In 1992 Robert Hayward, the Tory peer and polling expert, wrote a paper for his party contending that despite Labour’s narrow leads, the end result would likely be a majority for John Major’s Conservatives. This analysis was duly vindicated and the term “shy Tories” was born.
Hayward was proved right again in 2015 when he argued the polls were overstating support for Ed Miliband’s Labour. So when he argues, as he has, that the polls are understating the Conservative support it’s worth investigating.
In advance of the May local elections, Hayward forecast that the Tories would lose at least 400 council seats – they ended up losing 470. In spite of this, his post-election analysis finds that the Conservatives are winning back more undecided voters than surveys suggest.
If indeed more 2019 Tory voters are backing the party than previously suggested, then the Labour leader is likely to be narrower than the 21 points forecast. But narrow enough to change the election narrative? Not quite or, rather, not yet.
A 21-point Labour lead, with tactical voting and regional variation accounted for (as per the Britain Predicts model), would put the party on course for 400-plus MPs in the Commons with the Conservatives on 100-150. The expectation from most reasonable people is that, as they have nine times out of 10, the polls will indeed narrow. The Tory base will rally. Reform’s vote will be squeezed to some degree. All of this should bring the Conservative total up to around 150-200. A recovery from the current nadir? Sure. Enough to deny Labour a majority to govern? No.
I would contend that the polls are more unlike 1992 and 2015 than they’ve ever been. In both of these previous elections, while the Tories trailed behind Labour overall, they led comfortably on the economy and leader approval. Now, however, Labour’s 21-point lead is matched by its 13-14 point lead on the economy and its 12-point lead on leadership.
Without seeing the detail of Hayward’s analysis, I can’t contend whether his suggestion that the polls are understating the Tories is correct. But from my own aggregation of the May local election results, I can say this.
In seats as distant from Labour as Aldershot (the home of the British Army), the Conservatives found themselves in second place to Keir Starmer’s party. In Grimsby, where Labour failed to make headway last year and was demolished in 2019, the party topped the vote.
At the same time, in Keighley and Ilkley, a seat with a forecast Labour majority of more than 20 points, the local elections show a Tory lead of four points. Peterborough, with a vote fragmented between local independents and the Greens, also gave the Conservatives a narrow advantage over Labour despite it being an easy opposition target.
On the headline “who led where” metric from the May local elections, there were more Labour hits than misses. But there were enough misses to make this statistic worthy of note. A combination of split votes and swings away from Labour in seats with large Muslim populations, can explain most of this. The key question is whether these trends are borne out in the general election.
Were you to adopt a literalist reading of the local election results then, yes, it appears as if, in more than a few dozen seats, Labour will underperform. But the likelihood of voters behaving in the general election as they did in the locals is questionable. In Gloucester for instance, the council wards went 28 per cent Tory, 28 per cent Lib Dem, 26 per cent Labour and 5 per cent Green. But in a general election it is likely that, as they did in December 2019, voters will coalesce between the two main parties.
The old adage that people vote differently in local elections has often had relatively limited currency. But I wonder if it’s more relevant today than in decades gone by. I wrote in the aftermath of the local elections that Labour was underperforming its expected polling position. What remains uncertain is whether this underperformance reflects local anomalies – or a national polling miss.
[See also: The Tory plan for pensioners is expensive and pointless]
Content from our partners
Related
