Making Sense of the Census Revisited. Census Records for England and Wales, 1801–1901. A Handbook for Historical Researchers
London, Institute of Historical Research, 2005, ISBN: 300100396X; 244pp.; Price: £15.00
University of Edinburgh
Date accessed: 29 June, 2015
Eddy Higgs’s work on the census is much valued, not least because he is both a working, researching and publishing historian as well as an experienced archivist. This is a revised and extended version of an earlier book, and readers will recognise much from earlier versions and points Higgs has made in a variety of locations. However I want to review this as if it were a new book. This is because it should be read not only by people like me who ought to have listened more carefully when we were first told these things, but because the main audience ought to be the great numbers of graduate historians, community and family historians who come to the census enumeration manuscripts as the Holy Grail of modern history. The book needs careful reading and has implications far beyond the census itself. Indeed, it should have a place in the reading of the increasing number of ‘training’ courses for historians, for the book has lessons not just for those who would use the documents, printed and archived, of the census in England and Wales, but also for those who might use a wide variety of particular instance papers. The careful account of the creation and organisation of the records, their interpretation and analysis, and of issues of research design could be applied to many structured and semi-structured record types, such as wills, court records and poor law registers.
Higgs continually makes the reader aware that the census was never and could never be an objective account of the population. His account of the background makes brief reference to issues of ‘governmentality’ and ‘liberal citizenship’ but in the main the historian is invited to see the creation of the census with reference to very practical matters: the military strength of Britain, worries over the poor and poor law, the needs of actuarial science and of public health enquiries, and later of eugenics.
A large part of the book focuses on the problems of interpretation and the imperfections of the census. Sometimes people simply got things wrong. Many did not know their age or place of birth. Others would mislead, thinking that the document might be used to check up on a child working under age. Most interesting is and was the struggle of the census makers to place their information within ideological constraints of value and perception. The notion of a ‘house’ was fine in a middle-class terrace or even a back-to-back, but in tenements and sub-divided properties the notion of a ‘house’ became very flexible. Although Higgs feels that distinctions were easier in the countryside, the presence of living-in farm servants, especially in the north of England, probably created equally porous boundaries between house and house. The same went for family and household. It was not until 1861 that the notion of commensality was used to try and bring some order to the boundaries of families and households. Even this was an artefact of the census. We need to extend this discussion as our notion of family and kin is still very much dominated by census and household studies. My recent studies of wills, for example, indicated that the notion of the extended family was crucial to individual life course strategies but was combined with rather than being antithetical to nuclear family based strategies.(1) The census record can be a prison as much as a guide for the historian of family and household structures and cultures.(2)
A large and valuable part of this book is full of warnings of the imperfections and potential distortions of the census record. This is especially important as the census records still carry with them the aura of the statistical movement with its search for facts and objectivity. At the same time it is important not to be mesmerised by this. It is part of the historian’s task, with any class of record, to think through the issues raised by the distorting lens of the record and evaluate the risks taken. The social scientist’s notion of concept and indicator needs to be brought forward here.
This is especially true in the section on occupation and occupational classification. More emphasis needs to be given to the manner in which the census record does not give an occupation, it attributes an occupational title to an individual, perhaps as a result of a claim by that individual or a ‘head of household’. As Higgs points out this raised all sorts of issues, especially in relationship to seasonal work, multiple occupations, gendered ideologies of women’s occupational status as well as the meaning of the title ‘servant’. Although considerable work has been done on occupational titles since the 1970s, the issues raised again need to be taken further. When nominal information from a variety of records is linked, for example, from commercial directories, poll books as well as census records, it is clear that many records have a high probability of relating to the same person (all nominal record linkage is probabilistic), but show different occupational titles. The merchant or the solicitor became a ‘gentleman’ and the cabinetmaker became a ‘second hand furniture dealer’. The very necessary task of occupational categorisation or classification makes matters even more complex. The printed census itself provided some tempting help in the printed reports, but moved from groupings based upon materials worked with, to one based upon industrial output and upon a concept of status closely related to eugenic enquiries about fertility and death. Higgs offers a sympathetic critique of the so-called Booth-Armstrong classification which is still often used in local studies. We have come a long way since the 1970s in understanding occupational titles and classifications. It is crucial that the historian begins the choice of classification by defining his or her own purpose and interest, and by admitting that some sort of multi-dimensional coding may be required to get best value from the information implied by the occupational titles.(3) Only then should the possibility of adopting a widely used existing system be considered for comparative purposes.
There are many other goods things in this book. There is a list of the days on which the census was taken between 1841 and 1901 and a nice lecture on the administrative and geographical areas used in the census. All very tedious until the historian needs to know. If the impact of seasonality is a concern, which census was taken in June? (The answer is on p.167.)
This is an invaluable and very welcome book. It will be widely used for reference and in teaching. No-one should use the printed or manuscript records with any seriousness without having a copy.
- R. J. Morris, Men, Women and Property in England, 1780-1870 (Cambridge, 2005).Back to (1)
- See Leonore Davidoff and others, The Family Story. Blood, Contract and Intimacy, 1830-1960 (1999), esp. pp.1-50, for an account of the break out from this 'prison'. The work of anthropologists on the fluidity of notions of family and kin is also crucial, see ed. Janet Carsten, Cultures of Relatedness (Cambridge, 2000).Back to (2)
- R. J. Morris, ‘Occupational Coding: Principles and Examples’, Historical Social Research/Historische Sozialforschung, 15 (1990), pp.3–29; The Use of Occupation in Historical Analysis, ed. Kevin Shürer and Herman Diederiks (Max Planck-Institut für Geschichte, St Katharinen, 1993).Back to (3)
The author would like to thank Professor Morris for this review and does not wish to comment further.