Insight

Time for some predictions

Time for some predictions

Our final results are here and we’re forecasting close to a dead heat with the Conservatives leading on 35% to 34% for Labour.

How this translates to seats is almost anyone’s guess with the rise of the minor parties and the SNP’s surge in Scotland meaning that the relationship between votes cast and seats won may be looser than any election for decades.

With this in mind, here is what the Opinium team think will happen after the polls close in a few hours’ time:

Adam Drummond?s prediction

I’m going to repeat the form of my prediction for the Scottish referendum in saying “I think our poll is correct but if we’re wrong then it’ll be this way” so feigning confidence but getting my excuses in early. I expect a small Conservative lead in vote share and a very small lead in seats leading Labour by fewer than 10. This is based on reports of Labour having a stronger ground game with more activists knocking on doors in marginal seats and the the fact that, although the likely loss of their Scottish seats has mitigated the pro-Labour bias in the electoral system, Labour held constituencies tend to have smaller populations than Conservative held ones meaning that Labour votes are distributed more efficiently across the country and thus Labour can get more seats per vote than the Conservatives.

However given the margin of error technically we’d be correct with a tiny Labour lead as well so if things are very different to our poll then I’d expect it to be in the form of a larger Tory lead of around 3 points. The reasons for this are that 2010 was the first election for decades in which final polls did not overestimate Labour’s performance and that was because they overstated the performance of the other centre-left party, the Liberal Democrats. The groups that tend to vote more reliably are those that tend to lean to the right and the Conservatives have consistently done better than polls expected in this parliament, most notably in the European elections last year. Our method tries to take this into account but is not infallible so if the polls are all wrong then I expect that’s how they will be.

James Crouch?s prediction

Predicting this election for me has always depended on whether voters will act in a predictable manner, or truly break the mould. Last year UKIP became the first party other than the Tories or Labour to come top in a UK-wide election for over a century. Then there was a Green surge. Since September there’s been coverage of a SNP surge. In short, everything says this election will not be typical, but I’m still not convinced that voters will act totally out of character.

For the last 50 years, no governing party has gone into an election and lost when they have remained the largest party in local government. The Tories still have hundreds of councillors more than Labour. Despite their advantages narrowing, Cameron’s lead on who would be best PM and the Conservative’s lead on the economy are comparable to 1992. As are the polls.

I know a lot of things have changed since then, but I’d be prepared to stick my neck out for a 3-point Conservative lead in the share of the vote (plus or minus 1). I also think the swing in the Conservative seats with an array of first-time incumbents will be slightly lower than the average, so my prediction is that the Conservatives and Lib Dems should have enough to form a government. Just.