We are now into the home straight of the election race. For the circa 20% of the population who votes by post, decision day is effectively here with replies due in the post any day now. And for those who plan to vote in person, voters now have just 10 days to make up their minds as to where to mark their cross.
The MRP approach
Over the past week, the not so sassy sounding multi-level regression and post-stratification (MRP) election model stole much of the limelight. The poll had been eagerly awaited after its accurate prediction of a hung parliament in the 2017 race, after other major polling had failed to predict this outcome.
The MRP approach works by using a large sample to model voting intentions using factors such as socio-economic characteristics and pro ‘leave’ and ‘remain’ attitudes, allowing the pollster (YouGov) to predict individual seat results as well as voting shares. The model called 93% of seats correctly in 2017.
The results published last week, showed the Conservatives on course for a clear majority. Indeed, the YouGov analysis suggested that the Tories would win 359 seats (42 more than in 2017) and 43% of the vote (about the same as in 2017). This would be the Conservatives’ best performance in terms of seats won since 1987. Meanwhile, the model predicted that Labour would lose 51 seats – falling from 262 seats in 2017 to 211 now, taking 32% of the vote (a nine percentage point decrease). This would be the party’s worst performance (by seats won) since 1983.
The result of the MRP, which gives a prediction based on if “the election were held tomorrow” would likely have spread both optimism and nerves across the Tory ranks. Indeed, the Conservative party is likely wary that anything that makes them look too comfortably on track for a majority, could leave them vulnerable to Conservative voters not turning up to vote blue on the day.
These nerves will no doubt have crept higher over the week, as the six most recent polls suggested Labour were clawing back some ground against the Tories. Indeed, of the past six polls all showed Labour picking up between two and five percentage points whilst five of those showed the Conservatives losing one or two percentage points. According to the Financial Times’ poll of polls, the average lead is still just in double digits at 10 percentage points and enough, based on our Uniform National Swing election predictor, to expect a Tory majority. But, given the direction of travel in polling, a hung parliament is still an option not to be discounted.
Coming up in the next week
Over the coming week there are several factors with the potential to swing the election one way or another. One question is how the government and opposition parties’ responses to Friday’s terrorist attack near London Bridge carry through into the election.
The Conservatives have promised tougher prison sentences whilst Home Secretary Priti Patel sought to shift the blame to Labour, blaming legislation brought in by Labour in 2008. Security will be big theme for the week more generally as the UK plays host to the NATO Summit. PM Boris Johnson is set to meet with US President Trump as he attends the Summit, with the Conservatives no doubt praying Mr Trump does not say or tweet something that creates a big headache for them, in the wake of the meeting.
At the end of the week, the BBC will host the head-to-head to debate between PM Johnson and Labour leader Jeremy Corbyn. The TV debates so far have not shifted the dial much. However with the debate taking place less than a week before the 12 December vote and with this a direct clash of the two leaders, there is a serious chance this debate could be the one to spark some fireworks ahead of polling day.