The Strengths and Weaknesses of the Official Census
by K S Taylor
Queen Victoria reigned from 1837 to 1901. Early in her reign she discovered a serious concern with affairs of the state and domestic responsibilities. The typical Victorian City emerged, such as Birmingham and Manchester, along with rapid urbanisation, industrialisation and population growth The study of these is today becoming increasingly important to historians and genealogists alike They use various sources such as maps, commercial directories, poll books, registers, architecture and photographs, but, perhaps most importantly censuses
Census comes from the Latin phrase "to assess", and is the official and periodical counting and assessing of the people: demographically, socially and economically. In 1840 the responsibility was handed over to the GRO, General Register Office, resulting in a broader range of questions. The aim of the census was now to reveal the state of the nation and structure of the economy, in order to dispel radical discontent. The country was divided into registration districts in line with Poor Law Unions. The census was enumerated on one day in March or April every ten years. It was essentially a snapshot of society, intended to avoid problems of double counting. Data was collected by means of a standard printed household schedule completed by the householder or enumerator if illiterate. The information was then transferred into books and tabulated into report format.
The study of census data through location, evaluation, interpretation and processing is very useful in reconstructing Victorian society. The books remain the only complete analysis of the population and are particularly useful in identifying trends, shifts and aggregate numbers. In two respects, censuses are above average quality as historical evidence, in that they are universal, comprehensive and in standard format. Together with other contemporary material they help identify social structure, urban and economic history of the period. They bring new dimensions to previously neglected areas of study. Censuses can be used on different levels and scales, from individual family history, to a national study of spatial distribution. By 1941, the identification of people by name makes it possible to link information contained in censuses with other sources, such as parish registers and certificates. This enables a fairly exhaustive analysis of Victorian individuals, families and communities.
Farr has possibly achieved one of the most useful studies, showing the strengths of the census. He produced a series of census reports on levels of mortality, health was of primary concern to Victorians and seen as inversely related to population density. Farr constructed occupational life tables and theories of diseases based on the preconceived idea that health was related to occupation. Improvements in sanitation followed contributing in the lowering of mortality rates. Other conditions were also improved, such as overcrowding and, with interest in the social and economic structure of the nation, the introduction of factory acts. The primary uses of the census during and preceding the Victorian Era, tell us a lot about the concerns of government at that time. Indeed, changes and turmoil in Victorian times led to a conflict of interest over census management. Parliamentary, government, statistical, medical, charitable and social organisations, each made attempts to include suitable questions in connection with their concerns.
This in turn benefits us by providing a diverse source of information on key issues of health, housing, sanitation, employment and poverty.
However, there are, as with all sources, certain weaknesses, that need to be identified and taken into consideration. Problems stem from the data in manuscript returns, which is occasionally inconsistent or unreliable due to enumerators and householder, plus the problem of interpretation, questions were sometimes ill framed for such purposes There was concern for the increasing powers of the state; hence, a desire to conceal or distort certain information, even though it was agreed it should be confidential Records are indeed unavailable for a hundred years to protect such an agreement However, the ones that are available are several stages removed from the original household returns Changes in original data include factual corrections, additions to scripture and highlighted or marked information In reality, tabulated data has also become more of a summary, which was not intended.
The single enumeration district, in reality, contains several bureaucratic entities, as districts often retained outdated boundaries of previous censuses. Each locality varied according to individual circumstances, such as population changes, layout and efficiency of enumerators. Information set out on the page is, therefore, subject to the interpretation and beliefs of the enumerator and registrar. The information required changed, as acts were often introduced just before census taking. We must, therefore, make allowances for such variation and learn to spot two main weaknesses. Initially, errors of coverage, when figures were under or overstated and, secondly, errors of content, due to mistakes in reporting, recording and processing.
Registrars appointed enumerators aged between 18 and 65, They had to be literate, arithmetic, in reasonable health, as well as being, "respectable, orderly and temperate" Wages were insufficient to attract quality staff except those working out of social duty The responsibility fell mainly on local government officers, quality varied dramatically between districts, this was shown by the detail and perceived accuracy of their work Information was generally given from a lower middle class male perspective and may be blurred by a lack of judgement. Generally, analysis on the validity of returns suggests intelligent enumerators and a general willingness to co-operate.
The enumerator's task involved the distribution, collection and occasional completion of the household schedule and its recording. Special arrangements were made for inmates, armed forces, Intranets and night workers These arrangements were by no means consistent Some data was recorded in special institutional books, whilst others appeared among normal institutions. Certain confusion and distortion was caused in the analysis of occupation. For example, inmates recorded a trade even though they were not making a positive contribution to the economy. Also, aggregate numbers were distorted by "'intransit members of the community ", who were recorded over a period of time and often added onto the population of the nearest town.
Weaknesses can occur in the interpretation and collection of data. They are concentrated in three key areas: Housing, The Household and Individuals. Its "promise", however, is far more impressive than its "pitfalls". The interpretation of data relating to housing enables individuals to locate an ancestor or next of kin. In addition, the local historian can calculate the ratio of houses to people, indicating levels of overcrowding and building such was of particular interest in emerging Victorian towns.
The level of uninhabited buildings and, the number of people living in less than five rooms calculates spare capacity and determined the overall state of the housing market, which is a chief contributor to the economy. The social and economic historian may also be interested in; the layout of streets, social composition of cities and the extent of class mixing judged by the proximity of different classes. Statistics can be difficult to interpret, particularly in cities, as definitions of the words: "house", "uninhabited", and "room", were never adequate to form a consensus. These definitions and their recording changed both in detail and layout over time, being subjected to interpretation.
Records of people were taken where they were on census night but this cannot be assumed to be typical of their living arrangements. Indeed many buildings were subdivided to accommodate the increasing urban population. Many were clustered behind, or on top, of others, clouding overcrowding statistics. Census authorities were aware of problems, and privately admitted that the number of houses recorded and tenants were often "misleading and valueless". The use of other contemporary sources is, therefore, necessary, such as rate books, maps and street directories, helping with problems of misnumeration.
Addresses can be another useful commodity of study, especially for a family and social historian. No other source compares to the census for showing residential patterns and mobility in the city. In a period of rapid expansion though, the Victorians revolutionised the postal system.
This poses a problem for us, as we are dependent on the existence of proper addresses in 1847 the Town Improvement Clauses Act', led to the renumbering of houses and buildings by local authonties, although this was not on a uniform basis. The problem was acute in 'slum' areas, where boundaries of enumeration often split streets. The problem was compounded by the loss of certain numbers due to clearance, and individual builders erecting a small number of houses in their place.
The accuracy of class segregation is dependent on the enumerators entering the houses in the order they appear on the streets. The order to do this was not explicit and needs to be checked with other records. Indeed comparison of housing on a larger scale is far more justified than individual small-scale comparisons.
The second key study is that of the household, which gives a valuable insight into the workings of the Victorian family and is of unrivalled importance in the reproduction of population and social moves. Census information contains that of families. The family' was seen as an administrative artefact, a "non biological sum of relations with others in society" Such a narrow definition inevitably overlooks social and moral obligations, concerning the importance of relations between households
Victorian society is not solely, that, which was put down on the census. We should not be tempted to look at social connections, which are not implied by original definition. The question 'What is a family?' also caused problems for the enumerator, who had to identify the individuals involved 'Lodgers' and borders' posed particular problems and were often registered as part of a family, in reality many lived in quite separate social groupings The household being made up of a head, relatives, servants and apprentices did not exist exclusively, until the last census of the Victorian era. The ideal family was also reflected in the census, often failing to exist in working class urban areas, with other frequent social obligations.
For the purposes of historical research it is probably easier to view a household from head" to the next head, however, this does not remove deficiencies in original data Notice should be taken of schedule numbers and how enumerators interpreted instructions.
Individual information is mainly one of the most compelling and useful areas of interest in the Victorian City. The names are usually reliable, apart from variations in phonetic spelling.
However, these can usually be spotted by coincidence of other personal details and may be reflected due to the copying errors and interpretation of various enumerators. Usually, data gives a reliable approximation of the features of Victorian society. More of a problem is posed to the family historian who is trying to trace an individual Usually discrepancies in aggregate data are within reasonable limits and often cancel each other out However, certain entities were "glossed over" as a result of poorly trained officials, illiterate householders and occasional deliberate misgivings. Age is a key area of discrepancy and only a very rough guide to nineteenth century reality; it often reflected a general ignorance to the precision of information required and levels of illiteracy. The majority of ages strangely appear to be consistent from census to census with suggestions that only five percent in each parish were inconsistent by more than two years The greatest disparity occurs with children under fifteen, who were classified by the lowest term of five years within which their age fell and was further confused by a tendency to think in terms of their next birthday; in addition there was a deliberate attempt to conceal child labour in the family. Frequently unnamed and unchristened babies often went unrecorded. A few extra years was added by older women in domestic service and working class men in their fifties, hoping to obtain better wages and poor relief in old age. These deliberate misgivings bring into question the belief in confidentiality because people were suspicious of local government officers Sensitive information may be somewhat falsified, however, the actual ability to fill out information correctly was helped from 1870, by the Compulsory Education Act. 23 Terms such as 'relation to household head' were clearly placed in a different context to the use today and may cause some confusion over the exact status of individuals. The head as a social position did not necessarily reflect "biological descent". In reality the meaning of 'kin relations' differs and the term 'in law" was frequently omitted. In an attempt to fit into the Victorian ideal there was confusion about the word "servant", which included, apprentices, shop/ domestic workers and agricultural labourers, as well as concealing "unconventional relationships". Marital status also produced some doubtful statistics, so much so that in 1881, the number of wives exceeded the number of husbands by 61.064. The number of women who were engaged, under twenty accounted for this and mature separated women, who regarded themselves as married.
This confused the overall picture of family life, but now gives an insight into the thoughts of the Victorians and accepted normality.
Of paramount importance is birthplace data, particularly for genealogists and in the study of population, although there are few sources to check against and little work has been done on the accuracy of the data. Since 1851 people have been asked to indicate the country and town or parish of birth. Inevitably minor variations exist; the reasons are mainly due to: spelling anomalies, changes in parish boundaries, poor grasp of geography and the recording of the earliest residence remembered. Also there was a tendency to try and prove settlement in a parish, in order to obtain eligibility for Poor Law Relief, this in turn, reflects key issues of poverty and dependency in Victorian cities. It could be said that census records indicate a lower level of migration to the reality. Censuses are also limited in recording the final migration, rather than the number of movement's in-between. Indeed Anderson, in his study of Preston, suggested fourteen percent of the population showed discrepancy between two years, in half the cases, migrants had become non migrants and vice versa. This could show the lack of care of the enumerator, or householder. Often foreigners would fail to understand complicated instructions, and refugees falsified returns to avoid further persecution. Although such a source of information is vital in depicting the "make up" of Victorian cities and the urbanisation of the population, this could not be done so completely by another source, providing a series of snapshots' of: the condition and location of the entire population, at regular intervals.
In contrast, medical disabilities, as a record, are generally unreliable. The census authorities were very much aware of deficiencies and would no longer defend the level of accuracy but stated that the "levels of inaccuracy (were) consistent throughout the country", returns could therefore, be used for comparative purposes. Defects reflected the unwillingness of parents to confess to disabilities or even be ignorant of them, particularly relevant with younger children. Householders were, clearly unsure what sort of data was required, with entries such as "unwell". Aggregate analysis must, therefore,
be confined to comparison, and not as an indication of the extent of medical disabilities suffered by Victorians.
Finally, occupations are of fundamental importance in reconstructing Victorian society, providing the most detailed statistical study of the economy and social structure. Data has been used, primarily, as a basis for social classification' by the Registrars General Scheme and, aggregated into 'industrial categories' by Booth, showing the economic structure of England and Wales, during the Victorian era. Social classification is based, however, on the assumption that households can be assigned to specific social classes, according to the occupation of the head. This is particularly inaccurate amongst working class households, where prosperity was determined by family income. It may also be anachronistic to apply Armstrong's twentieth century classification, to Victorian data. Clearly census statistics must be used with sources such as rent books and trade statistics for verification. During that time, the concern was not economic but medical, with the belief that materials worked on affected the character and life expectancy of the individual. But today, however useful, we face a number of weaknesses in the data collected.
Primarily, until 1891, the term rank, profession, occupation', led people to think in social rather than economic terms. More serious problems, however, are caused by inclinations to omit information, or even falsify returns, usually with seasonal and casual labourers. Women were rarely recorded with much interest, as it was believed their place was in the home. The term 'domestic servant', covered a multitude of tasks. Children also suffered a similar bias and were often incorrectly described as a 'scholar', due to the pressure of factory and education acts. Census figures are incomparable to enrolment figures in schools, and many other scholars were not confined to one task. This is shown by school absences on Mondays and Fridays, also during harvest time. We must also look at school surveys. The General Register Office s interest also proved disappointing with the growth in the complexity of employment relations and a belief that many had inflated their true social position. Care must also be taken with abbreviations used by enumerators, particularly in the occupation column, for example, M' refers to manufacturers, whereas lower case m' implies a maker, such as a shoemaker.
In conclusion, all categories of historical evidence are, more or less, liable to inaccuracy. In certain respects, being universal, comprehensive and in standard format, census material is above average. We must be aware of certain weaknesses in the collection and interpretation of data and the analysis of census returns only gives us a picture of the Victorian City population, not necessarily the reasons behind it. Censuses do not provide indisputable facts; instead need to be used in conjunction with other sources. However, the strength of the census as a source for present day and the future, far outweigh its weaknesses.
Encarta Encyclopaedia “Census”
Higgs E, “Making Sense of the Census”
Lawton R,” The Census and the Social Structure”
Lumas S, “Making Use of the Census”
Knodel J, “Local Population Studies”
Szreters, “Family and Class Gender in Britain”
Skipp, “The Making of Victorian Britain”