The Hardest Election to Poll
by Andrew Hawkins, Chairman

The fourth national election in five years smashes several records, not least the first December poll in 96 years, called in the wake of the shortest parliamentary session since 1948 and the most rebellious House of Commons anyone can remember.

Several major factors combine to make this the hardest election to predict that as a pollster I can ever remember. The only certainty is uncertainty.

The ideal ingredient for modelling voting intention is to keep behavioural assumptions constant from one election to the next. Yet there is little that remains intact from 2017, when the Tories and Labour combined won 84.5% of British votes cast - the highest since 1970. They together now poll around 60%.  This means that one in three voters for one of the two main parties has since defected. Whether they return to the fold, stay away altogether or stick with their new choice is anyone’s guess.

Pundits used to rely on a simple formula for calculating how many seats could be imputed to a party by their vote share lead. In 2010, for example, the Conservatives needed an eight-point lead over Labour to win a majority of seats in the House of Commons. Such a simple calculation is now impossible because of the number of places where a voter’s support could end up if they switch vote.

Then there is the uncertainty of a Christmas ballot. Comparisons with 1923 are limited: despite the lack of Gore-Tex clothing, turnout then was an impressive 71.1%. The concern is not that weather puts voters off but that people will be unmotivated to vote when the polling station is competing with the office Christmas party. Also, when we last had a December election Christmas was not a sprawling materialistic free-for-all lasting two months. It was a short Christian feast day with most decorations not even going up until the 24th. So forget comparisons - this is unchartered territory.

What will be driving choice this time round? Not brand loyalty. By a ratio of two to one, voters are more likely now to say they define themselves in terms of Remain or Leave than by allegiance to a political party.

Yet it is a common mistake by pundits to assume that the public are as binary as the Westminster Village over Brexit. There are doubtless vast numbers of diehards on both sides, but there is a spectrum of attachment which has tended to lean towards Leave. This is because most voters, including a third of Remainers, believe the result should be adhered to, irrespective of how they voted in 2016. Leave voters themselves also tend to be more strongly motivated than Remainers to vote in line with their Brexit conviction.

As happened with Theresa May, non-EU issues may yet derail Boris Johnson, especially if the same miscalculation is made about the value of the leader as a talisman for the country’s future success. After all, leadership did not deliver for David Cameron in 2010 or Theresa May in 2017.

Then there is the sheer promiscuity of the electorate. Fewer than half the country now claims generally to vote for the same party in every General Election and around a third have switched parties in the last two contests. The parties whose support is especially soft are the Liberal Democrats and Brexit Party; they are also the parties for whose voters the Labour and Conservative Party respectively provide the most popular second choice.

This election has been described as 650 individual contests, each with different factors determining their outcome. But most of those battles are irrelevant to an outcome which will be decided by a far smaller number of marginal seats changing hands. The results in most constituencies are a foregone conclusion. Who forms the next Government, and the UK’s future relationship with the EU, will boil down to how motivated voters are in a small number of marginals. In many of those constituencies Leave voters will be as unmoved by predictions of economic Brexit hurt as Remain voters are by arguments of national sovereignty, so perhaps more traditional policy battles will tip the balance.

Either way, beware of pundits trying to forecast the unforecastable. Only someone who’s been on the eggnog would bet the mortgage on the outcome of this one.

This article was first published by The Telegraph