The BBC knows it’s doing something right when both Brexiters and Remainers shout “bias” during the news. The stakes are similar when producing a tactical voting tool in a general election. Hopes and fears are running high among the different tribes and, for them, any suggestion that another party might hold the key, anywhere, to a better outcome is sacrilege.
As a consequence, my team and I were unsurprised that parts of the Labour twitterati went into paranoid overdrive last week, when Best for Britain dared to suggest voting Liberal Democrat in a number of seats where Labour were first or second in 2017. This was despite the site overall recommending 375 Labour candidates to just 180 Lib Dems.
We were equally unsurprised to find Lib Dems suggesting that some of our analysis was wrong, and that they were able to win more than the 180 seats where we identified them as the strongest Remain force.
Some of the toughest seats to call are the Labour-Liberal Democrat marginals such as Leeds North West and Bermondsey. But these areas are the exception, not the rule: voters can choose Labour or the Liberal Democrats and know they won’t help a Brexiter Conservative win. We have tried, wherever possible, to back the sitting MP with a strong voting record on Remain. One place we haven’t done this, Kensington, is because Labour is so far behind the Lib Dems that backing them could give the seat back to the Tories. In most constituencies our objective is to identify the strongest opponent to the Conservatives.
While Labour has been quick to criticise, the model is more than fair to it. It has assumed Labour to be a Remain force on the same basis as the Liberal Democrats. Given that the latter is a full-throated pro-European party, and the leadership of the former has taken three years to get around to giving the public the final say, there are plenty of Remainers who accuse us of being too generous. Yet our mission is not to help any party. It is to stop Boris Johnson, stop the Brexit party, and – in turn – stop Brexit.
We are not relying on ancient data from 2017. Since then, politics has been turned upside down. Both the Conservatives and Liberal Democrats have new leaders; Labour has changed its Brexit policy; the two main parties sank to third place and fifth place at the last national election in May; the Brexit party has emerged. The political landscape has been repainted with Brexit in the foreground and previous party loyalties distant dots on the horizon.
To achieve accurate, up-to-the-minute recommendations, we have used a technique popular among pollsters called “multilevel regression and poststratification” – MRP for short. This method was deployed to accurately predict the election of Donald Trump, the 2017 hung parliament, and the “surprise” wins for Labour in Kensington and Canterbury that year. The MRP we have used was undertaken in the last couple of weeks, with a statistically enormous sample size of 46,000 voters. A normal national poll contacts less than 2,000 people.
With such a large sample, we are able to drill down into individual constituencies, many of which have changed considerably in their political weather since two years ago. Whereas the Conservatives and Labour polled 82.4% combined in 2017, the evidence is that they will struggle to get more than 66% between them this time. There are London seats that the Conservatives won in 2017, and where the Labour party came second. But our polling shows clearly that in such places as the Cities of London and Westminster, Chelsea and Fulham, and Wimbledon, the Liberal Democrats are now the party best placed to unite an anti-Conservative vote with a clear Remain message. That analysis may be inconvenient to Labour but it is the product of cold, hard, recent data. Labour remains the clear challenger in the vast majority of seats. We will update the MRP closer to polling day, to make sure we give the most accurate recommendations possible.
The party machines will inevitably rail against the parts of these conclusions they least like, while embracing parts convenient to them. Fortunately for the future of the country, the electorate is far more savvy than the political parties they vote for. This Christmas election is our last chance. Banish Johnson and Brexit from our politics for 2020 with tactical voting, or put up with both for another decade without it.
Naomi Smith is chief executive of Best for Britain.
