Not a prediction
This post isn’t for you, it’s for me.
We’re into the home straight of the general election. Time to set out what I expect is going to happen. Why am I doing this? In short, to learn. I’m not expecting to be right but I can learn from how I turn out to be wrong.
So, here goes.
The polling, and its flaws
We have had a nonstop waterfall of polls. They now show in broad terms a simple picture of a large Labour lead. The differences between them are, however, substantial. One shows Reform well clear in second place. That same poll has the Conservatives on 15% while one of the MRPs has the Conservatives on 28%. This is an enormous span.
Between those two extremes, many of the recent polls have settled into a pattern close to the Sky News polling tracker, which at the time of writing is:
Labour 40.7%
Conservatives 20.2%
Reform UK 16.6%
Lib Dems 10.8%
Greens 6.2%
Is there any reason not to work off this simple average?
In short, my answer is yes. For at least a year Reform have run far ahead in the polling of their actual results. As I have previously written, I believe that the polls are being skewed by the way in which they capture the unusually politically engaged.
One poll has tried to counter that, the recent Ipsos MRP. It’s worth quoting its description of its methodology in full:
“Panellists are recruited using random probability unclustered address-based sampling, the gold-standard in UK survey research, meaning that every household in the UK has a known chance of being selected to join the panel. Crucially, members of the public who are digitally excluded are given a tablet and provided with an email address. Ipsos interviewed 19,689 adults aged 18+, residing in Great Britain. Data collection took place between 7–12 June 2024, using the standard Ipsos voting intention and likelihood to vote question wording. This will be the first UK election when any voting intention data collected via a probability panel has been published, and this MRP is the first model of this type using probability data.”
Of course, it’s just one poll and like any other poll it could be skewed by the sample or by other hidden biases. Nevertheless, I place particular weight on it because of its attempts to overcome the inherent problems of conventional polling. Its implied poll shares are:
Labour 43%
Conservatives 25%
Reform 12%
Lib Dems 10%
Greens 6%
Both Labour and the Conservatives are higher than the average, and Reform lower. That accords with my sense of how the polls have been flawed in the past year, and given its unique methodology designed to address that flaw, I’m placing a lot of credence in it.
My expectation
Now it does seem that in the period since that fieldwork the Conservatives in particular and Labour to some extent have fallen further in the conventional polls. So I in fact expect something like:
Labour 41%
Conservatives 22%
Reform 15%
Lib Dems 12%
Greens 6%
Note, in my expectation Reform are some way adrift of the Conservatives in third, which is quite different from many of the polls we’re seeing. (If that eventuates, expect to hear endless discussion about the Putin story having broken Reform. I’m not at all sure that would be correct.)
Seat counts
What would this translate into seats, which is after all the important bit? On its higher predicted implied vote share, Ipsos expected the Conservatives to get between 99 and 123 seats, with an implied probability of 115. On a 3% lower vote share, we would expect to see them get considerably fewer. Even as its vote share dropped, we would expect Labour to be prime beneficiaries.
Scotland is really hard to predict. I expect the SNP to benefit a bit more from incumbency than the polls are perhaps picking up.
I guess we’d see something like:
Labour 475
Conservatives 75
Lib Dems 50
SNP 20
Reform 5
Plaid Cymru 4
Greens 2
Speaker 1
Northern Ireland 18
Sense check
Does this accord roughly with what the parties are doing? In short, yes. Rishi Sunak is campaigning in seats, including his own, which would not be safe on such projections. We see MPs like Danny Kruger posting leaflets in the pouring rain in their own constituencies that would be safe if the Conservatives were doing even a little better. We see MPs in Norfolk, Suffolk and Essex fretting like never before, campaigning on a niche but incendiary subject of pylons.
So, anyway, I’ve set out my expectation. It’s not a prediction: I will be wrong. But if I don’t set it out in detail in writing in public, I’ll backfill my own recollection after the event to pretend to myself that I was less wrong than I was. I encourage you to do the same.