Reading the Runes

It’s become commonplace to argue that opinion polls can no longer be trusted, after a series of electoral shocks like Brexit and Trump. However, it’s more accurate to say we can’t trust the conventional wisdom about how polls should be interpreted.

There were two rules of thumb that should have doomed Labour’s effort in 2017. One said that election campaigns only ever have a marginal impact on a party’s final performance. The other said that whenever there’s a large discrepancy between polling figures from different companies, the average score is the safest guide.

In the month when Theresa May called a snap election, the average polling lead for the Conservative Party was 18.5 percent. There was only one poll that month that gave the Tories an advantage of less than 10 percent. Even that outlier — with a 9 percent gap between Labour and the Conservatives — would have ensured May a decisive victory if it was borne out on voting day.

From the end of April, Labour steadily closed the gap, but the final weeks of the campaign still left a big margin of uncertainty. Of the last ten polls, just three gave the Tories a lead of less than 5 percent; three had them besting Labour by 10 percent or more. Yet the final result saw Theresa May’s party edge past its opponent by just over 2 percent. That was in line with the most optimistic polling forecasts for Labour.

There’s no reason to think the polling companies had it badly wrong when the election was called. Labour was well behind the Tories, by a margin that should have been insurmountable if precedent was any guide. Over the space of two months, that gulf closed to the extent that May lost her parliamentary majority and came close to being surpassed altogether by Corbyn’s party.

Explaining It Away

The result came as a huge shock to Britain’s political pundits, and they were visibly disoriented for some time to come. Eventually they began putting together a reassuring story about what had happened. That story made it possible to deny any credit to Corbyn, his allies, or the movement behind them.

According to this line of argument, it was May and the Conservative Party who had thrown away their advantage with a shoddy campaign that simply highlighted her flaws as a leader. Once this became the orthodoxy about the 2017 election, it was possible to add a further twist: far from deserving any credit for stripping May of her majority, Corbyn should be faulted for not winning the election outright against such a poor opponent.

This comforting tale is inaccurate and disingenuous in every respect. Before the 2017 election, very few people considered May to be a weak, incompetent leader: in fact, she was seen as a formidable operator who could deal a crushing blow to Corbyn’s Labour Party. For anyone who’s forgotten the tone of reporting on May’s premiership, it’s well worth reading the admiring profile by Jason Cowley, editor of the liberal New Statesman, which appeared just four months before her electoral shipwreck. The personality traits that are now depicted as May’s Achilles’ heel struck most commentators as an asset: she was meant to be a sober, serious politician, in contrast with the flashy PR man David Cameron.

The idea that May was the worst prime minister in modern times, heading the worst government, only took hold after the 2017 election. May wanted to face all the strains of implementing Brexit from a position of strength, with a big parliamentary majority. Instead she had to grapple with those contradictions while at the mercy of Tory rebels from either end of the party spectrum, who thought her approach was either too soft or too hard — not to mention the Democratic Unionist Party, whose MPs proved to be another thorn in May’s side.

Any leader in that position would have ended up looking weak. Nine times out of ten in contemporary politics, it’s circumstances that make the man (or woman), not the other way around.

Labour’s Achievement

If May’s election gamble failed, it wasn’t because of a Tory meltdown. Her party’s final result, 42.4 percent, was the best Conservative result in three decades. It was better than Labour’s performance under Tony Blair in 2001, which delivered a majority of almost a hundred seats. It was over 5 percent higher than David Cameron’s 2015 triumph, which also secured a majority in parliament.

May’s objective going into that campaign was to use Brexit as a wedge, winning over enough UKIP and Leave-supporting Labour voters to convert many traditional Labour strongholds into Tory seats. The regional breakdown shows how close she came to achieving this. In some of the key Leave-voting regions, the Tory increase was well above the national figure: 9.1 percent in North East England, 7.8 percent in Yorkshire & Humber, 7.3 percent in East and West Midlands alike, 6.3 percent in Wales. In every British region other than London, the Conservative vote share went up.

What put paid to May’s ambitions was the Labour performance. The vote share for Corbyn’s party went up by a bigger percentage than the Tory increase in every region apart from North East England — even there it was only half a percent smaller — and Scotland, where there was a different political dynamic. In Yorkshire & Humber the Labour vote went up by 9.9 percent, in Wales by 12.1 percent, in Eastern England by 10.7 percent (nearly double the Tory increase in that area). Labour increased its vote in every British region, and by double-digit scores in seven out of twelve.

The big story of the 2017 election was a Labour surge, not a Tory collapse. The party’s 40 percent vote share was higher than the winning performance in the previous three elections, and it hobbled Theresa May’s government in spite of her post-Brexit electoral windfall.

The oft-heard claim that Corbyn should have won outright cannot be taken seriously. Labour’s performance in the 2010 and 2015 elections had been so bad that it was virtually impossible for the party to win a majority in a single bound: there simply weren’t enough constituencies where Labour was in contention to gain a seat.

If Labour had won just 35 percent of the vote — still a big improvement on the results achieved by Gordon Brown and Ed Miliband — May would have come out of the election with her majority strengthened, and no doubt would have pushed through Brexit on her desired terms some time ago. Nobody would be talking now about the “worst government ever.”

Haywire

What does this mean for any snap election? The first lesson is obvious: Labour need not rely on a catastrophic Tory meltdown, since there was no such meltdown in 2017. If Labour could match its campaigning performance from two years ago, that would be enough to put a stop to Johnson’s gallop.

Of course, whether it can repeat that performance is now one of the biggest questions in British politics. The terrain has certainly shifted since June 2017, and the biggest change has come in the last few months.

For all that’s been rightly said about the fallibility of polls, they’re still the best rough guide we have to the state of public opinion. From the last election until the opening months of 2019, Labour’s typical polling was in a range between 35 and 40 percent, sometimes a little higher. The Tories were in much the same position, and the gap between the parties was usually in the low single figures. Then things started to go haywire, as the Brexit crisis took center-stage. Labour and the Conservatives both lost support to smaller groups that could take a harder pro- or anti-Brexit line: Nigel Farage’s Brexit Party, the Liberal Democrats, the Greens, or the Scottish and Welsh nationalists.

At first it was the Tories that were most affected by this trend, with their support dropping below 30 percent. But Labour soon took a major hit, especially after the European election in May. Since Boris Johnson became Tory leader, his party has managed to claw back some of the ground it lost, pushing past the 30 percent barrier for the first time since early April. Labour’s polling average since the start of August was still in the mid-20s. But the average Tory lead over that period has been lower than the smallest Tory lead in April 2017. Labour’s position going into a snap election is weaker than its supporters might have hoped for at the start of this year, but stronger than it was when Corbynism faced that previous moment of truth.

Complications

Can Labour do it again? An election held this autumn will differ from 2017 in one vital respect: Brexit will be absolutely central. One striking thing about the 2017 election was the gulf between the Tory and Labour electorates: not in the sense of being pro- or anti-Brexit, but in the importance that they gave to the issue. One survey found that Brexit was by far the most significant factor motivating Tory voters: 48 percent said that it was the most important issue for them in choosing how to vote. In contrast, just 8 percent of Labour voters said that Brexit was the most important issue (33 percent cited Britain’s health service).

It was almost as if two election campaigns were being held in parallel, one about Britain’s relationship with the European Union, the other about questions of domestic social policy. Those whose main priority was getting Britain out of the European Union voted Tory; those who cared most about repairing the damage after years of austerity voted Labour. Labour’s 2015 base split two-thirds to one-third between Remain and Leave voters during the Brexit referendum. Its electoral coalition in 2017, which was much larger, had the same two-to-one divide between Remainers and Leavers.

It will be harder for Labour to pull off the same maneuver this time around. To some extent it was inevitable that Brexit would rise to the top of the agenda, as the deadline approached and the parliament elected in 2017 proved unable to agree on the terms of a deal. But there was also a concerted effort by several political actors — Labour’s right-wing, anti-Corbyn faction; the party’s electoral rivals, especially the Lib Dems; and the People’s Vote campaign — to undermine Corbyn’s strategy of pushing for a soft-Brexit, Norway-style deal: the option that was most likely to secure a broad consensus.

The European election supercharged that effort and forced Labour to change direction over the summer, in a bid to win back support lost to the hard-Remain groups. The party now says it will hold a second referendum on the terms of any Brexit deal with Remain as an option on the ballot paper. In effect, the new policy means that Labour has more to offer Remain voters and less to offer Leavers, who formed a crucial minority of its electoral base two years ago. It’s a high-stakes gamble, but that’s the course the party leadership has chosen. Labour’s best hope is that its Leave-supporting electorate cares more about the party’s domestic reform agenda than it does about leaving the European Union. A snap election will put that theory to the test.

Boris Johnson’s team have placed a wager of their own, as Philip Stephens describes in the Financial Times:

During the 1960s, America’s right-wing Republicans embarked on what was called the “southern strategy” — a populist pitch to white working-class voters who were disenchanted with the civil rights liberalism of the Democratic Party. Mr Johnson has a “northern strategy”. By casting Brexit as a fight against foreigners and immigration he hopes to win an election by winning over anti-European white working-class voters in traditionally Labour areas. We are promised a campaign that might make even Mr Trump blush.

But the parallel being drawn here is misleading and (hopefully) delusional. Nixon’s southern strategy didn’t just appeal to white working-class voters: it appealed to white voters full stop. It was geared towards a region where a system of racial supremacy had been in place for generations, based on disenfranchisement of African Americans and violent repression of those who resisted. Any attempt by Johnson to flip the North and the Midlands — casually lumped together as “northern” England — will build on much weaker foundations.

Labour has a reform program that can appeal to working-class people of all races, however they voted in the Brexit referendum. It also has deep social roots in those areas, which the Tories mostly lack. Johnson’s “northern strategy” is sure to be every bit as foul as Stephens predicts, but it will still be very difficult for the Conservative leader to pull it off.

The electoral landscape will also be more crowded than it was in 2017, when the smaller parties were decisively squeezed. If there are strong performances from the Brexit Party and the Lib Dems, plus the nationalists in Scotland and Wales, Britain’s unrepresentative electoral system will play out in all kinds of confusing and unpredictable ways.

However, one thing remains firmly within Labour’s power. As in 2017, it can run an insurgent campaign with a manifesto based on popular, social-democratic policies, using its membership for large-scale canvassing and to push Labour’s message online, all of which proved vital two years ago. Britain’s political commentators badly underestimated the force of a campaign like that before. Corbyn and his supporters will soon have a chance to teach them the lesson again.