Census Question Under Fire

There are just six days to go before UK residents have to complete the household and individual questionnaire for the decennial population census.

But humanists are still simultaneously maintaining their attack on the voluntary question on religion while paradoxically encouraging people to answer it, ideally (from the humanist perspective) by registering as of no religion.

In a press release on 20 March the British Humanist Association (BHA) described the census question as ‘highly misleading’ and ‘fatally flawed for its intended purpose of planning public services’. The BHA’s evidence for this claim comes from new opinion polls conducted online by YouGov in England and Wales and in Scotland.

The English and Welsh survey, commissioned by the BHA, was undertaken on 9-11 March 2011 among a representative sample of 1,896 adults aged 18 and over. The Scottish poll was sponsored by the Humanist Society of Scotland and conducted on 10-14 January 2011 among 2,007 adults.

In England and Wales, when asked the census question ‘What is your religion?’, 61% ticked a religious box and 39% declared themselves to be of no religion. However, when asked ‘Are you religious?’, just 29% said ‘yes’ and 65% ‘no’, ‘meaning over half of those whom the census would count as having a religion said they were not religious’.

Responses varied somewhat according to demographics, most notably by age. Whereas 56% of 18-24s had no religion, the proportion fell steadily throughout the age cohorts, to stand at 25% among the over-55s. Similarly, while 70-73% of the three under-45 cohorts stated that they were not religious, this was the case with 68% of those aged 45-54 and 56% of the over-55s.

Marital status also appeared to make a difference, although this pattern doubtless conceals an age-related effect. The number professing no religion was highest among the never married (53%) and those living as married (52%). It was substantially lower among those who were currently married or in a civil partnership (31%) or had formerly been, 27% among the separated or divorced and 25% with the widowed.

The 53% of the English and Welsh sample who professed to be Christian were additionally asked: ‘Do you believe that Jesus Christ was a real person who died and came back to life and was the son of God?’ Fewer than half (48%) said that they did so believe, with 27% disbelieving and 25% unsure, BHA’s unspoken point presumably being that many so-called Christians have rather a shallow or unconventional faith.

It is also a generally inactive faith, in terms of attendance at a place of worship for religious reasons. Only 15% of the entire sample claimed to have been within the past month, with a further 16% going within the past year, 43% more than a year ago and 20% never. The never category was largest among the 18-24s (28%), with 32% for full-time students.

In Scotland, one-half of the sample was asked the Scottish census question: ‘What religion, religious denomination or body do you belong to?’ In reply, 56% of Scots professed some affiliation (with write-in responses available) and 42% none.

The other Scottish half-sample was initially asked: ‘Are you religious?’ 35% said that they were and 56% that they were not, with 8% uncertain. Those who answered that they were religious or who did not know were then asked: ‘Which religion do you belong to?’ At this point, 22% said that they did not belong to any organized religion.

The BHA press release and links to the data tables for both England and Wales and Scotland will be found at:

http://www.humanism.org.uk/news/view/771

These statistics serve to illustrate what is already generally well-known, that surveys on religious (and – indeed – all other) topics are inevitably informed or perhaps even shaped by question-wording.

The Office for National Statistics, which is overseeing the census, is fully aware of the sensitivities and ambiguities of investigating religion. It has gone to some considerable lengths to research and trial the merits of alternative wordings during its census preparations.

For fuller information about these deliberations and experimentation, see the October 2009 ONS report on Final Recommended Questions for the 2011 Census in England and Wales: Religion, which is available on the ONS website.