The first poll conducted after the start of the war reported a sharp rise in the prime minister’s rating, from +17 to +36 (from 55 per cent satisfied, 38 per cent dissatisfied, to 65-29 per cent.)
No, not Boris Johnson’s rating since Russia invaded Ukraine, but that of Neville Chamberlain in 1939. Much good did it do him. By the beginning of the following May, his rating crashed to -27 as the public, along with a growing number of MPs, concluded that he had to go. On 10th May 1940, Winston Churchill became prime minister. The rest, as they say, is history—including the facts that Churchill’s rating never fell below +63 for the next five years, and Labour still won the 1945 election by a landslide.
This trip down memory lane should reinforce the advice given seven decades ago by Adlai Stevenson, one of the ablest candidates never to be elected US president: polls should be taken but not inhaled. They are imperfect guides to the future at the best of times, and the early weeks of a war are far from the best of times.
However, that does not mean that polls tell us nothing. They tell us something—but what?
Here are three observations from recent polls in Britain and abroad.
First, British voters are ahead of the government in wanting to help Ukrainians. Opinium’s latest poll for the Observer finds that eight in 10 Britons either back the government’s plans to admit Ukrainian refugees (29 per cent) or want it to go further (52 per cent). Only 7 per cent think ministers have gone too far. Asking a slightly different question, Deltapoll finds that 51 per cent think “Britain should allow any refugees leaving Ukraine who wishes to live in Britain to do so.” A further 23 per cent would be happy to admit some refugees. Only 7 per cent want no more refugees to come here. (19 per cent say “don’t know.”)
Germans are even more generous. According to Infraset, 91 per cent agree that it is fundamentally right to admit refugees. Big majorities also support the decisions of Chancellor Olaf Scholz to send military aid to Ukraine and sharply increase defence spending.
In Poland, the country with far more Ukrainian refugees than any other, one poll conducted shortly after the invasion—when the coming influx was certain but not yet massive—found that more than 90 per cent wanted to admit either “all refugees” (58 per cent) or those “most threatened and in need” (35 per cent).
Second observation: the war has boosted Johnson’s rating a bit. An average of three recent polls by different companies suggests that his rating has recovered from a pre-war average of minus 41 (approve 22 per cent, disapprove 63 percent) to minus 30 (27-57 per cent). At the height of partygate, his ratings were catastrophic; they are now merely terrible. Johnson is now on a par with Chamberlain on the eve of his resignation.
Much of this limited recovery is due to Conservative voters rediscovering Johnson’s virtues. This explains why voters have not been flocking back to the Tories from other parties. Five polls in the past week put them on 35 per cent on average. In mid-February, before the Russian invasion, the same pollsters reported an average share of 34 per cent. Statistically, that‘s no change.
Why have the Tories not enjoyed a wartime boost? My guess is that domestic factors still determine party loyalties, and these have been put on hold. Without the war, we would have heard more about partygate (which definitely hurt the Tories), inflation (which might have hurt them) and the post-Covid drift back towards normal life (which would have helped them). Which of those forces would have been most powerful? We can’t tell. Maybe we shall one day; but not yet.
Abroad, Joe Biden has gained a tiny boost. According to the 538 website, which aggregates the vast range of polls conducted in the US, his net approval rating—which had slumped to its lowest figures of the presidency, minus 13, on the eve of the war—has recovered slightly to minus nine.
In Germany, there has been no perceptible movement in party loyalties: the latest polling figures are close to the result of the federal election six months ago.
Intriguingly, the one big democracy that has seen a significant shift is the one where it matters most: France. Emmanuel Macron, who faces voters next month, has received a pre-election boost at just the right time. His support has jump from 25 to 30 per cent when voters are asked who they would support in the first round of the presidential elections. Assuming he faces Marine Le Pen in the second round, as he did five years ago, his lead has doubled from 10 points (55-45 per cent) to 20 (60-40). He would defeat any other second-round rival even more convincingly.
Third observation: France apart, today’s polls tell us little about the impact of the war on the prospects for today’s leaders. Chamberlain’s downfall, months after basking in his reputation as a popular national leader, is merely the most dramatic example of a fall from grace. According to Ipsos MORI, Tony Blair’s rating recovered from minus 30 to plus two during the 2003 Iraq War and its immediate aftermath—only to drop back to minus 35 five months later when weapons of mass destruction failed to materialise.
What about the “Falklands factor,” which rescued Margaret Thatcher’s premiership, and which some have cited as possible precedent for Johnson’s fortunes? True, MORI found that Thatcher’s rating jumped from minus 21 before the war to plus 23 following the Argentinian surrender. But a closer look at the polls during the conflict tells us something significant. The reputation of Thatcher and her government rose significantly only after the first British victory, the recapture of South Georgia, three weeks into the war. The polls told a story less of wartime unity than of British military triumph—just as the collapse of Chamberlain’s standing in May 1940 followed the failure of British forces in Norway.
So forget the Falklands factor: the long-term consequences of the Ukraine war (including on prices and living standards here in Britain) have yet to play out. And in any event, unlike the Falklands, it is not being fought by British troops.
Instead, we should remind ourselves of the permanent truth about polls. They provide interesting, and sometimes useful, snapshots of the immediate past. They can tell us whether, at a moment of crisis, our leaders are broadly in tune with public opinion or dangerously at odds with it. They contribute to the vital dialogue between voters and politicians.
But at times like this, polls tell us nothing about what people will think six months or a year from now.
