Blog/Analysis

Does the rise of Reform UK pose an existential challenge to the Conservatives? And are economic issues behind this, or are social issues really driving differences in public opinion?

In this week’s Bi_Focal, which comes in two parts, we analyse two things. In Part 1 we look at the rise in polling numbers for Reform, and the degree to which this poses a challenge to the Conservative party support base. TL;DR - a great deal, but not 1993 Canada style. To answer this broad question we cover methodological differences in how we measure smaller parties like Reform and the Greens, the demographics of the Reform vote, and a look at the motivations of why people feel drawn to Reform in the first place. We’ve been able to do this, by conducting a survey of around 500 current Reform voters as well as a larger nationally representative sample.

In Part 2 - to be published later this week and also connected to the rise of Reform - we try to and really understand whether our politics is back to economics and issues of cost of living, or really whether our post-Brexit politics is operating on a social/culture war axis. TL;DR - politics is still about social issues and economic ones, but we’re not convinced that economics will be the primary axis of division until a Labour Government makes economic moves that putatively hit people at the higher end of the wealth and income spectrum.  

Where do voters currently stand?

Here at Focaldata, we ran a survey of 2,700 respondents (of which weighted n=171 Reform voters) along with a targeted booster poll of 468 Reform-intending voters, with data collected between 21/12/2023 and  16/01/2024 (note: the booster sample data was only used in the section on Reform pull/push and not in voting intention and attitude mapping).

How voting intention is asked differs across pollsters, and one methodological difference makes a large difference to Reform numbers: whether you prompt for smaller party names in the survey. To test how big this difference was we split our sample in half (i.e., a A/B test) where one ballot question for the next election had all the parties listed (prompted) and another with only the three major parties (Conservative, Labour and Liberal Democrat), and then smaller parties shown if 'Other' is chosen.

We find the vote % is around 2-3pp higher for the smaller parties when prompted vs unprompted. The Greens took a particular knock without prompting, more than halving their vote share from 6.5% to 3.2%, while Reform lost almost a quarter their support from 8.8% to 6.8%. The Conservatives are the main beneficiary of asking unprompted surveys - which is interesting: it’s not just about Reform-inclined voters picking the Conservative option when there are constrained minor party options, but other party voters as well. This is reflected in pollsters such as Savanta and Ipsos - who have a higher structural vote for the Conservatives than other pollsters.

For what it’s worth, we think that asking unprompted voting intention is a better metric this far out from an election - but that closer to election day it makes sense for pollsters to show all the parties that are going to be there on the ballot paper. Ultimately it’s a judgement call, and there’s good arguments on both sides.

Table 1: A/B testing prompted vs unprompted voting intention

In the analysis that follows, we’ve collapsed prompted/unprompted responses to give total and headline vote share as in the table below. This holds up against a recent poll of polls, while the effect of ‘unprompting’ is visible.

Table 2: Voting intention in our sample

Reform: Rising, and fast

So if we can agree there’s a distinct possibility that the Reform vote is being overstated by a couple of percentage points, this is made somewhat moot by the fact that their national vote share continues to increase into low double digits - and is now not far off their ancestor UKIP party’s 12.6% performance in 2015.

And all this, before a potential return of Nigel Farage. There’s been a lot of conversation about how little the party has been able to translate its vote share into real by-election and local election gains, so all eyes will be on the double by-election header coming later this month. A recent Survation constituency poll in Clacton, out earlier this month, certainly raised a few eyebrows with Reform polling only 18%. We would expect it to be structurally a lot higher than this, particularly in this constituency, if Reform were truly in the double digits.

To try and “de-Farage” a Reform voter analysis - there will be a lot of others polling his appeal -  we wanted to look at what Reform voters actually think on different issues, on public policy and public services. In other words to treat them with a bit more sophistication than just seeing Reform voters as people who respond to a Farage bat signal once every 4 years. There is a great detail and nuance behind the third surge for the radical right in the last decade; the first being in 2013-2015, the second in 2019, and the third now, shown below by tracking voteshare of parties over time.

Figure 1. National vote share, 2010-2014

Who are Reform voters?

We know from regular public polling who Reform voters are - largely similar to UKIP and Brexit party voters that preceded them; older, non-university educated, semi-rural, or rural voters - who are the sharpest edge of the 52%-strong 2016 Leave coalition. What we’re really interested in, in this Bi_Focal is what they think and how they're positioned, and how similar the Conservative vote is Reform today.  

To explain similarities and differences among people’s attitudes, a modelling approach we’ve seen success with in the past is to run a principal component analysis (PCA) on the different questions we asked respondents about their social, economic and political attitudes.

This technique teases out deeper held, core preferences survey respondents have, which then help explain patterns of responses to more specific questions. It’s a standard technique that most pollsters or agencies deploy when going beyond cross-tabs. What we’ve done for this Bi_Focal is essentially reconstruct the political compass - where the vertical axis represents economic attitudes with the right wing free marketers at the top and socialist left wing voters at the bottom - and the horizontal axis acting as a social axis - with socially Conservative pro death penalty, tough on crime, anti-immigration voters are on the right hand side, and liberal cosmopolitan voters on the left.

Below you can get a sense of what each quadrant represents in terms of economic and social skew. We can also visualise where groups of respondents sit given their position on certain political questions.

Here we see how binary questions about some issues are neatly split along our social identity axis: the social right skew towards the view criminals are bad people at heart, gender is determined by birth and making Brexit work, while the social left supports gender self-identification, structural perspectives on crime and rethinking Brexit.

Other issues are split more traditionally, either both socially and economically left-wing – support for strikers, pro redistribution and income equality – or right-wing – tough on strikes, laissez-faire and anti income redistribution.

Finally, some issues upend the traditional left- and right-wing lens: culture war issues – immigration and racial discrimination – and ‘the elites are out of touch’ issues – like Net Zero and ULEZ – that separate affluent, progressive-presenting liberals from traditional, working-class segments of the public.

What we can also do - shown in the figures below - is plot each party's supporters on these same axes, along with the average position by party. Taking a look at these figures, the attitude map of Britain today has eight interesting features:

1. British politics is loosely social - see how much variance there is between the left and right, Labour and Conservative is explained by the social attitudes (horizontal axis) - but where economics explains much of the difference between different parts of the right and left.
   
   
2. Each party’s platform isn’t that well calibrated to its voters. Vibes seem to be the order of the day
3. Green party voters show a good deal more tendencies to the economic right than Labour voters, but have a policy platform that is incredibly left wing (more on this analysis for the Spectator)
4. Reform are standing on a free market platform even though their voters are some of the most statist, economically-nationalist in the country
   
   
5. Reform voters are the furthest away from the average voter - but are also the party that’s showing the fastest rise in the polls (more on this paradox later)
   
   
6. The Don’t Know’s really look more Conservative than any other group - but they look slightly more economically right wing, and more socially liberal than the Conservative vote as it stands today
   
   
7. Reform defectors from the Conservatives look a lot like the Conservatives top-up “new voters” from the 2017 and 2019 elections - socially authoritarian and economically left wing, and operate almost exclusively in a part of the value space that is not being contested by other parties
   
   
8. Those opting out of the electorate “I would not vote” look working class, very left wing economically, and socially neither liberal nor authoritarian. Very similar to low turnout groups in the past
   
   
9. Left wing party supporters look attitudinally very similar - even if they are demographically not. This dislocation of belief-sets from demographics is really interesting and doesn’t get enough attention. The political compass when constructed using demographics only really looks different to the graph below by Liberal Democrat voters looking in many ways the opposite of traditional Labour voters - both in terms of geography and individual level demographic characteristics

Figure 2: Attitudinal mapping of each party's supporters

Looking at polling cross-breaks through this lens is a lot of fun. The age-gender curves plotted on these axes using our data really do back up the John Burn-Murdoch thesis on a global gender gap - which looks very acute for the youngest age cohorts.

Reform's triple threat

Shifting the focus to Reform directly, we see the party as a threat to the Conservatives in terms of their past, present and future voters:

Threat 1: Ancestral Conservative voters now voting Reform

Reform’s ancestry is mixed, but is not reliant on previous Brexit Party voters. Party support leans heavily on ancestral Conservative 2019 supporters who’ve shifted allegiance, with over half (4.7%) of Reform’s total voteshare coming from those voting Conservative in 2019. Interestingly, the Brexit party gained upwards of 2% voteshare in 2019, but only 0.7% has shifted to Reform: we find the remainder dispersed evenly across Labour, Tories, Lid Dems and even some Greens!

Figure 4: Reform voters' ancestry

Threat 2: Reform voters who did not vote Conservative but might otherwise

Not only is Reform pulling away a section of the Conservative voter base (4.7%, Threat 1), but it’s also standing in the way of the Government capturing new potential voters – those who didn’t vote Conservative previously but might otherwise have in the absence of a resurgent Reform party (Threat 2). This is highlighted in the useful - but complicated - graph below which visualises the different component parts of the current Reform coalition - and how many of Reform voters today who didn’t vote Conservative in 2019, look a lot like Conservative voters today.

Figure 5: Attitudes of Reform voters based on their 2019 General Election vote

Threat 3: Current Conservative voters whose attitudes resonate with Reform

The sheer overlap on social attitudes between Reform and Conservative voters today creates both large upside and downsides for the modern Conservative party. A look-a-like party that lends votes when it’s with you can balloon the party above 33/34%, but can crater you if they’re ascendant. We’ve seen both scenarios happen in just the last 5 years. For many reasons David Cameron’s 34-36% vote slice was a bit more demographically and attitudinally resilient, as it drew voters from across the political map. This may have lowered its potential ceiling for proper majorities, but it also gave it a floor.

Figure 6: Attitude overlap of current Tory and Reform voters

Not just a defector party

Another hypothesis we wanted to test was whether the Reform party was merely a vehicle for disgruntled Conservatives. We ran an additional poll of Reform-only voters (n=468) to combine with our existing sample (n=171), for a larger sample (n=634) that allows a deep dive into Reform voter attitudes. We’re interested in how reasons for choosing Reform interact with some of these social and economic attitudes we’ve been exploring.

The table below shows possible responses to a question of why Reform supporters are choosing the party. What’s clear is that it’s wrong to think that the rise of Reform is purely down to disaffected Tory supporters. Only around half (54%) of Reform supporters choose anti-Conservative “push” options, with the other half (46%) choosing pro-Reform party “pull” responses.

Table 3: Identifying Reform push vs pull voters

Thinking about breaking Reform out into two buckets - push and pull voters - and then seeing how these Reform sub-groups compare to two other voter groups - Current 2024 Conservatives (who also voted Conservative in 2019) and Conservative 2019 defectors to Labour - you can begin to understand the scale and problem the Government faces. Just to remind you that each group comprise of:

• Reform Pull (n=294): Reform voters whose primary viewpoint is pro-Reform identity fit or culture-war alignment with the Party
  
  
• Reform Push (n=340): Reform voters whose primary viewpoint is anti-Conservative: unhappy at ruling party performance, competence representativeness of everyday people
  
  
• Tory loyalists (n=459): Conservative-intending voters who also voted Conservative in the 2019 General Election
  
  
• Tory defectors (n=119): Labour-intending voters who voted Conservative in the 2019 General Election

For those of you who don’t want to wade through the 14 consecutive graphs below, there were 5 really interesting things that come from looking at the electorate through the lens of the above groups:

1. Each group looks pretty different to the other - with no obvious policy lever or political stance for Conservative recovery and maintenance other than: i) gripping immigration, ii) tackling the cost of living and iii) providing reassurance that Net Zero isn’t going to bankrupt the country. These are all quite difficult things to execute under the current timeframe.  
   
   
2. Current Conservatives are marginally - but clearly, across multiple measures - more economically right wing than people who’ve left the party for Labour or joined Reform. Classical low tax Conservativism - whatever its actual merits - is not currently the base from which to win an election
   
   
3. The Conservative defectors to Reform are very exercised about immigration control, and about Britain’s national identity  - but the equally large group of Labour defectors are pretty socially split - indicating that perhaps half of these Conservative-Labour defectors are unbiddable
   
   
4. The thing that unites all defectors to all parts of the political spectrum is the “time for change” argument and the idea that Britain needs radical reform. It is challenging to understand why the Government temporarily campaigned on a “change” agenda last autumn
   
   
5. The Conservative family is a very complicated beast when it comes to economics. We go into this in the next section - but basically the case for social democracy or classical low tax Conservatism is a pretty hard one to make on the right, opinions are deeply split, weakly held and paradoxical all around.

In Part 2, coming later this week, we analyse whether economic or social issues are driving public sentiment -- we'll see that politics is about both, and while economics is top of mind, social issues are the cheat code that continue to cleanly divide the public.

Figure 7: Reform 'push' and 'pull' voters, position on social and economic questions

Sample

Focaldata online polling of 2,746 respondents in Britain. Fieldwork was carried out between 21/12/2023 and 16/01/2024. Data is weighted by reported age, gender, education, ethnicity, region and 2019 General Election vote.

This blog comes out as a newsletter before we publish on our website. If you'd like to get ahead of the curve, why not subscribe and get the next one straight to your inbox. You can also follow Focaldata on Twitter and LinkedIn.

Feel free to check out our previous editions of Bi_Focal:

• Our MRP of the Voice referendum in Australia (we were one of the closest pollsters to the final vote tally) - link here
• A new theory of "elastic seats", a defining feature of pivotal areas with large variation in voter support over time - link here
• Stripping down how the British public see their leaders -- link here
• A deep dive into Scottish politics to see how much trouble the SNP were in when Yousaf started out -- link here
• Our (now year old) stress test of predictions about a Labour landslide -- still valid a year on, we think! -- link here

About Focaldata

We're on a mission to power the world’s understanding of what people think and do.

We research public opinion for data-driven organisations. With Focaldata, decision makers and researchers get the most rigorous data and analysis at 5x the speed — reducing the time to decision while delivering uncommonly actionable insights. We are a team of engineers, data scientists, product specialists and researchers building outstanding technology and next-generation services. Our services are global. We are non-political and non-ideological. 

Find our more: www.focaldata.com

Sign up to receive the next issue of Bi_Focal

If you have any thoughts or feedback, please email me at james@focaldata.com.