Prime Minister Keir Starmer speaks outside 10 Downing Street announcing his resignation as Labour Party leader and Prime Minister.
Prime Minister Keir Starmer speaks outside 10 Downing Street announcing his resignation as Labour Party leader and Prime Minister. Credit: Associated Press

British politics are roiling as Prime Minister Keir Starmer has announced his resignation, to be replaced by his fellow Labour politician Andy Burnham, who was elected as a Member of Parliament earlier this week in a by-election. Starmer is extremely unpopular because of how little his government has accomplished despite the party holding over 60 percent of the seats in the House of Commons. Burnham’s supporters hope the problem has been Starmer’s shortcomings, such that replacing him with a more charismatic, talented politician will allow the government to take the bold actions needed to turn around a country facing low productivity, massive debt, historically high taxes, and poor services. But it’s more likely that Burnham will be equally paralyzed by a long-term change in the electorate that is increasingly making bold actions nearly impossible.

Consider some political history. In the 24th year of the century, the Labour Party won a third of the vote in the general election, a disastrous performance that cost it 40 seats and led to a further descent into minority status. That was the nature of British politics in 1924, whereas in 2024 a comparable level of popularity with voters allowed Starmer-led Labour to gain 209 seats and achieve a stonking majority in parliament. If you had told politicians of previous generations that a massive majority in parliament could be secured in an election in which two-thirds of voters supported a different party, they would not have believed you.

Source: UK Election Statistics, House of Commons Library.

In the decades immediately following the war, governments of both major parties enjoyed levels of support that are unimaginable today (see Figure). Clement Atlee’s Labour Party won 48.8 percent of the vote in 1951—15 points higher than Keir Starmer’s Labour in 2024—and still lost. Confident of deep support, governments of this era could attempt bold things, whether that meant establishing the NHS or stymying the Soviet Union.

But by the 1970s, British voters began keeping their governments on a tighter leash. Almost incredibly, Ted Health’s Tories, who won election 55 years ago, were the last government to break 45 percent support. Harold Wilson’s governments of the 1970s scraped by with voter support substantially below that of the Conservative Party he defeated twice in the 1960s. Even Margaret Thatcher could not equal the vote share of the Labour Party leaders who ran and lost in the 1950s.

This century, multiple governments were put into power with levels of voter support that would have constituted crushing defeats in prior decades, including David Cameron’s two conservative victories in 2010 and 2015. Governments of this period, including Theresa May’s, indeed were also granted barely enough seats to govern, as voters fled the main parties for alternatives like the Scottish National Party, Greens, and Liberal Democrats (who did well enough in 2010 that Cameron had to accept them as the junior member of his coalition government) . Those governments were sometimes criticized for kicking knotty problems into the long grass or toward a referendum, but this was a predictable consequence of their pollsters making it clear to party leaders that a small shift in popular support could sink their governments. A brief revival of old-time popular support under Boris Johnson distinguished him in 2019, but only to the level of the victorious parties in the 1990s. The longer-term political trend of half-hearted voter endorsement reasserted itself in 2024 when Starmer’s Labour Party improved its share of the vote by less than 2 points yet netted a huge number of seats.

Parliamentary majorities don’t mean what they once did because of the decades-long decline in the voter support undergirding them. Tony Blair had almost the same number of Labour seats in his first two terms (418 and 412, respectively) as Starmer won in 2024 (411), but with millions more votes behind him. Labour won 13.5 million votes in the less populous country that was Britain in 1997 whereas the party secured only 9.7 million in 2024. Dips in the polls, difficult votes in Parliament, and the occasional scandal don’t scare MPs in a party with huge popular support in the same way they do those that squeak in to power, even when the number of seats is the same. The repeated revolts of Labour backbenchers are the acts of MPs who are hard to discipline because they feel their time in government is limited.

British voters have been frustrated by their governments for some time, but collectively they contribute to the problem. Government on a short leash of popular support is risk-averse and unimaginative. Voters will not trust political parties with much power and thus are likely to be disappointed by the increasingly fragile governments they put in place.

This is normally the point in a political essay where the author is expected to discuss the solution, but it’s not clear in this alienated time that there is one. Burnham is a more able politician than Starmer, but in a country facing very tough economic choices, he’s still going to have a very hard time persuading his party to make major policy decisions that some voters will dislike. Whoever wins the next general election may have it even worse. Britain now has four parties—maybe even five—that can break 10 percent popular support in a general election, but it’s quite possible none can even reach the low threshold of support that Starmer’s Labour Party reached. Coalitions of parties that can’t break a quarter of the vote on their own would be hard to achieve because of the potency of negative partisanship. Just as in the U.S., many British voters are now inclined to vote for one party (e.g., the nationalist, anti-immigration Reform Party) not so much out of liking it but out of hatred of another (e.g., the Tories), usually the one that would otherwise be the most likely coalition partner.

Expect the bleak prospect of a self-perpetuating cycle of weak governments with weak popular support instead. Parties will start their first day in office with a level of voter trust that would have represented a worrying decline from election day in the past. This only ends when the voting public decides they have to hand the keys to a driver without simultaneously tapping the brake.

Our ideas can save democracy... But we need your help! Donate Now!

Keith Humphreys is a Professor of Psychiatry at Stanford University and served as Senior Policy Advisor in the White House Office of National Drug Control Policy in the Obama Administration. @KeithNHumphreys