By Stephen Fisher, John Kenny, Paul Furey, and Polina Ryzhuk. 3^rd July 2024.

10am, 4th July. Updated to correct a minor typo. Forecasts published since the post below was published late last night but before polling stations opened this morning (and any we missed) will be incorporated in the evaluation after the results. So far it looks like those additions would make little difference to the tables below. Thanks to all the forecasters for their contributions.

The average of different kinds of seats forecasts points to a Labour majority of 194 in tomorrow’s general election. Changes since last week are: Conservative +8, Labour -3, LD +4, Reform +1, SNP -2, and no change for the Greens and PC. Those changes are partly due to moves in the betting markets, perhaps in response to polls that improved for the Conservatives, and new complex models (listed in the sources section below). However, MRP projections have worsened for the Conservatives. They are now forecasting an average of just 85 seats for the party.

By Stephen Fisher, John Kenny, Paul Furey, and Polina Ryzhuk. 26^th June 2024.

The average of different kinds of seats forecasts points to a Labour majority of 200; the same as last week. There is very little change for other parties, except that the average forecast for the Conservatives is down from 127 last week to 123 this week. That is despite a rise for the Tories in the simple-model projections (uniform change from opinion polls). The Conservative seat forecast is down in both the non-MRP complex models and the MRP average, which is now forecasting just 92 seats for the Conservatives. That average is now based on 9 different MRP models.

By Stephen Fisher, John Kenny, Paul Furey, and Polina Ryzhuk. 19^th June 2024.

All the different sources still point to a large Labour majority. On average across the different kinds of forecast, the predicted majority is, at 200, up eight from last week. The combined forecast for the Conservatives is down by nine seats, and that for the Liberal Democrats is up by seven seats since last week. Those changes have been largely driven by changes in MRP projections, but other sources have also typically changed in the same direction.

By Stephen Fisher, John Kenny, Paul Furey, and Polina Ryzhuk. 12^th June 2024.

All the different sources of seat forecasts still point to a large Labour majority. On average across the different kinds of forecast, the predicted majority is 192, down slightly from 197 last week. The Liberal Democrat forecast has crept up from 36 in our first forecast, to 37 last week, to 41 today. The betting markets are more optimistic for Reform UK winning seats than the poll-based models, as they were last week.

By Stephen Fisher, John Kenny, Paul Furey, and Polina Ryzhuk. 6^th June 2024.

All the different sources of seat forecasts still point to a large Labour majority. On average across the different kinds of forecast, the predicted majority is 197. That is up slightly from the 188 average majority in our first combined forecast this time last week.  

Table 1. Seat Forecasts

The increase is largely due to the higher Labour forecasts from the MRP models, which are listed in Table 2. They are remarkably different from each other, ranging from a Labour majority of 114 to as much as 324. The forecast number of Conservative seats in the MRP projections range from 66 to 180. A key part of the differences between the MRP projections for the Conservatives and Labour is the degree to which the models expect party performance to be proportional to the prior Conservative vote share. By comparison, traditional uniform change models (under simple models in Table 1) produce more modest forecasts for Labour gains and Conservative losses.

By Stephen Fisher, John Kenny, Paul Furey, and Polina Ryzhuk. 30^th May 2024.

Here we combine seats and vote share forecasts from a variety of sources including betting markets, polls, statistical forecasting models, and citizen forecasts. As well as updating weekly, incorporating new forecasts with different methods as they become available, there may be developments in the methodology. Comments welcome. 

All the different sources of seat forecasts in Table 1 point to a very large Labour majority, well over 100.