Disproportionality part I: Britain

We’ve had two very interesting elections in the same week, so there’s plenty to look at. Numbers settle down more quickly than politics (except in Australia and a few other places, where counting takes more than a fortnight), so this week I’ll try to tell a story about them, in three parts: first Britain, then France, then comparing the two. Next week we’ll get to the broader political implications.

There’s a general if hard-won awareness both that Britain has a very poor electoral system, and that last week’s election showed it at its worst. Here’s the table to demonstrate, with the share of vote and the seats won (“Actual”). The lack of any coherent relationship between the two is clear, but to make it more obvious I’ve also included the numbers of seats that would have been won under proportional systems, either Sainte-Laguë or d’Hondt (as usual, the two are very similar).*

Labour’s vote share is apparently the smallest for any majority government in the country’s history, yet it won not only a majority but a landslide. It improved by only 1.6% on its performance in 2019, when it lost in a landslide. And its vote last week was more than six points lower than it had been in 2017, an election that it also lost (albeit narrowly).

There is some method in this madness, although perhaps not a lot. In a two-party system, the key thing that matters is not the absolute shares of vote but the gap between the two parties. In 2017 Labour was two points behind the Conservatives – that’s why it lost. And in 2019 the Tories led by 12.5%, so it’s no surprise that they won a landslide. If you lead by ten points, it doesn’t matter very much whether it’s 45-35 or 35-25, although they reflect very different levels of public support.

But how then to explain the inequities further down the table? In particular, why did the Liberal Democrats, who finished two points behind Reform, win enormously more seats? Both are broad-based parties; the answer is not in the sort of extreme regional concentration that benefits our National Party, or that last time gave the Scottish Nationalists 7.4% of the seats with 3.9% of the vote. The answer, I think, comes down to what the voters were trying to do.

Those who voted either Labour or Lib Dem primarily wanted to throw the Tories out, so they naturally voted for whichever of the two was better placed in their constituency to achieve that. That mitigated some of the effect of first-past-the-post voting; it’s a smaller and less formal version of what happened in France. But Conservative and Reform voters, who superficially had a similar sort of relationship, did not share the same objective: people voted Reform not in order to beat Labour, but to express dissatisfaction with all the other parties, Conservatives included.

So Reform’s vote did not concentrate much in seats that it could win; its supporters knew that the Tories were going to lose anyway, so they were free to indulge in a protest vote regardless of where they lived. Conversely, the Conservatives did not regard Reform with the same sort of benevolence that Labour had towards the Lib Dems, and had no interest in throwing votes to it in seats where it may have been better placed to beat Labour.

At least there are now two substantial parties, at opposite ends of the philosophical spectrum, that have a strong interest in electoral reform. Nonetheless, its prospects remain slim; as I put it after the 2019 election:

Britain desperately needs electoral reform. But it won’t get it until there’s a party in government that’s badly disadvantaged by the existing system, and of course the whole point of that disadvan­tage is that such a party is unlikely to get into government.

Writing at the Conversation last week, Richard Shaw suggests that this is an opportune moment not unlike when New Zealand moved to proportional representation in the 1990s, but I’m afraid I can’t see it. Despite only winning a third of the vote, Labour seems in a much stronger position to resist reform than either major party was in New Zealand. But I’d be happy to be proved wrong.

.

* Technical note: Figures are taken from the BBC. Seven independents won seats, one of them the Speaker, who is from the Labour Party but by tradition is regarded as non-party. They represent somewhat more than two-thirds of the “other” vote. If you count them as a party (and some of the most successful ones did share a common orientation), then proportionally they should have won 13 seats, which brings the major parties’ totals down a little – Labour to 221 (Sainte-Laguë) and the Conservatives to 155.

.

PS: I should now also recommend this piece in the Guardian by Zoe Williams on the need for electoral reform – it’s very good. A sample:

The playing field, never exactly even, now looks more like a wildly irrational assault course. Twenty-four thousand votes would get you a Labour seat; a Reform candidate would need one million; a Green, 485,000; a Conservative, 56,000. It is a cute paradox that the Liberal Democrats, the longest-serving proponents of proportional representation, are the only party that, had we implemented it for this election, would have ended up with about the same number of seats.