Projecting belief - UK religion in 2031
Every ten years, the constituent countries of the UK hold a census, providing a comprehensive snapshot of the population that informs key policy-making and planning decisions over the next decade. In 2021, this snapshot revealed that the share of the UK population identifying as Christian had declined significantly since 2011, falling below 50% for the first time.
It is now four years since the last census and results from the next edition will not be published until 2032. However, the role that religion should play in our society is an active topic of debate. In parliament, bills to allow assisted dying and to remove the legal requirement for Christian worship in school assemblies are progressing through the houses of Commons and Lords respectively. Many have questioned the automatic right of 26 bishops to sit in the House of Lords. And in the wake of its handling of historical abuse cases, many others, including some of these bishops, are calling for the disestablishment of the Church of England altogether.
Our projection of religious affiliation aims to inform this debate, filling the gap between census snapshots by providing an accurate estimate of current and future religious affiliation.
Based on trends observed over the last three census periods, this projection indicates that non-religious worldviews recently overtook Christianity as the UK’s most widely-held belief category (most likely in 2024), and that the share of the UK population identifying as Christian will continue to fall to around 33% by the next census in 2031. This shift will be driven by a continued increase in the share of people identifying with no religion, which will grow to 50% by this point, while the total share of other religions will rise slightly to 12%.
Projecting belief: the model
Our model is composed of over 15,000 granular projections across geographic and demographic groups. For each of 387 local district authorities (LDAs) across the UK, separate projections were made for female and male populations within five-year age bands. These extrapolated the 2011–2021 trend using a weighted combination of linear, exponential, and transition-matrix-based approaches to give an ensemble model which achieved a high degree of accuracy when validated against 2021 results. That's as technical as this article will get, but if you'd like to dig further into the detail, the full methodology is explained here.
Every generation is getting less religious
A common assumption is that the overall change in societal religiosity is driven by a replacement effect as younger, less religious generations replace older, more religious ones. However, analysis of the granular models and aggregation of their predictions by age group reveals that the change is not primarily driven by replacement, but rather that every generation is itself becoming less religious.
Finer-grained analysis of these trends by five year age bands reveals that the most drastic change is visible amongst young adults filling in the census for the first time. In 2021, as in previous years, these young adults identified as having no religion in much greater numbers than their parents had done on their behalf ten years previously.
For example, in the 2011 census, parents were responsible for filling in the survey for their 10–14 year old children and, in aggregate across the UK, designated 28% of them as non-religious. Yet ten years later 51% of these young adults, now 20–24 years old, identified as having no religion when filling in the survey themselves.
Religion by region – our projections mapped
In addition to generational differences, religious identities vary widely across the countries and regions of the UK, and the categories that can be chosen in the respective census surveys reflect this. The survey for England & Wales has a single option for Christians, whilst Scotland’s survey breaks this category into three: Church of Scotland, Roman Catholic and Other Christian. Northern Ireland, the most Christian part of the UK, lists four separate Christian denominations as individual options on the census form, using a free text “Other” category to capture other Christian denominations and all other religions. Each survey also gives respondents the option to select "No religion" or "None".
The map below aggregates these categories into three main groups: Christian, Other religions and No religion. Each LDA is coloured according to its projected Christian population in 2031. Hovering or clicking on these will bring up a tooltip with more information - data for 2011 and 2021 is taken from the census, the 2031 figures are our projections.
Some patterns jump out when this data is visualised. In particular, the four countries of the UK map to three distinct regions. Every part of Northern Ireland will likely remain majority Christian in 2031, although this masks accelerating change, particularly amongst younger urbanites who increasingly resemble their peers in Britain. In contrast, Scotland was already majority non-religious as a nation in 2021 and this will be true of almost all of its regions by 2031, with only Na h-Eileanan Siar (the Outer Hebrides) retaining its Christian majority by then.
England and Wales are more of a patchwork, with greater ethnic and religious diversity (particularly in major cities) as well as rural areas which are projected to retain larger Christian populations. Nevertheless, when considering the largest belief category in each LDA, the scale of change between 2021 and 2031 will be significant: in the 2021 census, Christianity was largest worldview in 271 out of 329 LDAs in England and Wales; by 2031 we project this number will fall to just 12.
Ten years is a long time in politics
Recent events have shaken the assumption that politics in the global West will continue to secularise, one that seemed robust just a short time ago. In the US, Christian nationalism underpins the Project 2025 agenda, and advocacy for policy stances based on "Judeo-Christian values" is starting to happen more regularly in the UK.
Our projections suggest that such an agenda is unlikely to find broad support. On the contrary, it is likely that support for the separation of an established church from the activities of state will continue to grow as the UK population continues to become more plural, diverse and secular in its beliefs.
Coming next...
Next up, I'm planning extend my analysis to look at each of the parliamentary constituencies of the UK and ask if they held a first-past-the-post election, with religious affiliation in place of political parties, which would come out on top, and how is the balance shifting each year? If you've found this analysis interesting, please subscribe to make sure you receive that article.
