In the months after a decisive election, voting intention polls have no predictive value. That is why I have largely ignored them so far. But tracked over time, they do tell us something. That is why I am writing about them now. They confirm that this time the century-long Labour-Tory duopoly really does look to be on the way out.
The latest polls show a tight three-way race in mainland Britain, with Labour, Reform and Conservative all around 22 to 25 per cent. Compared with last year’s general election, the two big changes are Labour down around 10 points and Reform up about the same. Otherwise, the movements are slight, with the Tories down and the Greens and Lib Dems up, but none of them by much.
On the face of it, around one in 10 voters appear to have switched from Labour to Reform. In fact, just 2 per cent of last year’s 28m voters have done so. Excluding don’t knows, YouGov’s recent polls show that the two biggest shifts have been from Conservative to Reform (one in five of all of last year’s Tory voters; 5 per cent of all voters) and Labour to Liberal Democrat (12 per cent of Labour voters, 4 per cent of all voters.)
Those two subsets of the electorate point to the larger story. It is not the illusion of a Labour-Reform swing that matters most, but the continuing shift away from the mainstream towards the parties that used to struggle under the label “minor”. In July 2024, Labour and the Conservatives together received 59 per cent of votes, while the rest managed 41 per cent. In April 2025, Labour and the Tories are collectively polling at 46 per cent; the rest at 54 per cent.
This is not the first time supporters of the mainstream parties have been in a minority. It was the case for around six weeks after the Brexit Party won the election for the European parliament in May 2019. But that was a short-lived blip at an extraordinary time, when Theresa May was about to be deposed as prime minister, Jeremy Corbyn was Labour’s leader and parliament was convulsed by Britain’s relationship with the European Union.
This time, “the rest” have been in the majority for three months and Keir Starmer’s position is secure. In health terms, Britain’s two-party system in 2019 needed emergency treatment. Today its condition is chronic. Ukraine, slow growth, weak public finances and Donald Trump’s presidency all present tough challenges for years to come. We shall of course see fluctuations in party support. But while the Brexit Party bubble in 2019 was quickly burst, there is no obvious reason why today’s mainstream total, Labour plus Tory, should return to sustained dominance of the electorate.
Hence the likelihood that the century-long two-party dominance is over, in votes cast if not, yet, in MPs elected to the House of Commons. For what we are seeing is the continuation of a trend that started decades ago. In the 1950s, more than nine in 10 voters backed Labour or the Conservatives. In the 1960s, the share was still above 85 per cent. It slipped towards 70 per cent over the next three decades.
These have been the figures for the past 30 years.
1997: Labour plus Conservative: 74 per cent; the rest 26 per cent
2001: 72-28 per cent
2005: 69-31 per cent
2010: 67-33 per cent
2015: 67-33 per cent
2017: 82-18 per cent
2019: 78-22 per cent
2024: 59-41 per cent
The two elections that stand out are, of course, 2017 and 2019. And yet, in a way those two contests help to explain the long-term story. The saga of Brexit polarised the electorate, not between left and right but between in and out. There was a clarity to the fundamental choice. A re-elected Conservative government would implement the result of the 2016 referendum; an anti-Conservative government would think again
The story of polarisation holds the key to understanding the threat to the Labour-Tory dominance. “Polarisation” is usually associated with rival extremes. Here it refers simply to choices that are clear-cut, not necessarily at the outer edges of left and right. In 1950s and 60s, when Labour and the Conservatives each won at least 40 per cent of the vote at every election, social class was the dominant influence on party support. Working-class voters identified overwhelmingly with Labour, middle-class with the Tories. According to the long-running British Election Study, more than four in five voters identified with one of the two big parties. In practice, governments of both parties were pretty moderate. Polarisation was a social fact, not an ideological observation
Those were the days when “working-class” and “middle-class” had clear-cut meanings, with millions of (mainly male) manual workers employed by the owners of large factories, mines, shipyards, and steel works. Most belonged to large, powerful trade unions. (When British Steel was nationalised in 1967, it had almost 300,000 employees. Its last remnant, the Scunthorpe plant, employs fewer than 3,000.) These industries gave Labour its citadels, while most non-manual, generally Tory-voting, workers had clerical, managerial or professional jobs, some in industry sectors and—increasingly—many in services such as finance, education, health or local and national government.
To reduce what has happened to a single sentence, that world has disappeared. We now have an utterly different economy. The proportion of workers employed in manufacturing and construction has fallen from around 40 per cent to 15. Sixty years ago, just over half of all employees worked in services; today the share is more than 80 per cent. Britain now has almost as many women as men at work. Trade union membership has halved—and most of those that remain work either in the public sector or formerly nationalised industries (such as energy and transport).
All that has eaten away at Labour’s core vote. But the Conservatives have had their headaches, too. In the early 1960s, just 4 per cent of school leavers went to university; today the figure is nearly 40 per cent. Few white-collar workers in the past were exposed to the radicalising influence of higher education. Now, most are. The impact has been enormous.
According to YouGov, progressive parties (Labour, Lib Dems, Green and SNP, Plaid Cymru) accounted for 70 per cent of votes from graduates, while right-of-centre parties (Conservative and Reform) accounted for 27 per cent. Among non graduates, votes were equally split between right- and left-leaning parties.
To repeat: that is a broad-brush account of how Britain’s economy and electorate have changed over the past six decades. A full account would include other twists and turns. The key point remains: the traditional foundations of both Labour and Conservative support have crumbled. To some extent both have attracted new sources of support—notably Labour with university graduates. But overall, the loyalties of voters across the spectrum are far weaker. As the British Election Study data show, the long-term decline in the number of people voting for either of the two main parties is matched by the falling numbers who identify with them.
It's not that those loyalties have switched to other parties (though some have), but that far fewer people identify strongly with any party. This is why election swings have become more volatile—and why the apparent revival of the old duopoly in 2017 and 2019 was the consequence of that volatility rather than a return to firm, historic loyalties.
As long as we retain first-past-the-post (FPTP) for elections to parliament, Labour and the Conservatives (and, intermittently, the Lib Dems) will enjoy an advantage over the other parties. But fragmentation is now built into our democratic arrangements. Labour’s victory last year may well be the last landslide won by either main party, and possibly one of the last elections to deliver an overall majority of any size.
That would, of course, be certain if our voting system changes. But even if it doesn’t, FPTP won’t protect the duopoly for ever. Last year 118 constituencies elected an MP who was neither Labour nor Tory. This was easily the largest number since 1923. Only a brave punter would bet against another big jump at the next general election, with Reform winning lots more seats, the Greens a few, and the SNP recouping some of last year’s losses.
The latest poll numbers—like, in all probability, next month’s local elections—matter not because they are a shock but because they confirm a remorseless trend that will end up transforming our politics. The only question is: when?
