Forecasting any election is a challenge. There’s always a risk that analysts see what they expect to see and discount data points that contradict those expectations.
We may be seeing the beginnings of that in the run up to next week’s UK elections. If it does happen, it won’t be the first time.
During the BBC Election Night broadcast of 1992, Malcolm Rifkind, then-Secretary of State for Transport and contender for most important Scottish Conservative, was interviewed. The Scottish Tories had been predicted for a decimation, going from 10 seats in 1987 to a projection of 2 or 3. The Tories had gained one seat at that point in the night, and held their most marginal seat, but the BBC forecast still had them on 4 seats. When asked about that forecast, Rifkind unloaded; “how can you say we’re doing badly in Scotland when we’ve just won back South Aberdeen from the Labour Party and when we’ve held on to Sterling which all the experts confidently claimed we would lose … all your predictions have been turning out to be wrong, therefore I think you were very unwise to make that prediction.”
Rifkind would turn out to be correct – the Tories would walk out of the election with 11 Scottish members, and it was indeed a very unwise prediction. The problem was that the forecasters in the BBC’s data office, and the pundits, couldn’t fathom that, after the Poll Tax fiasco and the general air of discontent with the Tories in Scotland, the Scottish Tories could make a net gain of seats there. The narrative got away from the math, because the narrative made more sense, but the idea that Michael Forsyth would get an increase in his majority but half of the Scottish Tory caucus would lose should have been seen for what it was – and what it ended up being. It was laughable, and worthy of the contempt that Rifkind showed it.
This brings us neatly to the current election, where IPSOS-Mori released some polling this week (although, no top-line vote intention numbers). One of their questions was expected result of the election, where by 44% to 38% voters expect this election to end with one side ending up with a majority as opposed to a hung Parliament, and only 34% believing the Conservatives will get the majority they want. The narrative of this election – at least, as seen by the expectations voters have of the likely outcome – is completely divorced from the reality of the polls. The LeanTossup model currently projects a Tory majority of 96 seats – 373 total Tory seats – and a Tory majority 80.8% of the time. Now, even if one were to accept those numbers were high – for which the only basis would be wishcasting from denialist Remainiacs – the data is a complete and utter destruction of the narrative that this election is close. Nothing in the current data says a Hung Parliament is anything but an unlikely, but still possible outcome (if not particularly plausible). A moderate polling miss – one on the order of the UK’s 2015 miss would not be enough to cause such an event, if the miss went in Labour’s favour – which it didn’t in 2015.
Does it mean this election is over? No, but another outcome will require massive reversal over the remaining seven days of the campaign to a trend that’s been growing since July’s change of Conservative leader.
Labour’s problem is that their old coalition was split between socially liberal Remainers, mostly in cities, the South of England, and Cardiff, and socially conservative leavers in the rest of England and the rest of Wales. Last time, the trick they pulled off was stitching together both ends of the coalition to the point they would win enough seats to stop the Tories winning a majority. However, last time they were a party committed to leaving the European Union and ending Free Movement of People into the UK, so they could credibly say to Leave voters that Brexit was safe in their hands. They are now a pro-Second Referendum Party, with many in the party arguing they are a Remain party, getting ~14% of the Leave vote on average instead of the ~30% they got last time, and suffering swings against in the North, the Midlands, and Wales – making their bevy of former heartland Labour seats now varying shades of Tory blue. They may make advances into the very posh portions of London, including stunning wins into places like Cities Of London and Westminster, Finchley and Golder’s Green, and Wimbledon, while repeating Blair-era advances into Putney and Enfield Southgate, but those advances won’t be nearly enough. For every Remain Tory seat in London Labour will probably win at this point, the model points to a half dozen Labour Leave seats the Tories should win. Two seats of former Labour leaders – Sedgefield, Tony Blair’s constituency for 24 years, and Ed Miliband’s current seat of Doncaster North – are both vulnerable, and the old seat of Blair’s confidant and Cabinet Minister Peter Mandelson in Hartlepool has a sizeable Tory lead. If Labour is going to win this election, they need to return Hartlepool to their column, and they’re frankly running out of time.
For the Tories, many people on the Remain and/or liberal axis of politics cannot fathom their support levels. To those kinds of people, Boris Johnson is a Trumpian figure – an unserious grifter whose unseriousness should be so obvious it should be immediately apparent he cannot win. While to an extent the comparisons may be dubious, there are enough similarities between the two men for the comparison to be worth making. Both are men who come from family wealth and power who attended the best schools available, who flailed around the world of private business and then entered the world of politics. Both are controversial, loose speakers with stories of affairs littered in their pasts, and both took over political parties whose establishment was contemptuous of them. And both of them are breaking the political map by outperforming their more conventional predecessors with the new swing voter.
The media conception of who the swing voter in the UK is much the same as in the US – the fiscally conservative, socially liberal suburbanites who like tax cuts, “responsible” spending decisions, and who would feel perfectly comfortable living in any major US or European city. The media conception of this as the swing makes sense – these voters are in most metro areas, and if one spends enough time in well off circles, with professionals – lawyers, doctors, accountants, bankers, et all – it’s a viewpoint one would see quite frequently. Consequently, this is the swing voter the media thinks of – as New York, DC, or London media would frequently encounter that view in their work – either in their private lives after hours or in the seemingly endless amounts of charity dinners, functions, and galas that occur. The problem is, that isn’t the swing voter anymore, at least not the one that swings UK elections. Labour could do even better with that voter and still lose this election worse than any since the Second World War.
The new swing voter in the UK is the reason the Tories are on track for a large majority government. The new swing voter is the fiscally liberal, socially conservative voter who wants more money spent on northern towns and health care in regional areas and less money spent on “elites”, which routinely means whoever that voter isn’t a fan of. They’re wary of immigrants, mad at the Blair-era broken promise of only 13000/year net immigration from the 2004 EU Accession states – 250000/year would come in the decade after – and is annoyed by social issues that grip the modern left. This class of voter was staunchly Labour for decades, especially in the aftermath of Margaret Thatcher’s term in office. These voters were tempted by Theresa May last time, but went home to Labour because Corbyn did enough to reassure that Brexit would happen. With Labour policy now being a second referendum with an option to Remain – and every senior Labour politician outside Corbyn saying no possible deal is better than staying in – their likelihood of repeating their 2017 trick is somewhere between small and non-existant.
Put another way, here’s the 2019 UK Election in a nutshell – per the most recent YouGov polling, the Tories are at 72% with Leave voters and 16% of Remain voters, on average, with Labour on 48% with Remain voters and 14% with Leavers, and those broad findings are backed up by the findings of Survation, ComRes, and Opinium. The Tories are winning with their good group over Labour’s good group by 20-25%, and with their bad group over Labour’s bad group by about 5%. There may be an election where someone has won despite such a bad hand, but it’s not immediately apparent. Predictions of a close race – and anyone who wants to suggest a Hung Parliament is in the offing – should be met with extreme caution at this stage, and the wise words of Sir Malcolm Rifkind should be remembered; as of now, any such prediction would be very unwise, indeed.