4 March 2015updated 05 Oct 2023 8:31am

‘Experts’ don’t think SNP will fade, but inexplicably predict they will only win 29 seats

More than 500 pundits, mostly academics, have offered their election predictions. Their forecasts don’t quite add up.

(Photo By Pundits seem to be underestimating the SNP surge. Photo: Getty.)

Read this piece – and keep up to date with all the latest predictions – on our election site May2015.com.

Who will win the election? More specifically, how many votes and seats will each party win? This is the question we try to tackle here on May2015. We’re tracking all the latest polls and turning them into an election prediction each day.

You can make your own prediction. Or keep tabs on the other models being run by academics, the bookies and other papers.

Now the Political Studies Association, the UK’s association of political academics, has added to this picture. They have gathered the predictions of more than 500 pundits – nearly 90 per cent of whom were academics – and averaged together their opinions.

By bundling together many estimates, the “errors of individual guesses [are] cancelled out”, the theory runs (the ‘average’ pundit will guess the weight of the ox perfectly).

Labour or Tories?

Their findings confirm current forecasts: Labour are predicted to win 282 seats (24 more than in 2010), versus 278 for the Tories (29 fewer). That’s because of the weight of academic opinion. In contrast, albeit not a statistically significant contrast, the 45 journalists asked had the parties the other way around: 286 Tory, 282 Labour.

Our Friday round-up of the five forecasts tracked on May2015 put the Tories on 278 and Labour on 275.

Lib Dems

The Lib Dems are predicted to win only 24 seats by academics. Pollsters (28) and academics (29) are slightly more favourable to them, but no group thinks the party will hold onto more than 30 seats, as we recently suggested they might.

No group thinks the Lib Dem will hold onto more than 30 seats.

This doesn’t make complete sense. The PSA report also asked people to predict vote share, and the Lib Dems were forecast to poll 10 per cent in May. They are currently on around 7 per cent. A sub-30 seat prediction makes some sense if that’s where they remain (but what really matters is how the Lib Dems are doing in the 40 or so seats where they are actually competing), but it makes less sense if these experts think the Lib Dems will recover to 10 per cent.

If you give the Lib Dems 10 per cent on May2015’s seat calculator, the coalition’s minor party wins 32 seats.

SNP

But this is small fry compared to predictions for the SNP. The PSA’s experts think the party will win 3.8 to 4.7 per cent of the GB-wide vote in May. That would translate into a high-40s/mid-50s vote share in Scotland. If the SNP win that many votes, they will almost certainly win at least 40 seats, and maybe more than 50.

No one thinks the SNP surge will fade, which would mean 40 or 50 – not 29 – seats.

Yet the PSA’s experts predict the party will win just 29. One forecaster recently summarised the way people predict the SNP as “That number looks high. Let’s half it.”

That seems to be what’s happening here. Academics and journalists don’t think the SNP surge will fade before May. But they still can’t believe the party will go from winning 6 seats in 2010 to 40 or 50 in 2015. Understandably – nothing like that has ever happened before in the history of British elections.

Other parties

As for Ukip and the Greens, people seem to think the latter will win a second seat in May. We don’t see how that’s going to happen; they are distant seconds in their target seats.

The former are forecast to fall further in the polls before May. They’re currently on 14 per cent and predicted to dip to 11 per cent. That would be in line with the trend: they were on 16 per cent in November.

The Greens seem unlikely to win a second seat.

But if that happens you would expect the Tories to do better than the PSA’s experts are forecasting they will. Roughly half of Ukippers voted for the Tories in 2010: their defections are the biggest reason the Tories have lost votes since then. (The Tories have lost next to no voters to Labour.)

So if Ukip fade, you’d expect the Tories to notch up. But the PSAers don’t think the Tories will escape the 32 per cent mark they’ve been stuck on for three years (although that may finally be changing this week). Maybe because they think the Lib Dems are going to recover slightly.

So who wins?

Finally, the report – a great read, put together by Chris Hanretty of Election Forecast and Will Jennings of Polling Observatory, whose election model will shortly be launching on May2015 – looks at the probability of various election outcomes.

It shows that 7 in 10 journalists think the Tories will win more votes than Labour, while just 47 per cent of academics agree. Journalists and pollsters can’t decide who will win the most seats, but academics think it’s twice as likely to be Labour (the bookies think it will be the Tories).

As for that majority? If you hadn’t heard, no one thinks that’s going to happen. We’re 64 days away from our next hung parliament. Then what? Here’s what happens next, and here’s we already know about the 24 hours after the election.