The End of the Political 20th Century (And Why PR Won’t Save It)
John Harris is right about the diagnosis. He’s wrong about the cure.
John Harris has written a piece in the Guardian about Thursday’s Gorton and Denton byelection that is, by the standards of British political commentary, unusually honest. Labour came third. The Greens and Reform took a combined 70% of the vote. Starmer responded by writing a letter to his MPs that read like a landlord complaining about tenants moving out. Harris calls it “the belated end of the political 20th century.”
He’s right. But he can’t explain why it’s happening, and the solution he reaches for — proportional representation, via Andy Burnham — would make the problem worse.
Let me explain.
The bit Harris gets right
Something structural has changed in British politics, and it isn’t that voters got stupider or angrier. Harris spent three days in Gorton and heard the same thing everywhere: people want something radically different. A thirtysomething resident put it simply: “Politics needs to change in some huge way, doesn’t it?”
Harris identifies the symptoms with precision. Labour’s “landslide” in 2024 came with barely one in five of the total electorate. Its bond with old industrial heartlands has been fraying for years. Now it’s losing the urban base that was supposed to be the replacement coalition. The party, he says, looks “dangerously like an archaic legacy party.”
All of this is correct. Where it gets interesting is why.
The bit economics explains
Here’s what happened to politics, and it has nothing to do with ideology.
Until roughly the end of the twentieth century, political attention was an input. You got noticed in order to win office. Office was the prize. Attention was the means. Under those conditions, broadening your appeal, moderating your message, and building coalitions all made sense — because that’s how you won the thing that paid.
Then the economics changed. Direct monetisation of attention — through media platforms, subscription models, audience-funded distribution, and the vast infrastructure of digital engagement — turned attention from an input into an output. You no longer need to win office to extract value from politics. You just need to be visible, polarising, and consistent.
This is the structural break Harris is describing without having the mechanism. When attention itself pays, the old party system — designed around the assumption that office is the only prize worth chasing — stops working. Not because people are disillusioned (though they are), but because the economic logic that sustained two-party dominance no longer holds.
Labour and the Conservatives are legacy parties in the same way that Kodak was a legacy company. They’re not failing because they make a bad product. They’re failing because the market moved and their entire business model was optimised for conditions that no longer exist.
Starmer’s response is the model working
Harris is outraged by Starmer’s reaction — the arrogance, the dismissal of Green voters, the asymmetry of chasing Reform defectors while scolding leftward ones. But from an incentive perspective, Starmer’s behaviour is entirely predictable.
Labour treats rightward defectors to Reform as “hero voters” to be won back, while leftward defectors to the Greens are scolded as dupes of “divisive, sectarian politics.” Why the double standard?
Because Labour’s leadership instinctively understands — without ever framing it this way — that Reform operates in a different economic regime. Reform extracts value from attention directly. It doesn’t need office to thrive. Competing with Reform for voters means competing with an actor whose business model doesn’t require winning. Labour can’t outbid that.
The Greens, by contrast, still look like they’re playing the old game — building coalitions, winning council seats, taking parliamentary seats. So Labour treats them as a conventional electoral threat to be fought with conventional electoral weapons: attack their policies, question their credibility, dismiss their voters.
The problem is that Green voters in Gorton didn’t leave Labour because the Greens had better policies. They left because Labour’s identity container — the thing that used to make people feel like Labour was their party — has emptied out. You can’t attack people back into a container they’ve already decided doesn’t represent them.
Why PR won’t fix it
Harris ends by reaching for Andy Burnham’s 2022 call for proportional representation, elected regional senates, and maximum devolution. This is framed as the route to “a more honest and collaborative way of doing politics.”
It sounds lovely. Here’s why it doesn’t work.
Look at Germany. The German system is the poster child for proportional representation. It’s designed to ensure that the Bundestag reflects the electorate, that coalition-building enforces moderation, and that fringe parties face meaningful barriers to entry through the 5% threshold.
And what has it produced? A permanent centrist coalition that pays a continuous moderation tax — because every compromise, every difficult negotiation, every messy deal is visible, narratable, and monetisable by actors outside the coalition. The AfD doesn’t need to win. It just needs to clear 5%, gain institutional legitimacy, and then use every coalition muddle as content.
The cordon sanitaire — the agreement by mainstream parties not to coalition with the AfD — is supposed to be a penalty. It’s actually a subsidy. “They won’t let us in” is the most profitable political narrative in Germany. Every exclusion is content. Every refusal is proof that the system is rigged. The penalty has become the product.
And the AfD’s strongest support is in the former East Germany, where it activates pre-existing identity capital — decades of perceived second-class citizenship, economic underperformance, and cultural alienation from the reunification settlement — by positioning itself as pro-Russia. This isn’t ideology. It’s market positioning. Russia is the identity signifier that activates a specific, undervalued capital stock. Both the AfD and Russia extract value from this arrangement at negligible cost to either.
Now look at France. Somehow, worse.
Macron built En Marche as a pure identity startup — a political vehicle designed around a single person, bypassing the legacy parties entirely. He destroyed the Parti Socialiste, hollowed out Les Républicains, and proved that you could build a new political identity container from scratch if the old ones were sufficiently degraded.
Then the French system broke him. The two-round electoral system is designed to enforce moderation: you need broad second-round support to win. But in practice, every time the “republican front” activates to block the Rassemblement National in round two, it confirms the RN’s identity narrative. “The system is rigged against us” is the content. The penalty is the product. Again.
The result? A hung parliament, a government that lasted three months, and a president who — constitutionally barred from running again — has rationally pivoted from trying to win office to building post-presidential identity capital through international positioning and attention-generating domestic interventions. Macron is now optimising for visibility, not governance, because the economics of his situation have changed.
France, Germany, and the UK: three different electoral systems, three different institutional designs, all converging on exactly the same outcome. Legacy party decline. Retail-political insurgency. The structural failure of moderation. Coalition paralysis.
The institutional furniture is different. The economic gradient is identical.
The moderation tax, institutionalised
This is the core point, and it deserves saying plainly.
Proportional representation doesn’t eliminate the moderation tax — the economic penalty that moderate politics pays under current attention market conditions. It institutionalises it.
Under first-past-the-post, the moderation tax is paid at elections. Governments at least get periods of relatively unchallenged implementation between votes. Under PR, the moderation tax is levied continuously, in real time, because coalition dynamics are permanently visible, permanently contested, and permanently available as raw material for actors whose business model is attention rather than governance.
Every coalition is a compromise. Every compromise is a moderation event. Every moderation event is an opportunity for someone outside the coalition to generate attention by attacking it. PR doesn’t change the underlying incentive structure. It just ensures the tax is collected daily instead of every five years.
What Harris is actually describing
Harris calls it “the belated end of the political 20th century.” That’s a good line. But it’s a description, not an explanation.
The explanation is economic. Something changed about how political attention converts into value, and that change made the old party system — the one designed around two big coalitions competing for the prize of office — structurally unsustainable. Not because people stopped caring about politics. Because the economics of caring changed.
The residents of Gorton and Denton who told Harris they wanted “seismic political change” aren’t wrong. They’re responding rationally to a system that no longer serves them. But the change they need isn’t a new voting system or a new party leader or a better candidate. It’s a change in the incentive structure that makes polarisation profitable and moderation economically irrational.
That’s a harder problem. But at least it’s the right one.
The Rent Theory of Political Identity, the formal framework behind this analysis, is currently under review at Constitutional Political Economy. A book-length treatment, The Outrage Dividend: How Political Identity Became a Business Model, is in development. None of the equations were harmed in the writing of this article.