A social diffusion model has found much favour, in which smaller family sizes were pioneered by the upper class, were subsequently adopted by the middle classes, and eventually percolated down to the working classes, a model which fits snugly into more general concepts of social change as a process in which changes spread downwards from the top to the bottom of society by imitation and emulation. There is indeed good evidence that the British aristocracy, from the sixteenth century at least, consistently went in for smaller family sizes than did the generality of the population, and that these began to contract from about the 1830s, a good generation in advance of the rest. Thus the average aristocratic family was down to four children in the third quarter of the nineteenth century from its high mark of five children fifty years earlier, at a time when six children was the national average; a steep fall in aristocratic family size, however, did not set in until after 1875, at much the same time as the general decline. The persistently lower levels at which aristocratic families operated were the outcome of later and less frequent marriages, which can be explained by the importance of property questions in marriage arrangements and by the reluctance of many to marry beneath themselves in the social scale. Eldest sons almost invariably did marry, in order to continue the line and because they could readily attract women who were both their social equals and were backed by satisfactory marriage portions. Younger sons, however, might find it difficult or impossible to make such good matches, or to live at the standards to which they had been brought up in childhood, and might thus tend to stay single, thereby depriving a similar number of aristocratic daughters of eligible marriage partners and obliging them, on the social parity principle, to remain spinsters. The smaller family sizes were also, in part, the result of family limitation, and this appears to have been more practised with increasing effect from about the mid-1820s onwards. This could well have been a lagged response to the decline in child mortality which had set in some thirty to fifty years before, since it would have become obvious that the chances of children surviving to maturity had risen markedly and that it had become necessary to have fewer babies in order to achieve a target number of grown-up children, and prudent to do so to avoid the mounting costs and responsibilities of supporting larger families of survivors. The decline in child mortality may well have included a fall in infant mortality, although that cannot be separately measured; the growing unfashionability of putting aristocratic babies out to wet nurses would certainly have produced such a fall. Most of the decline, however, was probably due to better child care, improving nutrition, and better home conditions, all matters in which aristocratic resources were clearly likely to put them comfortably ahead of the masses. Further, such material improvements could well have been accompanied by, and indeed have helped to foster, a change in parental sentiments leading to growing attachment to each individual child and hence to a more caring, and careful, kind of parenthood.
There are clear suggestions, although no hard statistical proof, that the urban middle classes were beginning to follow suit in the 1850s and 1860s. It could even be that the middle classes had for long had families that were small in relation to the national average, and conceivably smaller than those of the aristocracy. One strand in middle-class opinion which was already well established by the 1830s, after all, was a puritanical disapproval of aristocratic extravagance, indulgence, frivolity, and excess. This was openly expressed by criticism of luxurious and improvident styles of living, and of moral laxity in sexual behaviour; but it is not impossible that this contained an unspoken criticism of improvidence in the begetting of children, with the implication that the middle-class critics behaved differently. The improvident marriages which were loudly condemned, however, were those of the labouring classes. The thrust of the Malthusian case on the pressure of increasing numbers of mouths upon the means of subsistence, leading inexorably to growing impoverishment, was directed at imprudent and youthful marriages of the poor which produced insupportably and undesirably large families. Criticism of the system of poor relief, which was held to encourage such imprudent marriages by subsidizing them at the ratepayers’ expense, was a well-publicized special application of this view in the run-up to the reform of the Poor Law in 1834. The validity of the case is one thing, and it has been demolished by historical research; but those who held these views were presumably satisfied that they themselves were not guilty of imprudence in their own family affairs. The middle classes had, indeed, every reason to exercise moral restraint in delaying their own marriages until the bridegroom was sufficiently established in his career to be able to afford to keep a wife and family in the style considered suitable to his station in society. That this was the ideal to aim at was taken for granted in guidance literature and fiction alike, and was presumably largely observed in practice.
The postponement of marriage by middle-class men to their late twenties or early thirties did not necessarily affect the ultimate size of their families, although the delay is commonly held responsible for the Victorian’s ‘double standard’ which connived at or even stimulated the sexual activities of bachelors while insisting on chastity for unmarried women. Other things being equal, the number of children born in a marriage depended on the age of the wife at marriage; the husband’s age, provided let us say he was under fifty when he started, made little difference. Given that middle-class daughters were brought up to regard marriage and motherhood as their main purpose in life – although generally kept in ignorance of the mechanics of procreation – there was nothing in their upbringing to suggest that they, or their parents, had a duty to exercise restraint by delaying marriage, unless sexual ignorance and fears of childbirth may have nourished anxiety to postpone the start of the long haul of childbearing. Middle-class standards, in other words, may simply have led to an unusually large difference between the ages of husbands and wives, with women starting married life at much the same age as in any other section of society. There are, unfortunately, no studies and no statistics of any of the particular social groups – apart from the peerage – which go to make up the national averages. On the other hand, the prudence which made middle-class men feel they were too poor or insecure to marry until they had reached some target level of income and independence, must also have made them feel too poor to support in acceptable style the indefinite, or very large, number of children with which God was all too likely to bless a very young wife.
It is, therefore, entirely possible that normal middle-class patterns were for above average ages at marriage for both sexes, which would have given them below average sized families. Another possibility is that middle-class intercourse was infrequent: male abstinence or weakness of sex drives, or female adroitness in rebuffing or avoiding encounters, would not have been inconsistent with prevalent evangelical attitudes to carnal pleasures, indeed pretences that sex was not pleasurable at all, however hard it may be to believe that intimate private practice did not render such attitudes mere public hypocrisy. A third possibility is deliciously ironic. Middle-class couples may have been practising family limitation within marriage, by coitus interruptus or otherwise, at the very time in the 1820s and 1830s that middle-class, and clerical, opinion was vehemently denouncing radical campaigns for birth control among the working classes, in the name of public decency and morals.
It is thus plausible to speculate that middle-class fertility and family sizes were already functioning on a comparatively low level, for a variety of social and economic reasons, before demographic evidence begins to surface that shows they were on a declining trend from the 1860s. This evidence, mainly from the first census of fertility taken in 1911, which recorded information by social and occupational classes from all women then living, supplemented by a few studies using mid-Victorian census enumerators’ books, indicates that middle-class family size had fallen to about 2.8 children by 1911 and had virtually closed any earlier gap there may have been between them and the upper class, since aristocratic families were only slightly smaller, at 2.5 children. Beatrice Webb has been cited as a dramatic illustration of the experience of the late Victorian generation of the wealthy middle classes: born in 1862 as one of the ten children of an industrialist and railway director, her marriage to Sidney in 1892, though it lasted over fifty years, was childless – whether by design or from infertility is not known. A less ambiguous illustration of upper-middle-class habits, if Oswald Mosley’s account