Our ever-popular voter transition graphic shows visually the extent of movement between parties. It also includes 'non-voting' as an option, which is an important reservoir of voters.
The transition graphic below shows the results of our recent MRP polling, plus national vote share changes, up to August 2026. Each figure represents one per cent of those who voted in the 2024 general election, which is equivalent to about 280,000 actual voters. The colour of each figure indicates which party they voted for in 2024.
The migration flows, indicated by the arrows, show how voters have changed their minds since 2024, and indicate the likely transfer of votes at the next election. Parties which are losing voters, such as Labour and the Conservatives, have their ranks thinned by lost voters (coloured in grey) who supported them in 2024 but are switched away now. Parties which are gaining votes have their ranks augmented by figures bearing a plus sign (+) who are new supporters of that party since 2024.
Also important are voters who will likely not vote. These are indicated by the big grey figure in the middle. Some voters did not vote in 2024 but will likely vote next time, and they are represented by figures coming out of the non-voting area. Other voters did vote in 2024 but are unlikely to vote next time are shown going into the non-voting area.
The 'Other' group represents Restore Britain, the SNP, Plaid Cymru, Gazan independents and others
In 2024, Labour won 35pc of the vote, with the Conservatives on 24pc, Reform with 15pc, the Liberal Democrats on 13pc and the Greens on 7pc. This is shown in the sizes of the party blocks (including grey voters but ignoring 'plus' voters). The remaining 7pc of voters voted for the SNP, Plaid Cymru and minor parties.
In terms of voter migration, the main stories are that Reform is gaining voters (though losing one to Restore), and voters are leaving Labour in several directions. Seven out of the 35 Labour voters no longer support the party. Three have moved leftwards to the Greens, two have chosen not to vote, and one each has gone to the Liberal Democrats and Reform.
The Conservatives have also lost some ground. Of their 24 voters, five have switched to Reform, but they have gained two more voters who did not vote in 2024. and two have decided not to vote. That leaves them with only around 18 voters out of one hundred. The Liberal Democrats have lost net one voter (winning one from Labour but losing one each to the Greens and non-voting), and the Greens have gained five from various directions.
The biggest two winners are currently Reform, which has gained eleven new voters, and the Greens who have gained five.
You can compare the current migration with the 2019-2024 migration.
A striking fact is that although 25 voters out of one hundred have changed allegiance in some way, only one voter has crossed over between left and right (from Labour to Reform). This polarisation of politics could have important consequences for party strategists as it may make more sense for parties to seek to dominate their "half" of politics, rather than compete over the centre ground.
Also watch our Podcast on this topic on either YouTube or Spotify.
Data were produced using a combination of MRP polling and up-to-date average of national opinion polls. The transition data in matrix form are:
| CON | LAB | LIB | Reform | Green | Other | Non-Voting | Total | |
|---|---|---|---|---|---|---|---|---|
| CON | 15.5 | 0.7 | 0.6 | 5.6 | 0.1 | 0.3 | 1.5 | 24.3 |
| LAB | 0.9 | 21.9 | 1.8 | 1.3 | 3.3 | 0.9 | 4.5 | 34.7 |
| LIB | 0.6 | 0.8 | 7.4 | 0.6 | 1.1 | 0.4 | 1.7 | 12.5 |
| Reform | 0.9 | 0.2 | 0.1 | 10.8 | 0.1 | 1.7 | 0.9 | 14.7 |
| Green | 0.3 | 0.5 | 0.3 | 0.2 | 4.3 | 0.1 | 1.2 | 6.9 |
| Other | 0.1 | 0.5 | 0.1 | 0.4 | 0.7 | 3.5 | 1.5 | 6.8 |
| Non-Voting | 3.3 | 2.8 | 1.1 | 7.0 | 1.9 | 2.2 | 48.9 | 67.2 |
| Total | 21.5 | 27.4 | 11.6 | 26.0 | 11.5 | 9.0 | 60.2 | 167.2 |
Here the rows represent how each party did at the general election in July 2024, and columns show how each party is doing as at August 2026. For example, the Conservative row has 24.3 voters, 15.5 of which are sticking with the party, but 5.6 are switching to Reform. The Conservative column has 21.5 voters, with the Conservatives gaining small numbers from other parties and 3.3 voters from the non-voting bloc.
The rounded version of this table, with opposite flows (eg Conservative to Labour and Labour to Conservative) netted out to a single flow, is shown here:
| CON | LAB | LIB | Reform | Green | Other | Non-Voting | Total | |
|---|---|---|---|---|---|---|---|---|
| CON | 19 | 0 | 0 | 5 | 0 | 0 | 0 | 24 |
| LAB | 0 | 28 | 1 | 1 | 3 | 0 | 2 | 35 |
| LIB | 0 | 0 | 11 | 0 | 1 | 0 | 1 | 13 |
| Reform | 0 | 0 | 0 | 14 | 0 | 1 | 0 | 15 |
| Green | 0 | 0 | 0 | 0 | 7 | 0 | 0 | 7 |
| Other | 0 | 0 | 0 | 0 | 0 | 7 | 0 | 7 |
| Non-Voting | 2 | 0 | 0 | 6 | 1 | 1 | 57 | 67 |
| Total | 21 | 28 | 12 | 26 | 12 | 9 | 60 | 167 |
A summary way of seeing this data is the final table. Here there is one row per party, and the columns show respectively: the GE 2024 vote shares; the number of "crossover" voters who change from left to right or vice versa; the number of "same-side" switchers who change party but not side; "other" switchers in-and-out of other parties and the non-voting bloc; the total net change for the party; and the current support as at August 2026.
| Party | GE 2024 | Crossover Switch | Same-side Switch | Other Change | Net Change | Aug 2026 |
|---|---|---|---|---|---|---|
| Reform | 15 | 1 | 5 | 5 | +11 | 26 |
| CON | 24 | 0 | −5 | 2 | −3 | 21 |
| LIB | 13 | 0 | 0 | −1 | −1 | 12 |
| LAB | 35 | −1 | −4 | −2 | −7 | 28 |
| Green | 7 | 0 | 4 | 1 | +5 | 12 |
| Other | 7 | 0 | 0 | 2 | +2 | 9 |
| Total | 100 | 0 | 0 | 7 | +7 | 108 |
Note that the "units" are equal to one percent of the 2024 turnout. The predicted turnout has increased, so the total number of voters has increased to 108.