The story so far

Our campaign so far

Recognising problems with the Census

Over the last several years in meetings and briefings with parliamentarians and the Office of National Statistics (ONS), the key arguments the BHA raised against the Census question on religion were:

  • The question is leading
  • A single question is asked, rather than first asking whether respondents have a religion or not
  • The question is framed to capture the loosest possible religious affiliation
  • The question is placed in a group with questions of nationality and ethnicity
  • Administering the question through heads of household, many of whom may answer for their family members

The BHA argued strongly that the question on the census should be altered to capture a more accurate picture of religion and belief in England and Wales.

We made a detailed submission and a supplementary submission to the House of Commons Public Administration Select Committee (PASC) on the problems with the Census question on religion proposed for the 2011 Census.

In these submissions the BHA argued that the inclusion of the question concerning belief in the 2001 Census had been added at an extremely late stage with minimal consultation or deeper consideration. The figure yielded from the question was 72%, which in comparison to the levels found in social surveys and polls seemed exceptionally high.

The 27th report of the highly respected British Social Attitudes survey, published in 2011, found that 51% of those questioned professed having no religion. In the 26th report, published in 2010, 62% stated they never attended any religious services, and only 18% attend religious services at least once a month.

Similar levels of belief are indicated in a variety of other polls or surveys, for example a 2006 Guardian/ICM poll found 63% said they were non-religious.

Looking at church attendance figures in further detail, it is clear that the level of religious belief indicated in the 2001 census appear exaggerated. According to ‘Religious Trends No 7 (2007-2008)’ published by Christian Research In 1990 5,595,600 people, representing 10% of the UK population, regularly attended Church, by 2005 this number had reduced to 3,926,300, equating to 6.7% of the UK population. The Church of England’s attendance figures indicate between 2002 and 2008, average Sunday attendance figures have diminished from 1,005,000 to 960,000.

A measure of cultural identification?

The 2001 Census question on religion actually measured cultural identification (and even then very crudely). The ONS stated in their white paper on the 2011 census that the ‘evidence suggests that the 2001 question [on religion] provides a reasonable proxy for Sikh and Jewish ethnic groups’.

The ONS also tested the question ‘Do you regard yourself as belonging to a religion?’ and this produced statistically significant changes in the proportion of Christians, Sikhs and the non-religious. As would be expected, the proportion of Christians decreased, and the proportion of non-religious people increased.

The BHA was eventually told by the ONS that they had decided not to propose this question for the 2011 Census because it produced a lower figure also for those following the Sikh religion. This was presumably because it did not capture non-religious Sikhs in England and Wales. A lower figure for the Sikh category was deemed unacceptable by the ONS. See What does the religion question really measure?

What happens next

Through The Census Campaign the BHA is encouraging individuals with no religious belief to tick the ‘No Religion’ box on the census form. In the 2001 census, 7.7 million people, (14.6 per cent in England and 18.5 per cent in Wales) stated they have no religion. This formed the second largest group, after Christianity. See Our recommendation.


Comments are closed.