The 'Empty' Church Revisited

Robin Gill

• Publisher: Routledge
• Year: 2018
• ISBN:
  9781351775984
• Edition: 2
• Topic: Theology & Religion
• Subtopic: Religion

This title was first published in 2003. When did churches start to appear more empty than full - and why? The very physicality of largely empty churches and chapels in Britain plays a powerful role in popular perceptions of 'religion'. Empty churches are frequently cited in the media as evidence of large scale religious decline. The Empty Church Revisited presents a systematic account of British churchgoing patterns over the last two hundred years, uncovering the factors and the statistics behind the considerable process of decline in church attendence. Dispelling as myth the commonly held views that the process of secularization in British culture has led to the decline in churchgoing and resulted in the predominantly empty churches of today, Gill points to physical factors, economics and issues of social space to shed new light on the origins of empty churches. This thoroughly updated edition of Robin Gill's earlier work, The Myth of the Empty Church, presents new data throughout to explore afresh the paradox of church building activity in a context of decline, the patterns of urbanisation followed by sub-urbanisation affecting churches, changes in patterns of worship, and changes within the sociology of religion in the last decade.

1

The Data

Until about ten years ago sociologists of religion tended to pay more attention to sects and to new religious movements than to long-established denominations or churches. With a few important exceptions at the time,¹ mainstream churches seldom received the sort of detailed attention that one might expect given their prevalence and greater size. In part this may have resulted from lingering suspicions of ‘Religious Sociology’ and to the ecclesiastical control often thought to lurk behind it in France. It may also have been a consequence of the sheer difficulty of analysing amorphous religious institutions. Small scale religious bodies, in addition to being frequently more exotic and deviant, are also perhaps more sociologically controllable. Whatever the reason, long-established denominations until recently remained surprisingly unresearched.

Even when sociologists did show an interest in long-established denominations they typically relied upon generalized data. As has been noted already, Churches and Churchgoers² invaluable as it may be as a source of historical information about national church membership, is notoriously lacking in information about Sunday-by-Sunday churchgoing and tends to put forward factors deemed responsible for church decline in a highly impressionistic manner. And the vast literature on secularization tended to avoid statistical data altogether; or else it used both statistics indicating church decline in Europe and statistics showing persisting (but supposedly epiphenomenal) churchgoing in the United States as indications of secularization.³

In contrast, it was initially a number of younger social historians who showed an interest in churches as social phenomena. Stephen Yeo’s detailed study of Reading in the nineteenth and early twentieth centuries, Religion and Voluntary Organisations in Crisis,⁴ proved to be a pioneer work and is well known to sociologists of religion. So did Hugh McLeod’s Class and Religion in the Late Victorian City.⁵ Rather less well known is Jeffrey Cox’s The English Churches in a Secular Society: Lambeth, 1870-1930,⁶ James Obelkevich’s Religion and Rural Society: South Lindsey 1825-1875,⁷ and Callum Brown’s The Social History of Religion in Scotland Since 1730.⁸ Together with a number of research dissertations and doctorate theses that they have often served to inspire,⁹ they suggested that there is considerably more statistical evidence available on churches as social phenomena than is often imagined, and that it is directly relevant to understanding the decline of most British churches throughout the twentieth century. Far from being generalized studies, they each show that an intense analysis of churches in specific areas yields insights that cannot be deduced from national church membership statistics.

This use of church statistics has dramatically changed among sociologists of religion in the last few years. I will not dwell on this since I have analysed it elsewhere,¹⁰ but it is worth noting the impact of recent works using a wide range of statistics such as Grace Davie’s Religion in Modern Europe: A Memory Mutates,¹¹ Steve Bruce’s God is Dead: Secularization in the West,¹² and again Callum Brown in his The Death o f Christian Britain.¹³

1

The accumulative effect of these studies has been to undermine the approach championed by Currie, Gilbert and Horsley’s Churches and Churchgoers. The latter argued that churchgoing statistics collected through various local censuses ‘have contributed little to the formation of time series’.¹⁴ Instead, they insisted that national church membership statistics are the most valuable resource:

But religious statistics [of local attendance] are, however useful, of secondary importance compared with those collected by the churches themselves. Just as, for example, it would be impossible to study the growth of trade unions or political parties without reference to their own membership series, it is impossible to study church growth unless the churches’ own membership returns are used; and these returns are probably unique among those available for organizations in Britain, both in quantity and in variety.¹⁵

Although this claim might at first seem plausible, on closer inspection it faces a number of formidable difficulties. Not the least of these is that the very concept of ‘church membership’ is a Free Church concept. Comparison of, say, varieties of Methodist membership at a single point of time might have real value. In all probability they shared an understanding of ‘church membership’ and even had similar procedures for measuring this membership. But the Church of England has never had a comparable concept of ‘church membership’, nor has the Catholic Church.

This is an obvious, but very long-standing, deficiency. Those attempting to compare the relative strengths of British denominations have been struggling with it for more than 150 years. In what has been termed the first specifically English survey of the relative strength of denominations, the Congregational Union survey of 1834 distinguished between ‘hearers’ and ‘communicants’.¹⁶ However, in the process the survey clearly imposed Free Church concepts on other denominations. Although, as is seen in Chapter 5, the evidence produced in the Congregational Union survey is not worthless, it does have to be approached with considerable caution.¹⁷

To remedy this deficiency, Churches and Churchgoers treated Easter Communions as indicators of Church of England ‘membership’ - on the basis, presumably, that Easter Communion is supposed to be a requirement of belonging. In reality, despite many attempts by bishops and clergy to make it a requirement (questions about communicants have continuously featured in bishops’ questions to their clergy since the eighteenth century), Easter Communion has not been a serious indicator of Anglican belonging, or more accurately ‘conformity’, since the seventeenth century. Even then, seasonal ‘conformity’ was compatible with regular attendance at ‘dissenting’ chapels.

In the pages that follow abundant evidence is provided to show that Easter communicants in the Church of England represented a very small proportion of regular attendances in both rural and urban areas in the mid-nineteenth century. Furthermore, the ratio between regular attendances and Easter communicants show considerable local variations today. In rural areas of England, Easter communicants represent as much as three times the average Sunday congregation; in urban areas they are seldom more than half as much again. Thus, Easter communicants vary as an indicator of Anglican belonging both over time and within time - or, to express this more technically, variations are diachronic as well as synchronic.

Confusion about how the notion of ‘membership’ applies to Catholics is even more obvious. Churches and Churchgoers relies primarily upon ‘estimated Catholic populations’. These are based on estimations by local Catholic priests of the number of Catholics living in their parishes, regardless of whether or not they are practising Catholics. Often estimated from baptisms, they are, in most Free Church understandings, affiliation rather than membership statistics. Aware of this difficulty, early versions of MARC Europe’s UK Christian Handbook¹⁸ relied instead upon regular Mass attendances (measured on the last Sunday every October) to produce statistics that can be compared with ‘membership’ in other denominations. Unfortunately this approach confuses membership and churchgoing statistics.

It is evident from this that ‘church membership’ is a very odd concept for the two largest denominations in England and Wales. Because a number of Free Churches (especially Methodists) have kept national ‘membership’ figures for such a long period of time, it is understandably tempting to try to squeeze Anglicans and Catholics into the same mould. However, in so doing a minority criterion is transformed into a yardstick for measuring all. As a result, some very curious statistics emerge. Sadly, it cannot be considered to be a serious means of assessing relative interdenominational strengths and weaknesses.

Membership may not even be a very consistent measure of Free Church belonging. Later I show that the ratio of members to quarterly Communion attenders in a number of Free Churches changed significantly over time. The twentieth century has seen a widening gap between official communicant membership rolls and numbers of people actually attending seasonal communion services. Furthermore, the urban churchgoing statistics reviewed later suggest that national Methodist membership statistics at the Union of 1932 considerably exaggerated the strength of Methodism at the time. Many chapels already had a higher proportion of largely dormant ‘members’ than they would have had in the mid-nineteenth century.

What emerges from this, I believe, is the realization that an intense study of churches at a local level and over a sufficient period of time (using records of local attendances rather than national membership figures) is a prerequisite for investigating the social factors underlying church decline.

This would be a standard premise in the sociological study of sects or new religious movements. Ever since the 1950s, and in particular Bryan Wilson’s pioneering Sects and Society,¹⁹ the detailed local study of small-scale religious bodies has been preferred to generalized discussions of them as national, or international, institutions. Presumably this is based upon the realization that national information is often partisan and may not accurately represent the way a religious body functions in practice.

But this applies a fortiori to churches. The claims that a church makes at a national level may or may not be based upon the way it functions in practice at the local level. Indeed, part of the skill of the social scientist involves comparing claims with actual behaviour. To make such comparisons there is no substitute for detailed empirical research; and it has, perhaps, been the most obvious failing of the literature on secularization that it has seldom been based upon such research.

2

In this book I examine a great deal of data, much from local censuses and most previously uncollated. It is astonishing that so many generalizations about church decline and secularization were made without anyone surveying this evidence before. To be honest, I had little concept myself of its extent until I started this research in the late 1980s. I had assumed for years that longitudinal statistics about regular churchgoing patterns were largely unattainable. Certainly I had assumed that little was available outside the three nationwide censuses of church attendances: namely, the 1851 Religious Census,²⁰ and the 1979 and 1989 (and now the 1998) MARC Europe/Christian Research censuses.²¹

In fact, there is a wealth of data from at least six major sources, as well as an invaluable source guide to most of the data produced by Clive Field.²² There is the 1851 Religious Census which covered the whole of England and Wales and (with more gaps and no surviving original returns) Scotland. There are independent newspaper censuses for the 1880s, and occasionally for the 1890s and...

Table of contents

• Cover
• Title Page
• Copyright Page
• Contents
• Preface
• Introduction
• 1 The Data
• 2 Rural Churchgoing Decline
• 3 Glan-llyn: A Rural Case Study
• 4 Urban Churchgoing Decline
• 5 General Increase: 1821 to 1851
• 6 Free Church Increase and Anglican Decline: 1851 to the 1880s
• 7 General Decline: 1880s to 1919
• 8 Continuing Decline: 1920s to 2000
• 9 York: An Urban Case Study
• 10 The Future of the Churches
• Tables
• Index