10 August 2005

2001 Census Figures On Religion Ludicrously Exaggerated

In its submission to the Office of National Statistics consultation about the 2011 Census, the National Secular Society has criticised the religious question used for England and Wales as being “imprecise to the point of being unprofessional” and resulting in grossly exaggerated figures for the extent of religious belief. The Society proposes an alternative based on the questions used in the 2001 Scottish Census.

Keith Porteous Wood, Executive Director of the National Secular Society commented: “If the objective was to maximise the figure for religious affiliation and minimise it for non-belief, it is difficult to think of any method that would achieve this more effectively than that used in the 2001 English Census.

1. The question was so vague: What is your religion?, rather than make it clear it is the religion (if any) to which you belong at the time of the census, rather than the much higher figure, that of your upbringing. Respondents predominantly opted for this latter interpretation.
2. Even those who no longer practise any religion are used to tick Christian, CofE or RC as a default option on official forms, as for hospital admissions
3. The “none” option was presented in an unnecessarily bald way which put non-religious people off ticking it.
4. The head of the household completes the forms, and is likely to be more religious and put down the religion he would like his/her household to be, rather than as they are. Some householders may not wish to disclose to the head of the household that they no longer share the predominant religion in the household.

Keith Porteous Wood added: “The Office of National Statistics acknowledges that, in effect, the question was interpreted by respondents as being ‘upbringing’. In Scotland, both questions were asked. For those of no religion the proportional difference was a third: 28% (now) v. 18% (upbringing). For the religious, the current figure was 7% lower for religious affiliation than the ‘upbringing’ figure (67% v. 74%).”

“The England and Wales results were so misleading that it would be better to have no such question in the 2011 Census, unless substantial improvements are made. Instead, we propose that the Scottish method be adopted, which asks about religion or belief to which the respondent belonged in upbringing and belongs currently. We believe it would also be helpful to know how many of those belonging to a religion currently practice it.

Having this information with its ability to give geographic analyses will assist Government in understanding much more about religion and ethnicity and for schooling and faith based welfare, and in connection with monitoring discrimination against minorities.