Why the Tories are wrong to hope for a Thatcher-style poll recovery

Unlike Thatcher in the 1980s, the current Conservative Party does not enjoy the advantage of a divided left.

Having trailed Labour in every poll since 18 March (three days before George Osborne's fateful Budget), the Conservatives are consoling themselves with the thought that they have been here before. Margaret Thatcher, they recall, often lagged behind in the polls but twice recovered to win landslide victories in 1983 and 1987. At the Conservative conference earlier this month, David Cameron told a 1922 Committee/ConservativeHome reception that between 1983 and 1987, the Tories averaged just 24 per cent in the polls but went on to win a majority of 102 seats at the election. While he would "settle for less than that", he believed that "boundaries or no boundaries", the Tories could win. It's not just Cameron who is seeking to strike a more optimistic tone. Ken Clarke recently told the cabinet that "he had never been in a government that had been more popular at mid-term." With unemployment and inflation down, and the economy out of recession (although for how long remains to be seen), some Tories are beginning to dream of a majority again. But a closer analysis of the polls suggests that they're wrong to be so sanguine.

First, it's unclear which polls Cameron is referring to when he claims that the Tories averaged support of 24 per cent between 1983 and 1987. As UK Polling Report's comprehensive archive of polls shows, only once (on 12 August 1985) did backing for the party fall this low. Cameron probably meant to say that support for the Tories averaged 24 per cent at this stage in Thatcher's second term (as his personal pollster Andrew Cooper is reported to have told the cabinet), but even this claim doesn't stand up. In 1985, support for the party more often stood at around 33 per cent. The Thatcher recovery was not as great as the Tories suggest.

Second, unlike Thatcher, the current Conservative Party does not enjoy the advantage of a divided left. One of the biggest obstacles to a Labour majority in the 1980s was the strength of the SDP-Liberal Alliance, which won 25.4 per cent of the vote in the 1983 election and 22.6 per cent in 1987. It is the present weakness of the Lib Dems that is one of the biggest obstacles to a Conservative majority.

While it is the Tories who are in second place in most Lib Dem seats (38 compared to 17 for Labour), any gains they make from Nick Clegg's party are likely to be outweighed by the gains Labour makes as Lib Dem defectors carry the party to victory in Tory marginals (see Rob Ford's recent post "Who benefits from a Lib Dem collapse?" for more on this). We are seeing this trend at work in the Corby by-election, where a recent poll by Lord Ashcroft found that support for Labour had risen from 39 per cent to 54 per cent since the general election, while support for the Lib Dems had plummeted from 15 per cent to five per cent. Corby is one of 38 Labour-Tory marginals where the third place Lib Dem vote is more than twice the margin of victory, showing the potential for Miliband's party to make significant gains even if Clegg's party partially recovers before 2015. In addition, while existing Lib Dem MPs, many of whom enjoy large local followings, are likely to benefit from an incumbency effect, it is the Tories, not Labour, who will suffer as a result (as I noted, Cameron's party is in second place in 38 of the Lib Dems' 57 seats).

Finally, while the Conservatives' core vote has held up better than many expected (the latest YouGov poll puts them on 33 per cent, down just three per cent since the general election), there is no evidence of the party advancing beyond this. The Tories are still in retreat in those areas – the north, Wales, Scotland – that denied them a majority at the last election. Rather than drawing false comfort from history, the Conservatives should focus on adopting the policies needed to change this, something they currently show little sign of doing.

The Conservatives hope that David Cameron, like Margaret Thatcher, will overturn Labour's poll lead before the general election. Photograph: Getty Images.

George Eaton is political editor of the New Statesman.

Photo: Getty
Show Hide image

What will the 2017 local elections tell us about the general election?

In her timing of the election, Theresa May is taking a leaf out of Margaret Thatcher's book. 

Local elections are, on the whole, a much better guide to the next general election than anything the polls might do.

In 2012, Kevin Cunningham, then working in Labour’s targeting and analysis team, surprised his colleagues by announcing that they had lost the 2015 election. Despite gaining 823 councillors and taking control of 32 more local authorities, Cunningham explained to colleagues, they hadn’t made anything like the gains necessary for that point in the parliament. Labour duly went on to lose, in defiance of the polls, in 2015.

Matt Singh, the founder of NumberCruncherPolitics, famously called the polling failure wrong, in part because Labour under Ed Miliband had underperformed their supposed poll share in local elections and parliamentary by-elections throughout the parliament.

The pattern in parliamentary by-elections and local elections under Jeremy Corbyn before the European referendum all pointed the same way – a result that was not catastrophically but slightly worse than that secured by Ed Miliband in 2015. Since the referendum, thanks to the popularity of Theresa May, the Conservative poll lead has soared but more importantly, their performance in contests around the country has improved, too.

As regular readers will know, I was under the impression that Labour’s position in the polls had deteriorated during the coup against Corbyn, but much to my surprise, Labour’s vote share remained essentially stagnant during that period. The picture instead has been one of steady deterioration, which has accelerated since the calling of the snap election. So far, voters buy Theresa May’s message that a large majority will help her get a good Brexit deal. (Spoiler alert: it won’t.)

If the polls are correct, assuming a 2020 election, what we would expect at the local elections would be for Labour to lose around 100 councillors, largely to the benefit of the Liberal Democrats, and the Conservatives to pick up around 100 seats too, largely to the detriment of Ukip.

But having the local elections just five weeks before the general elections changes things. Basically, what tends to happen in local elections is that the governing party takes a kicking in off-years, when voters treat the contests as a chance to stick two fingers up to the boost. But they do better when local elections are held on the same day as the general election, as voters tend to vote for their preferred governing party and then vote the same way in the elections on the same day.

The Conservatives’ 2015 performance is a handy example of this. David Cameron’s Tories gained 541 councillors that night. In 2014, they lost 236, in 2013 they lost 335, and in 2012 they lost 405. In 2011, an usually good year for the governing party, they actually gained 86, an early warning sign that Miliband was not on course to win, but one obscured because of the massive losses the Liberal Democrats sustained in 2011.

The pattern holds true for Labour governments, too. In 2010, Labour gained 417 councillors, having lost 291 and 331 in Gordon Brown’s first two council elections at the helm. In 2005, with an electoral map which, like this year’s was largely unfavourable to Labour, Tony Blair’s party only lost 114 councillors, in contrast to the losses of 464 councillors (2004), 831 councillors (2003) and 334 councillors (2002).  This holds true all the way back to 1979, the earliest meaningful comparison point thanks to changes to local authorities’ sizes and electorates, where Labour (the governing party) gained council seats after years of losing them.

So here’s the question: what happens when local elections are held in the same year but not the same day as local elections? Do people treat them as an opportunity to kick the government? Or do they vote “down-ticket” as they do when they’re held on the same day?

Before looking at the figures, I expected that they would be inclined to give them a miss. But actually, only the whole, these tend to be higher turnout affairs. In 1983 and 1987, although a general election had not been yet called, speculation that Margaret Thatcher would do so soon was high. In 1987, Labour prepared advertisements and a slogan for a May election. In both contests, voters behaved much more like a general election, not a local election.

The pattern – much to my surprise – holds for 1992, too, when the Conservatives went to the country in April 1992, a month before local elections. The Conservatives gained 303 seats in May 1992.

What does this mean for the coming elections? Well, basically, a good rule of thumb for predicting general elections is to look at local election results, and assume that the government will do a bit better and the opposition parties will do significantly worse.

(To give you an idea: two years into the last parliament, Labour’s projected national vote share after the local elections was 38 per cent. They got 31 per cent. In 1985, Labour’s projected national vote share based on the local elections was 39 per cent, they got 30 per cent. In 2007, the Conservatives projected share of the vote was 40 per cent – they got 36 per cent, a smaller fall, but probably because by 2010 Gordon Brown was more unpopular even than Tony Blair had been by 2007.)

In this instance, however, the evidence suggests that the Tories will do only slightly better and Labour and the Liberal Democrats only slightly worse in June than their local election performances in May. Adjust your sense of  what “a good night” for the various parties is accordingly. 

Stephen Bush is special correspondent at the New Statesman. His daily briefing, Morning Call, provides a quick and essential guide to British politics.

0800 7318496