The Religion Counts team looks at how religious background and practice is likely to influence who people vote for. 15/05/2024

It has been over 10 years since Theos published its Voting and Values in Britain: Does religion count? report, examining the relationship between religious and political commitments in Britain. As in 2014 so today, religion remains an understudied variable in British political life, and important questions – especially in this General Election year – on the existence and significance of the religious vote remain unanswered.

In our previous article, which launched our Religion Counts series looking at how different aspects of religion affect voting patterns in the UK. We looked at whether people intended to vote at all; now we ask how they intend to vote. This article first looks at voting intentions by religious affiliation over the past 10 years, and then concludes with a focussed cross–sectional analysis of the state of affairs today.

As in our previous paper, our source of data is the British Election Study (BES), a national representative survey which has been conducted at every General Election since 1964. More information is available in the notes on the data below.

Voting intentions over time 2014–2023

Picking up where our previous 2014 Theos report stopped, we look at voting intentions by religious affiliation between 2014 and 2023.

We find that between 2014 and 2019, there was slightly more support for the Conservatives than for Labour in the population as a whole. This support increased significantly at the time of their landslide victory in the 2019 general election. However, this increase in support reversed dramatically after the first year of the pandemic, with voting intentions for the Conservative Party plummeting ever since.

We also note that in early 2019, half a year before the snap election, we can clearly see a crisis, when traditional party preferences and voting habits broke down, with the Lib Dems, Labour, and Conservatives all sharing around 15% of the vote. In fact, it appears that the Brexit Party (now Reform UK) could have won a general election with a plurality of the vote at the time the BES survey was conducted. This can be explained by the political climate at the time: the then Prime Minister Boris Johnson was unable to persuade Parliament to approve his Brexit withdrawal agreement. The extraordinary show of support for the Brexit party six months before the 2019 General Election, called by Boris Johnson to strengthen the Conservative majority in Parliament and get his deal through, can be interpreted as a proxy vote of support for the Conservative push to ‘get Brexit done’.

Anglicans consistently vote Conservative

While the Anglican vote has fluctuated over the past 10 years, it has consistently leaned to the right, with a clear preference for the Conservative Party. As with the general population, the then newly formed Brexit party attracted a large share of the Anglican vote in 2019 – as much as 30% – mostly stolen from the Conservatives, but also from Labour. Today it has fallen to under 10%, in line with the rest of the British electorate.

Support for the Conservative Party peaked twice over this time period: once immediately after the Brexit referendum, for the 2017 General Election, and once immediately after the 2019 General Election, which was also run on a platform of ‘getting Brexit done’ – both times when national sentiment ran high.

Anglicans also show a clear preference for the two main parties, with overall little support for the Liberal Democrats or any of the smaller parties.

Roman Catholics have become floating voters

Roman Catholic voters have traditionally been seen as a source of Labour votes, but there is little evidence of this in the data over the past decade. While support for the Conservatives among Catholics was 17 points lower than among the wider population in 1979, by 2019 Catholic support for the Tories was two points higher. As the Catholic ‘anti–Tory sentiment’ has eroded, it is possible that their marriage to Labour has finally come to an end. The reasons for this may include the many social changes that have taken place over the last 40 years within the British Catholic community, in particular the weakening of working–class identity. It at also includes the multiple policy standoffs between the Catholic church and Labour governments of the past.

Until the late 20th century, British Catholics were mainly of Irish descent, predominantly working class, so that their ‘natural party’ was Labour.[i] Even before the early 2000s, the British Catholic community’s link to Labour began to weaken as more Catholics became middle class. Over the last 20 years, with the significant influx of Catholic migrants from southern and eastern Europe and elsewhere (e.g. the Philippines and India), it has become even more ethnically pluralistic and further removed from its Irish working–class roots.[ii]

While it is true that Roman Catholics have been intending to vote Labour again since 2022 – ostensibly for the first time since 2016 – this is entirely consistent with the national decline in Conservative support and preference for Labour in the wider population. In fact, Roman Catholic voting intentions have shadowed those of the general population closely since 2016.

We do note that, as for the Anglican vote, usual trends broke down around the time of the 2019 GE, with Roman Catholic voting intentions fairly evenly split between the Conservative, Lib Dem, Labour and Brexit parties. There has also consistently been greater support for the Lib Dems and other smaller parties among Roman Catholic voters in the UK, compared to Anglican voters.

The ‘Other Christian’ group is difficult to analyse because of its denominational diversity. Nevertheless, we can pick up a few things.

There is no clear common party preference among the various Presbyterian, Methodist, Evangelical, Pentecostal and other Christians. The Conservative Party has led voting intentions for most of the last decade, but only by a small margin. And no more than can be interpreted as average, in line with general trends. There is also some evidence that this group is not by definition Conservative–leaning, given the very significant peak in Labour support around the 2015 GE. During the past decade, Christians in this group have also supported smaller third parties more consistently than either Anglicans or Roman Catholics, and support for Lib Dem is higher among ‘Other Christians’ than among any other Christian group.

It is also significant that both around the time of the Brexit referendum and before the 2019 GE, where the main issue was ‘getting Brexit done’, the vote was very divided between a number of parties, similar to the Roman Catholic vote. In the aftermath of the referendum in 2016, ‘Other Christian’ voting intentions have also followed national trends quite closely – though not as closely as Roman Catholic intentions.

Muslims vote Labour

The Muslim population has more consistently preferred a single party – Labour – than any other religious group, even despite divisive events such as the Brexit referendum. It is also the only group not affected by the 2019 crisis when party preferences collapsed for most religious affiliations. It stands out in this respect.

Of course, as this dataset is almost a year old, it does not reflect changes in attitudes towards Labour within the Muslim population since the Gaza–Israel war, but it is unlikely that any other party has won the sympathy of a majority or even a plurality of Muslim voters in the last eight months.

The Muslim population is also the only group where the Brexit party failed to win over a significant proportion of the electorate, probably in part because of the Brexit party’s policy priorities and anti–immigration (and indeed anti–Muslim) rhetoric.

Тhe ‘Other religious’ group is very heterogeneous because of the large number of smaller religious faiths that make it up – from Hindus to Wiccans. It is therefore also even more difficult to analyse meaningfully. Nevertheless, as with the other Christian group, some aspects are worth noting.

Since the Brexit referendum in 2016, with the exception of a small Labour peak in June 2017, the voting intentions of other religious have been virtually indistinguishable from those of the population as a whole. However, they have shown a marginal preference for Labour for most of the last decade, and have also supported smaller third parties more consistently than either Christians or Muslims.

Today, voting intentions appear to have returned to pre–Brexit levels, with around 35% of other religious people intending to vote Labour, just over 20% preferring the Conservative Party, and around the same number intending to vote for the Lib Dems or one of the other smaller parties.

Non–religious Britons vote more consistently Labour

Finally, the non–religious vote. While overall strongly following the general national trend, nones have consistently intended to vote Labour more than average since June 2016. Even when they were hit by the same crisis in June 2019, from which only Muslims were spared, they maintained above average support for Labour.

Conservative support has steadily been decreasing over the past 10 years, and smaller parties feature very prominently. The Liberal Democrats in particular seem to find in the Nones their greatest support base. Non–religious voters have also been affected by the Brexit Party in 2019, who gathered support both at the expense of Labour and Conservative. It is also interesting to note that Reform UK attracts today a similar level of support of 5% across all religious affiliations – with the exception of Muslims.

Voting intentions today

Having observed voting trends over the past ten years, let us now focus on the latest available data.

Christian voters are more likely to vote Conservative

First, the total religious vote brings national Conservative voting intentions up 6%, at the expense of Labour and smaller parties. This is primarily due to Christian voters, who are overall most likely to vote Conservative (despite the denominational differences noted above).

Where only 14% of non–religious voters intend to vote Conservative, 30% of Christians do so. This is largely due to Anglicans, who are more likely to vote Conservative (34%) than for any other one party – confirming that the Church of England, at least in terms of affiliation alone, is still an important provider of Conservative votes.

Roman Catholics are most likely of all Christian groups to vote Labour (36%). They are also strikingly close to the national average (35%). Other Christians are somewhere in between, but tend to lean towards Labour (30%).

Overall voting intentions for the Liberal Democrats and other smaller parties appear to be less affected by religious affiliation, suggesting that factors other than religion predominantly affect voting intention for these.

Still, Christian voters are divided, as only 2 percentage points separate Christian Conservative and Labour voting intentions overall. There is also a great deal of uncertainty, and over a fifth of all Christian voters – or indeed all voters – say they “don’t know” who to vote for.

Muslims and other religious voters are more likely to vote Labour

Muslims overwhelmingly support Labour (58%) as do other religious voters (34%). Muslim voters are the only ones with majority support for a single party, making them (in light of the data examined in our previous article) both the most uncertain about whether to vote at all and the most certain about who to vote for. Of course, as this dataset is almost a year old, it does not reflect changes in attitudes towards Labour within the Muslim population since the Gaza–Israel war, but it is unlikely that any other party has won over a majority of Muslim voters in the meantime.

As for turnout in our previous article, we have broken down affiliation into – first – communal practice measured by frequency of attendance and – second – self–reported religiosity when respondents were asked to say how much of a difference religion makes in their life.

More frequently practicing Christians are more likely vote Labour

Religious voters who intend to vote Labour or Lib Dem are more likely to practise.

This is especially true for Christians: practising Christians tend to prefer Labour (35%), while non–practising Christians tend to prefer the Conservative Party (33%). This is true even of otherwise strongly Conservative–leaning Anglican voters. Amongst Anglicans, support for Labour correlates with a higher average frequency of communal practice and Conservative voters are on average the least frequently practicing Anglicans.

Of course, this does not mean there is a large reservoir of Anglicans who are more likely to vote Labour, since only 17% of Anglicans in the BES sample worship weekly or monthly. In absolute terms the Anglican vote remains Conservative–leaning.

The data also hints at the remnants of a ‘Catholic vote’ for Labour, but only among more frequently practising British Catholics. ‘Cultural’ Catholics – those who rarely or never attend services – are likely to vote in line with the general population. Catholics who still intend to vote Labour (36% of all Catholics) are, on average, still more committed to their faith, as evidenced by their higher frequency of communal practice.

Very counter–intuitively, the trend is reversed for Muslim and other religious voters, where those more sympathetic to the Conservative Party appear to be significantly more frequently practising. However, we should bear in mind that only a small proportion of Muslims (7% of the Muslim sample) are Conservative–voting. In overall numbers, support for Labour remains overwhelming across the board: 48% of practising Muslim and other religious voters intend to vote Labour, as do 45% of those not practising their faith. Additionally, we should bear in mind that the last wave of the BES included in our analysis took place in May 2023, predating the latest Gaza conflict and the controversy over Labour’s policy positions on that issue. We would expect to see movement in Muslim support away from Labour, but of course not to the Conservatives. It is more likely to go to other parties and, as we saw in the 2024 local elections, independents.

Voting intention by self–reported religiosity

Unlike for turnout, which we explored in our previous article, likely voting patterns when analysed by self–perceived religiosity yields different results versus measuring religious practise.

We especially note how Conservative–voting Christians of all denominations are the least religious.

We see an increase in preference for smaller parties (nationalist or otherwise) among more devout Roman Catholic voters, as well as among Anglican voters. Lib Dem–voting and other party–voting Roman Catholics also overtake Labour–voting Roman Catholics in terms of self–reported religiosity.

Lib Dem–voting Muslims perceive themselves as significantly more religious than other Muslim voters, with Labour supporters the second most religious overall.

For all ‘Other religious’ voters, it is difficult to discern a trend. This may be due to the diversity of the ‘Other religious’ group and to an overall more homogeneous religious self–identification.

While objective and subjective measures of religiosity show a similar picture for turnout – as explored in our previous article in this series – this is not the case for voting intentions. In part, this may be because turnout is a matter of general civic–mindedness (as a citizen and a person of faith, do I have a duty to vote on election day?) whereas voting intention is much more nuanced and a matter of political ideology, worldview, and priorities.

The discrepancy in voting intentions between the two measures of religiosity reveals a qualitative difference between being religious and frequently practising one’s faith – a difference that also affects political views. It raises the question of what the practice of our beliefs does to our thinking about our society and its needs, as opposed to merely holding those beliefs.

Conclusions

We know from British political history that affiliations between particular religious groups and political parties have emerged and sometimes submerged as part of wider processes of social change. Of course, it is reductive to treat these affiliations as a pure religion effect – the demographic and socio–economic characteristics of religious groups play a large part in those affiliations. However, it is also reductive to suggest that religion is not playing a role. For example, policy conflicts (or sympathies) over issues that are religiously freighted will also have an effect.

In this data, we see that some things have changed (the disappearance of the Roman Catholic – Labour affiliation), but that somethings have stayed the same (the Anglican Conservative lean). We can also speculate that some things may be in the process of change (a Muslim turn away from Labour). As yet, we do not have the data to know one way or the other.

Beneath the broad claims about these affiliations, there are important details to observe. First and foremost, a religious affiliation plays a different role from religious practice and self–perceived levels of religiosity. This needs to be factored in to any discussion about the role of religion in elections.   

Note about the data

Our longitudinal analysis uses data from the 2nd, 6th, 9th, 13th, 14th, 16th, 17th, 20th, 21st, 23rd and 25th waves of the BES Internet Panel. ‘Don’t know’ has been omitted from all longitudinal graphs. Our focussed cross–sectional analysis of the last available data uses both the 23rd and 25th waves of the BES. All data was collected for BES by YouGov via a representative internet panel using an online sample of YouGov panel members. All figures are weighted respective to their own waves of collection and percentages are rounded to the nearest unit.

For data looking at affiliation alone, only the 25th wave of the BES is used (N = 28,964). When looking at communal practice and self–reported religiosity, data from the 23rd and 25th waves has been combined, keeping only participants who were common to both waves (N = 17,446).

Variables were coded so as to be consistent with previous Theos research on religious voting patterns.

On Theos’ Religion Counts series

This paper is part of a larger series of work written by experts exploring the links between religion, values, identity, and democracy in countries around the world. Theos working on a further briefing papers on the impact of religion on voting patterns in the UK. Watch this space and follow Theos on social media to be updated on their release.

Could you help uncover the impact faith can make in this election year by giving to our Religion Counts election appeal?