This week, almost exactly ten years after the Brexit referendum, Keir Starmer announced his resignation. Andy Burnham, who has spoken before about wanting to scrap First Past the Post and replace the House of Lords with something elected, is the overwhelming favourite to succeed him. It is the right week to ask why a Prime Minister so rarely arrives with the genuine consent of the country they lead, and why that keeps happening regardless of which party benefits.

Start with the numbers, because they are not really in dispute. At the 2024 general election, Labour won 63% of the seats in the Commons on 33.7% of the vote. Fifty-eight per cent of voters backed a candidate who lost. Their ballot, for the purposes of deciding who governs the country, may as well not have been cast. This is not a one-off statistical quirk. It is what First Past the Post does, with grim reliability, almost every time. The same mechanism that handed Labour a “landslide” on barely a third of the vote spent most of the 1980s handing the Conservatives artificial dominance, and kept Labour itself out of office for the better part of two decades. The system has no loyalty. It simply rewards whoever happens to win the geographic lottery that election, and calls the result a mandate.

Political scientists have a useful way of separating two things that get muddled together whenever this argument comes up. There is Executive Durability– the ability of a government to stay in office and get its programme through– and there is Systemic Legitimacy, which is the harder, slower thing: whether people actually trust the rules of the game, win or lose. First Past the Post (FPTP) is supposed to deliver the first at the cost of the second, a trade that defenders of the status quo present as somehow inevitable. But look at the last decade and the trade has not even worked on its own terms. Between 2016 and 2024 the UK got through five Prime Ministers and several snap elections. An 80-seat Conservative majority, won in 2019 on the kind of “strong and stable” mandate FPTP is meant to deliver, did not stop the governing party imploding through its own internal factionalism within two years. Labour’s 2024 majority has already been nicknamed a “sandcastle majority” by commentators who can see how thin the underlying support actually is.

Systemic Legitimacy should worry us more, because that is where Brexit lives. In 2015, UKIP took 12.6% of the national vote and won one seat. Millions of people who wanted a real alternative to the two main parties discovered, in the starkest possible terms, that the parliamentary route was closed to them. They found another one instead, and the country has been negotiating the consequences for a decade. The Brexit referendum was less a cause of national rupture than a symptom of a parliament that had, for years, simply stopped reflecting what a meaningful slice of the country actually believed. Reform UK’s current trajectory looks uncomfortably similar. A party can be kept out of the Commons. It cannot be kept out of politics. Suppressing representation in Parliament does not suppress the underlying opinion– it just removes the normal democratic channel for expressing it, and pushes people towards whichever channel remains open, referendum or otherwise.

There is a deeper, almost embarrassing irony sitting underneath all of this. Britain maintains an unelected second chamber, stuffed with appointees and a shrinking rump of hereditary peers, largely to act as a check on a House of Commons that routinely fails to represent how the country actually voted. Britain has built an undemocratic institution specifically to compensate for the democratic failures of another institution, and has somehow normalised this as a sensible constitutional arrangement rather than an admission of guilt. Fix the Commons properly–make it genuinely proportional– and an enormous amount of the case for the Lords as currently constituted simply evaporates. A further irony is that the Commons rejected a House of Lords proposal to introduce STV (one of the PR systems) in 1918.

None of this is news to anyone who has looked seriously at the comparative evidence. Countries that use proportional systems consistently outperform Britain on voter turnout, on women’s representation in parliament, on public trust in the political process, and on several other measures, without paying any of the economic price that FPTP’s defenders imply is the cost of fairness. Germany has had nine Chancellors in seventy-five years under a proportional system that somehow also manages to combine a strong constituency link with near-perfect proportionality between votes and seats. New Zealand switched from First Past the Post to a proportional system in 1996 and, when asked again in a referendum fifteen years later whether to go back, decided overwhelmingly to keep what they had. The fear that voters cannot cope with anything beyond a single cross on a ballot paper does not survive contact with the evidence. Irish voters have ranked candidates on STV ballots since 1922. German voters have managed two votes for seventy-five years. The complexity, such as it is, lives in the counting room, not the ballot box. The fear survives mainly because it is convenient for the two parties who currently benefit from the system staying exactly as it is.

What is worth resisting is the idea that any form of PR will do. Some of the systems on offer are genuinely better than others, and the choice between them is not a matter of taste so much as a matter of mathematics – which formulae are used to allocate seats, how thresholds are set, how the inevitable trade-off between a strong local link and genuine national proportionality is handled. These are technical questions with evidence behind them, not questions to be settled by whichever compromise gets a deal over the line in a hurry. The Alternative Vote (AV), occasionally floated as a halfway measure, keeps single-member constituencies and does not reliably produce proportional outcomes at all; it solves the problem of tactical voting while leaving the deeper unfairness almost completely untouched. If Britain finally gets the chance to change its electoral system, after decades of false starts, it would be a tragedy to spend that one chance on something that merely looks like reform.

The public does not need to relitigate seat allocation formulae any more than it needs to understand monetary policy to have a view on interest rates. What it deserves is a system chosen properly, on the evidence, by people who have actually done the work – and then a straightforward, honest choice put to it: do you want this, or not? That is not a technocratic stitch-up. It is simply matching the right kind of decision to the right kind of process, instead of asking sixty-eight million people to adjudicate a maths problem most professional politicians could not solve either.

Ten years after the referendum that this same broken system helped produce, and in the same week that a Prime Minister has resigned without ever having won the support of even four in ten voters, it feels like a reasonable moment to ask whether Britain is finally ready to fix the machine rather than just keep changing the person sitting on top of it.