By Stephen Fisher, John Kenny and Rosalind Shorrocks.
Our combined forecast for a Conservative majority has dropped yet again this week. It now stands at 71.
In line with the changing forecast seats the changing probabilities of a Conservative landslide and even a Conservative majority have been declining
Looking in more detail at the seats forecasts, today saw the publication of the PSA Expert Survey. Sadly the forecasts the experts provided are potentially already too old given that most response were from 16th May, before the launch of the Conservative manifesto and from when forecasters were suggesting 400+ seats for the Tories. Probably the experts would change their forecasts if they were asked again today. Similarly the volunteered forecasts from the Times Red Box sweepstake were last updated on the 17th May so they have been dropped. The overall average is therefore calculated without the volunteered or expert forecasts.
The betting markets and both kinds of models all have lower predictions for the Conservative majority. The average across the complex models has come down most because of the introduction of YouGov’s MRP estimates that point to a hung parliament. The complex models forecast Conservative seat tallies vary from 317 to 396, suggesting considerable uncertainty.
|Seats||Betting Markets||Complex models||Simple models||Experts||Average
The combined probability of a Conservative majority is now 74%, showing another small change from 87% last week. The probability of a 100+ majority Conservative landslide is now just 11%. Again we have excluded the PSA Expert Survey average forecast of an 65% chance of a 100+ Conservative majority. Conversely, no polls in the last week have shown the Conservatives with the lead of 16 points or more that they would need to get a landslide.
However, more polls this week are now suggesting the outcome might be a hung parliament, or at least that Theresa May’s majority is on a knife edge. Both the betting markets and the forecasting models that publish probabilities are somewhat more sceptical about the chances of a hung parliament. Overall our combined forecast gives it about a 1 in 4 chance.
|Betting markets||Models||Polls||Citizen forecast||Average|
The forecast vote shares help explain why the betting markets and models give higher probabilities for a hung parliament than our pseudo-probability from the polls. The markets and models think that Labour’s eventual performance is being considerably overestimated in recent polls. So too do they think that the Tories will do less well than polls suggest, but not by much. Overall our combined forecast this week (again without the somewhat dated expert survey) now suggests the Conservatives will win with a 10.2 point lead over Labour, down 4 points since last week.
|Poll aggregates||Betting markets||Models||Experts||Average
The main changes in method this week are as follows. Since the polls have been showing dramatic movements and we are within a week of the election we are restricting our pseudo-probabilities from the polls to the latest poll from each pollster conducted within the last week.
Nigel Marriott now only publishes one model rather than 4; we have re-included polling averages and seat forecasts from Principalfish and Adrian Kavanagh because they now appear more in-date; and we have added in YouGov’s model shares and seats even though they say it is not a forecast.
We have two citizen surveys, conducted since the middle of May so we exclude older ones. The citizen forecast for a Conservative landslide just includes one poll from YouGov, conducted 30th-31st May.
The Political Studies Association (PSA) has published an Expert Survey of the election, which forecasts seats, shares, and probability of a Conservative landslide (by taking the percentage of respondents who said they thought the Conservatives would get a majority of 100+ seats). Most of the fieldwork for this was conducted in early- to mid-May, prior to the Conservative manifesto launch, making it potentially out of date in the context of the Conservatives’ tightening poll lead. As a result, this source is not included in the combined average even though it is presented for comparison.
There now appears to be no usable betting market for UKIP vote share. To get a point estimate for vote share there needs to be betting options both above and below some more popular middle option(s).
Our basic approach is to combine forecasts by averaging them within each category and then taking the average across categories. Since the different sources do not all present equally clear figures that can be averaged on a like for like basis we have made various judgement calls on how to treat the data.
Historically the idea of combining forecasts from different sources has had a good track record, though it has to be admitted that our attempt to do one for the EU referendum did not work out well. Most recently the pollyvote.com combined forecast of the US presidential election last year was 2 points out on the share of the vote.
For vote shares, we use the various available polling averages, or ‘polls of polls’, and take their average. We exclude polling averages for whom the most recently published polling average is more than a week old. There are seven different polling averages. They are in truth nowcasts rather than forecasts, but we are in effect treating them as forecasts. There are seven different polling averages. Some admittedly are quite sophisticated, allowing for pollster (aka house) effects, but they are nonetheless estimates of current public opinion and not future votes.
We do not attempt to say what seats outcome is implied by polls (that is the job of the modellers). However, since statistical models are rarely if ever clear about the probabilities their models place on key events like a Conservative majority and a 100+ majority, we have included in the probabilities table some pseudo-probabilities from the polls. Taking the most recent poll published by each pollster in the last week, we calculate the proportion showing a Conservative lead over Labour of more than 6 points as the pseudo-probability of a Conservative majority, but we allow polls with exactly a 6 point lead to contribute a 0.5 to the average. Using the same polls, we use the proportion showing a Conservative lead over Labour of 16 points or more as the pseudo-probability of a Conservative landslide. These thresholds of 6 and 16 points are based on what would be required under uniform swing assumptions for the Conservatives to win a bare majority and a 100+ majority respectively.
There are numerous betting markets for the various outcomes in the election. We have taken those that are most helpful for the four forecasts we want to produce. For seat shares, we take the mid-point of the spread as the seat share, and average these mid-points between different sources. Note that the markets imply fewer seats forecast than there are actually are in the House of Commons. This is because the markets are separate for each party and do not need to be consistent collectively.
For vote share, we use betting markets for the Conservatives, Labour and UKIP. (Vote share markets for other parties are unavailable on the betting market aggregation site Oddschecker.) Odds are given for 5-point ranges of vote share. We take a weighted sum of the mid-points of these ranges where the weights are the implied probabilities. For the top and bottom options we use 2.5% above/below the upper/lower bound (e.g. 52.5 for “Above 50” and 27.5 for “Below 30”). The weighted sum is calculated just using the three categories with the largest implied probabilities, because the probabilities for other categories are so small and unstable. For UKIP we use just the two most likely categories.
For the probability of a Conservative majority we give an average of the implied probability from sites offering this market. For the probability of a Conservative landslide, we use the combined prices PredictIt that the Conservatives will win 370-379 seats, 380-389 seats, and 390 or more seats. This really represents a majority of 90 or more but that was as close as we could get to 100.
There are numerous statistical forecasting models this year (and more to come). We have divided them into two categories: simple (poll average plus uniform swing seats projection) and complex (anything more elaborate than the simple models, although they are not necessarily particularly complex). Some make adjustments for long run differences between pollsters, for constituency variation, and some estimate by how much things will differ between current polls and the eventual result. Chris Hanretty’s forecast at electionforecast.co.uk does all of these things. Within these categories we simply average the available estimates of seats and shares.
We should note that not all of the models are the modellers’ favourite. Some are counterpoints to their main models for comparison. We have included these on the basis that they are still talked of and expected to be reasonable estimates, for example, at Electoral Calculus, Martin Baxter has a local election results based model. We have not excluded any models based on our judgement of quality, but they do have to be statistical models as opposed to personal guesses.
Some polls ask people what they think that the outcome will be on June 8th. Different pollsters use different survey questions but they can be combined to generate pseudo-probabilities. We use the proportion of poll respondents who think the Conservatives will win/there will be a Conservative majority, excluding don’t knows and re-percentaging, as the pseudo-probability of a Conservative majority. We similarly use the proportion of poll respondents who think that there will be a Conservative landslide/the Conservatives will win more than 100 seats, for the probability of a Conservative landslide. Due to the limited number of polls these questions are asked in, we take results from the last two months.
Note: Estimates come from around lunchtime on 2nd June 2017. In all seat estimates, the Speaker’s seat is counted as Conservative. Since preparing the figures we have seen that Lord Ashcroft’s forecasts for the Conservatives have dropped substantially. These are not reflected in our figures but as just one model of several it would not change the average dramatically.
There sources we used are listed below in no particular order. Please let us know of any that you think we have missed or misclassified. Some polling averages we know of were not included because they were more than a week old.
Complex Forecasting Models:
ElectionForecast.co.uk (Chris Hanretty)
Electoral Calculus (main and local election forecast)
PME Politics (Patrick English)
Chris Prosser (GE vote shares from Local elections vote shares)
The three authors are equal contributors and our names are in alphabetical order.