The evaluation of baptism as a form of birth registration through cross-matching census and parish register data: A study in methodology

Abstract

A method of evaluating the accuracy of Anglican baptism as a form of birth registration during the period 1760-1834 is outlined and discussed in the light of the substantive application of the method to a sample of 45 parishes. The method fundamentally consists of comparing information in the 1851 Census with that in baptism registers for each individual parish. A first stage of the research tests an assumption that the 1851 Census is a reliable source of information on birthplace and age statements, and this is achieved primarily through comparisons of the 1851 and 1861 censuses for individuals in each parish. Subsequent research reported in the Appendix outlines a near-perfect method of testing this assumption: the comparison of 1851 census data with information about births collected under civil registration, which when applied to the parish of Kingston-on-Thames further confirms the reliability of 1851 census statements on birthplace and age. Having established this, an analysis of the underregistration of births in the baptismal registers is undertaken by comparing expected births according to the 1851 Census with actual births registered in the parish registers for each parish in the sample. Certain substantive conclusions are reached from this cross-matching process: there was little change in the reliability of baptism as a form of birth registration during the period 1760-1834, although there were marked variations from one parish to another. Estimates of birth and death rates for England and Wales were constructed on the very tentative basis of taking the omission rate of births from baptisms in the sample as typical of the country as a whole, and although the sample was not a random one and there were a number of uncertainties about the absolute degree of reliability of Census data, certain independent checks suggest that the findings from the sample are in fact fairly representative. About one-third of all births were omitted from baptism registration during the whole period and there is some evidence that this figure was also applicable to much earlier periods; literary evidence indicates that most underregistration was a function of the very inadequate process of registering baptisms and burials within the Anglican Church. Further research using the methods outlined in this paper is required in conjunction with family reconstitution techniques before definitive conclusions about the demographic history of the period of the Industrial Revolution can be reached. © 1972 Taylor & Francis Group, LLC.