the ability to do more or less whatever it likes for five years. The whole distance between "build a coalition" and "untrammelled power" is manufactured by the voting system, not granted by the voters. That is the figure to challenge me on, and I have chosen it precisely because it survives the methodological objection that the +180 invites.

The mirror image of Reform's bonus is the Greens' penalty. A party supported by roughly one in eight voters wins nine seats, when even a larger-party-friendly proportional system would give it around 55, and pure proportionality nearly 80. The Conservatives, historic masters of the system, find themselves losing some 80 seats to the distortion; Labour loses around 55. Two of the parties that have governed Britain for a century would, on these numbers, be actively penalised by the system they have always defended.

The Additional Member System used in Scotland and Wales - which keeps a directly elected constituency MP and tops up with regional lists to correct the distortion - removes roughly 80 percent of First Past the Post's disproportionality, according to the Electoral Reform Society's modelling, while preserving the constituency link that defenders of the current system rightly prize. The Single Transferable Vote, used in Ireland, does something similar through multi-member constituencies and ranked ballots. None of these systems is perfect, and each over-rewards larger parties somewhat - which is exactly why I have modelled the distortion against them rather than against a purist ideal. But none of them, on any plausible set of assumptions, hands a party with 31 percent of the vote a majority of 112. That outcome is unique to First Past the Post.

We can put a single number on all of this. Political scientists measure the gap between votes and seats using the Gallagher Index - a least-squares calculation where zero is perfect proportionality and higher numbers mean greater distortion. Most established democracies using proportional systems score below 5. Ireland's STV system scored 1.96 in 2020. Sweden, using a modified Sainte-Laguë method, scored 0.64 in 2022.

Now consider the United Kingdom. Here is the trend across recent general elections, using Michael Gallagher's own dataset:

In 2017, when Theresa May's Conservatives and Jeremy Corbyn's Labour between them hoovered up over 82 percent of the vote, Britain's electoral system was almost proportional - 6.47 was the lowest score since 1955. That was the system working as designed, in the two-party conditions it was built for. Then the vote fragmented, and by 2024 the index had nearly quadrupled to 23.64 - the highest score in British electoral history, beating the previous record of 20.6 set in 1983. (Published figures for 2024 range from 23.6 to 24.4 depending on how independent candidates are grouped; Gallagher's own authoritative figure is 23.64.) As Make Votes Matter pointed out, there is only one other national election in any advanced Western democracy since the Second World War - France in 1993 - with a higher disproportionality score.

When you run the More in Common projection through the same formula, it produces a Gallagher Index of approximately 23 - essentially tying the all-time record set in 2024. More in Common's claim that its scenario rivals 2024 as the most disproportionate result in modern history is not rhetorical. It is arithmetically exact.

The crucial point is what that trend line means. The disproportionality is not a fixed feature of First Past the Post. It is a function of fragmentation. The more parties compete seriously, the more violently the system distorts. We have spent the last decade steadily breaking the two-party conditions that made the system tolerable, while leaving the system itself untouched. The 2024 election was the most disproportionate ever - until the next one beats it.

Part Three: The Price of a Seat

There is a more visceral way to feel the distortion than indices and counterfactuals, and it comes from the 2024 election - a real result, not a projection, that we can examine in full.

Labour won that election with 411 seats (412 if you count the Speaker) on 9.7 million votes. That works out at roughly one MP for every 24,000 votes. Now compare the other parties, using the Electoral Reform Society's figures: - Labour: one MP per ~24,000 votes - Liberal Democrats: one MP per ~49,000 votes - Conservatives: one MP per ~56,000 votes - SNP: one MP per ~80,000 votes - Greens: one MP per ~485,000 votes - Reform UK: one MP per ~820,000 votes

Read that list again from top to bottom. A Labour vote in 2024 was worth, in terms of seat-buying power, roughly thirty-five times a Reform vote. The Greens, who won nearly 1.9 million votes, took four seats. Reform, who won 4.1 million votes - more than the Liberal Democrats - took five. The Liberal Democrats, on 3.5 million votes, took 72.

Sit with that last comparison, because it is the cleanest illustration in modern British politics of how the system actually works. Reform UK won more votes than the Liberal Democrats in 2024 - 14.3 percent against 12.2 percent - and ended up with 5 seats against the Liberal Democrats' 72. The difference was not a single vote of popular support. It was geography. The Liberal Democrats had spent years carefully cultivating clusters of winnable seats in the south-west and the home counties, concentrating their vote where it could clear the winning post. Reform's four million votes were spread evenly and thinly across the whole country, coming a strong second almost everywhere and first almost nowhere. Under First Past the Post, coming second is worth precisely nothing. The system does not reward support. It rewards the geographic efficiency of support, which is a different thing entirely, and a thing that has nothing to do with the will of the electorate.

The aggregate consequence is staggering. The Electoral Reform Society calculated that in 2024, 73.7 percent of all votes cast - 21.2 million ballots - did not affect the result. Some 16.6 million people voted for candidates who lost; another 4.6 million were piled up as surplus votes for winners who had already won. Nearly three-quarters of the electorate, by this measure, might as well have stayed home. One in three voters told pollsters they had voted tactically - that is, for a party that was not their first choice - simply to influence an outcome the system would otherwise have decided without reference to them.

Now project this forward. In the More in Common scenario, the beneficiary changes but the mechanism does not. Reform becomes the party winning seats cheaply, picking up constituency after constituency on pluralities in the low-to-mid thirties because the opposition to it is split four ways. The Greens, on 12 percent of the vote, are once again left with a handful of seats. The party that was the system's biggest victim in 2024 becomes its biggest beneficiary in 2029 - and absolutely nothing about the underlying injustice has changed. Only the direction of the distortion has flipped.

A necessary admission

I need to stop here and answer an objection, because a fair-minded reader - and certainly a Reform-supporting one - will have noticed something. I have described Labour's 2024 result (33.7 percent of the vote, 63 percent of the seats) as a "distortion," soberly, as the baseline that shows how the mechanism works. I have described Reform's projected result (31 percent, 59 percent) with words like "untrammelled power" and "do whatever it likes." These are arithmetically near-identical outcomes. Why the different temperature?

The honest answer has two parts, and I want to separate them, because one is legitimate and one is not.

The illegitimate part first: some of the asymmetry is simply that I find the prospect of a Reform government more alarming than the Labour government we got, on policy grounds that have nothing to do with electoral mechanics. I am not going to dress that up as arithmetic. If you support Reform, you should discount the adjectives accordingly, and you would be right to. The mechanism does not care who it crowns, and a principled objection to it has to hold whether the beneficiary is a party you like or one you loathe. A Reform voter who says "you only mind now that it's our turn" has caught me fair and square on the parts where I let policy distaste leak into a structural argument, and those parts should carry no weight with anyone.

But there is a legitimate part, and it is structural rather than partisan. The two results are not quite the arithmetic twins they appear. Labour in 2024 was one of two parties that between them still took 57 percent of the vote; its 33.7 percent led a field in which the runner-up was close behind in vote terms. Reform's projected 31 percent leads a field so fragmented that it sits more than ten points clear of the second party, converting a plurality into a supermajority of seats precisely because the opposition is shattered into four or five pieces that cannot coordinate. The more fragmented the field, the larger the seat bonus the system hands the leader - which is exactly why the Gallagher Index, as we will see, keeps climbing. A 31 percent majority in a six-party field is a more extreme artefact of the mechanism than a 34 percent majority in what was still residually a two-and-a-half party field. That is a difference in the distortion itself, not in my sympathies, and it is a fair basis for treating the projected result as the more severe case.

So: dock me for the policy leakage, but the structural point stands. And it is the structural point I am resting the argument on.

Part Four: The Irony of Reform

This is the moment to confront the central irony of the whole situation, because it tells us something profound about why the system survives.

Nigel Farage spent the better part of two decades as the loudest advocate of proportional representation in British public life. He had to be. Under First Past the Post, his vehicles - UKIP, then the Brexit Party, then Reform - were systematically robbed of representation. In 2015, UKIP won 3.9 million votes, 12.6 percent of the national total, and was rewarded with a single MP. It was, at the time, perhaps the most disproportionate outcome for a single party in British history, and Farage was entirely right to rage against it. He understood, better than almost anyone, that the system was rigged against insurgents and in favour of the established duopoly.

And now the same system is poised to make him Prime Minister with a three-figure majority on a vote share that would have left him in opposition under any proportional arrangement on earth.

What is fascinating - and instructive - is how Reform's own voters have responded to this reversal of fortune. As recently as January 2025, around 55 to 56 percent of Reform voters told YouGov they favoured changing the voting system to a more proportional one. By early 2026, that had fallen to 45 percent, while the proportion of Reform voters actively favouring First Past the Post had risen from 17 percent to 31 percent. The British Social Attitudes survey recorded the same shift even more sharply: support for electoral reform among Reform supporters fell from 79 percent to 61 percent in the space of a year. As the system began to work for them, Reform's supporters discovered they rather liked it after all.

This is the iron law of electoral reform, and it explains almost everything about why Britain is stuck. Parties and their voters do not hold principled positions on voting systems. They hold positions that track their immediate self-interest, and they revise those positions the moment the arithmetic changes. Farage was for PR when PR would have helped him and is conspicuously quiet about it now that First Past the Post is about to hand him the keys to Downing Street. There is no hypocrisy unique to Reform here. It is simply the clearest current example of a universal dynamic - one that, as we are about to see, has trapped the Labour Party in exactly the same way, just with the polarities reversed.

Part Five: Why the Turkeys Won't Vote for Christmas

If a clear majority of the public has consistently supported proportional representation - and according to the British Social Attitudes survey it has, by margins running from comfortable to overwhelming, for years - then why does Britain still use First Past the Post?

The answer is not public opinion. The answer is that the only party in a position to deliver reform is the Labour Party, and the Labour Party will only deliver it when it cannot win under the existing rules, which is precisely when it lacks the power to change them. This is the trap, and it is worth walking through its mechanics carefully, because it is the real reason nothing changes.

Start with the precedent everyone reaches for: the 2011 referendum on the Alternative Vote. The result was decisive - 67.9 percent voted No, 32.1 percent voted Yes, on a turnout of 42 percent. The No campaign won in 430 of 440 counting areas. Defenders of First Past the Post cite this result constantly, as though it settled the question for a generation.

It did no such thing, and it is important to be precise about why. The Alternative Vote is not a proportional system. It is a majoritarian one - voters rank candidates, and if no one wins 50 percent the lowest candidates are eliminated and their votes redistributed until someone does. It can actually produce results more disproportionate than First Past the Post, not less. Nick Clegg, who was forced to campaign for it as the price of coalition, famously called it "a miserable little compromise." The 2011 referendum was a vote on a system almost nobody actually wanted, held at the nadir of the Liberal Democrats' popularity, framed by both main parties' machines as a verdict on Clegg personally. To treat its result as the British public's considered rejection of proportional representation is a category error. The British public has never been asked about proportional representation. On the question it has been asked - the British Social Attitudes question about changing the system "to allow smaller parties to get a fairer share of MPs" - it has answered yes, by growing margins, for years.

So the barrier is not the voters. It is the Labour Party leadership, and here the story becomes one of deliberate choice.

In September 2022, Labour's annual conference voted, overwhelmingly, in favour of proportional representation. It was no fringe gesture: 129 motions on changing the voting system were submitted that year, and at least 140 constituency parties backed PR motions, making it the most popular single issue on the conference floor two years running. The Labour Campaign for Electoral Reform has cited grassroots support running above 80 percent. A Survation poll of Labour members in early 2026 found two-thirds backing PR. The party's own affiliated unions - Unite, Unison and others - have lined up behind reform. On paper, the Labour movement is a pro-PR party.

And yet. As that very conference opened, Keir Starmer told The Observer that electoral reform was "not a priority for me." The conference vote was noted and ignored. Proportional representation did not appear in the 2024 manifesto. And then, in July 2024, First Past the Post did for Starmer exactly what it had done for every Labour and Conservative leader before him who happened to lead the largest party in a fragmented field: it converted 33.7 percent of the vote into 63 percent of the seats and a 172-seat majority. The system that the Labour membership wanted to abolish had just delivered the leadership near-absolute power. The incentive to abolish it evaporated on the instant.

This is the trap in its purest form. The New Statesman captured the leadership's calculation with brutal clarity: for the people who run the Labour Party, "long periods of no power are better than consistently sharing power." Proportional representation would mean Labour almost never governs alone again - it would mean coalitions, negotiation, the permanent sharing of power with Greens and Liberal Democrats and whoever else. First Past the Post offers a different bargain: long stretches in opposition, punctuated by occasional spells of total, unconstrained control. Faced with that choice, the Labour machine has consistently preferred the lottery to the partnership. It would rather roll the dice on winning everything than guarantee itself a permanent share of something.

The Conservatives are, if anything, less ambiguous about it. They campaigned for No in 2011, reaffirmed their commitment to First Past the Post in their 2024 manifesto, and in office changed the voting system for elected mayors from the Supplementary Vote back to First Past the Post. They have always been the party least sympathetic to reform. The one flicker of movement is telling: after their own catastrophic wipeout in 2024, the share of Conservative voters supporting PR doubled, from 24 percent to 52 percent. Even Conservative voters, it turns out, discover the appeal of proportionality the moment the system stops working for them - the same iron law that governs Reform, running in the opposite direction.

So we arrive at the structure of the trap. Reform out of power was for PR; Reform near power is cooling on it. Labour out of power was for PR; Labour in power abandoned it. The Conservatives out of power are warming to it; in power they entrenched the opposite. At every turn, the party with the power to change the system is the party with the least interest in doing so, and the party with the most interest in changing it lacks the power. The only configuration in which reform could happen is one where a pro-PR party wins a majority under First Past the Post and then chooses to abolish the very system that just rewarded it - an act of principled self-sacrifice that no governing party in Britain has ever attempted.

But here I have to check my own pessimism against a fact I would prefer to ignore, because it is the single most relevant precedent in the world and reformers and sceptics alike tend to skate past it. A Westminster country with a near-identical constitutional inheritance to Britain's did exactly this, within living memory. New Zealand used First Past the Post, inherited from the same imperial source as we did, and in the 1980s found itself in precisely Britain's predicament: two successive elections in which a party won more votes nationally but lost the seat count, and a growing sense that the system no longer represented the country. A Royal Commission recommended the Mixed-Member Proportional system in 1986. The question went to a non-binding referendum in 1992 and a binding one in 1993, and New Zealanders voted to abolish First Past the Post and adopt MMP. The 1996 election was the first under the new rules. The National Party, which had governed under and benefited from the old system, contested the change, lost the argument, and then won power under the new system anyway.

This matters because it punctures the fatalism - including some of my own. It is not a law of nature that Westminster systems cannot reform themselves. One did. But notice how it happened, because the mechanism is the whole lesson. Reform did not come from a governing party volunteering to fall on its sword. It came from a Royal Commission that took the question partly out of the hands of self-interested incumbents, followed by referendums that let the public override the parties whose careers depended on the status quo. The self-interest trap was not defeated by appealing to politicians' better natures. It was defeated by routing around them - by building an institutional process that could reach a decision the parties themselves would never have reached unaided. That is the difference between New Zealand and Britain so far: not that New Zealand's politicians were more high-minded, but that its constitutional machinery offered a path that did not require them to be.

Part Six: Cracks in the Wall

And yet the situation in 2026 is not quite static, because the trap depends on First Past the Post reliably benefiting one of the two big parties, and that assumption is now breaking down.

This is the genuinely new thing. For a century, the deal made sense for Labour and the Conservatives because, between them, they could count on the system to deliver one of them a majority. They were the joint beneficiaries of a duopoly. But a system that rewards whoever leads a fragmented field only benefits you if you are the one leading it. With their combined vote share down around 37 percent, neither Labour nor the Conservatives can rely on being that party any more. The instrument they preserved to protect themselves is now cocked and pointing at both of them. In the More in Common projection, it is Reform that collects the seat bonus of roughly 136, while Labour loses some 55 seats to the distortion and the Conservatives around 80. For the first time, the two architects of the system are simultaneously on the receiving end of it.

That changes the political calculus, and you can see the change beginning. In early 2026, 64 Labour MPs backed an amendment to the Representation of the People Bill calling for a national commission on the voting system - a small number, but not nothing, and a sign that self-interest is starting to pull some of the parliamentary party towards reform now that the system threatens to annihilate it. Andy Burnham, the Greater Manchester mayor and a long-standing supporter of PR, has argued that reform would need to be put to the country in a manifesto. Zack Polanski, who took over the Greens in September 2025 and oversaw an extraordinary surge in membership - past 200,000 by March 2026, with the party leading among voters under 25 - has been explicit that he wants enough seats to act as a "kingmaker" who could force electoral reform onto the agenda as the price of any future arrangement.

I should be honest that this last route sits awkwardly with my own argument, and I would rather flag the awkwardness than hope nobody notices. A Green party with, say, 12 percent of the vote using coalition leverage to extract a wholesale change to the constitution is precisely the disproportionate influence by a small party that defenders of First Past the Post warn about - the very thing I am about to take seriously in the next section. I cannot cheer it as a clever path to reform and then wave away the identical concern when an opponent raises it. So let me be consistent: small-party leverage in coalition formation is a real cost of proportional systems, it is a cost I think is worth paying, and a Green-brokered PR deal would be an instance of exactly that cost, not an exception to it. If it happened it would be reform achieved by the mechanism I am criticising, which is an irony I will own rather than launder.

The deeper context is a Labour Party in open crisis. Starmer's net favourability hit minus 57 in January 2026, joint-lowest of any prime minister on record bar Liz Truss. The May 2026 local elections were a rout: Reform gained around 1,450 councillors and took 14 councils from a standing start, while Labour lost nearly 1,500 councillors and 38 councils, and the Conservatives lost more than 560 councillors. On the same day, Labour fell to third in the Welsh Senedd behind Plaid Cymru and Reform, ending more than a century of Labour dominance in Wales. Within weeks, more than 95 Labour MPs had reportedly called on Starmer to resign or set out a timetable for doing so; the Health Secretary, Wes Streeting, resigned; and Andy Burnham manoeuvred towards a return to Parliament as a potential leadership challenger.

A wounded governing party, staring down a Reform majority manufactured by the very system it once defended, may yet rediscover its enthusiasm for proportional representation. But notice what that would be: not a principled conversion, but the same self-interest that has always governed the question, finally pointing in the direction of reform. And notice the cruel timing. Labour would need to legislate for PR while it still holds its 2024 majority - that is, while the system is still notionally working for it - in order to protect itself from the next election, when the system is projected to destroy it. That requires the leadership to act against the short-term logic of incumbency in anticipation of a future defeat. It is exactly the act the trap is designed to prevent. Whether a party in crisis can summon that kind of strategic foresight, against every instinct of self-preservation that got its leaders where they are, is the open question on which the next decade of British democracy may turn.

Part Seven: The Case for the Defence

Honesty requires engaging with the strongest arguments for First Past the Post, because they are not frivolous, and the people who make them are not fools.

The first and best argument is stable, decisive government. Proportional systems, the case runs, produce coalitions, and coalitions produce horse-trading, instability, and the empowerment of small parties who extract concessions wildly out of proportion to their support. First Past the Post, by contrast, manufactures clear majorities that allow governments to govern, to take difficult long-term decisions, and to be held cleanly accountable at the next election. A government with a working majority can act; a government dependent on coalition partners often cannot.

I want to give this argument its strongest form rather than its weakest, because the weak form is what reformers usually knock down. The weak form points at Italy's revolving-door governments and Israel's hostage-taking by fringe religious parties and says: there is your proportional representation. But those are the worst advertisements for PR, not representative ones - Italy's instability is a product of its particular party culture and constitutional history, and Israel's of an ultra-low electoral threshold applied to a deeply polarised society. The honest comparison is with Germany, Sweden, Denmark, the Netherlands, New Zealand - proportional democracies that have produced stable, accountable, long-lived governments for decades without anything resembling Italian chaos. If you want to defend First Past the Post, point at those countries and explain what Britain gains that they lack, because they are the real benchmark. Cherry-picking Israel is the mirror image of my own temptation to cherry-pick the +180 figure, and it should be resisted for the same reason.

And the strongest form of the stability argument is not really about coalition breakdown at all - it is about incentives, and it is the one objection in this whole debate that genuinely gives me pause. The argument runs like this. First Past the Post forces parties toward the median voter, because winning a constituency means assembling the broadest possible coalition of support within it; you cannot win from the fringe. Proportional representation weakens that pull, because every additional vote converts directly into a sliver of a seat, so a party can prosper by consolidating and exciting its base rather than by reaching for the centre. Over time, the claim goes, PR rewards differentiation and purity, FPTP rewards breadth and moderation, and the long-run effect on a country's political culture - not just on a single election's seat maths - is to make politics more centripetal under FPTP and more centrifugal under PR. That is a serious argument, made by serious people, and it is not answered by any amount of disproportionality arithmetic, because it is a claim about behaviour, not about seats.

I think it is overstated, and here is why, but notice that I have to actually argue the point rather than wave it away. The centrist discipline FPTP supposedly imposes is exactly what has just collapsed: a system that was meant to pull parties toward the median has instead produced a fragmented field in which a party can win power from 31 percent by exciting a base and letting the opposition split, which is the centrifugal outcome the argument warns about, arriving via FPTP rather than despite it. Meanwhile the coalition-formation stage under PR imposes its own centripetal discipline - parties that want to be in government have to be tolerable to potential partners, which is a moderating pressure of a different kind, applied after the election rather than before it. The empirical literature comparing polarisation across electoral systems is genuinely mixed, not a clean win for FPTP. But "mixed" is the honest verdict, and a defender of the current system who leads with the incentive argument rather than with Israel is making the case I find hardest to dismiss.

The second argument is the constituency link. Under First Past the Post, every one of the 650 seats has a single, identifiable Member of Parliament, tied to a defined patch of the country, who can be lobbied, blamed, and thrown out. Many proportional systems, especially closed-list ones, sever this link, turning MPs into anonymous names on a party slate accountable to the party machine rather than to voters. The constituency relationship is one of the genuine strengths of the British system, and any reform that destroyed it would lose something valuable.

The third argument is that First Past the Post keeps out extremes by setting a high bar for fringe parties to win representation. The fourth is simplicity: one cross, most votes wins, results overnight, no transferable-vote arithmetic or list calculations that voters struggle to follow.

These arguments deserve respect. But notice what has happened to them in the conditions of 2026.

"Stable single-party government" now means a single party with the support of fewer than one in three voters wielding total power for five years, having been actively opposed by the majority in most of the seats it holds. That is not the considered stability the argument imagines. It is the concentrated, unaccountable power of a faction. A 112-seat majority on 31 percent of the vote is not a strong mandate; it is a weak mandate amplified into an overwhelming one by an accident of vote distribution.

"Keeps out extremes" is the argument that has aged worst of all. Whatever one thinks of Reform - and reasonable people think very different things - the claim that First Past the Post protects Britain from insurgent parties is now precisely inverted. The system is not keeping Reform out. It is on course to hand Reform an outright parliamentary majority that the party could not win under any proportional system. If your defence of the electoral system is that it guards against a single faction seizing disproportionate power, and that system is about to deliver a single faction disproportionate power, the defence has refuted itself.

The constituency-link argument is the one that survives, and it is the reason serious reformers do not advocate closed national lists. The Additional Member System and the Single Transferable Vote both preserve a real, local representative relationship while correcting most of the distortion. You do not have to choose between the constituency link and proportionality. Systems exist that deliver both, and Britain already uses them in Scotland, Wales, and Northern Ireland. The choice the defenders pose - local accountability or fair representation - is a false one.

Part Eight: What the Numbers Actually Say

Step back from the detail and the picture is stark.

A system designed for two parties is being used to adjudicate a contest between six. The result is that vote share and seat share have come unmoored from each other to a degree without precedent in British history. In 2024, the gap between them produced the most disproportionate election ever recorded here. The next election, on current projections, is on course to match or beat it - and to do so in a way that hands near-absolute power to a party that fewer than a third of voters support.

The distortion is not ideological. It does not favour left or right, progressive or populist. In 2024 it manufactured a Labour landslide; in 2029 it may manufacture a Reform one. It favours whoever happens to lead a fragmented field, and it punishes whoever has the misfortune to have their support spread evenly rather than piled into the right postcodes. It is, in the most literal sense, arbitrary - an artefact of geography and vote-splitting rather than an expression of the national will.

The public has worked this out. Support for proportional representation reached a record 60 percent in late 2024, against 36 percent for the status quo - though it is worth being honest that this figure has since softened: the British Social Attitudes survey published in early 2026 found support had eased back to 53 percent as the immediate shock of the 2024 result faded and as Reform supporters, sensing the system might soon serve them, cooled on reform. The headline number moves with the political weather, exactly as the iron law of self-interest predicts. But across every reading, a clear plurality and usually a majority of the public wants change, and has done for years.

What stands in the way is not the electorate. It is the structure of incentives facing the only party that could deliver reform. Labour wanted proportional representation when it was losing and abandoned it the moment First Past the Post made it win. The Conservatives have never wanted it except, fleetingly, in the aftermath of defeat. Reform wanted it desperately right up until the point the system began to work for it. At every moment, the party able to change the rules has been the party least motivated to do so. That is the multi-party trap: a six-party electorate yoked to a two-party machine, sustained not by public consent but by the calculation of whichever party the machine happens to be rewarding at the time.

The 381-seat projection may not come to pass, and here I have to separate the two arguments I promised at the outset to keep apart. The crisis argument - Reform is about to win an outright majority on 31 percent - depends entirely on a specific number, and that number is already moving. More in Common itself called it a baseline, not a forecast; the next election need not come until 2029; and the firm's own April 2026 model had Reform falling back to 324, one seat short of a majority, with the June polling averages lower still. If you came for the crisis argument alone, you should hold it loosely, because by the time you read this the headline figure may have drifted from 381 to something less cinematic. I have built the piece around the January high-water mark because it is the vivid, legible illustration, but I will not pretend it is a fixed prediction. The structural argument is the one I am actually resting on, and it does not depend on whether Reform ends up with 381 seats or 297. It depends only on the fact - true at every reading, in every model - that a fragmented electorate run through First Past the Post produces seat counts wildly unmoored from vote shares, in whichever direction the fragmentation happens to point.

Which brings us to tactical voting, the most serious objection to my own case, and one I am not going to bury in a clause. If a fifth of Labour, Green and Liberal Democrat voters vote tactically against Reform, More in Common calculates that Reform loses 46 seats; if twice as many do, the majority vanishes altogether. Caerphilly in October 2025 showed it can happen: an anti-Reform electorate coordinated behind Plaid Cymru and delivered a result the raw first-preferences would not have predicted. So the system is not quite the inexorable distorting engine my main argument implies. It contains a correction mechanism - voters can, with effort, organise around it.

But look hard at what that correction actually is, because it is a damning thing to have to rely on, not a reassuring one. Tactical voting is the electorate performing unpaid labour to partially repair a system that is failing them - millions of people voting for parties they do not want, gaming a mechanism they did not choose, simply to prevent an outcome the mechanism would otherwise impose. A defence of First Past the Post that depends on voters routinely lying about their preferences to make the result tolerable is not a defence of the system. It is a confession that the system does not work as advertised and that only the voters' corrective dishonesty keeps it standing.

There is a more sophisticated version of the objection, and it deserves a straight answer because it cuts at my conclusion. Under proportional representation, the incentive for tactical voting largely disappears - you can vote sincerely, because every vote counts toward seats - but so does the voters' capacity to block. If Reform has 31 percent and the remaining 69 percent is split across five parties, PR does not magically produce a government the country chose. It relocates the coordination problem from voters in constituencies to MPs in coalition negotiations, where it can be every bit as fraught, and where the bargaining may produce a government no voter actually voted for. I concede this fully. PR does not abolish the problem of assembling a governing majority from a divided electorate; nothing can, because the division is real. What PR does is conduct that assembly in the open, after the election, through negotiations voters can see and judge, rather than disguising it inside a seat count that converts 31 percent into 59 percent and calls the result the settled will of the nation. The coordination problem is genuine under both systems. The difference is whether it is solved transparently by named politicians who can be held to account for the deal they strike, or silently by a mathematical accident that nobody chose and nobody can be blamed for. I would rather have the argument in daylight.

So the system is volatile, and tactical voting is its safety valve, and PR would bring problems of its own. All true. But notice that the volatility is itself the indictment. In a fragmented electorate, tiny shifts in vote share produce enormous swings in seats, and a few hundred thousand tactical voters can be the difference between a Reform majority and a hung parliament. We have built a system in which the composition of the entire government can hinge on the geographic accident of where a handful of votes happen to fall, and on whether enough voters can be persuaded to vote against their own preferences to stop it. That is not a feature. And whether the final figure turns out to be 381 or 324 or 297, the underlying truth holds: First Past the Post, applied to the Britain that actually exists in 2026, no longer translates the will of the country into the shape of its Parliament. It manufactures outcomes the country did not choose, and then calls them a mandate.

The arithmetic is not hidden. It is sitting in plain sight, in every MRP and every Gallagher Index and every votes-per-seat calculation. New Zealand showed the trap can be escaped, but only by building a process - a commission, a referendum - that reaches the decision the parties never will. In Britain the one thing missing is a governing party willing to act before, rather than after, the system turns against it, or willing to hand the question to a process that does not depend on its own self-interest. On current form, that is precisely what the multi-party trap is built to prevent. The numbers have made the case for change. Whether anyone with the power to act will read them before the next disproportionate landslide - in whichever direction it falls - is now the only question that matters.

Sources and data: More in Common January 2026 MRP (fieldwork 27 November to 16 December 2025, n=16,083); YouGov MRPs (June and September 2025); Electoral Calculus / Find Out Now modelling (October 2025, January 2026); House of Commons Library 2024 general election results; Electoral Reform Society, "A System Out of Step" (2024) and subsequent briefings; Michael Gallagher's electoral indices dataset, Trinity College Dublin (updated June 2025); British Social Attitudes surveys 42 and 43 (NatCen, fieldwork 2024 and 2025), "Britain's Democracy: A Health Check," co-authored by Professor Sir John Curtice; Make Votes Matter PR modelling of the 2024 result; PollCheck polling averages (June 2026). The New Zealand transition to Mixed-Member Proportional representation followed the 1986 Royal Commission on the Electoral System and binding referendums in 1992-93, with the first MMP election held in 1996. The proportional counterfactuals are the author's own estimates: the "realistic" column applies the d'Hondt method on a regional basis (approximating the system the UK used for European Parliament elections to 2020 and the top-up principle used in Scotland and Wales), and the "ceiling" column applies pure national proportionality. Both are approximate; precise totals would vary with regional boundaries, threshold levels, and the specific proportional system adopted. Gallagher figures for the projected scenario are likewise calculated from published vote and seat shares and are approximate.