Religious Census 2011 – England and Wales

There have been some marked changes in the religious composition of England and Wales during the past decade, according to the first results from the 2011 census of population which were released by the Office for National Statistics (ONS) on 11 December 2012.

A statistical bulletin containing the principal findings, with analysis and commentary, is available at:

http://www.ons.gov.uk/ons/dcp171776_290510.pdf

A spreadsheet of statistics for each local or unitary authority, expressed as numbers, percentages and rankings for nine main religious categories, can be found in Table KS209EW at:

http://www.ons.gov.uk/ons/publications/re-reference-tables.html?edition=tcm%3A77-286262

At the same URL is Table QS210EW which contains the same information for local authorities broken down by a 58-category classification, utilizing the write-in responses given by those claiming to be of ‘any other religion’.

A podcast featuring audio and graphical animations is on the ONS YouTube channel at:

http://www.youtube.com/watch?v=tXdZJoXuxC8

The census was taken on 27 March 2011. The question asked on religion was the same as in 2001: ‘What is your religion?’ This formulation, which some interpret as implying that a respondent would/should have a faith, is known to give higher positive answers than the more open sort of question asked in the British Social Attitudes (BSA) Surveys, which is: ‘Do you regard yourself as belonging to any particular religion? If yes, which?’ The 2011 BSA reported 46.1% as having no religion on this basis.

The census reply options (in this order on the self-completion schedule) were: no religion, Christian, Buddhist, Hindu, Jewish, Muslim, Sikh, and any other religion (write in). However, it should be noted that ‘no religion’ replaced ‘none’ (which had been used in 2001), ‘for clarity’ and consistency with other questions, as ONS explains. Where appropriate, ONS reassigned the ‘other religion’ given to one of the main religious categories (including ‘no religion’).

The religion question was entirely voluntary, and 7.2% chose not to answer it, slightly down on the 7.7% in 2001. It cannot be assumed that the religion of these people will exactly match the distribution of those who did reply, so it is not recommended that the data are rebaselined to exclude non-respondents.

Reasons for refusing to answer were not sought, although, clearly, some will have been motivated by an aversion to ‘the state’ enquiring into such a private matter as religion, while others will have proceeded on the principle of never volunteering information which is not mandatory.

Another methodological consideration to be borne in mind is that the census question on religion will often have been answered by the head of household on behalf of all members of the household and thus may reflect his/her interpretation rather than their actual views, especially if the head of household held strong religious convictions one way or another. 

Finally, we should note that there was some apparent ‘trivialization’ in the responses. The social media campaign to have ‘heavy metal’ inserted as a religion did not have much impact, since only 6,000 eventually declared themselves as such on the census form. However, the ‘Jedi knight’ (from the Star Wars films) campaign, which had been so influential in 2001, continued to have an effect, with 177,000 adherents, albeit very many fewer than a decade previously (when there were 390,000).   

Moving to the picture which emerges from the census, there has been a big increase (over two-thirds), from 14.8% to 25.1% (the latter equating to 14,100,000 individuals), in those professing no religion. It is hard to say whether the alteration in the wording of the reply option contributed to this or not, or the reassignment to ‘no religion’ of 32,000 agnostics, 29,000 atheists, and 15,000 humanists.

At ‘regional’ level, the proportion with no religion peaked in Wales (32.1%), once renowned for its Nonconformity, and at local authority level in Norwich (42.5%), closely followed by Brighton and Hove (42.4%). Regionally, the North-West had the fewest with no religion (19.8%). Local authorities with unusually low numbers of the non-religious superficially correlated with the strong presence of Catholics, Jews, Muslims, and Sikhs, albeit this requires further investigation. 

There has been a correspondingly big reduction, almost one-fifth, in the number of people claiming to be Christian, from 71.7% in 2001 to 59.3% in 2011, although Christianity still remains by far the biggest religious group. The fall has occurred across all local authorities, with the largest drop (16.8%) in Kingston-upon-Hull. The North-East (67.5%) and North-West of England (67.3%) had the highest proportion of Christians, reaching 80.9% in Knowsley (perhaps due to the Roman Catholic heritage). London had the least (48.4%), linked to the fact that only a minority of Londoners now describe themselves as white British.

The census did not seek to differentiate between the various denominations which make up Christianity, and this is also true of other Government sources. However, it is clear from other sample surveys that the main reason for Christianity losing ground is the collapse in nominal allegiance to the Church of England.

A crude comparison of the BSA data for 2001 and 2011 shows that the Church of England’s ‘market share’ of the adult population fell by 28.1% during the decade. According to a very large (64,000) sample of adult Britons interviewed by YouGov in April-May 2011, i.e. about the same time as the census, one-third affiliate to the Church of England (and its Anglican equivalents in Wales and Scotland); in 1947-49 it had been 51% in a series of Gallup polls.

The next largest faith after Christianity in the census was Islam. Muslims now account for 4.8% of English and Welsh residents, up from 3.0% in 2001. The increase, to 2,706,000, is broadly in line with expectation. It probably substantially reflects the youthful profile of the Muslim community, with a greater concentration in the child-bearing cohorts (although age by religion breakdowns have yet to be published by ONS, so this assumption cannot be empirically validated). Migration may also have played a part.

London had the highest proportion of Muslims (12.4%), and the fastest rate of growth since 2001, Tower Hamlets being the local authority with the largest concentration (34.5%, 7.4% more than Christians), and Newham the next (32.0%). More generally, the capital was the most diverse region, with 22.4% identifying with a faith other than Christianity.

All other religions, apart from Christianity and Islam, comprised 3.6% of the population, against 2.8% in 2001. They include Hindus (1.5%), Sikhs (0.8%), Jews (0.5%), Buddhists (0.4%), and others (0.4%). Sikhs were disproportionately resident in the West Midlands, but Buddhists, Hindus, and Jews were most prevalent in London. Among the 0.4% of others, Pagans (57,000) and Spiritualists (39,000) were most numerous.

A second post by me, summarizing public comment on the English and Welsh religious census data, will appear over the next few days. BRIN’s co-director, Professor David Voas, hopes to post his own analysis on BRIN in due course.

Religion results for the 2011 census in Scotland (which was organized by the National Records of Scotland) have not been published yet, but are expected to be released shortly (and will be covered by BRIN as soon as possible). Those for Northern Ireland (where the census is run by the Northern Ireland Statistics and Research Agency) are summarized at:

http://www.nisra.gov.uk/Census/key_stats_bulletin_2011.pdf