¹ An overview of this work may be found in Coale, A. J. and Watkins, S. C. eds., The decline of fertility in Europe (Princeton, 1986)Google Scholar. Findings for Britain are published in Tietelbaum, M. S., The British fertility decline: demographic transition in the crucible of the Industrial Revolution (Princeton, 1984).CrossRefGoogle Scholar

² See, for instance, Anderton, D. L. and Bean, L. L., ‘Birth spacing and fertility limitations: a behavioural analysis of a nineteenth century frontier population’, Demography 22, 2 (1985)CrossRefGoogle Scholar, Knodel, J., ‘Starting, stopping, and spacing during the early stages of fertility transition: the experience of German village populations in the eighteenth and nineteenth centuries’, Demography 24, 2 (1987)CrossRefGoogle Scholar or David, P. A. and Sanderson, W. C., ‘Measuring marital fertility control with CPA’, Population Index 54, 4 (1988)CrossRefGoogle ScholarPubMed. The latter cites several works based on the Stanford Project on the History of Fertility Control, a project which has contributed a great deal of knowledge to the ‘stopping’ versus ‘spacing’ debate.

³ Child spacing, by means of extended breastfeeding and post-partum abstinence, in order to better preserve infant health and welfare was until very recently a reportedly common practice in certain parts of Africa. ‘Modernization’ is now breaking down the practice, leading to increased fertility rates in some areas. See Bongaarts, J., Frank, and Lesthaeghe, R., ‘The proximate determinants of fertility in Sub-Saharan Africa’, Population and Development Review 10, 3 (1984).CrossRefGoogle Scholar

⁴ See, for instance, Anderton, and Bean, , ‘Birth spacing and fertility limitation’.Google Scholar

⁵ See, for instance, Tilly, C. ed., Historical studies of changing fertility (Princeton, 1978)Google Scholar, particularly the chapter by Easterlin, R. A., ‘The economics and sociology of fertility: a synthesis’, pp. 57–133.Google Scholar

⁶ For a discussion of the uniqueness of the French experience see van de Walle, E., ‘Alone in Europe: the French fertility decline until 1850’Google Scholar, in Tilly, , Historical studies of changing fertilityGoogle Scholar or Wrigley, E. A., ‘The fall of marital fertility in nineteenth-century France exemplar or exception’, European Journal of Population 1 (1985)Google ScholarPubMed; also in Wrigley, E. A., People, cities and wealth (Oxford, 1987).Google Scholar

⁷ Coale, and Watkins, , The decline of fertility in EuropeGoogle Scholar. See also van de Walle, E. and Knodel, J., ‘Europe's fertility transition: new evidence and lessons for today's developing world’, Population Bulletin 34, 6 (1980).Google ScholarPubMed

⁸ The whole debate as to whether birth control was a completely new innovation, or an adjustment in behaviour utilizing previously known but unused methods, was opened by Carlsson, G. ‘The decline of fertility: innovation or adjustment process?’, Population Studies 20 (1966).CrossRefGoogle ScholarPubMed

⁹ Notestein, F. W., ‘Population: the long view’, in Schultz, T. T. ed., Food for the world (Chicago, 1945).Google Scholar

¹⁰ van de Walle, and Knodel, , ‘Europe's fertility transition’.Google Scholar

¹¹ Banks, J., Prosperity and parenthood (London, 1954).Google Scholar

¹² Becker, G. S., ‘An economic analysis of fertility’, in National Bureau for Economic Research Demographic and economic change in developed countries (Princeton, 1960)Google Scholar. In England and Wales the first Education Act was passed in 1870, but it was not until a further act was passed in 1880 that the full impact of compulsory education was felt.

¹³ The argument that higher income groups would adopt fertility limitation earlier than their social inferiors, as they would feel the conflict between the expense of child-rearing and conspicuous consumerism at an earlier stage in the process of industrialization, can be traced from the writings of Becker and Leibenstein. See Woods, R. I., Theoretical population geography (London, 1982), 102–10Google Scholar for a succinct overview of their arguments. See also Carlsson, , ‘Innovation or adjustment process’.Google Scholar

¹⁴ Nissel, M., People count: a history of the General Record Office (London, 1987).Google Scholar

¹⁵ Wohl, A. S., Endangered lives: public health in Victorian Britain (Cambridge, 1983).Google Scholar

¹⁶ Nissel, , People count.Google Scholar

¹⁷ The results of the 1911 census were published in two parts, the intervention of the First World War delaying publication of Part 2 until 1923. The full references are: Parliamentary Papers, Census of England and Wales, 1911: fertility of marriage, vol. 13, pt. 1, 1917–1918 (Cd. 8678), 36Google Scholar. Registrar General of Births, Deaths and Marriages, England and Wales, Census of England and Wales, 1911. Fertility of marriage, vol. 13, pt. 2 (London, 1923).Google Scholar

¹⁸ See, for example, Roberts, E., A woman's place: an oral history of working class women, 1890–1940 (Oxford, 1984).Google Scholar

¹⁹ Longmate suggests that the factories were replacing the church as a mass meeting place for women who would otherwise have had a constrained social circle centred on the home. Mill yards, he further argues, would have allowed far more secular topics of conversation than those countenanced in church pews. See Longmate, N., The hungry mills (London, 1978).Google Scholar

An observer in 1832, Dr James Blundell, also believed that factory employment assisted the dissemination of information. ‘Where individuals are congregated in factories,’ he commented, ‘I conceive that means preventative of impregnation are more likely to be generally known and practised by young persons.’ Quoted in Roberts, E., ‘Working wives and their families’, in Barker, T. and Drake, M., Population and society in Britain 1850–1980 (London, 1982), p. 153Google Scholar. Dr Blundell was giving evidence before the Committee on the Factories Bill of 1832.

²⁰ Hewitt, M., Wives and mothers in Victorian industry (Connecticut, 1958).Google Scholar

²¹ There are many volumes dealing with the details of worsted production. A selection includes: James, J., A history of the worsted manufacture in England (London, 1968 reprint)Google Scholar; Jenkins, J. G. ed., The wool textile industry in Great Britain (London, 1972)Google Scholar; Johnstone, C., ‘The standard of living of worsted workers in Keighley during the nineteenth century’ (unpublished Ph.D. thesis, University of York, 1976)Google Scholar. Worsted manufacturing processes were quite distinct from those employed in the manufacture of woollen cloth; Lawson, J.Progress in Pudsey (Sussex, 1978 reprint) describes the latter in some detail.Google Scholar

²² See, for example, Thompson, E. P., The making of the English working classes (Harmondsworth, 1963)Google Scholar. Smith, G. M. ‘Robert Clough Ltd., Grove Mill, Keighley: a study in technological redundancy 1835–1865’ (unpublished M.A. thesis. University of Leeds, 1982)Google Scholar discusses the causes and effects of men's reluctance, and also inability, to obtain jobs in the new ‘manufactories’ of the particular town dealt with in this paper.

²³ Feather, G. A., ‘A Pennine worsted community in the mid-nineteenth century’ Textile History 3, 66–91.Google Scholar

²⁴ Johnstone, , ‘Worsted workers in Keighley’.Google Scholar

²⁵ Johnstone, , ‘Worsted workers in Keighley’.Google Scholar

²⁶ The census enumerators' books need to be treated with some caution as they are subject to errors and omissions, both at the time of compilation and during interpretation. We cannot tell, for instance, how many married women had no occupation returned while they were actually out at work, nor do we know how many may have returned an occupation despite having left work for a period of time. Women may also have had part-time, or casual, employment which was not deemed, either by the women themselves or by the enumerators, to be ‘an occupation’. Examples might include baking, brewing, dressmaking, cleaning, taking in lodgers, childminding, washing, as well as ‘home-work’ such as carding buttons or making matchboxes or artificial flowers. See Lawson, , Progress in PudseyGoogle Scholar and Roberts, E., Women's work 1840–1940 (Basingstoke, 1988).CrossRefGoogle Scholar

In times of high unemployment men too may have returned their last occupation rather than the fact that they were unemployed, although the Keighley enumerators' books showed a marked increase in the proportion of men listed as being ‘labourers’ or having no occupation in times of economic stress.

A further problem lies in the fact that ‘occupation’ does not always indicate ‘industrial sector’. Textile mills often employed carpenters, carters and mechanics to ensure the smooth running of their production processes. Without information on place of employment it was not possible to assign all those actually employed in the mills to the ‘textile’ category.

The percentages of males and females employed in the various occupational categories used in this paper are subject to these caveats.

These, and other problems associated with the use of the census as a data source, are discussed in Lawton, R., The census and social structure (London, 1978)Google Scholar and Wrigley, E. A. ed., Nineteenth century society: essays in the use of quantitative methods for the study of social data (Cambridge, 1972).CrossRefGoogle Scholar

²⁷ The fertility measures used here were calculated following the methods outlined in the following works: Coale, A. J. and Trussell, T. J., ‘Model fertility schedules: variations in the age structure of childbearing in human populations’ Population Index 40 (1974)CrossRefGoogle ScholarPubMed, Coale, A. J. and Trussell, T. J., ‘Technical note: finding the two parameters that specify a model schedule of marital fertility’, Population Index 44 (1978)CrossRefGoogle ScholarPubMed and Grabill, W. H. and Cho, L.-J., ‘Method for the measurement of current fertility from population data on young children’, Demography 2 (1965)CrossRefGoogle Scholar. Some modifications to the methods were made and these are discussed in Garrett, E. M., ‘Before their time: employment and family formation in a northern textile town, Keighley, 1851–1881’ (unpublished Ph.D. thesis, University of Sheffield, 1987).Google Scholar

²⁸ Hinde, P. R. A. and Woods, R. I. argue in their article, ‘Variations in historical natural fertility patterns and the measurement of fertility control’, Journal of Biosocial Science 1, 6 (1984)Google Scholar, that the parameters of Coale and Trussell's standard schedule of marital fertility do not fit accurately to nineteenth-century British experience. They therefore proposed a new standard schedule which they claim lies at a more realistic level. The results of the Keighley study certainly support this claim.

²⁹ Garrett, , ‘Before their time’.Google Scholar

³⁰ It must be remembered that infant and child loss are ‘invisible’ in the census. Although calculation of the fertility measures includes an inflation factor to ‘replace’ children not appearing in the census, this inflation factor is assumed to be the same for all classes and occupational groups and for all age groups of wife. It is much more likely that lower-status textile workers and their wives suffered greater infant and child mortality than their higher-status workmates. If this were particularly so in the younger age groups, then levels of M amongst the former group might well be lower than amongst the latter. In the discussion of child spacing which follows, allowance should be made for the fact that the pattern observed may be the result of high mortality in the very young age groups rather than of deliberate ‘spacing’ of children by the parents. The implications for a couple's labour market participation might well be the same, but through ill-luck rather than carefully considered strategy.

³¹ Hewitt, , Wives and mothers.Google Scholar

³² Roberts, , A woman's place.Google Scholar

³³ Roberts, , A woman's placeGoogle Scholar, gives examples of women working in order to supplement their husband's income, or the money supplied for housekeeping by their husbands. Tilly, L. A., Scott, J. W. and Cohen, M., ‘Women's work and European fertility patterns’, Journal of Interdisciplinary History 6, 3 (1976)CrossRefGoogle ScholarPubMed suggest that in many working-class families the earning power of daughters was also of considerable importance to the household budget.

³⁴ It is unlikely that this hypothesis will ever be proven although circumstantial evidence, such as that provided by Hewitt, in Wives and mothersGoogle Scholar, suggests that couples often did find time together hard to come by. In A woman's place Roberts quotes one of her respondents as saying: ‘I'm certainly not saying we tried to get any more family. We certainly didn't do anything to stop it. But we never tried very hard’, suggesting that low coital frequency was playing a part in the reduction of fertility.

³⁵ The original categorization of husband's occupation was based on that suggested by Armstrong, W. A., ‘The use of information about occupation’Google Scholar, in Wrigley, ed., Nineteenth century societyGoogle Scholar; including ‘industrial grouping’ and ‘social rank’. Amendments made to Armstrong's scheme are detailed in Garrett, , ‘Before their time’Google Scholar. The six broad groupings used here are condensed from more specialized occupational categories. The working-class groups have not been distinguished by skill level.

³⁶ Figures 8A, B and C are all read in the same way. The larger, heavily outlined squares refer to the husband's occupational group and may be read as in Figure 7; the columns representing the situation in the first census in which a couple are identified, the rows showing what had become of the men 10 years later. The ‘unsure’ and ‘dead’ categories of Figure 7 have been amalgamated into one ‘husband not present’ category in the matrices of Figure 8. Again the shaded squares show where men appear to have remained in the same occupational group over the intercensal interval. Looking down the ‘other trades’ column in Figure 8A, for instance, we can see that 47 men were in this category in 1851. By 1861 four of these men had moved into the white collar category, four had become metal-mechanical workers and six had been lost from view. On looking along the ‘other trades’ row it becomes apparent that more men moved into this category than out of it over the 1850s – by 1861 58 of the husbands were included. The intersection of the columns with the ‘other trades’ row indicates that one man entered from the white-collar sector, two men came from metal-mechanical work, two from high-status textiles and twenty from low-status textiles. Thirty-three men remained working in ‘other trades’.

Within each large square are nine smaller squares representing possible combinations of wife's occupation at succeeding censuses.

Looking again at the shaded square where the husbands were in ‘other trades’ at both censuses in Figure 8A we can calculate, by summing the small columns within the square, that eight of the men had wives who were in textiles in 1851, 24 had wives at home and one man had a wife working in a trade other than textiles (see endnote 26). Occupational change amongst wives can be observed by reading off the intersections of the thin columns and rows. Thus in the ‘husbands remain in other trades’ square discussed above 3 wives remained textile workers 1851–1861, 23 remained housewives, I moved out of the home and into textiles, 5 left the factory to become housewives and the wife employed outside the textile industry also moved into the home.

³⁷ Surviving children per woman (s.c.p.w.) is a very crude indicator, measuring the number of a woman's children who survive until the census after their birth, and designed to show the average number of living children for which women in a particular occupational group might be expected to have responsibility at a particular census. We can note the number of children enumerated with their mother at, say, the 1851 census. Through linkage with the 1861 census we may discover a further child, alive in 1851, but enumerated at an address other than that of his or her parents. This child would be included in the s.c.p.w. calculation for the 1851 census, not the 1861 census. The number of an individual woman's surviving children is the minimum number of children she could have borne in the period of interest: we have no indication of her experience of miscarriage, stillbirth or infant or child mortality in that period. It is also possible that a child may go unobserved in succeeding censuses and thus not contribute to the s.c.p.w. Children surviving until the 1851 census may have died by the 1861 census, or may have left home, therefore we cannot use the sum of the 1851 and 1861 s.c.p.w.s to indicate the number of children a woman is caring for in 1861.

³⁸ Watterson, P. A. has addressed this question at a national level in her article ‘Infant mortality by father's occupation from the 1911 census’, Demography 25, 2 (1988)CrossRefGoogle ScholarPubMed. Female textile workers were, in fact, notorious for the number of their offspring who perished – see, for example, Taylor, R. W. Cooke, ‘The employment of married women in manufactures’, Transactions of the National Association for the Promotion of Social Science (1873), 605–613Google Scholar and ‘What influence has the employment of mothers in manufactures on infant mortality’, Trans. Natl. Assoc. P.S.S. (1874), 569–84Google Scholar; also Thompson, B., ‘Infant mortality in nineteenth century Bradford’, in Woods, R. I. and Woodward, J., Urban disease and mortality in nineteenth century England (London, 1984)Google Scholar – but the validity of such claims cannot yet be proven in the context of Keighley.

³⁹ The ‘Census of religious worship, England and Wales’, Parliamentary Papers 1852–1853, Ixxxix, 494Google Scholar reported that Keighley Registration District (R.D.) encompassed 14 Church of England churches in which, on the 30 March 1851, 133 people attended Sunday worship. The Wesleyan Methodists had 10 chapels and 1,604 worshippers and the Baptists had 9 chapels and 1,173 worshippers.

⁴⁰ In vol. 40 of the journal, Local Population Studies, the editors reported that a government Working Group on the Registration Services had, in their report, recommended that access to civil registers over 75 years old be made more directly available as befits ‘historical documents’. It must be widely hoped by researchers from a wide variety of disciplines that these recommendations are soon implemented.