English Church Census, 2005

Data Archive > International Surveys and Data > Single Nation Surveys > Summary


The fourth English Church Census was carried out on 8 May 2005. Comparable studies had been conducted in 1979, 1989 and 1998. All were organised and led by Dr Peter Brierley, executive director of the organisation Christian Research prior to his retirement in 2007. The goal of the study was to enumerate a complete census of Trinitarian Christian churches in England and their attendance, along with gathering data on a number of questions relating to age and racial makeup, evangelistic ministries, and mission-related activities. A similar attendance survey in Scotland was conducted in 2002.

Data File
Cases: 18,633
Variables: 195
Weight Variable: None
Data Collection
Date Collected: May 8, 2005
Original Survey (Instrument)
English Church Census 2005
Funded By
Financial support was provided by Ansvar Insurance Company, the Baptist Union of Great Britain, the Bible Society, Christian Aid, the Church Mission Society, the Church Pastoral Aid Society, the Churches’ Child Protection Advisory Service, the Economic and Social Research Council, the Evangelical Alliance, Holy Trinity Brompton, the Methodist Church, Moorlands College, Operation Mobilisation, the Salvation Army, Tearfund and World Vision.
Collection Procedures
An initial mailing to all the 38,000 churches on the Christian Research database took place in February to let churches know the Census would take place on 8 May 2005, and asked them to let the organisers know of any new churches that had been started in their locality or any that had closed. A huge and very helpful response to this letter was received. At the same time every senior leader (Bishop, Archdeacon, Vicar-General, Regional Minister, Chairman, Superintendent, Moderator, District Commander, etc.) was informed that the Census was taking place.

The crucial mailing took place in April 2005, with a covering letter indicating that the information provided would be treated as confidential. No information on any individual church would be released without its approval. Likewise churches were able to tick appropriate boxes to indicate whether or not they wished to receive future mailings,
according to the Data Protection Act.

The thousands of returned forms were then individually checked, and entered on to the Christian Research computers, an exercise that was finally completed in October 2005. The analysis then began. Almost three percent of English churches did not hold a service on 8 May, some 973 churches in total. This figure is likely to be understated since many churches not holding a service that day will not have replied. These were mostly churches in remoter rural areas whose practice is to hold a service every fortnight or month, some members of whose congregations will perhaps have attended a service in a neighbouring hamlet or village.

These churches that had a zero congregation on Census Sunday were not assumed to have closed, but rather were amalgamated with the other linked churches in recording the results. Thus if there were three linked rural churches, two of which had no service on 8 May, and the third church had 75 people who were a mixture from all three congregations, the church where the service was actually held was recorded as having 75 present and the other two as zero, but not closed. This means that the correct number is included in the overall total, and the size of that congregation is also correctly recorded, and not put as, for example, three congregations of 25 each.
Sampling Procedures
The Census covered all Christian denominations accepting the Trinitarian doctrine (that God exists as Father, Son and Holy Spirit in one being). It thus included all Free, Protestant, Anglican, Roman Catholic and Orthodox Churches, and excluded the Jehovah’s Witnesses, Mormons, Christian Scientists, Christadelphians and Unitarians. There was no coverage of non-Christian groups such as Jews, Muslims and Hindus. The actively religious people thus excluded represented four percent of the UK population in 2005.

A ‘church’ was defined for the purpose of this Census as a body of people meeting on a Sunday in the same premises primarily for public worship at regular intervals. The congregations approached thus included normal church buildings, and also those meeting for worship in school chapels, Armed Forces chaplaincies, and those in hospital chapels.
Congregations that meet on Saturdays (like the Seventh-day Adventists, or Catholics taking Vigil Mass) were counted in. Churches whose services are held fortnightly or monthly were also counted in, but not those used less frequently. Congregations that do not own a building but hire a local school or civic hall for their meetings were included. Some people meet for worship during the week, and these numbers were also counted, but not included in the Sunday total.

The study was limited to England; a similar attendance survey in 2002 covered Scotland. In common with previous studies, the fourth English Church Census included both the Channel Islands and the Isle of Man. It is of course recognised that these two territories are separate jurisdictions legally, but most of the major denominations include supervision of their churches within an appropriate Diocese or Region that is based in England. These two
territories are therefore analysed along with the 47 counties in England and included in the commentary where appropriate.

Altogether information was received from 18,720 churches or congregations by the time we stopped entering data on our computers, and details from a further 39 were received subsequently. Several dozen other churches returned details more than six months after the Census. Three-quarters of those supplying information returned one of our forms. We were also able, however, to input data on total attendance, sometimes with additional information, that was very kindly supplied by 10 Church of England and eight Roman Catholic Dioceses, the Baptist Union of Great Britain, the Fellowship of Independent Evangelical Churches, the Salvation Army and 91 Methodist Circuits for the remainder.

This gives a total overall response rate of 50.02%, a very much higher response to a postal survey than most research companies expect to achieve today. We are grateful to the army of people who completed the forms or helped us get so much information. That the response rate was so high again shows something of the interest in this Census. Half of the forms ultimately received were returned during the first six weeks; a reminder was sent out after two months, after which a few thousand came in quickly, followed by a long trickle over the following months.

In terms of the overall percentage, the 50 percent response achieved in 2005 compares to 33 percent in 1998, and to 70 percent in 1989. In the 2002 Scottish Church Census we had 52 percent.
Principal Investigators
Dr Peter Brierley
Christian Research (formerly MARC Europe)
Related Publications
Brierley, Peter William. 2006. Pulling out of the Nosedive: A Contemporary Picture of Churchgoing - What the 2005 English Church Census Reveals. London: Christian Research.

Brierley, Peter William. 2006. Religious Trends No 6 (2006/2007 edition). London: Christian Research.