Final polling projection: Pay attention, vote tactically—and everything is possible

There are fewer than 100 hours remaining until the polls open. We have just released the findings of our final Datapraxis model on the state of the election race, in a new report entitled “Tory Landslide or Hung Parliament?” It is based on advanced analytics of an extraordinarily large YouGov data set of almost half a million polling responses nationwide.

The race has narrowed slightly since I wrote about our mid-campaign model findings in Prospect two weeks ago. Then we had Boris Johnson’s Conservatives on track for a thumping majority of 48. This has now been slightly reduced, to a Tory majority of 38 seats.

Yet everything is still to play for. We have never seen as many undecided voters this late in the campaign. As many as 80-90 constituencies are still up for grabs. A much larger Conservative landslide is still possible—but so is a hung parliament. Under first-past-the-post, it is the decisions of just a couple of thousand swing voters in each of these seats that will determine the final outcome.

The new model finds the Conservatives winning 344 seats, a total they reach by snatching dozens of Labour seats that have always been thought of as Labour heartlands—across the midlands, the north, Wales and coastal towns: the crumbling of the so-called red wall. We estimate that Labour could drop to 221 seats as a consequence of this, 41 down on its 2017 result. The party has reclaimed eight seats since our last projection—mostly from the Conservatives—but a Corbyn majority is now clearly impossible.

The Scottish National Party is set to win more seats from both the main parties, but its new tally of 47 seats is two less than the mid-campaign Datapraxis model predicted. The Liberal Democrats’ campaign continues to stagnate, with them picking up just 14 seats, two more than they won in 2017 and significantly less than they had at the end of the last parliament. Many prominent defectors to their party, such as Chuka Umunna, Sam Gyimah, Sarah Wollaston, Luciana Berger and Phillip Lee, are currently projected to lose.

This is the final Datapraxis public model, but not the final call. If the Conservatives outperform the model and win a landslide, it will be first and foremost because a critical number of currently undecided Labour Leavers and Conservative Remainers ultimately resolve that they would rather tolerate the stench of Johnson’s incipient banana republic than embrace the bogeyman of Jeremy Corbyn.

If this strikes you as a highway to hell, do not give up hope. There is still time to take the off-ramp to a hung parliament. This is the nightmare scenario for Johnson and Dominic Cummings, and was the focus of their YouTube masthead advertising on Saturday. Senior Conservatives are already briefing about the fateful potential of combining “the two Ts,” youth turnout and anti-Tory tactical voting. The new Datapraxis report analyses what the effect of these two factors might be in 12 bellwether seats, illustrating vividly how these and dozens more seats could yet shift.

We are already seeing more people than ever before saying they will vote tactically, as traditional party loyalties erode. The final days will turn on a contest of anxieties. If the undecideds are in the end more anxious about an outsize Johnson majority, the threat of US-style privatisation of the NHS and the social damage done by another five years of the Tories than they are about “getting Brexit done” or the supposed “cost of Corbyn,” then they will use their votes to deny Johnson a majority.

Cummings and the “Wizards of Oz”—the Antipodean consultants running the Tory campaign —are finally unleashing their blitzkrieg of digital advertising. But face-to-face conversations and peer-to-peer sharing of politically persuasive content is far more powerful than dodgy attack ads. Non-party campaigns, many still crowdfunding support, will play a big role in the final days. Depending on their strategy, they can spend up to £9,750 each in battleground seat—around two-thirds of the parties’ own spending limits.

Momentum was Labour’s secret weapon in 2017. This year it has stepped up its efforts massively and received approval from the central party for much more spending. By combining data, social media and doorstep efforts, it is working to persuade voters and get them out to the polls.

But there is an enormous diversity of progressive efforts underway. Productions by the crowdfunded Real Change Lab (run by Corbyn’s former press secretary Matt Zarb-Cousin and Arun Chaudhary, formerly creative director for both Bernie Sanders and Barack Obama) include the viral Rob Delaney NHS video seen by almost 15m people. Hope Not Hate is working to stop the Brexit Party from winning any seats, and to increase turnout from under-represented groups.

More United is supporting moderate opposition candidates “who will put country over party,” and the anti-Brexiteers at Led By Donkeys are preparing their final deluge of ads. A new campaign called Boris Johnson Uncut is producing viral videos such as “Fact-chucker,” featuring an anonymous but remarkably good Johnson impersonator, and disseminating his confessions of his many political porkies over Facebook and Messenger. Just launched, it is well worth following.

At the start of the campaign it looked like analytical disagreements between progressives about who was best placed to challenge the Tories in which seats could drown the call for tactical voting in confusion. But thankfully, there is now 90-plus per cent alignment between the various sites, which have advanced plans to reach millions more in the final days. Tactical.vote, Gina Miller’s Remain United, People’s Vote, Best For Britain and VoteSmart2019 are the key players here.

Tiny civic startups like SheVotes, a non-partisan effort to turn out a million more young women by linking up with brands and influencers, has achieved huge reach on a shoestring. Meanwhile, RizeUp is funding street teams working to turn out poorer and ethnic minority younger voters in 30 marginals.

On the other side, the Conservatives and their consultants are tending their own grass-roots—or, at least, laying astroturf. The recent scandal around a pro-Brexit campaign called “Mainstream Network,” which was anonymously targeting swing voters in Labour-held marginals, provides one warning sign. Thomas Borwick, a senior digital influencer who has worked at both Vote Leave and as a consultant with the Cambridge Analytica network, is running a breathtakingly cynical effort called “3rd Party Limited,” which promotes Green votes in constituencies where only Labour can win, and supports the SNP against Jo Swinson.

All this creates a fierce battleground in the bellwether seats. The new Datapraxis report analyses their specific dynamics. By working with 475,323 polling responses from our partners at YouGov, we can typically tap into 700-800 polling responses in any given constituency—an unprecedentedly large data set, which allows us to distil much firmer signals about how different types of local voters are moving during the campaign.

The new bellwethers to watch include seats in the old Labour heartlands such as Workington, Don Valley (despite the “Get Brexit Done” sympathies sometimes shown by local MP Caroline Flint) as well as Wrexham and Vale of Clwyd in Wales, all of which the Tories seem strongly placed to take. The fight has a different dynamic in Peter Mandelson’s old seat of Hartlepool, where a strong campaign by Brexit Party chair Richard Tice is throwing everything at a shock victory, and has inched ahead of the Tories. Labour could hold on, though, unless a lot more Brexiteers lend Tice a tactical vote.

At the start of the campaign, progressives held high hopes for snatching back the bulk of the 13 seats the Conservatives won in Scotland last time. But the scale of these gains is far from guaranteed. The Scottish National Party and the Tories are battling closely in many seats. We highlight Lanark, an SNP seat which we calculate the Tories have a good shot at taking. Also worth watching is East Dunbartonshire, where an SNP surge could yet do for Swinson.

South of the Watford Gap, the dynamics are very different—and more promising for progressives. We showed the vulnerability of seven Tory “big beast” Brexiteers in the first Datapraxis general election report, “Seven Seats That Could Change Britain.” All seven still look vulnerable in our latest data.

Dominic Raab could yet lose Esher and Walton to Monica Harding, his Liberal Democrat challenger. In Chingford and Woodford Green, Iain Duncan Smith is even closer to being defeated by the spirited campaigner and champion of the Green New Deal, Faiza Shaheen.

The Liberal Democrats could take Winchester off the incumbent Conservatives, but it is very tight indeed. In Putney, formerly held by the ex-Tory moderate cabinet minister Justine Greening, Labour is running the Conservatives close, helped by significant tactical voting and youth registration.

Finally, in Beaconsfield the newly-independent Dominic Grieve is lagging well behind the new Conservative candidate. But this former attorney general, and most forensic of Remainers, could yet benefit from a late surge of undecideds and tactical votes, making this one of the toughest of all seats to predict.

With both the main parties pushing to rewrite Britain’s social contract in different ways, the stakes could not be higher. But, as the race draws to a close, there is an asymmetry in the threat they pose.

Johnson’s Conservatives seem hell-bent on delivering a hard Brexit at the end of 2020, negotiating a Trump trade deal and executing sweeping changes to Britain’s fragile unwritten constitution. It may be that nobody will be able to stop them.

Meanwhile, Corbyn and John McDonnell want to move the UK decisively away from the Anglo-Saxon turbo-capitalist model. While some voters will be thrilled by that prospect, others might be terrified. But their fears should not weigh as heavily as they might have done in our “contest of anxieties,” because it is now crystal clear that there is zero chance of Labour winning a majority. The party’s best hope is a minority government or coalition, whose first priority would be to deliver a “final say” Brexit referendum; other parties could wield a veto on any domestic reforms. This fact alone may end up tipping some hesitant Conservative Remainers or natural Lib Dems into voting tactically against an untrammeled Johnson majority.

As we brief our campaign clients in real time on the dynamics of the closing days, all sorts of possibilities could still play out. The Conservatives could yet be held to 310 seats, which would render them unable to govern and see their hard-Brexit platform crumble. Alternatively, Johnson could surge to 100-plus majority if Labour’s red wall completely caves.

On Tuesday night, our partners at YouGov will release their own “final call” prediction, based on their own proprietary model, which crunches the same data set but applies different assumptions to it. It will be fascinating, but still won’t be the final word. Some voters won’t make up their minds until election day—perhaps in the polling booth.

No data model can be completely confident in its predictions—and this is as it should be. No digital mastermind in any Cambridge Analytica-style operation or the Downing Street bunker can completely override the unpredictability of human agency.

I believe in the people of the United Kingdom. Despite a chilly and frequently grim election season, I look forward to them making their free decisions on Thursday. There are many great and spirited campaigns of committed citizens that deserve support. And there is still just enough time to counter the blitz of disinformation and put swing voters in the know about how to hold a capricious government in check. This is a moment for optimism of the will.