Considering Reform’s Potential ‘Ceiling’

Viability and Intensity of Preference: Considering Reform’s Potential ‘Ceiling’

Jane Green and Marta Miori

Understanding British politics in 2025 requires recognising the context set by the General Elections in 2017 and 2019. Throughout this period, the Conservatives became increasingly the party of ‘Leave’ voters (who had voted to leave the EU in 2016, or continued to wish to ‘stay out’), and Labour the party of ‘remainers’, and these patterns persisted through to the General Election in 2024. By 2024, the large amount of vote switching in evidence took place primarily within two blocs; with left-liberal (Remain) voters switching between parties on the left and the nationalists, and right-socially conservative (Leave) voters primarily switching from the Conservatives to Reform.

Since 2024, the Conservative vote has continued to drop and Reform UK have continued to rise in opinion polls, but this seems to have plateaued in recent months, suggesting, perhaps, that Reform have hit their ‘ceiling’ and can’t expect to do much better. If this is the case, we would expect the potential for greater switching within the right bloc to have run its course, or perhaps for more current Reform supporters to switch back to the Conservatives. Which is more likely?

In bloc-based competition, there are two critical factors which – at least for now – we expect to be critical in by-elections, in May next year, and in a General Election to come. These, we suggest, are viability and intensity of preference, and how they inter-relate. We saw each at play in the Caerphilly by-election: voters in the left/nationalist bloc appeared to vote tactically for the most viable party in that bloc (in this case, Plaid Cymru) to keep out the most viable party on the right (Reform). This is also about the party perceived to be best-placed to win on the left being more desirable – or much less undesirable – than the most viable party on the right, related to how much clarity there is about ideological position between parties across the blocs.

All of this points to the importance of information signals to voters about which right-party (and which left-party) appears to have the greatest chance of beating a party in the opposite bloc, nationally, or in constituencies, and how important this ‘viability’ question becomes. This is why the May 2026 elections could be important for Conservative fortunes. Will Reform increasingly be seen by right-bloc voters as the more viable choice, rather than the Conservatives, and so could more right-bloc voters who want to stop a more undesirable left-bloc party vote in greater numbers for Reform UK? Or – do the Conservatives who have not yet switched to Reform see Reform as too undesirable?

We turn to British Election Study panel data to help us examine these questions.

The following graphs use the British Election Study internet panel which follows the same people over time and can also be used in each time point as a cross-section of the British voting eligible population, showing the ‘propensity to vote’ for Reform among three groups. This uses a scale where 0 = very unlikely and 10 = very likely, with those who said ‘don’t know’ removed. The first (left-hand-side) is our comparison group. These are the ‘propensity to vote for Reform’ scores for 2024 Conservatives who eventually did switch to Reform in May 2025. This tells us what these ‘propensity to vote’ scores may indicate, given the caveat that people don’t always know or report what they will do later and much is changing politically in a short period. Around 70% of 2024 Conservatives who later switched to Reform (by May 2025) had reported that they were likely to do so (gave scores between 5-10). However, a smaller but significant proportion of those Conservative to Reform switchers (around 15%) had said they were very unlikely to do so and yet reported a Reform voting intention by May 2025. So this type of data is informative, but can underestimate subsequent switching.

Turning to the middle panel, we focus on 2024 Conservatives who still hold a Conservative vote intention (by May 2025). 37% of those people say they may vote for Reform, and half of the group who moved from Conservative to undecided (16% of 2024 Conservatives) may vote for Reform. Together, we see that the remaining Conservatives currently have a lower likelihood of voting for Reform than last year’s switchers, but still very high likelihoods. If we assume that these 5-10 scores can underestimate switching, the potential for Reform appears to be at least 40% of remaining Conservatives (49% of 2024 Conservative voters), and over half of those who have switched from the Conservatives in 2024 to undecided by May 2025 (16% of 2024 Conservative voters).

The above can be contrasted with likelihoods in part of the ‘left bloc’. This shows just how much of Reform’s vote could still come from the right purely by comparison.

The next graphs focus on 2024 Labour voters who either switched to different destinations in May 2025 or who stayed with Labour. Note the different Y axis, with very high zero responses (‘very unlikely to vote for Reform’). We can ignore those who have switched from Labour to the Conservatives – this group is just too tiny to be reliable (2.2%). However, the Labour to undecided group is where Labour could lose some more of its 2024 voters to Reform; but the vast majority of these are probably more likely to vote for Labour or another party on the left.

Finally, there’s the question of ‘non-voters’. Here we remain much more sceptical for now, because so many non-voters are younger, on average, and will still be more likely to vote for a left-bloc party, and so many truly alienated non-voters are so disengaged they may still be very unlikely to vote for any party. However, Reform may be doing better among former Conservative ‘non-voters’; people who voted for the Conservatives in 2017 and 2019 (when turnout was also higher) but who couldn’t vote for the Conservatives or a left-party in 2024. Here, again, Reform’s viability is critical, because their 2024 viability was very low, and many people wouldn’t have thought they had a chance in 2024, but that is now changing as Reform shows success in polls and in local elections.

The following graphs again use the British Election Study internet panel, but we note that these data are not ideally suited for examining turnout (for which the BES post-election random probability is much more reliable, but not suitable for our purposes here). Among the non-voters in our BES online sample, non-voters switched to Reform at higher rates compared to other parties between the general election and this year’s local elections. This can be seen in the far right column of each panel, which shows the percentage of 2024 non-voters the parties had converted to a vote intention by May 2025. Despite this, looking at the propensity to vote scores, the proportions of likely voters for each of the main parties are broadly similar across the board.

Together, this suggests that the ‘right-bloc’ is probably only slightly growing disproportionately to the left, which would confirm years of evidence that turnout doesn’t tend to be differential. One exception regarding clear differential turnout was the 2016 EU referendum when Leave voters (the right bloc) turned out in higher proportions than Remain voters (the left bloc). A repeat of this is the main risk to left parties in a forthcoming General Election; if their voters stay at home disproportionately, even with the ‘threat of Reform’, and Reform’s voters are more mobilised. Here, again, the increasing viability of Reform will also matter. Reform appearing likely to win would be a very different scenario to Remain voters feeling confident that Leave would not.

On the right, the battle for which party can defeat a party on the left side comes down to a battle between intensity of preference and who is the most viable party. Ultimately, this dynamic may shape whether Reform's current support represents a ceiling or a platform for further gains.