I have been asked to reply.
The information requested falls within the responsibility of the UK Statistics Authority. I have asked the authority to reply.
Letter from Karen Dunnell, dated July 2009:
As National Statistician, I have been asked to reply to your question asking what estimate has been made of the Muslim population of Britain in each year since 1979. (284443)
The Census is the most commonly used source for statistics on religion by the Office for National Statistics (ONS), with more recent information on religious affiliation (that is the identification with a religion irrespective of actual practice or belief) from the Labour Force Survey (LFS). The way people respond to questions on religion is sensitive to what question is asked and how it is asked. Therefore because of differences between the Census and LFS in terms of question design and also coverage of the type of establishments in which people reside, data from these two sources cannot be compared.
The 2001 Census was the first time that a religion question had been asked in the England and Wales (and Scotland) census, and revealed that 1,589,000 people in Great Britain identified as Muslim in 2001.
A question on religious affiliation was introduced into the LFS in Great Britain in Spring 2002 using an extended version of the wording used in the England and Wales Census, but was not asked of people aged 16 and under until 2004. Consequently the earliest suitable estimates are from 2004 which cover people of all ages, and are shown in Table 1 attached.
As with any sample survey, estimates from the LFS are subject to a margin of uncertainty. A guide to the quality of the estimates is included at Table 1.
Thousand 2004 *1,870 2005 *2,017 2006 *2,142 2007 *2,327 2008 *2,422 1 All ages Coefficients of Variation have been calculated as an indication of the quality of the estimates, as described: Guide to Quality: The Coefficient of Variation (CV) indicates the quality of an estimate, the smaller the CV value the higher the quality. The true value is likely to lie within +/- twice the CV - for example, for an estimate of 200 with a CV of 5 per cent. we would expect the population total to be within the range 180-220. Key Coefficient of Variation (CV) (%) Statistical Robustness * 0 = CV≤5 Estimates are considered precise ** 5 = CV ≤10 Estimates are considered reasonably precise *** 10 = CV ≤20 Estimates are considered acceptable **** CV 20 Estimates are considered too unreliable for practical purposes Note: It should be noted that the above estimates exclude people in most types of communal establishment (e.g. hotels, boarding houses, hostels, mobile home sites etc. Source: Labour Force Survey