What does the UK Office for National Statistics publish about religious affiliation trends and how do they compare to these projections?

1. ONS’s headline findings: Christianity down, “No religion” up

The ONS’s published census outputs show that for England and Wales the proportion identifying as Christian fell from 59.3% in 2011 to roughly 46.2% in 2021 (about 27.5 million people), making “Christian” still the largest single response but no longer a majority; simultaneously the “No religion” category grew to a substantial minority ^{[1] [3]}. The ONS has released complementary datasets — including age and sex breakdowns and experimental inter-censal estimates for population by religion from 2014–24 — to provide a timelier picture between censuses ^{[4] [1]}.

2. How the ONS measures religion — affiliation, not belief or practice

ONS documents explicitly define the census religion question as measuring “religious affiliation” — the religion people connect or identify with — and note it does not directly measure belief or frequency of worship; the question was voluntary in 2021 and differences in question wording across the UK (affiliation in England/Wales versus “belonging” in Scotland/Northern Ireland) complicate strict UK-wide comparisons ^{[2] [5] [3]}. The Government Analysis Function and ONS guidance stress that surveys capture three dimensions — affiliation, practice and belief — and the census focuses on the first, meaning decline in self-identification does not automatically quantify decline in cultural or familial religious practices ^[3].

3. Age, sex and regional patterns the ONS highlights

ONS breakdowns show the religious shift is strongly generational and geographically uneven: younger cohorts are much more likely to report no religion, while older cohorts remain predominantly Christian; the ONS age-and-sex tables and bespoke datasets reveal, for example, that most people under 65 reporting “no religion” are concentrated in younger age groups ^{[2] [1]}. Parliamentary and constituency-level summaries using census data also demonstrate regional variation — some Scottish constituencies top the “no religion” list while parts of the North West show higher Christian affiliation — and Northern Ireland and Scotland use different census question forms, affecting comparability ^{[6] [5] [7]}.

4. How ONS evidence compares to external projections and narratives

Some media and faith-focused outlets have extrapolated the ONS trends into sharper predictions — for example, commentary suggesting non-religious majorities within a decade or projecting Christianity below 40% by 2031 — but those claims are often external extrapolations rather than ONS-produced long-range projections ^{[8] [9]}. The ONS itself publishes inter-censal estimates and experimental series (2014–24) and stresses data limitations; it does not present single definitive long-term forecasts of the religious landscape in the same way commentators do, and warns about interpreting affiliation changes as equivalent to changes in belief or practice ^{[4] [1] [3]}.

5. What the ONS data can and cannot tell policymakers and the public

ONS outputs are authoritative for measuring self-reported affiliation at census moments and for producing demographic breakdowns useful to planners, but the office repeatedly cautions users that the question’s voluntary nature, differences in how “religion” is asked across UK nations, and the distinction between affiliation and practice limit causal claims about secularisation; therefore while ONS data firmly documents the direction and scale of recent change, claims about the pace and societal impact of future change rest on assumptions outside the ONS’s core descriptive outputs ^{[2] [5] [3] [4]}.