“Gee, I’m glad I didn’t sell my car.” Kim Campbell was only half joking. Canada’s prime minister had not only been voted out of office; she had lost her own seat. Her party had slumped from an outright majority to just two MPs.
Thirty-one years later, could Rishi Sunak suffer the same fate? Probably not: when Labour returns to power next month, it is likely that the Conservatives will be the biggest opposition party. What is remarkable is that the question is even worth asking.
A brief summary of the downfall of Canada’s Progressive Conservatives (PC) in 1993 can give little comfort to their British cousins today. With five weeks to go, the race with Jean Chrétien’s Liberals was close. Then Campbell’s party started to slip. She stumbled in a TV leaders’ debate. The Liberals gained ground, and so did a right-wing populist party. Its name? Glad you asked. It was Reform. In the final week it overtook PC. On the day, PC’s support was just 16 per cent, down from 43 per cent at the previous election. Reform won 19 per cent and picked up 52 of the country’s 295 seats.
Canada has the same first-past-the-post system as Britain. This means it shares three of the characteristics that are relevant to our current election. First, it exaggerates success. At the previous election, the PC’s 43 per cent of the vote rewarded it with 57 per cent of the parliamentary seats. (In 2019, Boris Johnson’s Tories won 45 per cent of the Britain-wide vote and 58 per cent of the seats.)
Secondly, first-past-the-post is also kind to smaller parties with geographically concentrated support. In 1993, the Bloq Québécois won fewer votes than PC, but ended up with 54 seats because—like the SNP in Scotland—it fought only in Quebec, where it came top. But thirdly, the system crucifies parties whose support is spread across the country and slips badly. That was PC’s fate—and, in just over three weeks’ time, could be the Conservatives’ .
Under first-past-the-post, nationwide parties in decline face a tipping point. Two or 3 per cent above that point, and they retain a fair number of seats, albeit with small majorities. Two or 3 per cent below and they lose most of them. The precise tipping point depends on the size and distribution of the votes of other parties. How big will Labour’s vote be next month? How much tactical voting will take place? How will Reform fare in different areas? We can’t predict the tipping point precisely, but for the Tories this year it’s likely to be in the region of 25 per cent.
Polls conducted in the past week put Tory support at 19-26 per cent. The next few days could see even lower ratings. In short, the party could be heading for the wrong side of its tipping point. Dozens of MPs defending normally safe seats would end up losing narrowly—some of them just a few hundred votes short of the winning post.
Because so many contests could be close, any prediction is subject to a wide margin of error. But—as an illustration, not a prediction—if the Conservatives end up with 20 per cent, they would probably have well under 100 MPs and possibly as few as 30. Rather than count the cabinet ministers who would lose their seats, it’s quicker to list the three likeliest survivors: Kemi Badenoch, Claire Coutinho and Laura Trott. Below 20 per cent and even they might be gone.
At this point, a confession is in order. Over the past year I have been more cautious than others about the depths the Tories might plunge. This was because I expected Labour’s lead to slip as the election approached, as many of the people who deserted the Conservatives in mid-term returned to their political home. I also expected Tory MPs in normally safe seats to benefit most from this. As a result, I assumed that the traditional pattern of uniform swing would return—as it did in 1997, when the Conservatives last lost power. I was sceptical of large scale MRP polls that predicted massive Labour majorities, because they assumed huge swings in strong Conservative areas.
The facts have changed, so I have changed my mind. (Or rather, the facts have not changed as they have in the runup to past elections, and I thought they would again.) There is no sign that the Conservatives are rebuilding their core vote, and hence no sign of a return to uniform swing. Unless things change significantly between now and 4th July, the assumption of proportional swing at the root of MRP’s basic methodology may well be right this time. It would be the first election that swings between Labour and Conservative have worked like this since 1945—and it opens the door to disastrous Tory losses.
Of course, we must not exclude the opposite possibility: that instead of the Tories tumbling towards oblivion, Labour’s strategy of reassuring middle Britain will come unstuck, the Conservatives will win back the voters they have lost and the gap between the parties will narrow. Last week I recalled Theresa May’s disastrous campaign in 2017, when her policy on social care blew up in her face. I argued that the biggest threat to a big Labour victory was not that voters would warm to Sunak but that Keir Starmer would have a May moment and put voters off.
That remains possible, though so far Labour’s leader has not messed up—unlike Sunak, who is giving a masterclass in how to lose friends and alienate people. Which is why, even though I still expect the Conservatives to end up with more than 100 seats, the notion that they won’t is no longer ridiculous.
This blog was first published by Prospect
Peter Kellner. The Politics Counter 