This is a pre-publication draft of a chapter that has since been published in Conservative Governments in the Age of Brexit (Palgrave Macmillan, 2023), pp. 47–64.
Introduction
At first glance, the Conservative party’s performances and results in the general elections of 2015, 2017, and 2019 could hardly have been more different from each other. Three different leaders, David Cameron, Theresa May, and Boris Johnson, achieved respectively (1) a surprise majority when a hung parliament was widely expected; (2) a hung parliament when a majority was expected; and (3) a majority when a majority was expected. The first election brought about the Brexit referendum; the second caused stalemate in implementing its result; and the third cut the knot, ended the parliamentary crisis, closed the Brexit debate for any but academic purposes, and left the Conservatives with an 80-seat majority until (nominally) 2024.
It is my intention here to narrate the performance of the Conservatives through these three elections, but with special attention to voting patterns. The branch of political science that we call electoral studies is today professional and sophisticated, which makes it impossible to find anything new to reveal about voting behaviour in elections that happened more than sixth months ago. Here I have therefore sought primarily to present only a ‘distillation’ of what the reader can inspect in greater detail elsewhere, should he wish. For almost all students of politics, the volumes of the British General Election series contain everything that could be wanted, so for the reader’s convenience I have referred to the three relevant volumes in that series wherever possible.
However, I will also target some of the received wisdom about these elections, in order to overturn a number of common myths about them—myths which have usually been birthed by election night pundits trying to understand results as they’re coming in, which they often do by positing a single cause. Such explanations always oversimplify voter behaviour. Yet they also tend to enjoy a long afterlife in political commentary. The first of these ‘myths’ is that the Conservatives’ surprise majority in 2015 owed to David Cameron’s ‘disingenuous’ offer of an in/out referendum on membership of the EU (see Cameron, 2020: 398–417). The second is that the 2017 election showed that voters did not, after all, support Theresa May’s vision of a ‘hard Brexit’. The third is that the 2019 result reflected the greater popularity of Johnson compared to May, and the superiority of his campaign, especially in Labour’s former ‘red wall’ in the Midlands and North of England and Wales.
Each of these myths is false, and psephological evidence shows us what has been overlooked. First, Cameron’s referendum policy did not win a significant number of new Conservative voters in 2015. Second, in 2017, although the party’s slender Commons majority was lost, the election was in fact very successful for the party outside of the House of Commons, and laid the foundation for its 2019 landslide. And finally, the party leader and his Brexit policy were hardly more decisive issues for Conservative voters in 2019 than they had been in 2017; it was really the collapse of support for Labour outside of the big cities in 2019 that dramatically redrew the electoral map.
The 2015 General Election
Of the three elections discussed here, only 2015 was held as envisaged by the Fixed Term Parliaments Act, and was the only one that was planned well in advance. Its date had been public knowledge since it was baked into the agreement that founded the Conservative/Liberal Democrat coalition government in 2010. In the intervening years, the Coalition’s project of reducing the government budget deficit in the wake of the 2007–08 global financial crisis may have been broadly recognized as necessary, but each part of the ‘austerity’ method for achieving it—cuts to government spending, public sector pay freezes, reforms to benefits, quantitative easing—provoked criticism (see Cowley and Kavanagh, 2016: 1–40; Lee, 2011). These measures also seemed to have had side-effects, including rising unemployment, riots in several English cities in 2011, a ‘double-dip’ recession in 2012 (which on later analysis turned out not to have happened), and a downgrading of the UK’s credit rating. The Coalition had also performed a number of ‘U-turns’, including on proposed cuts to schools and coastguards, and on the ‘pasty tax’. For its critics, such episodes suggested that the ‘Con–Dem’ Coalition was ideologically callous and practically incompetent. There had also developed a general public distrust of Westminster and the press. Neither the expenses scandal in 2009, nor the phone-hacking scandal in 2011, had left the Conservatives untouched—though neither seemed to have helped other parties either.
Also during these years, the three issues of Islamist terror, immigration, and membership of the EU would become more closely connected in public opinion—while they continued to be treated as separable by mainstream politicians. In the earlier years of the Coalition there seemed to be only a very tenuous connection between the ‘home-grown’ Islamic fundamentalism of Abu Hamza al-Masri, the killers of Fusilier Lee Rigby, and those conflicts in the Islamic world which terrorists used to justify their crimes. But in the light of the European migrant crisis, which began in 2014, such factors became increasing more politically salient as parts of a single issue.
By this time, Cameron was under pressure from many of his own MPs to see off the UKIP threat. In the 2009 European Parliament elections, UKIP—then the only party committed to withdrawal—had won just 2.5 million votes. Four years later, in the 2014 European elections, it topped the national poll with nearly 4.4 million votes, winning in every English region from Cornwall to Yorkshire, with the sole exception of London. UKIP was now unambiguously the natural recipient of Eurosceptic votes. But its leader, the charismatic and straight-talking Nigel Farage, had also repositioned himself as Britain’s leading critic of the Westminster consensus on immigration. He thereby won for UKIP the sympathy of those also concerned about other aspects of the liberal Westminster consensus on social and cultural issues: negligent leniency towards criminals, terrorists, and the illegal drugs trade; the growth of dependency culture and the sentimental futility of ‘hug a hoodie’; the legal innovation of same-sex marriage, and the transformation of London into an impoverished babel and millionaires’ playground. In 2014 UKIP had pushed the Conservatives into third place, behind Labour—by popular vote, vote share, and seats. Their coalition partners, the Liberal Democrats, had lost ten of the eleven seats they were defending. If this pattern of voting, or anything like it, had been repeated in the general election a year later, the Conservatives would find themselves back in Opposition.
In January 2014, the Conservatives hired Lynton Crosby to lead their election campaign. I cannot improve on Cowley and Kavanagh’s sketch of the Australian campaign strategist:
Known as the “Wizard of Oz” because of his role in guiding John Howard in four successive Australian elections, he had also managed Boris Johnson’s successful mayoral campaigns in 2008 and 2012 in largely Labour-voting London. He had a reputation in Australia as a right-winger because of his hard line on immigration … Crosby’s reputation rested on his success in identifying the issues and language which would appeal to key voters and enforcing message discipline. (Cowley and Kavanagh, 2015: 61)
The Conservatives’ own research on voters’ perceptions and values, combined with Crosby’s commitment to ‘message discipline’, produced two primary campaign themes: (1) the economy: i.e. the Conservative party’s ‘long term economic plan’); and (2) leadership: i.e. competence versus chaos (see Cowley and Kavanagh, 2016: 61–5). The themes interlocked neatly: ‘economic competence with the Conservatives, or chaos with Ed Miliband propped up by the SNP’. No attack on the Liberal Democrats was deemed necessary, since their polling was consistently poor, and they might anyway still be needed following another hung parliament.
It may surprise some readers that Crosby did not make key messages out of immigration or the in/out referendum on the EU that Cameron had already promised in January 2013—surprising, that is, given Crosby’s popular reputation, and given that the referendum is now generally seen as the most significant outcome of the Conservatives’ victory in 2015. But in 2014 and 2015, those issues did not reflect the priorities of the Conservatives’ target voters. It may be that Cameron had misjudged those priorities in 2012, when he had decided to offer voters the opportunity to have their say on the renegotiated terms of EU membership. It seems anyway that Cameron genuinely believed that it was right policy (Cameron, 2019: 398–417), which contradicts those who have dismissed the policy as a disingenuous election promise intended to be bargained away in a new round of post-election coalition negotiations—an interpretation that has survived among staunch Remainers. The chief electoral benefit of the referendum policy was that it aided party discipline. First, it parked the issue of Europe within the party, and therefore neutralized potentially off-message Eurosceptic Conservative MPs who had long been pressuring Cameron to deal with what they saw as the threat of UKIP. And second, it neutralized otherwise difficult questions about Europe so that the Conservatives could quickly return to talking about the long-term economic plan, and the prospect of Labour and the SNP wrecking the hard-won economic recovery.
Meanwhile, Labour had replaced Gordon Brown with Ed Miliband. But by the seventh of May 2015, the party, and Miliband personally, were not polling well enough to win a majority. The Conservatives’ chief concern now was not that Labour would win a majority, but that it could recover enough to lock the Conservatives out of government with the help of other parties, even if the Conservatives emerged as the largest party. It was widely expected that the next government would be another coalition, in the shape either of another term for the governing parties—which were, however, expected to suffer some losses—or of a Labour Prime Minister with some kind of support from other parties, probably including Alex Salmond’s Scottish National Party.
At 10 o’clock in the evening on the 7th of May 2015, the BBC’s exit poll revealed that the Conservatives would be the largest party in the Commons. In itself, this was no great surprise. Labour’s advances in England were poor, and in Scotland Labour saw its worst result since 1918. The Liberal Democrats, as predicted, lost support almost everywhere, but unevenly, and more dramatically than many in the party had expected. It was the Lib Dems’ worst result since 1970. Their voters had split four ways, and those sticking with the Lib Dems were not the largest group. (More had switched to Labour.) The Lib Dems had been overtaken on popular vote by UKIP. Indeed, more of UKIP’s new voters had voted Lib Dem in 2010 than had voted for UKIP that year.
But it wasn’t until later in the night that it became clear that a Conservative majority was actually still in play. The projected swing to Labour was smaller than expected in their target seats; indeed, in many, the swing was against Labour. Overall the Conservatives gained twenty four seats, giving them a majority in the Commons of twelve, their first majority win since 1992. So what exactly had happened?
First, the Conservatives’ surprise majority in the Commons disguised electoral stagnation and even decline in some areas. They had gained 24 seats (nett) with the addition of just 630,472 more votes than they won in 2010, a vote share increase of just 0.8 pp. Labour had collapsed in Scotland, but the Conservatives had not advanced there at all. Of Scotland’s fifty nine MPs, all but three went to the Scottish National Party. This may not seem surprising, given the dramatic rise in turnout in Scotland off the back of the independence referendum the previous year. But those Conservatives who had expected some sort of corresponding advance as ‘No’ voters swung behind the Conservatives would be disappointed.
Turnout in England had barely increased on 2010, and in the large cities of the North, and in Lancashire and Yorkshire more generally, the Conservatives had actually lost support. Other parties were also advancing across the UK in areas with higher unemployment, and those with higher numbers of ethnic minority voters. Further, although the Conservatives had benefitted from the collapse of the Liberal Democrats (more on which in a moment), in many former Lib-Dem strongholds Labour were the chief beneficiaries—perhaps as some disaffected Lib Dems in these areas switched to Labour tactically or in protest. Indeed, more of 2010’s Lib Dem voters voted for Labour in 2015 than voted Lib Dem again.
However, the Conservatives had held most of the seats they’d taken from Labour in 2010. This ‘incumbency effect’ would also feature in 2017 and 2019. It was especially pronounced where the Conservative candidate was defending the seat for the first time, and where the leading opponent was a new candidate. Here there was an average swing to the Conservatives. But even in such seats where there was a swing away from them, they usually still held the seat (Curtice et al, 2016: 397–9).
It was the gains, however, that won the election, and the Conservatives made thirty five of them. Just eight of these seats were taken from Labour: Bolton West, Derby North, Gower, Morley and Outwood, Plymouth Moorview, Southampton Itchen, Telford, and Vale of Clwyd—a diverse group of constituencies, evincing few general patterns, and several local factors.
The remaining twenty seven seats gained by the Conservatives were taken from the Liberal Democrats. These were mostly economically buoyant Southern towns, in London suburbs, such as Kingston and Surbiton, Sutton and Cheam, and Twickenham; and those more rural constituencies in the Westcountry where the Liberal Democrats normally gather the anti-Conservative vote. Constituencies fitting this profile included Bath, Cheltenham, Chippenham, Taunton, Torbay, Yeovil, Wells, and several in Devon and Cornwall. In such areas, household income was higher than average, and there were fewer public sector employees.
It seems, then, that in 2015 the Conservatives probably benefitted from the combination of two main factors. First, voters who had experienced economic recovery in their towns and in their private sector work rewarded the Conservatives for it. And second, in areas where Labour was traditionally weak, the Lib Dems had lost control of the anti-Conservative vote, which was now split between other parties, allowing the Conservatives to win seats without necessarily increasing their vote share.
Voting patterns also seriously problematize the view that Cameron’s referendum policy had made much of an impact in this election. First, as we have seen, the Conservatives’ key gains were made in affluent former Lib-Dem strongholds in the South of England. In the EU referendum the next year, these same areas would lean towards Remain, or at least less heavily towards Leave. Meanwhile, those ‘more Brexity’ areas in the Midlands and North of England and Wales, where average levels of wealth and education were lower, saw no Conservative advance in 2015. Second, it had been supposed before the election that UKIP would be a headache for the Conservatives, and taken as a whole this turned out to be true, despite the referendum policy. UKIP advanced significantly in 2015: they won 12.6% of the popular vote, their best ever performance in a general election. Very few of 2010’s UKIP voters switched to the Conservatives in 2015; far more stuck with UKIP, while the Conservatives lost more voters to UKIP than they won from any other single party.
However, because support for UKIP grew less where the Conservatives were defending marginal seats, the party in Parliament was insulated from the popularity of UKIP nationally. Elsewhere UKIP also took voters from Labour. This affected, above all, those areas which had rejected Labour most heavily in 2010: areas with relatively few university graduates and more public sector employees, those most affected by the economic crash of 2008, and those with higher numbers of white working-class voters.
Though it has very commonly been supposed that immigration was the leading issue for these voters—an issue on which approval of Labour in these areas was now dangerously low—in fact UKIP did much better in areas relatively unaffected by immigration. It might be supposed that UKIP’s success was therefore due to the fear of immigration, rather than to first-hand experience of it. This is the story told by Danny Dorling and Sally Tomlinson, for example (2020). However immigration is just one point in a complex of cultural questions on which Labour was diverging from the white working class. More than any other indicator, including class and ethnicity, the strongest predictor of Conservative and Labour support in 2015 was newspaper readership. Voting for Labour was now correlated less strongly with being working class, and more strongly with reading The Guardian or The Independent—papers that take a consistently liberal or even anti-British position on cultural questions. Thus, while Labour was holding and gaining support among what David Goodhart would later called ‘Anywhere’ voters, whose work and cultural attachments are geographically transferable, it was already losing them in working-class constituencies. The cultural realignment of the Labour vote that would later be exacerbated by the party’s position on Brexit was already affecting its electoral performance in 2015, while at the same time, the Conservatives were advancing not so much in the areas that Labour would later so dramatically lose—poorer, culturally conservative towns—as in richer, more culturally liberal towns.
The 2017 General Election
[Author], in this volume, has already explained the campaign, results, and fallout of the 2016 Brexit referendum, so there is no need to repeat that story here. The Home Secretary, Theresa May, had contributed almost nothing to the Remain campaign; her experience with the European Court of Human Rights had probably hardened her attitude to the Continent’s institutions, and she was suspected by many Conservative colleagues of planning for a leadership challenge after the referendum (Cowley and Kavanagh, 2018: 43). By the summer of 2017, and within a year of May’s acceding as party leader, the debate over Brexit within the Conservative party had moved on from ‘Leave or Remain?’ to what kind of Brexit would be preferable. As usual, despite a complex number of options being viable, popular political debate reduced the options to two, in this case labelled ‘hard’ and ‘soft’. In political practice, these unhelpfully vague terms could be ignored. But if it came to an election, there would be little political capital to be gained in openly advocating a ‘soft’ Brexit. May therefore ‘hardened’ her stated aims. Announcing in Downing Street that there would be a general election on the 8th June, she said, ‘Britain is leaving the European Union, and there can be no turning back … That means we will regain control of our own money, our own laws, and our own borders, and we will be free to strike trade deals with old friends and new partners all around the world’.
May’s stated aim in calling the election was to defeat the obstructions and ‘gameplaying’ at Westminster, to replace them with ‘strong and stable leadership in the national interest’, and to strengthen her hand in negotiations with the EU. Those negotiations would first centre upon the terms of withdrawal, and then on the future relationship. But, May argued, the British Government’s position would be weaker in the run-up to the next scheduled election in 2020. As many Conservative MPs were opposed to Brexit, May could expect sizeable rebellion in the Commons, which could block the process completely—with economic consequences for the UK and political consequences for the party.
The Conservatives were widely expected to enlarge their majority in June 2017, and this probable complacency among voters was identified early on as a real problem for the campaign. Voters might think that, since Conservative victory was inevitable, that they could safely vote for non-Conservative local candidates whom they liked more, or register a ‘protest’ vote, or that they could use their votes to promote whichever exciting Labour policies had particularly attracted their eye. Growing Conservative support seemed to mean making Corbynite Labour a genuine challenge and danger, while positioning May as a national leader who needed a powerful mandate to strengthen the UK in the context of a complex international process. May’s campaign team needed to paint her as a unifying figurehead who could be supported by reasonable people who, regardless of past party allegiances, accepted the referendum result, even if they didn’t vote for it, and who wanted terms of withdrawal that served the national interest. The Conservatives could not allow May to be assessed as merely a domestic party leader whose policies offered little in reply to Labour’s generous spending programme. For this reason, the campaign strategy emphasised May herself, her ‘team’ (somewhat de-branded), and the great timely need for ‘strong and stable government in the national interest’.
It is not unusual for MPs to be secretly critical of their leader’s election strategy, especially in light of a bad result. But Theresa May had the added disadvantage of an extremely short planning period, and a reputation among her colleagues for making decisions with the help of a very small group of close advisors. Commentating for ITV’s election night coverage, George Osborne, by now the editor of the London Evening Standard, claimed that the Conservative manifesto had been ‘drafted by her [May] and about two other people, was a total disaster, and must go down now as one of the worst manifestos in history’. In fact it was not the ‘two’ that Osborne probably had in mind, Nick Timothy and Fiona Hill, who had most shaped the Conservatives’ campaign: it was, again, Lynton Crosby. May had also appeared over-protected during the campaign. In contrast to Cameron, who in 2010 and 2015 had taken part in live television debates and appeared, sleeves rolled up, among public audiences, May seemed somewhat too obviously to be avoiding any situation in which she might struggle to respond to challenge, or which might take her off message. Three days before polling, May told an ITV interviewer that the naughtiest thing she had ever done was to ‘run through fields of wheat’.
Although, as we will see, the Conservative campaign certainly seems to have grown support for the Conservatives dramatically, it might be argued that the narrowness of the message left the party vulnerable on domestic policy.
It quickly became clear from the exit poll, released at 10pm on 8th June (2017), that May’s gamble had failed. Though still the largest party in the Commons, the Conservatives were projected to win just 314 seats, twelve short of a majority. At worst, they could be out of office, and Labour able to form an agreement with other parties to make Jeremy Corbyn Prime Minister. By 10:05pm, dozens of commentators in national media had questioned May’s ability to stay on as party leader. In the end the Conservatives would win three crucial seats more than projected, and were able to continue in office thanks to a “confidence and supply” arrangement with Northern Ireland’s Democratic Unionist Party. But still May’s authority, at home, within the party, and also on the Continent, appeared damaged. Far from increasing the Conservatives’ 2015 majority, May had thrown it away, and Parliament was hung again. The 2017 election looks, on the surface, like a reversal of 2015, and a reversion to 2010—only this time it was the more culturally conservative DUP that would prop up a Conservative Prime Minister, rather than the culturally liberal and uniformly Europhile Liberal Democrats. In 2010 there was agreement between the coalition parties on the big issue of the time, which was the economy. But now, on Brexit, the DUP would effectively have a veto on whatever Withdrawal Agreement the Conservatives were able to negotiate, which could be a problem if any compromise seemed to have been made over Northern Ireland’s constitutional status within the UK.
What had gone wrong for “Theresa May’s team”? As usual, commentators looked to the “turning points” of the campaign for an explanation. May’s personality; the empty manifesto; Crosby’s excessively narrow message: “strong and stable in the national interest”. Later, in an interview with Emma Barnett for the BBC, even May herself would express regret that she had not made more of her own positive Conservative case for opportunity and social mobility, and had allowed herself to be guided by Crosby’s strategy. At first, it appeared that the Conservatives’ offer of strong leadership through the Brexit process had been rejected. There had been much speculation about whether the Conservatives could enlarge their majority by winning back former UKIP voters, whose party now seemed superfluous. Because that majority had been lost, this looked, in retrospect, like it had been wishful thinking. Clacton had been taken from UKIP, as expected. But everywhere that UKIP support had collapsed, Labour also seemed to have benefitted. Perhaps Crosby was wrong? Perhaps voters had changed their minds on Brexit, or preferred a weaker government forced to build consensus with other parties?
Analysis of voting patterns tells a different story, and seems more to vindicate Crosby’s strategy—certainly in the long run.
The Conservatives had also taken one of the few surviving Liberal Democrat seats. But their gains had been very few. They had taken only six from Labour—far fewer than they’d hoped—and those lost to Labour numbered twenty eight, and included Brexit-supporting areas. Four of Cameron’s gains from the Liberal Democrats had been taken back in areas where support for Brexit was more mixed.
However, the Conservatives’ losses in the Commons disguised the significant advances that were made across the country, especially in Scotland, and among working-class voters in the English towns that would be known in 2019 as Labour’s “Red Wall”—towns such as Workington. This went well beyond the “incumbency effect”. Theresa May had won more votes than the Conservatives had won at any election since 1992, and more even than Tony Blair had won for Labour in 1997. The Conservatives had not enjoyed such a large share of the vote (42.4%) since 1983. Most historically, as Curtice et al have noted, the vote share increase on 2015 was “the largest increase in support enjoyed by any incumbent government that had won the previous election since the 1832 Reform Act” (Curtice et al, 2018: 452). However this performance was unusually uneven (see below).
Because they were focusing on May’s disaster in the Commons, the consequences for the Brexit process, and Corbyn’s surprise success, many commentators did not give much attention to the Conservatives’ significant advances in Scotland, where it won a higher share of the vote (28.6%) than at any election since 1979. The growth in Conservative support was strongest in areas where support for Leave had been strongest. However, because the primary issue in Scottish politics was now Scottish independence, following the divisive 2014 referendum, the Conservatives also benefitted from tactical voting among Unionists, especially where the Liberal Democrats now seemed unlikely to unseat Scottish National Party MPs (Curtice et al, 2018: 468–9). In other areas, however, a significant number of former Conservative voters voted tactically for Labour.
These successes, however, were poor consolation for Conservatives the morning after the election. Their problem was that Labour had gained thirty seats overall. While still behind the Conservatives on seats, vote share, and popular vote, Labour had increased its vote share by a dramatic 9.8 pp—more than in any election since 1945. Labour had won more votes than in any election since 1997, and had nearly matched its 2001 vote share (40%). At no election since 1970 had the two main parties claimed such a large proportion of the vote: in England, the two-party system seemed to have returned dramatically.
For some time it was believed that Labour’s performance was due to a “Youthquake”: a mass mobilization on election day of young voters enthused for Jeremy Corbyn. It now seems, however, that the relationship between voter age and turnout changed very little from 2015; indeed 2017 may have seen lower turnout than 2015, but both possibilities are well within the margin for error (Prosser et al).
Another overlooked fact was a statistical development in the Conservatives’ support base in England and Wales—a development which it is now customary to call “realignment”. The Conservatives had lost support in strong Remain areas, which tended to be more affluent. But they had gained significantly where the Leave vote was strongest, where levels of formal education were lower, and which were less economically buoyant. This did not translate into seats—at least, not yet (Curtice et al, 2018: 453). But the differences it implied between the Conservative supporters of 2015 and 2017 are significant: many new Conservative voters were less cosmopolitan, less educated, in more routine manual employment (Curtice et al, 2018: 461), and poorer. For Labour, the statistical significance of Leave/Remain support was much weaker in 2017. Perhaps this was because Labour had identified that voters still cared more about domestic spending priorities than they did about Europe; or perhaps Labour’s ambiguous Brexit stance worked to accommodate different views. In retrospect, however, we can say that the Conservatives had taken an early lead in a race to realign party support. The Conservatives were claiming Leave voters, including in the Midlands and North of England; Labour would need, then, to claim more Remain voters, including in the South of England. The prize for winning the realignment race would be (as it turned out) a large majority in the general election of 2019.
The 2019 General Election
In view of the Conservatives’ now obvious need to attract the full range of Eurosceptic voters—from those who always hated the EU, to those who voted Remain but now supported leaving on good terms—it is perhaps surprising that, as late as June 2019, there were still leadership candidates who were seen as Remainers—one (Rory Stewart) even arguing that a ‘no-deal’ withdrawal should be taken ‘off the table’, though in legislative terms it was not clear what this meant. Sajid Javid told the first leadership hustings that ‘You don’t beat the Brexit Party by becoming the Brexit Party’. Sam Gyimah was perhaps the most vocal Remainer to enter the leadership contest: he advocated a referendum on the withdrawal agreement, with the option of remaining in the EU on the ballot paper. Those who elected Boris Johnson, at whatever stage of the party leadership election process they participated, can hardly be expected to ignore consistent polling data which were by now common knowledge: that far more respondents said they would vote for a Conservative Party led by Boris Johnson than for one led by someone else.
Johnson’s campaign slogan, ‘Get Brexit done’, was even simpler than May’s had been two years earlier, and it had the advantage of reflecting the impatience and exasperation with the wrangling at Westminster that many voters now felt (see Ford et al, 2021: . This simple message also lent itself to blunt visual metaphors: Johnson was photographed in blue “Get Brexit done” boxing gloves; driving a “Get Brexit done” tractor through a wall labelled “Gridlock”.
The release of 2019’s exit poll was the final whistle for those who had spent more than three years trying to prevent Brexit and for those who had sought to make a Prime Minister out of Jeremy Corbyn. Although polling had consistently shown a Conservative lead throughout the campaign, and especially a personal lead for Johnson ahead of Corbyn, most politicians and commentators were surprised by the scale of the Conservative victory. Johnson had won 43.6% of the popular vote—the highest share for any party since Margaret Thatcher’s victory in 1979. Labour had won 32.1% of the vote, which is more than it won under Ed Milliband in 2015 (30.4%). However, because of the way Britain’s electoral system works, this gave Labour the lowest number of seats since 1935. It is perhaps worth pointing out that Johnson’s majority in the Commons was also smaller than historical precedent might have made him feel he deserved: in 1983 Margaret Thatcher won just 42.4% of the vote, but won 397 seats. If anything, Johnson had this in common with May: results in the Commons seemed to underrepresent the popularity of both in the country. The Conservatives had lost some seats, but most of these were to the SNP in Scotland, which had also taken six seats from Labour. The much bigger story was the fifty four seats that the Conservatives had won from Labour. These cut deep into the “red wall” of formerly safe Labour seats in the North of England and Wales—many of which had not elected Conservative MPs since the 1950s or earlier. Labour’s 202 seats comprised a worse result for them even than 1983; it was in fact the worst since 1935. The political map had been dramatically redrawn in the North of England, and with it the arithmetic of the House of Commons. Johnson’s new majority of 80 gave him the numbers he would need to get his “oven-ready” Withdrawal Agreement through the Commons, and the UK out of the EU within two months. It also gave him that stronger position that May had sought for the next stage of EU negotiations, and it gave him another five years of Conservative rule with a strong mandate and high number of Conservative MPs elected under his leadership.
At 4:13am on election night, BBC journalist Andrew Neil asked Theresa May the obvious question, if a very awkward one:
When you called an election two years ago, you lost your majority. Boris Johnson has called an election and achieved a quite substantial majority. What has he done right that you did wrong?
Johnson had won more votes for the Conservatives than Theresa May, but not many more—though Neil’s interviewee would not have known this at the time. Johnson had added just 329,770 to the Conservatives’ popular vote; Theresa May had added 2.3 million.
What had made the difference in 2019 was that, although Johnson had lost 2017 voters to other parties and to abstention, he had more than compensated for these losses by attracting new voters from Labour. A third of those 2017 Labour voters who had voted Leave in 2016 had now backed Johnson. (Perhaps more surprisingly, he also attracted 14% of those who voted Liberal Democrat in 2017.) In percentage terms, this is not the largest group of Conservative switchers: that position is taken by the 45% of Leave-voting 2017 Liberal Democrats. However this is a larger percentage of a much smaller number of voters. Only in North Norfolk does this Lib-Dem–Leave group seem to have caused a seat to change hands, and even there local factors were probably more significant. Labour had also lost 12% of their Remainers to the Liberal Democrats, 3% of them to the Conservatives, and another 6% of their Leavers to the Brexit Party. Between 2017 and 2019, Labour under Corbyn lost 2.6 million voters nett. The combined effect was extremely damaging: fifty four seats were lost to the Conservatives (only one, Putney, went the other way); six were lost to the SNP. The Corbynite explanation for this was simple: this was a “Brexit election”, and the party’s complicated Brexit positioning confused voters. In fact, Corbyn had long been polling poorly with working-class voters. Even worse for Labour, the Brexit Party contested Labour seats but not Conservative ones.
For Labour, then, the collapse of support in the “red wall” seemed sudden and dramatic compared to 2017. But on the Conservative side of the election, 2019 was a story of remarkable consistency. The switchers I’ve just described only comprised 15% of Conservative voters. The great majority (85%) of the Conservatives’ 2019 voters had also been 2017 voters, and many of these were “red wall” voters who had made their switch back then, and not only now. In many seats (such as Bolsover) the Conservatives had increased their support by more in 2017 than they did in 2019, even though it was only in the latter election that the seat was actually won.
Johnson had, then, very effectively retained the backing of 2017 Conservatives. Labour’s retention was weaker, at just 72%. Indeed, of those who had voted Leave (2016) and Conservative (2017), 92% had stuck with the Conservatives in 2019—which we would expect. But the defection of 2017’s Conservative Remainers was also smaller than we’d expect: 65% of this group still voted Conservative in 2019, and of the rest, only 8% switched to Labour. Far more Conservative Remainers (22%) switched to the Liberal Democrats—which, ironically, may well have allowed the Conservatives to hold or gain crucial Con–Lab marginals, and increase the majority they needed for getting Brexit done.
There was also nothing very new about Conservative voting patterns, demographic trends, or what I have called the “realignment”. As is 2017, the 2019 election also saw a stronger swing from Labour to the Conservatives where there were more working-class voters, and where the proportion of Leave-voters was higher. Social class really has ceased to be a significant indicator of Conservative support—certainly relative to age-group. It is also noteworthy that the Conservatives had the lowest proportion of voters of any party who reported voting tactically: just 17% of their voters did so; the other 82% (the figures are rounded) wanted them to win. For Labour those figures were a considerable 31%, and just 67% respectively (Ashcroft, 2019).
The realignment of Leave-voters towards the Conservatives that had already begun in 2017 thus continued in 2019. With 74% of 2016’s Leavers voting for them in 2019, the Conservatives are now unequivocally the preferred party of government for most Leavers. Conversely, Labour has not become the preferred party of government for Remainers. Only 49% of Remainers voted for Labour in 2019; 19% of the rest voted for the Conservatives.
Since the trigger of that realignment, Brexit, has now subsided, it seems unlikely that the realignment has any more mileage. Retaining these new voters without losing the old ones is now the challenge. “Patriotism” may be the natural successor theme to Getting Brexit Done, but without concrete political disputes to divide voters into the patriotic and the “unpatriotic”, it is unlikely to be a powerful device. This has left the Conservatives with the social composition of a more “One Nation” party than was the case in 2015. They now draw support more evenly from every social class, and their MPs represent more people in the Midlands and North of England and Wales than many can remember. Their voters also now have a set of priorities distinct from other parties. Where other parties typically attract those whose priorities are (1) politicians’ good motives and (2) domestic spending promises, the Conservatives have attracted those who between 2016 and 2019 prioritized (1) the Brexit outcome, and (2) economic competence (Ashcroft, 2019). It seems probable that 2019’s “Remainer Conservatives”, who still form a significant portion of their vote base, had either accepted the necessity of executing the referendum result, or prioritized a competently-managed economy.
Labour’s realignment has been going on for longer, but its character is still cultural. While retaining much of its traditional white working-class supporters in poorer areas, Labour has also lost a significant number of them to the Conservatives, especially over Brexit and Corbynism. Meanwhile, Labour has piled up support in areas that are above-averagely culturally liberal, affluent, and Muslim (Curtice et al, 2018: 458). At time of writing, it seems that this may prove a difficult coalition to maintain so long as questions of nationality and culture remain high on the political agenda, especially as it is primarily middle-class people in the UK’s larger cities that are joining the Labour party and steering its policies. Labour’s best hope is that the political wind blows questions of tax and spending back to the top of that agenda. It seems likely that the economic destruction of Covid-19 will leave behind exactly such questions.
References
The British General Election series:
Cowley, P., and Kavanagh, D. (eds) (2016) The British General Election of 2015 (Basingstoke: Palgrave Macmillan).
- Curtice, J., Fisher, S. D., Ford, R. (2016) Appendix 1: The Results Analysed. 387–431.
Cowley, P., and Kavanagh, D. (eds) (2018), The British General Election of 2017 (Basingstoke: Palgrave Macmillan).
- Curtice, J., Fisher, S., Ford, R., English, P. (2018). Appendix 1: The Results Analysed. 449–95.
Ford, R.; Bale, T.; Jennings, W.; and Surridge, P. (eds) (2021) The British General Election of 2019 (Basingstoke: Palgrave Macmillan).
Other texts cited
Cameron, D. (2019) For the Record. William Collins.
Dorling, D., Tomlinson, S. (2020) Rule Britannia: Brexit and the End of Empire (London: Biteback).
Lee, S. (2011) No Plan B: The Coalition Agenda for Cutting the Deficit and Rebalancing the Economy, in Simon Lee and Matt Beech (eds), The Cameron–Clegg Government: Coalition Politics in an Age of Austerity. Palgrave Macmillan. pp. 59–74.
Lord Ashcroft Polls, “How Britain voted and why: My 2019 general election post-vote poll”.
Prosser, C; Fieldhouse, E. A.; Green, J.; Mellon, J.; Evans, G., “Tremors But No Youthquake: Measuring Changes in the Age and Turnout Gradients at the 2015 and 2017 British General Elections” (January 28, 2018). Available at SSRN: https://ssrn.com/abstract=3111839 or http://dx.doi.org/10.2139/ssrn.3111839
YouGov, “How Britain voted in the 2019 general election”.
Dr Christopher Fear 