The population of Victorian and Edwardian Kent - Part II

THE POPULATION OF VICTORIAN AND EDWARDIAN KENT (II) NATURAL INCREASE: BIRTHS, MARRIAGES AND DEATHS W.A. ARMS TRONG The population of Kent during the period was influenced considerably by migratory flows. As the tempo of internal migration increased, towns grew at the expense of the smaller villages and there was a marked tendency for the balance of the population to shift towards the north and west of the county. There was an appreciable overseas outflow of Kentish-born persons, but losses to the other English counties - and eventually even to London - were more than offset by in-migration which gradually increased the heterogeneity (in terms of birthplaces) of the resident population.1 Yet, the aggregate growth of the county's population, which was of the order of 156-216 per cent between 1831 and 1911 - depending on the boundaries chosen - owed much less to net gains by migration than to natural increase, i.e. the balance of births over deaths. It is the aim of the second instalment of this article to examine movements in the underlying rates of mortality, marriages and procreation. The chief sources used are censuses and the Registrar-General's annual reports on births, deaths and marriages, which commenced at the beginning of the Victorian period. These data provide us with a comprehensive picture of changes in vital rates in the registration county.2 They also afford a basis for comparison with national trends and a framework for the interpretation of a handful of more localized studies, which in this county have not been numerous, so far. 1 W.A. Armstrong, 'The Population of Victorian and Edwardian Kent, (I) Growth, Migration, Distribution', Arch. Cant., xcii (1993), 1-16. 2 Otherwise known as 'extra-metropolitan' Kent. For the differences between Kent so defined, the ancient or civil limits, and the administrative county boundaries see Armstrong, op. cit., 2. 17 W.A. ARMSTRONG MORTALITY (I) THE INFLUENCE OF EPIDEMIC DISEASES Throughout the period, variations in the incidence of epidemic diseases could give rise to fairly dramatic mortality fluctuations at the local level, and Kent was not spared these frequent and unpleasant experiences. Historians have tended to give particular attention to the cholera which first appeared in England in 1831-32. On that occasion, Kent was not very deeply affected: there were in all eleven outbreaks, principally at Rochester, Chatham, Dartford, Sheerness, Dover and Faversham, but they produced only 135 deaths between them.3 The impact of the 1849 epidemic was much greater, producing 1,208 deaths in all, which was sufficient to add about 10 per cent to the normal level of mortality prevailing in the county.4 The districts most seriously affected - in accordance with the national experience - were coastal settlements, notably Gravesend, Medway, and Thanet where mortality was 'more than double the average' as well as, to a lesser extent, Sheppey and Dover. However, the inland towns were not immune; at Canterbury 562 persons were seen at the Kent and Canterbury Hospital 'with premonitory symptoms' and 45 cholera fatalities were recorded. Mortality also rose significantly at Maidstone, where in September there were '43 deaths from cholera in Loose, one of its sub-districts .. . all except two or three were among vagrants, chiefly Irish, who came into the parish of East Farleigh to pick hops.' 5 The next major national epidemic extended into Kent in 1854, again producing over a thousand deaths (1,056), a figure exceeded only in London and Lancashire. Once more, the impact was greatest in the coastal settlements, most notably at Thanet where the cholera death rate (20.6 per thousand) was as high as in 1849, and at Sandgate; while Canterbury on this occasion generated 60 cholera deaths.6 Tonbridge, although still a comparatively small place (population 3,948) was dramatically affected by an outbreak of cholera and diarrhoea in conjunction: here, over a thousand persons out of a total population of 1,131 living in the confined district lying between the railway and the Great Bridge were affected, and of 170 cases ascribed to cholera, 39 3 C. Creighton, A History of Epidemics in Britain, 2nd edn., ii (1965), 821-2. 4 Ibid., 843-4. For a more detailed discussion of cholera epidemics in Kent see P. Hastings, 'Epidemics and Public Health' in N. Yates, R. Hume and P. Hastings, Religion and Society in Kent 1640-1914, Kent History Project, ii (1994), 1923. I am indebted to Dr Hastings for allowing me to see his work, prior to publication. 5 Registrar General, 12th Annual Report, 1853, 69; A. Bateman, 'Public Health in MidNineteenth Century Canterbury', Local History Dissertation, Kent University, 1988, 24, 28. 6 Creighton, op. cit., 852-3; Registrar-General, 17th Annual Report, 1856, Quarterly Return No. 4, 84; Bateman, op. cit., 28. 18 THE POPULATION OF VICTORIAN AND EDWARDIAN KENT proved to be fatal.7 The last significant visitation of cholera was in 1866, but produced on this occasion only 264 fatalities which contributed only a small proportion (some 2 per cent) to the total deaths that occurred in Kent during that year. 8 Of all the fatal diseases of the period, none were more dramatic in their effects on the public mind, and the cholera has continued to fascinate historians, perhaps to an extent out of proportion to its statistical significance. There were many other diseases, both epidemic and endemic, to which the population of Kent was susceptible. Smallpox, though commonly thought of as having been vanquished after Jenner's discovery of vaccination in 1798, still occurred from time to time: at the beginning of our period, in 1837-39 it accounted for 735 deaths in extra-metropolitan Kent and a further 226 in the registration district of Greenwich; and in 1871-72, the occasion of another national outbreak, the registration county of Kent contributed a further 537 deaths.9 Unusually hot summers were liable to give rise to further problems. That of 1857 induced marsh fever (ague) on quite a large scale in districts such as Strood and Romney Marsh, an experience which tended to debilitate and render people vulnerable to 'diseases of a general character'; while the blazing summer of 1911 gave rise in Kent, as it did nationally, to a sudden rise in infant mortality of about one-quarter (compared to the previous year), which was caused by diarrhoea.1° Kent (but more especially the Deptford, Strood, Medway and Tonbridge districts) appears in common with other south-eastern areas to have suffered much from diphtheria through much of the second half of the nineteenth century; as late as 1896-98 the death rates prevailing from this disease were close to those of London and 80 per cent above the national average.11 A spectacular outbreak of typhoid visited Maidstone in August-September 1897 when over 1,800 cases were detected and 132 deaths occurred. This was attributed to contaminated water supplies and an extraordinary incidence among girls aged 14-24 was explained by the presumption that women were 'more regular water-drinkers than men' .12 7 M. Barker-Read, 'The Public Health Question in the Nineteenth Cent ury: Public Health and Sanitation in a Kentish Market Town, Tonbridge 1850-1875', Southern History, iv (1982), 180. 8 Creighton, op. cit., 852. 9 Registrar-General, 2nd Annual Report, 1840, 181; Creighton, op. cit., 615. 10 Registrar-General, 20th Annual Report, 1859, xxx; 74th Annual Report, 1913, xxxiii, 80. 11 Registrar-General, 61st Annual Report, 1900, xxii, xxvi. 12 F.B. Smith, The People's Health, 1830-1910 (1979), 247. See al so I. Hales, 'Maidstone's Typhoid Epidemic September 11, 1897-January 29, 1898', Bygone Kent, v (1984), 217-23. 19 W.A. ARMSTRONG MORTALITY (II) THE LEVEL AND COURSE OF DEATH RATES In view of this melancholy list of afflictions, which is illustrative rather than exhaustive in its scope, it may come as something of a surprise to discover that overall, Kentish mortality rates were low by the standards of the time, if not by those of today, while they were clearly falling from the 1870s. The early years of civil registration revealed that average death dates (deaths per thousand at risk per annum) for the registration county of Kent in 1838-44 stood at 20.9 (for males) and 18.6 (for females). Broken down by individual registration districts, there was a predictable difference between the predominantly urban and rural districts. For example, Canterbury showed an average death rate of 23.5 (male) and 22.1 (female), Maidstone 23.7 and 22.1, Dover 23.8 and 19.8, while the highest Kentish death rates occurred in the Medway district - 25. 7 for males (even when deaths taking place in military hospitals were excluded) and 21.8 for females. By contrast, death rates of the order of 17-19 were recorded, for both males and females, in such districts as Elham, Bridge, East and West Ashford, Hollingbourne, Malling, Sevenoaks, Tenterden and Cranbrook.13 Taken in the round, Kentish mortality rates were lower by about 8 per cent for males and 12 per cent for females than the corresponding national figures.14 Table 1 surveys the data on general and infant mortality across a sixty-year period extending from 1850 to 1910. The reduction in the general death rate that occurred between 1866- 70 and 1906-10 (some 36 per cent) echoed a national improvement in both its extent and timing, while with respect to the infant death rate, despite a setback in the 1890s, there was actually a rather greater improvement in Kent (32 per cent) than in the nation at large between these dates. Set against national standards, Kent remained favourably placed, and this is reflected in an above-average expectation of life on the eve of the First World War, as is shown in Table 2. It goes almost without saying that the county death rate continued to conceal considerable variations from year to year and at the local level. In the aggregate, in 1913 there were 11.2 deaths per thousand population in the urban districts, and 10.3 in the rural.15 Yet, among the urban districts suburbanized local authority areas, such as Bromley, Beckenham, Chislehurst and Foots Cray, showed mortality well below the Kent 13 Registrar-General, 9th Annual Report, 1849, 180, 197-8. 14 As given in B.R. Mitchell and P. Deane, Abstract of British Historical Statistics (Cambridge, 1962), 36. 15 Registrar-General, 76th Annual Report, 1915, 144-5. These rates were standardized to adjust for variations in age and sex composition. 20 THE P OPULATION OF VICTORIAN AND EDWARDIAN KEN T urban average despite their very rapid growth rates - as if to vindicate a suggestion made long before by the Registrar-General, that the highest attainable health either was, or might come to be, found in 'a happy combination of both states, rus in urbe' .16 On the other hand, prominent among the places exhibiting disturbingly higher rates, more than ten per cent above the Kent urban average in that year, were Gravesend and Chatham, the latter featuring an infant death rate still as high as 106 per thousand live births. TABLE 1 General and Infant Mortality in Kent, compared to England and Wales, 1851-1910 (A) (B) General mortality Infant mortality (deaths per 1,000 (deaths under one, population) per 1,000 live births) Kent Ratio, Kent Ratio, Kent: Kent: England England and Wales and Wales 1851-5 20.2 89 132 85 1856-60 19.8 91 134 89 1861-5 19.9 93 131 87 1866-70 20.0 89 135 86 1871-5 18.1 91 124 81 1876-80 17.5 86 123 84 1881-5 17.2 89 114 82 1886-90 16.0 85 119 82 1891-5 15.7 84 123 81 1896-1900 15.3 86 135 87 1901-5 14.5 90 118 86 1906-10 12.8 87 92 79 Sources: Calculated from Registrar-General's Annual Reports for specified dates. Figures for England and Wales are drawn from B.R. Mitchell and P. Deane, Abstract of British Historical Statistics (1962), 36-7. 16 Registrar-General, 23rd Annual Report, 1862, xxi. 21 W.A. ARMST RONG TABLE 2 Expectation of Life at Several Ages according to the Mortality Experience of 1911-12 Age Kent England Ratio of (Admin. and Wales survivors to County) specified ages, Kent: England and Wales male female male female male female 0 55.5 59.7 51.1 55.0 (1) 102.3 102.2 5 59.6 62.7 57.0 59.8 (5) 104.3 104.1 25 42.2 45.2 39.9 42.7 (25) 105.0 105.1 45 25.9 28.4 23.8 26.2 (45) 106.9 106.9 65 12.0 13.6 10.9 12.2 (65) 117.6 115.0 Source: Registrar-General, 75th Annual Report (1914), Supplement No. 2, 34, 50. MORTALITY (Ill) EXPLAINING THE IMPROVEMENT IN MORTALITY The progress of the reduction in mortality rates is easy to document, but explaining the favourable trends is more difficult and in the last resort a matter for speculation, in the present state of knowledge. Historians are inclined to classify the reasons for the improvement under three headings: medical intervention, environmental improvements, including sanitary reform, and nutritional gains arising from advances in the material standard of living. There is no agreement - at the national level - about the weight to be assigned to each of these improving influences, and no way, short of devoting a lifetime's research to the subject, of providing even a provisional estimate of their relative importance in Kent. This said, there can be no doubt that all three improving influences were at work in the county, and each will be discussed. Evidence drawn from directories leaves little room for doubt that, in the nineteenth century, seaside resorts were very attractive to medical men: for example, Dover could boast as many as 13 physicians and surgeons and Deal and Walmer 11, in 1851.17 From that time on, the census occupational tables ought to offer a reliable guide: however, it is likely that the authorities became more discriminating as to which practitioners qualified for inclusion, as 17 J. Whyman, 'Rise and Decline: Dover and Deal in the Nineteenth Century', Arch. Cant., lxxxiv (1969), l l 8. 22 THE POPULATION OF V IC TORIAN AND EDWARDIAN KENT time passed.18 On the face of things, the ratio of Kent's population to each phy sician or surgeon deteriorated from 882: 1 (1851) to 1,339 (1911). On the other hand, if nurses and all para-medical personnel are included, the ratio improved over the same period, moving from 283: 1 to 126: 1 and this may be the better guide to the expansion of medical services. Following the establishment of the first hospitals in the county, the Kent and Canterbury ( 1794) and the Margate Seabathing Infirmary (1796), the first half of the nineteenth century saw the provision of a number of dispensaries for the poor at Ramsgate, Tunbridge Wells, Maidstone, Sheerness, Folkestone, Canterbury, and elsewhere, and new institutions in the second half of the century included sundry cottage and isolation hospitals. In all, by 1911, the administrative county of Kent hosted 86 assorted hospitals and 50 convalescent homes: indeed, no fewer than one in 136 of the Kent population were enumerated as inmates in such institutions, against one in 614 in England and Wales at large.19 However, there are limits to the extent to which mortality reductions can be attributed to medical intervention, either outside or inside these establishments. Some of the institutions in the county (particularly the convalescent homes) ministered to the needs of a non-Kentish, mainly London clientele, such as the Metropolitan Convalescent Institution and the Wainwright Convalescent Home maintained by the Lewisham and Wandsworth Unions, both at St. Peter's. Secondly, it is widely accepted that the capacities of nineteenth-century medical men to diagnose accurately and their ability to effect cures for most conditions were very limited at any time before the advent of sulphonamides in the 1930s and later, antibiotics. The most decidedly sceptical view of nineteenth-century medical practitioners and their achievements is associated with the work of Thomas McKeown, prepared in the 1960s and 1970s.20 In recent y ears, some doubts have been raised, chiefly on the grounds that he defined the role of medicine in somewhat restricted terms. Even in the absence of effective drugs, it is maintained, medical expertise was increasing, the importance of isolating infectious cases was increasingly appreciated, and the restoration of full health by the simplest forms of therapy, such as rest and the proffering and acceptance of sound advice 18 Especially after the passing of the Medical Act of 1858 which created the concept of the 'registered medical practitioner'. See W.J. Reader, Professional Men. The Rise of the Professional Classes in Nineteenth Century England (1966), 66-7. 19 Census of England and Wales, 1911, Area, Families, Occupiers and Population, I, Administrative Areas, 494, 525-9. 2° For a summary view incorporating the findings of earlier articles see T. McKeown, The Modern Rise of Population (1976), eh. V. 23 W.A. ARMSTRONG may have raised resistance to other, perhaps more fatal diseases.21 Moreover, doctors certainly played a leading role in the encouragement of public health measures. Whenever the environmental issue is raised - either in national or local studies - it usually gives rise to a narrative stressing the forces of inertia, the parsimoniousness of ratepayers, the prevailing ethos of laissez-faire, the various engineering set-backs that were experienced, and so on. McKeown, for one, took the view that advances in public health played only a subsidiary role in reducing death rates from disease.22 That progress was always an uphill struggle there is no doubt, yet it has been argued recently that 'it was ... in the last 30 years of the nineteenth century when most of the significant improvements and works of construction and concrete applications of preventive health measures went forward and were actively occurring on the ground throughout the provincial cities and towns of Britain.'23 Recent work on Kent carried out by Hastings, mainly using newspaper evidence, suggests that securing urban improvements was indeed a tortuous business in the middle decades of the nineteenth century. However, a number of towns ( or at any rate some of their citizens) benefited from piped water supplies as works were either established or improved at, among other places, Folkestone (1848), Dover (1854), Chatham (1855), Margate (1859), Broadstairs (1860), Maidstone (1861), Sevenoaks (1863) and Canterbury (1870).24 Following legislation of the early 1870s, the 1872 Sanitary Act and the 1875 Public Health Act between them gave Britain 'a common and comprehensive sanitary code unrivalled in Europe' ,25 and gave to local authorities new powers and new responsibilities. In late nineteenth-century Kent, the problems experienced certainly did not vanish but there was a certain amount of demolition of old, insanitary property, especially in the resort towns. Further improvements in water supply were made, partly as a result of promptings from the Local Government Board, notably at Whitstable and Margate, extending after 1890 even to some of the larger villages, such as Biddenden, Wingham, Eastry and Cranbrook. Sewerage schemes were usually more contentious, but improvements were achieved 21 J.M. Winter, 'The Decline of Mortality in Britain 1870-1950' in (Eds.) T. Barker and M. Drake, Population and Society in Britain 1850-1980 (1982), 111; J. Woodward, 'Medicine and the City: the Nineteenth Century Experience' in (Eds.) R. Woods and J. Woodward, Urban Disease and Mortality in Nineteenth Century England (1984), 66-77. 22 McKeown, op. cit., 125, 142, 153, 159-60. 23 S. Szreter, 'The Importance of Social Intervention in Britain's Mortality Decline: A Reinterpretation of the Role of Public Health', Social History of Medicine, i (1988), 22. 24 Hastings, op. cit. in n. 4, 202-5. 25 Winter, op. cit., 113. 24 THE POPULATION OF VICTORIAN AND EDWARDIAN KENT in many places, notably with the implementation of the West Kent and Darenth Valley schemes of 1877-78 (serving, among other places, Bromley and Sevenoaks); at Canterbury (1881); Hythe and Broadstairs (1884); Tonbridge and Tunbridge Wells (1886); Margate (1889); Folkestone (1892); Faversham (1893); Sheerness (1894); Gillingham (1895); Herne Bay (1897) and Deal (1898). Hastings concludes that, although many problems remained, 'the period 1871-1914 saw considerable improvement in the drainage, water supply and housing of some Kent communities' ,26 and in this respect his well-documented findings support the current rehabilitation of the environmental contribution to mortality reduction. Having relegated medical care and public health to a comparatively minor role in explaining the natural decline in mortality, it was the view of McKeown that the major share of the credit should go to improved nutrition, particularly because of its bearing on the incidence of tuberculosis, the major killer in the nineteenth century.27 In Kent, where the tuberculosis death rate was modest, it would be unwise to rule out this factor as a partial explanation of the comparatively low mortality and of the decline in death dates from the 1870s.28 The most systematic modern comparison of wage variations at the national level acknowledges that wages in the Home Counties tended to fall away with increasing distance from London where they stood highest, but nevertheless includes Kent - with Middlesex, Surrey and Essex - in a region featuring above average wages. In 1867-70 agricultural workers' weekly earnings stood at 17 s. per week, one of only eleven counties where this level ( or higher) was reached; and in 1898 the comparable figure was 19s. 10d. and exceeded only in two. In 1890, bricklayers' wages in Margate were only Id. an hour below London rates and 2d. above those ruling in Halstead (Essex) which, though nearer London, was in some respects further removed from its influence.29 And Kentish wage-earners, like their counterparts 26 Hastings, op. cit., 215. 27 McKeown, op. cit., 92, 142. 28 Data from Registrar-General, 25th Annual Report, 1864, lxiii, lxix indicates that in 1851-60, for England as a whole, the death rates per million living from phthisis were 2580 (male) and 2770 (female) and for diseases of the lungs 3320 (male) and 2730 (female). In Kent, the corresponding figures were 2140, 2410, 2530, 2001 (i.e. lower by 17, 13, 24 and 26 per cent respectively). This comparative advantage was maintained. Data for 1912 given in Registrar-General, 75th Annual Report, 1914, 418 yields an overall tuberculosis death rate of 1388 per million for England and Wales, and 1190 and 1147 for the aggregates of urban and rural districts of Kent. For a fuller discussion of this disease see G. Cronje, 'Tuberculosis and Mortality Decline in England and Wales, 1851-1910', in (Eds.) R. Woods and J. Woodward, op. cit., 79-101. 29 E.H. Hunt, Regional Wage Variations in Britain, 1850-1914 (Oxford, 1973), 14, 62, 104. 25 W.A. ARMSTRONG elsewhere, must have benefited considerably from gains in their purchasing power arising out of the considerable fall in prices, especially those of foodstuffs, that occurred in the period 1873-96.30 None of this is to suggest that cases of under-nutrition, caused by poverty, were rare in late nineteenth- and early-twentieth centuries Kent: in the end, much would depend on prudent budgeting and the skills of the individual housewife and mother; and of course, life would usually be a struggle for any family where the chief bread-winner was sick, unemployed, or had passed on. Yet, the potential was undoubtedly present for widespread gains in nutritional standards, and oral recollections of the period attest to a lively and at times discriminating demand for foodstuffs on the part of working people - while references to hunger and absolute want are conspicuous by their absence.31 Whatever balance is struck between these improving influences, the fact remains that Kentish mortality was low throughout the period, maintaining that superiority even when national rates began to improve, and this is perhaps the most striking and distinctive feature of nineteenth-century Kentish demography. However, the study of natural increase naturally requires as much attention to be given to birth rates and the factors which influenced these. It is convenient to divide the ensuing analysis into three sections which will discuss, in turn, some fragmentary data from the era before the advent of civil registration, then the early and mid-Victorian patterns of procreation and finally, the late nineteenth- and early twentieth-century decline in fertility. MARRIAGE AND FERTILITY (I) SOME EVIDENCE FROM PARISH REGISTERS The picture presented by Wrigley and Schofield suggests that the main contribution to national population growth after 1750 was made by rising fertility promoted by a falling age at and rising incidence of marriage, coupled with increasing numbers of births outside marriage. Birth rates were running at their peak in 1801-10 (39.5 births per thousand population), 40. 9 in 1811-20 and 40.6 in 1821-25, but remained as high as 35.8 in 1831-40.32 Comparatively little work has been done on the parish registers of Kent in this era, but what is available tends to suggest that the county must have shared in the 30 M. Winstanley, Life in Kem at the Tum of the Century (Folkestone, 1978), 23. 31 Idem, The Shopkeeper's World, 1830-1914 (Manchester, 1983), 147-8, 153, 159 and 'Rural Kent at the Close of an Era' in (Ed.) G.E. Mingay, The Victorian Countryside, ii 0981), 627-31. 32 E.A. Wrigley and R.S. Schofield, Tlze Population History of England 1541-1871 ( 1981), 534-5. 26 THE POPULATION OF VICTORIAN AND EDWARDIAN KENT general buoyancy of fertility. At Ash, the ruling (median) age at marriage in 1780-1840 was 24 for males and only 22 for females; these early marriage-ages along with a tendency for illegitimate births to rise (peaking at 7 per cent of all births in 1801-25), sustained high baptismal rates, estimated at 35.8 (1801), 42.1 (1811), 38.7 (1821) and 33.3 (1831).33 Such rates are not dissimilar, in their general order of magnitude, to the birth rates computed for eleven Kentish nonallowance giving parishes (34.7) and eighteen allowance giving places (36.0) in 1821-30.34 Comparatively low female ages at marriage have been also calculated for Hernhill and Dunkirk (22.3 and 21.6, respectively, for 1800-50) in an investigation which has also disclosed a great deal of new information about rates of premarital sexual activity. Of 202 known first births in Hernhill, 1780-1851, 22 were illegitimate and 86 were pre-nuptial pregnancies, leaving only a minority (94, or 46.5 per cent) conceived within wedlock; although only about five per cent of all births were recognized as illegitimate.35 More work along these lines on the parish registers would reveal a great deal more about the springs of population growth and the cultural context of courtship and marriage. However, there is no reason to expect that the population of Kent was in any way exceptional in regard to procreation. Moreover, the first returns becoming available under the new system of civil registration revealed an overall birth rate of 31.5 per thousand at risk,36 and in Kent - as in the nation at large - it is likely that the momentum of fertility was beginning to subside from peaks attained a little earlier. MARRIAGE AND FERTILITY (II) THE EARLY AND MID-VICTORIAN EXPERIENCE Under the new system of civil registration, all the material exists for close examination of nuptiality and procreation by local areas and for particular social groups. Unfortunately, the Registrar-General does not 33 A.E. Newman, 'The Old Poor Law in East Kent, 1606-1834. A Social and Demographic Analysis', Ph.D. thesis, University of Kent, 1979, 299, 323. 34 J.B. Huzel, 'The Demographic Impact of the Old Poor Law. More Reflexions on Malthus', Economic History Review, 2nd series, xxxiii (1980), 377. 35 B. Reay, 'Sexuality in Nineteenth Century England: the Social Context of illegitimacy in Rural Kent', Rural History, i (1990), 223, 242. 36 Figures based on data given in Registrar-General, 7th Annual Report, I 846, 108, covering births of 1839-43, and relating to the registration county of Kent, excluding Greenwich. 27 W.A. ARMSTRONG permit free access to the individual marriage and birth certificates which would make possible the application of family reconstruction techniques analogous to those practised on parish registers. Very occasional exceptions to this rule have been made, however, and one of these relates to mid-nineteenth century Ashford. Under the influence of the railway which arrived in 1843, Ashford's population more than doubled in thirty years (from 3,082 in 1841 to 8,455 in 1871). Here, 38.4 per cent of brides were pregnant when they arrived at the altar (45.2 among the daughters of labourers); female marriage ages were on the low side ranging from 22.4-23.2 among the wives of unskilled labourers for the marriage cohorts of 1840-44, 1845-49, 1850-54, 1855-59; and birth rates as high as 36.9 (1846-56) and 39.2 (1856-66) were upheld.37 The Ashford findings serve to remind us, too, of just how large early and mid-Victorian families might be: looking at all Ashford marriages traced to subsequent maternities in 1840-64, on average a couple could expect to produce 2.3 in the first five years of marriage, 4.4 after ten, 6.2 after fifteen and, where a couple had been married for twenty years, an average of 7.4 children,38 These were the demographic patterns of a Kentish community experiencing considerable in-migration. A contrast might be drawn with Canterbury (a net loser by migration) where the crude birth rate in 1861 at 31.8 stood as much below the national average as Ashford's was above it, and where on the evidence of the census females were marrying at 26.5 years of age.39 For the county as a whole, the early and mid-Victorian experience of marriage and fertility shows some minor, but interesting variations from the national norms of the period. There are signs that a more cautious approach towards marriage had come to prevail by the midcentury. This is revealed in the pattern of marriage ages which (so far as can be judged from a comparison with the isolated parish statistics quoted earlier) were somewhat higher than those characterizing the early decades of the century. While the data on marriage ages cited in the previous section were mined directly from individual parish registers, the information given in Table 3 is estimated from census data on the proportions ever married in successive age-groups: this is a well-established method of approach and, of course, has the 37 C.G. Pearce, 'Expanding Families', Local Population Studies, x (1973), 22, 23, 26, 29, 32. 38 fbid., 33. Newman, op. cit., 324 calculates cumulative family sizes for Ash in 1800-40: these were 2.1 (after five years), 3.8 (10), 5.6 (15) and 6.9 (20). 39 C.E.D. Bell, 'The Pattern of Fertility Decline in Twelve English Towns, 1850-1900', Ph.D. thesis, University of Kent, 1980, 564, 591-2. 28 THE POPULATION OF VICTORIAN AND EDWARDIAN KEN T advantage that it gives rise to county average figures.40 The outcome of these calculations suggests that Kentish males by the midnineteenth century were marrying a shade later than in England as a whole, and females rather younger, and that these differences were consistent across estimates drawn from successive censuses. TABLE 3 Patterns of Nuptiality in Kent and England and Wales, 1851-71 Age at first marriage Proportions still single at Kent England ages 40-44, in Kent (i) Males 1851 27.4 26.9 16.7 1861 27.2 26.4 13.5 1871 27.2 26.4 15.3 (ii) Females 1851 25.3 25.8 15.6 1861 24.7 25.4 15.9 1871 24.6 25.1 15.3 Sources: For Kent, tables of civil condition by age-group, given in successive census volumes. The English data, computed on the same basis, is drawn from E.A. Wrigley and R.S. Schofield, The Population of England 1541-1871 (1981), 437. Other things being equal, one would expect the marginally lower female ages at marriage in Kent to support higher birth rates. However, considered as a proportion of the total population, women in the fertile age groups (15-44) were somewhat less numerous in Kent than in the nation at large. These differences were significant, if not enormous. In 1851, the comparative proportions were Kent 22.2 per cent, England and Wales 23.6; in 1861, 21.9 and 23.5; and in 1871, 21.9 and 23.1. The reasons for this consistent shortfall may be associated with a bias towards male dominated employment in Kent (barracks, dockyards, etc.), or with a particularly heavy drain of young female domestic servants out of the county - or both. At all events, this factor would run counter to any tendency for marginally earlier female ages at marriage to promote higher crude birth rates. Another, more minor feature which would reduce recorded Kentish birth rates was a tendency for a small number of the confinements of Kentish mothers to take place in 40 See J. Hajnal, 'Age at Marriage and Proportions Marrying', Population Studies, vii (1953). See especially pp. 118, 129-31. Wrigley and Schofield, op. cit., deploy the same technique to arrive at the table of singulate mean age at marriage given in The Population History of England, op. cit., 437. 29 W.A. ARMSTRONG London, thus artificially transferring a few (properly Kentish) births to the metropolitan figures. Teitelbaum's estimates of birth rates for England and Wales and for Kent, given in Table 4, take account of this factor and also of the problem of general under-registration which affected both the county and the national figures, particularly in the early years of civil registration.41 TABLE 4 Crude Fertility Rates in Kent and England and Wales 1840-80 (Births per 1,000 population) 1841-50 1851-60 1861-70 Registered Births Kent 32.0 32.0 33.8 England and Wales 32.5 34.1 35.1 Ratio, Kent: England & Wales 98 94 96 Estimates after correction Kent 33.0 32.9 34.6 England and Wales 34.5 35.0 35.8 Ratio: Kent: England & Wales 96 94 97 1871-80 32.4 35.3 92 32.9 35.5 93 Source: M.S. Teitelbaum, The British Fertility Decline (1984), 66, 69. See note 41. T he outcome of the interplay of these various factors was thus a marginally lower crude birth rate for Kent than in the nation at large. It should be noted that the figures given in Table 4 include illegitimate births which would vary according to local circumstances and over time. In Canterbury, for example, there was a noteworthy rise in illegitimacy following the arrival of large numbers of soldiers at the barracks in the late 1850s and the same effect would be noticed in the 1880s.42 In the nation as a whole, illegitimate births accounted for 6.4 in every hundred in 1842 and 4.8 in 1870-2. Again, these figures were not exceptional, for in a ranking of 41 English counties, Kent occupied the 34th and 33rd places at these 41 M.S. Teitelbaum, The British Fertility Decline. Demographic Transition in the Crucible of the lndustrial Revolution (Princeton, 1984). The author discusses the steps necessary to compensate for both factors on pp. 56-74. The adjustment factors for Kent are 1.029 (i.e. the addition of three births per thousand) in 1841-50, and subsequently 1.028 (1851-60), 1.025 (1861-70), 1.015 (1871-80), 1.010 (1881-90), 1.032 (1891-1900) and 1.037 (1901-10). 42 Bell, op. cit., 570. 30 THE P OP ULATION OF VICTORIAN AND EDWARDIAN KENT dates.43 For reasons already touched on, illegitimacy ratios offer a poor guide to the extent of extra-marital sexual activity. However, some 94 out of every hundred births occurred within the context of holy (for most) or civil wedlock - or at least were subsequently legitimized by it. MARRIAGE AND FERTILITY (III) THE LATE NINETEENTH AND EARLY TWENTIETH CENTURIES From the 1880s there was a pronounced fall in birth rates in Kent, with a corresponding decline in England and Wales. Indeed, if the figures for 1871-80 (Table 4) and 1901-10 (Table 5) are compared, the percentage fall in Kentish crude birth rates is actually slightly greater (27 per cent) than that recorded for the nation as a whole (25 per cent), notwithstanding that the mild depressant of the crude fertility rate identified in the previous section (namely, the relatively low proportion of the Kentish population made up of females aged 15-44) was gradually eliminated.44 TABLES Crude Birth Rates in Kent and England and Wales, 1881-1910 (Births per 1,000 population) 1881-90 1891-1900 1901-10 Registered Births Kent 30.6 26.2 23.6 England and Wales 32.4 29.8 26.6 Ratio, Kent: England and Wales 94 88 89 Estimates after correction Kent 30.9 27.1 24.5 England and Wales 32.3 29.9 27.0 Ratio, Kent: England and Wales 96 91 91 Source: Teitelbaum, op. cit. See note 40. 43 Registrar-General, 6th Annual Report, 1845; 33rd-35th Annual Reports, 1872-4, for illegitimacy ratios. For rank ordering of counties see (Eds.) P. Laslett, K. Oosterveen and R.M. Smith, Bastardy and its Comparative History (1980), 30. 44 Females aged 15-44 as a proportion of the total population of Kent rose as follows:- 1881, 22.0 per cent; 1891, 22.9; 1901, 24.2; 1911, 24.8, by which date the figure was closely similar to that of England and Wales (24.9). 31 W.A. ARMSTRONG We should first examine trends in marriage habits. As can be seen from Table 6, further computations of Kentish ages at first marriage show (at least from the tum of the century) a small rise for males. With females - even though all the marriage ages quoted remained fractionally lower than the corresponding national figures - the more significant feature is the late nineteenth-century increase. If the figures for 1871 (Table 3) and 1891 (Table 6) are compared, fully a year was added to the female age at first marriage. Such an increase may not appear very dramatic until it is appreciated that this is likely to have meant on average a reduction of about 0.4 of a child per couple, according to the best available estimates.45 Added to this was the effect of a rise of one-third between 1871 and 1911 in the population of Kentish females still single at the age of 40-44. TABLE 6 Changes in Nuptiality, Kent and England and Wales 1881-1911 Age at first marriage Proportions still single Kent England at specified ages, in Kent At40-44 At 35-44 (i) Males 1881 26.7 26.6 15.4 1891 26.0 27.1 15.5 1901 27.4 27.3 16.6 1911 28.6 27.7 15.0 (ii) Females 1881 24.9 25.3 16.7 1891 25.7 26.0 18.5 1901 25.7 26.3 21.1 1911 25.7 26.3 20.5 Sources: As Table 3. The aggregate number of births (and hence the crude birth rates) was also influenced by a further fall in illegitimacy. Between 1870-72 and 1900-02 illegitimate birth rates per thousand unmarried and widowed 45 H.J. Habakkuk, Population Growth and Economic Development since 1750 (Leicester, 1971). 38, 101-2. This estimate is roughly consonant with the conclusion of Wrigley and Schofield, op. cit., 436, that the national increase in female marriage ages, 1871-1901 (1.14 years) would have reduced the completed fertility of women surviving to the end of the child bearing period by at least 7 per cent. 32 THE POPULATION OF VIC TORIAN AND EDWARDIAN KEN T females virtually halved, falling from 14.7 to 7.5, again in conformity with a national trend; in 1901-10 Kent stood 47th among the 55 registration counties of England and Wales on this kind of index.46 However, the most significant and novel development - in terms of its implications for the future - was a noticeable downturn in procreation within marriage. This is identified in the movement of marital fertility rates which are calculated in such a way as to remove distortions arising from variations in age-sex composition, or in ages at marriage.47 In Kent, according to recent calculations carried out by Teitelbaum, marital fertility fell by 35 per cent (against 32 per cent for England and Wales) between 1871 and 1911.48 The beginning of regular and systematic control of births within marriage - for this is what these figures attest to - has long fascinated historians. Not a great deal is known about the techniques involved - although it is suspected that mechanical methods were far from common - and most discussion of the phenomenon has been conducted at the national level, focusing on changing attitudes to child-bearing and strategies of family formation. Prominent among the factors involved in the literature are demographic necessity (parents were reacting to falling infant mortality, which confronted them with a new situation, namely that the number of children conceived would equate more closely with the number they had to bring up); economic trends, mediated through rising aspirations, initially among the 'middle classes' who led the way to smaller families; a range of immeasurable social factors, including a rise in the status of women and the declining hold of religion; reduced opportunities for children to earn incomes contributing to the family purse (as a result of state intervention in child labour and of more - eventually compulsory - schooling); or, conversely, enhanced opportunities for mothers to find jobs (in some counties); greater literacy (facilitating the spread of knowledge); and the way in which professional careers were becoming increasingly meritocratic in their patterns of recruitment, encouraging among parents a sense of 'future time perspective' and a disposition to plan their families carefully.49 46 Registrar-General, 68th Annual Report, 1907, xxxv; 75th Annual Report, 1914, Supplement, Pt. II, xx-xxi. 47 Calculations of marital fertility rates take account of legitimate births to married women in successive (usually quinquennial) age groups between 15 and 44 (or 49). The summation of these rates gives the appropriate marital fertility. 48 M.S. Teitelbaum, op. cit., 128. 49 See, inter alia, Habakkuk, op. cit.,; Teitelbaum, op. cit.; J.A. Banks, Prosperity and Parenthood (1954); idem, Victorian Values (1981); N.L. Tranter, Population and Society, 1750-1940 (1985), 49-61, 92-123; R. Woods, The Population of Britain in the Nineteenth Century (1992), 39-55. 33 W.A. ARMSTRONG Returning to the case of Kent, we have reasonably firm information on some of these variables, but almost none on others. In the present state of research, the most that can be done is to engage in a certain amount of - grounded - speculation. There were some factors which probably tended to hurry on the reduction in marital fertility in Kent, and others that would have worked to delay it. In the first category we can place (i) The comparatively low infant death rate, already documented (see Table 1); (ii) Proximity to London. It has been shown in comparative studies covering virtually the whole of Europe, that great cities (including London) tended to lead the way in the reduction of marital fertility.50 The assumption here is that Kent, by virtue of its proximity and good communications, would have been more open to metropolitan cultural influences than most other English counties; (iii) A reasonably high rate of literacy. This can be inferred from the proportions signing the marriage register with signatures (as opposed to marks), going from data regularly collected and published by the Registrar-General. National data, broken down on a county by county basis suggest that 'the lower the illiteracy the earlier and more substantial the irreversible decline in fertility' ;51 and throughout the century Kent stood reasonably high on this index. For example, on the basis of the marriages of 1881-90 and 1891-1900 the county was placed fourteenth and seventeenth among the counties of England and Wales (males) and seventh and seventh equal for females;52 (iv) A substantial 'professional' population. Banks has identified the elements of the professional population which, at the national level, pioneered the move to smaller families. They included gentlemen of private means, military officers, physicians and surgeons, artists, the legal profession, civil engineers, accountants, authors and journalists, clergymen and ministers and a little later, bank staff, dentists, chemists, architects and school teachers.53 As a proportion 50 See, for example, D. Kirk, Europe's Population in the Jnterwar Years (League of Nations, Geneva, 1946) 48, 55, 57; United Nations, The Determinants and Consequences of Population Trends (New York, 1953), 85-6. 51 Teitelbaum, op. cit., 213. 52 Data by county is given in D.E. Baines, Migration in a Mature Economy (Cambridge, 1985), 328-31. 53 Banks (1981), op. cit., 98-100 provides a more complete list of pioneers of the smaller family, covering altogether 24 census occupational groups. 34 THE POPULATION OF VICTORIAN AND EDWARDIAN KENT of the male population aged ten and over in 1891, these groups comprised 4.4 per cent in Kent: a minority of course, yet a figure over half as high again as in England and Wales at large (2.9 per cent). Conversely, factors working to delay the reduction in marital fertility in Kent may have included (v) The weakness of incentives arising from a strong and regular demand for married female labour, which in some counties (Lancashire is the obvious case) may have entailed a conflict between the claims of motherhood and maintaining a job. In Kent, the 1891 census shows that the proportion of females aged ten and over who were economically active - in the sense that they held jobs recognised by the census authorities - was a little higher than in the nation at large (30.5 against 25.5 per cent). However, much of this employment was in domestic service, overwhelmingly the province of single women, and there are some indications that this proportion was rising; at Canterbury the number of domestic servants per thousand population was 102 in 1861, 117 in 1901.54 If domestic service is discounted, the proportion of economically active females was rather lower in the Kent of 1891 than in the nation as a whole - 14.9 against 16.7 per cent; (vi) A high proportion of labourers. From studies conducted at the national level, it is clear that general, dock and agricultural labourers were comparatively slow to join in the move towards smaller families.55 In 1891, Kent showed well above average proportions of labourers: as many as 19.5 per cent of all males aged ten and over were placed in one or other of these categories (against 9.9 per cent in England and Wales) and in the specific case of agricultural labourers, 11.8 against 5.2 per cent. If we assume that factors (i)-(iv) would tend to be cancelled out by (v)-(vi), at least at the county level, there would be no reason to expect Kent to be in the vanguard of the late nineteenth-century decline in marital fertility. Fortunately, we have some information on that point as s4 Bell, op. cit., 95. For the general background see E. Richards, 'Women in the British Economy since 1700: An Interpretation', History, !ix (1974). ss J.W. Innes, Class Fertility Trends in England and Wales, 1876-1934 (Princeton, 1938), 43-7; W.A. Armstrong, 'The Influence of Demographic Factors on the Position of the Agricultural Labourer in England and Wales, c. 1750-1914', Agricultural History Review, xxix (1981), 73. 35 W.A. ARMSTRONG a result of the publication of Teitelbaum's impressive study. He employs two approaches to identifying the significant turning point in marital fertility: (i) the date by which a decisive 10 per cent decline from 1871 levels occurred, and (ii) the date of the achievement of a low absolute fertility rate equating to 60 per cent of the highest levels of human fertility ever reliably recorded.56 On the first of these measures, the critical date for Kent is 1891, one shared with or preceded by as many as twenty other English counties. Alternatively, if the second approach is preferred, the decisive date is 1893 and on this basis Kent was accompanied by, or followed, seventeen other counties.57 Quite evidently, Kent was not a prominent leader in the national decline in marital fertility although it will not do to infer a significant lag, either, for Teitelbaum does emphasize the high degree of homogeneity of the various English and Welsh counties in this respect. Furthermore, the indications are that, once adopted, practices of birth control were taken up with some enthusiasm in Kent, for between 1891-1900 and 1901-10 marital fertility was falling slightly faster than in the nation at large.58 CONCLUSIONS At no point during the entire period considered here - not even in the worst epidemic years such as 1849 when the county death rate rose by about 10 per cent - did the absolute number of deaths reach anywhere near the level of births. This remained true even after both rates were significantly reduced. As Table 7 indicates, natural increase was responsible for adding around 5,000 persons to the population of Kent each year in the early years of Victoria's reign, and around 10,000 in the later Victorian and Edwardian years. This investigation has failed to uncover any sensational revelations about the rate of natural increase, or about the pattern of birth and death rates that underpinned it. For the most part, the Kentish experience changed in close conformity with national trends. It remains to add that behind these somewhat bland statistics lies a richly varied world of 56 This standard is the fertility of the Hutterite community in the United States in 1921-30, a sect which was strongly opposed to birth control. It is used as a yardstick throughout the multi-volume Princeton series of which Teitelbaum, op. cit., forms a part. 57 Teitelbaum, op. cit., 135, 140. 58 Registrar-General, 75th Annual Report, 1914, Supplement, Part II, xvii. Between these two dates, the ratio of Kentish to national marital fertility rates fell from 94 to 93. 36 THE POPULATION OF VICTORIAN AND EDWARDIAN KEN T TABLE 7 Natural Increase in Kent (Births minus Deaths), 1841-1910 Annual averages of births and deaths at selected periods (A) (B) Average no. Average no. Natural Increase of births of deaths (A-B) 1841-50 14,440 9,592 4,848 1851-60 16,463 10,293 6,170 1861-70 19,804 11,543 8,261 1871-80 21,666 11,956 9,710 1891-1900 22,680 13,498 9,182 1901-10 23,564 13,562 10,002 Source: Registrar-General 6th-73rd Annual Reports, 1842-1912. human experience where suffering and anxiety, hopes, aspirations and fears were constantly in flux, the full complexity of which can only be hinted at in the course of a few printed pages. Historical demography for its own sake is and will probably remain a minority taste, but its findings provide a springboard from which we can reach out for a more complete understanding of our social past. Indeed, searching questions about such issues as the dynamics of change in towns and the countryside, the role of public authorities, the process of family formation and the texture of human relationships can scarcely be framed properly without its aid. The findings summarized in this and the preceding article59 will, it is hoped, lend perspective to our existing knowledge of aspects of Kentish history, and be found useful in suggesting new lines of enquiry. 59 Arch. Cant., xciii (1993), 1-16. 37