2. THE IDENTIFICATION OF DIFFERENT TYPES OF HOUSEHOLD

When classifying households, most historians have adopted (sometimes with modifications and additions) the classification scheme developed by E. A. Hammel and Peter Laslett.Footnote ²¹ This classification terms households containing just one conjugal family group – either a married couple (with or without unmarried children) or a lone parent and an unmarried child – as ‘simple family households’. Alternatively, if households included not only a conjugal family unit but also other relatives, they were identified as ‘extended family households’ if the additional relatives, such as grandchildren, siblings or a parent, did not themselves form a separate conjugal family unit and as ‘multiple family households’ if they did. Households without conjugal family units were of two types: first ‘solitaries’ and, secondly, ‘no family households’. These latter consisted of co-resident relatives such as siblings or more distantly related or unrelated persons who were not members of a conjugal family unit (as defined above).

This classification has been widely used to chart geographical variations in household forms by region, type of local economy and country.Footnote ²² It has in fact become customary to characterize populations depending on whether they appear to have favoured the formation of simple or more complex households or, in particular, multiple family households,Footnote ²³ and Laslett himself used the classification in this way.Footnote ²⁴ Yet comparatively little attention has been given to whether members of these various societies were aware of the fact that they were living in households that centuries later have been designated as ‘simple’, ‘extended’ or ‘multiple’. let alone in a ‘no family’ household, even though the members of a no family household were usually relatives.Footnote ²⁵ When defining the boundaries of the household, Laslett was keen to stress that it was important to take account of where contemporaries thought these boundaries should be: whether, for example, to include or exclude the retired parents of the household head.Footnote ²⁶ It thus seems logical that the same principle should be extended to the identification of different types of household.

The inhabitants of different communities would have known that households could be larger or smaller and more or less complex, since all the historical populations that have so far been studied – even when dominated by a particular type of household, for example by simple family households in England or by multiple family households in many Russian populations – also included some households that were of a different type (complex households in England and simple family households and solitaries in Russia). It does not necessarily follow, however, that such households were viewed at the time as aberrant or different. If, for example, we look at the ways households in seventeenth- and eighteenth-century England were understood at the time, we find that not even the most perceptive observers classified households on the basis of their kinship structure. Terms such as ‘simple’ and ‘extended’ family (or anything similar) were never used. Most informed observers knew (correctly) that English households were usually small, containing on average, between four and five persons.Footnote ²⁷ However, even when they exaggerated the size of a household this was because they included in it too many children, servants and lodgers, as opposed to large numbers of relatives.Footnote ²⁸

The views that residents (as opposed to outsiders) might hold of the differently structured households within their community are much harder to identify. The fact that the census authorities in 1851 found it necessary to reclassify lodgers as householders when they were analysing the composition of households indicates that the preference of the census authorities was to define households as economic units that maintained separate budgets, while the householders and enumerators had placed greater emphasis on the occupation of a distinct living space.Footnote ²⁹ However, for detailed evidence on how the inhabitants of particular communities accounted for the presence of married children in some households, it is necessary to move forward in time and examine the sociological and anthropological studies of both rural and urban populations undertaken in the middle decades of the twentieth century.Footnote ³⁰ The earliest of the communities to be surveyed was Llanfihangel yn Ngwynfa, Montgomeryshire, in 1939 and 1940, by Alwyn D. Rees. In Llanfihangel, farmers and their male relatives undertook 80 per cent of the work on the farms carried out by males while the wives, female occupiers of holdings and their female relatives did 90 per cent of the work undertaken by females.Footnote ³¹ Fifteen per cent of the households of the farmers contained additional relatives such as an unmarried sibling, uncle, aunt, nephew, niece or cousin, and there were a number of three-generational households. However, almost half of these three-generational households were regarded by Rees as temporary or makeshift arrangements. Two of the households contained married daughters whose first child had arrived before a farm had been found for the new family (the young husbands lived with their own parents). Another two households included widowed daughters with children and a further two three-generational households were established when the grandparents took into their households the illegitimate child of a daughter. Instances where a married son lived with an aged father formed a distinct minority even of the three-generational households.Footnote ³²

There is a certain similarity between these households in Llanfihangel and the interpretation I have offered of the role of the household as a provider of welfare in order to explain the presence of relatives in the households of the inhabitants of Colyton, Devon (only a minority of whom were farmers), in the middle of the nineteenth century.Footnote ³³ In Llanfihangel, there seems to have been no conscious preference expressed in favour of a particular type of household other than that if there were no unmarried children, an aged father should live in the household of a married daughter. Most residence patterns thus appear to have been governed by unexpressed, perhaps unconscious, preferences.Footnote ³⁴ There is a sharp contrast in this respect with the acknowledgement of the higher status of men once they were married: they would no longer be referred to by their forenames and the name of their father's farm but by their forename and surname or as ‘the husband of a particular farm’. On the other hand it is possible to infer a decided preference against the co-residence of a married son with his parents, given that so few did so even though, in a large number of families, one son (usually the youngest) inherited the parental holding.Footnote ³⁵ For this, there could be an economic explanation in that hill farms would not generate sufficient income to support a large household. Rees, however, felt that many of the residence patterns were shaped by medieval Welsh law, according to which the father retained control of the land of the kindred until his death and was then succeeded by his youngest son.Footnote ³⁶ Another inference that could be drawn is that it was the influence of the mother of the inheriting son that would prevent his marriage during her lifetime, because she would never willingly surrender the management of the household to a daughter-in-law.

In Gosforth (Cumberland), studied by W. M. Williams between 1950 and 1952, the usual practice was for a farmer to retire to a house he had bought in the district and pass the farm to a son, usually the eldest.Footnote ³⁷ It seems plausible that inheritance by the eldest rather than the youngest son (as in Llanfihangel) was feasible because the father was prepared to retire. On the other hand, in both Llanfihangel and Gosforth, the co-residence of a married heir and a parent was unusual. In both communities, therefore, there were very few stem families of the classic type, even though many farms were inherited by a son. In Llanfihangel, most heirs did not marry until their parents were dead; in Gosforth it was until they had retired. Both in Llanfihangel and in Gosforth it was also customary for the widow of a farmer to retain possession of the farm for the rest of her life (unless, in the case of Gosforth, her late husband had already retired). When, in a few cases, a son did inherit the holding immediately following the death of his father and his mother was still alive, the inheriting son would be obliged to support her for the rest of her life, as long as she did not remarry.Footnote ³⁸ The latter limitation appears also in arrangements governing the transfer of customary land on some medieval English manors (although the widow did not have to forfeit the land if her new husband paid a high entry fine) and in retirement contracts in the Nordic countries.Footnote ³⁹ Williams was keen to stress the difference between retirement arrangements in Gosforth and those in some other rural communities. In Gosforth, on retirement the farmer and his wife relocated to a new house in the district whereas in medieval England and in early-twentieth-century Ireland a separate room would have been reserved for them on the farm, or in Germany they would have moved to an Altenteil, which was still on the farm but often consisted of a separate dwelling.Footnote ⁴⁰ This distinction, however, places too much emphasis on the element of residential proximity when the essence of the retirement arrangement was the same.Footnote ⁴¹ It seems very likely that even after retirement an elderly couple in Gosforth would continue to assist with the farm work just as their peers did elsewhereFootnote ⁴² and just as did the sons of farmers in Gosforth who were not destined to inherit the parental farm and who moved off the farm after they had married.Footnote ⁴³

The children of the residents of Gosforth whose fathers were neither farmers nor craftsmen were allowed greater independence once they reached working age. Nevertheless, the villagers informed Williams that when families had been larger at the start of the twentieth century, the practice had been to reward whichever child had taken care of the elderly parent by ensuring that it was this child who received the larger share of any moveable property after their death, as there was no land to inherit.Footnote ⁴⁴ Both J. M. Mogey (for Oxford) and Michael Young and Peter Willmott (for Bethnal Green) have also indicated that, even in urban areas in the 1950s, the ties between working-class parents and their adult children, particularly between mothers and daughters, remained strong after they had established separate households. Indeed, as mothers and married daughters spent so much time in each other's households and ate and drank there, Young and Willmott argued that to some extent the households were merged, since eating at the same table helps define what constitutes a household.Footnote ⁴⁵ In reporting the proximity of the households of married daughters to those of their mothers in the St Ebb's district of Oxford, Mogey termed them ‘stem families’ even though the generations did not co-reside, there was no land to inherit and the link was between mothers and daughters rather than between fathers and sons.Footnote ⁴⁶ It is significant that Mogey had to borrow the term ‘stem family’, used by Frédéric Le Play to identify peasant families in which the son who was designated as the heir co-resided with his father. Mogey's adaptation of the term implies that there was no suitable expression in use locally to describe this type of pattern. In both Oxford and Bethnal Green it was possible to observe a clear preference on the part of both generations to live near each other rather than to share a household, with the persons interviewed emphasizing the negative aspects of a shared household with the parents.Footnote ⁴⁷

3. THE HOUSEHOLD AS AN INCOME-POOLING UNIT

A number of lists of the poor in English populations in the past record both the earnings of different members of the households of the poor and the proportion of their income provided by the community in the form of poor relief. For poor widows with wage-earning children, the earning power of these children could be crucial. For example, children provided 65 per cent of earnings in nine such households listed in 1597 in Ipswich, 48 per cent of earnings in six households in Salisbury in 1637, 89 per cent of nine households in Corfe Castle, Dorset, in 1790, and 73 per cent in 21 households in Ipswich in 1906.Footnote ⁴⁸ It would, however, be unwise to assume that all the earnings of the children were placed at the disposal of the household in general. Evidence from other communities indicates that some children retained a proportion of their earnings to spend as they saw fit. One reference to this practice occurs in one of the budgets of labouring families in 1789 compiled by David Davies, where the eldest son (aged 14) in the family of a married labourer from Holwell, Somerset, earned £6 10 shillings per annum but spent 15 per cent of that on his clothing. This was considerably more, sometimes double, what most labouring families spent on clothing one person.Footnote ⁴⁹ There may also be a hint in a reference to an enumeration of Sandwich, Kent in 1776 that older children may simply have used their earnings to meet the costs of their maintenance and retained the surplus for themselves, in that unmarried children above the age of 30 who still lived in the parental home were classed as lodgers.Footnote ⁵⁰ At the start of the twentieth century, B. Seebohm Rowntree also argued that older children living with their parents in York contributed only that part of their incomes to the family purse that any other lodger would pay for board and lodging.Footnote ⁵¹

Further evidence as to the willingness of children to contribute to the household budget is also available for the households of 12 widows in Ipswich in the early years of the twentieth century.Footnote ⁵² On average, the children of these widows had handed over to their mothers 74 per cent of their earnings. There was, however, considerable variation, with one widow receiving all the earnings while at the other extreme another widowed mother received just under half of the earnings. In the households of two married men, their children were contributing two-thirds of their earnings to the household budget. Apart from rare exceptions, these children, it should be noted, were considerably less generous than were their fathers who, even although most retained some of their earnings for beer and tobacco, made available a much larger proportion of their earnings to meet the expenses of housekeeping.

These findings need careful consideration. In the first place they would appear to indicate that widows and their non-earning children might be somewhat poorer than would first appear from a consideration of the earning power of their households. However, this would not be the case if the mothers were thereby spared the cost of maintaining their earning children. There is some merit to this latter argument but it is evident that widows would have lost some potential income since their children retained some of their earnings for additional personal expenditure. In the rural districts of Suffolk in 1909, investigated by the Royal Commission on the Poor Laws,Footnote ⁵³ adult sons living in the parental home were said to be meeting the costs of their board, lodging, mending and washing. Some sons were more generous in that they also paid most of or even all the rent. However, there were also said to be some spendthrift sons who expected a parent to support them even when they were out of work, if that parent was receiving poor relief. Level of earnings, custom and the character of the son were the three factors mentioned that determined what proportion of their earnings sons were prepared to make available for the household as a whole.

The second implication is that the economic circumstances of both widows and married couples with earning children are likely to have been very variable, depending on what proportion of their earnings children made available. It is also necessary to modify the standard understanding of how household income varied according to the age of the household head. As set out by Rowntree,Footnote ⁵⁴ this indicated that individuals fell below the poverty line between birth and age 15 and then experienced a second descent beneath the line between about 30 and 40 when any children in their household would be too young to earn much, with a final descent into poverty after the age of 65. Rowntree was thus ignoring his own observation that older children would not be contributing all their earnings to the household. If due account is now taken of this, then in households where the head was between the ages of 50 and 65 (and which would be most likely to contain adult children) the parents would be somewhat poorer than he, Rowntree, envisaged. It also follows that the descent into poverty in old age or on widowhood might be less severe.

However, it should be emphasized that these remarks apply only to wage-earning households. Studies of the economies of family farms in the 1940s and 1950s suggest that the situation of sons who remained, unmarried, at home to help run the farm were very different. These accounts, from Llanfihangel yn Ngwynfa, Montgomeryshire in 1940 and from Gosforth, Cumberland between 1950 and 1952, indicate that the sons of labourers had considerably greater independence and more cash than the sons of farmers.Footnote ⁵⁵ In return for all this largely unpaid work, one of the farmer's sons could expect eventually to inherit the parental farm, sometimes jointly with an unmarried sister, although they could be well into middle age before this occurred.Footnote ⁵⁶ Once a suitable farm became available, a father would provide his sons who did want to marry with most of the stock and implements. Daughters would also receive some stock but principally household goods.Footnote ⁵⁷ This would imply some depletion in the resources of the parental farm as sons and daughters married, but this was very likely not until many farmers were at least in their late fifties, given that the mean age at first marriage of the sons of farmers was over 31 and that of daughters of farmers was over 27 in Llanfihangel yn Ngwynfa and over 25 in Gosforth.Footnote ⁵⁸

4. SUPPORT FROM OUTSIDE THE HOUSEHOLD

A number of cases of children supporting elderly widowed or disabled parents even when they did not live with them, have been documented for English populations. One elderly man in Corfe Castle in 1790 was, for example, at least in part supported by his children who lived elsewhere.Footnote ⁵⁹ On the other hand, Frederic Morton Eden documented from Seend, Wiltshire in 1796 much more limited assistance to a widow from a son aged 18 living independently, who was only giving his mother sixpence a week as payment for her washing and mending his clothes. The rest of his earnings of ten shillings a week as a bricklayer were devoted to his own maintenance. From the information given by Eden, sixpence constituted 5 per cent of his earnings and just over 3 per cent of the income of his widowed mother, who had several wage-earning children and was in receipt of poor relief.Footnote ⁶⁰

Examination of some family budgets from Corsley, Wiltshire in 1906 reveals the variety of arrangements that might be made. One aged widow living on her own was largely supported by a daughter who paid her mother's rent of £5 and sent her bacon, butter, potatoes and firewood should her store be exhausted.Footnote ⁶¹ Analysis of the housekeeping budget indicated that the daughter provided 61 per cent of the food budget in the first week of observation and 81 per cent in the second.Footnote ⁶² A niece who lived nearby looked after the widow, and a great-niece recorded the expenditure and the nature of the meals that her great-aunt was eating. Three children helped to support another widow in Corsley: two married sons each contributed two shillings per week and a married daughter made a home for her mother, taking three shillings and sixpence (88 per cent of the four shillings) to meet the cost of her keep. In the summer the mother undertook some housework while her daughter was working in the garden and in return the daughter gave her a share of the profits from the sale of garden produce as pocket money, amounting in all to 12–13 shillings.Footnote ⁶³ The experience of this family thus demonstrates not only co-operation between siblings in the provision of care to an elderly parent but also the way in which an elderly widow could still make a contribution to the family economy and see that work rewarded by a transfer of cash from daughter to mother.

In general, however, married children with families of their own could only provide limited support to their parents. In 1906 in Market Drayton, Shropshire married sons were typically paying six pence or one shilling to the Poor Law Guardians as partial compensation for the amount of outdoor relief the Guardians were providing to their widowed parent.Footnote ⁶⁴ These amounts represented between 3 and 7 per cent of their earnings and between 17 per cent and half of what had been granted in poor relief. Many married children with households of their own did not provide any financial support to their parents. Analysis of the information that was provided on some of ‘typical cases’ selected by the Royal Commission on the Poor Laws from the list of outdoor poor in Govan, Glasgow, again in 1906, shows that only about half of their married children provided any assistance to the widowed mothers.Footnote ⁶⁵ The evidence collected by Charles Booth on the economic circumstances of the elderly in rural areas of England in 1892 also indicated that support from family members (identified simply as ‘relatives’), when available, was usually combined with other sources of income. The amount of assistance was not specified but the sources of support were given. Only 5 per cent of the elderly were maintained solely by relatives while almost a quarter supported themselves from their own earnings and close to a further quarter from their savings and property.Footnote ⁶⁶ A quarter of all rural elderly people were reported as receiving some assistance from relatives. This compares with the almost half of the elderly who relied on their earnings to supply a portion of their income, the third that used savings and property and the fifth who received some assistance from the Poor Law. Another fifth had no other assistance than that provided by a relative. Looking specifically at those receiving support from their relatives, a third were also assisted by the Poor Law, almost a third by charity, a fifth from earnings and 14 per cent from property or savings.

Family budgets also indicate that on occasion some support might be provided by siblings who resided elsewhere, by more distant relatives and even by neighbours.Footnote ⁶⁷ However, such support (often provided in kind rather than as cash) was usually limited and – taking into account the number of potential donors, whether neighbours or relatives – was not often available.Footnote ⁶⁸ Employers (or former employers) and local landowners might also provide some assistance. This could involve the provision of food while at work, allowing their labourers to buy grain at below the market price and permitting the gathering of firewood. The value of such assistance is difficult to estimate, given the many forms it might take, but it was almost certainly less than the pensions some employers were paying to elderly former employees or their widows in the Thingoe Poor Law Union in Suffolk in 1906. These pensions of two shillings a week, paid by agents employed to manage the estates of the landowners, represented 80 per cent of the amounts granted as out relief by the Poor Law Guardians. The payments of out relief were, however, apparently not curtailed nor were the pensions even recorded in the records of the Poor Law because these agents, many of whom also served as Poor Law Guardians, thus ensured higher incomes for their pensioners at very little cost to the landowners.Footnote ⁶⁹

5. DEMOGRAPHIC DETERMINANTS OF THE KINSHIP NETWORK

Important new insights into how the composition of kinship networks of societies in the past have been transformed have now become available with the development of the technique of micro-simulation, which uses the rates of fertility, nuptiality and mortality for different populations to estimate the number of kin an individual man or woman would have as they aged. Micro-simulation has demonstrated that the impact of demographic change on the size and composition of the kinship network could be considerable. For example, according to J. E. Smith and J. E. Oeppen, using the micro-simulation program CAMSIM, a man born in 1550 could expect to have 2.2 children still alive and 5.2 live grandchildren when he reached the age of 65. A man born a century later, on the other hand, could expect to have at age 65 only 1.9 children and 2.8 grandchildren. A century later, the numbers had recovered to leave such a man at age 65 with 2.6 children and 5.7 grandchildren.Footnote ⁷⁰

More recently, the micro-simulation technique has been extended to cover sub-groups of the population such as the widowed, revealing evidence of substantial changes to the kin universe of widows in England over much shorter periods of time, and specifically between the second half of the seventeenth and the early nineteenth century. As the mean age at which a woman was widowed rose from 47.4 to 53.3, widows in the early nineteenth century had greater numbers of close kin such as children, siblings and grandchildren and potentially more effective kin support as, on average, the relatives were older. The percentages of widows with at least one surviving child rose from 81 to 88 per cent, of those with a sibling from 76 to 87 per cent and of those with a son or daughter-in-law from 22 to 46 per cent.Footnote ⁷¹ On the other hand, as there is reason to consider that ties of kinship often involved obligations as well as dependence, it is also likely that in the early nineteenth century there were more kin whom an individual might need to assist and a greater possibility that family resources would have to be shared between members of different generations during their lifetime rather than their being received in the form of an inheritance.

Variations in the availability of kin over time can also be profitably compared with rates of co-residence with kin and the numbers of kin living locally as a way of assessing the strength of kinship ties in different populations.Footnote ⁷² For example, Ruggles has calculated for the United States that in 1860 virtually all elderly persons with surviving children would live either with them or next door, with just a small decline (11 per cent) by 1900 in rates of co-residence between elderly non-married women and their children.Footnote ⁷³ Kinship ties seem to have been weaker in England in that estimates suggest that in 1891 a third of widows with surviving children did not co-reside with them.Footnote ⁷⁴ Rates of co-residence of a widow and at least one of her children in England in previous centuries were also considerably below the maximum that micro-simulation indicates was possible. In the second half of the seventeenth century, 64 per cent of widows co-resided with a child while the simulation suggests that 81 per cent of widows would have had children. In the early decades of the nineteenth century only 62 per cent of widows lived with a child, although the percentage with living children had risen to 88 per cent. On the other hand, calculations by Lyn Boothman for Long Melford, Suffolk in the late seventeenth century indicate that 63 per cent of men in their mid-sixties had at least one child who also lived in Long Melford, representing 90 per cent of the men estimated from micro-simulation to have at least one surviving child.Footnote ⁷⁵ This proportion of elderly men with surviving children where at least one lived in Long Melford is close to the proportions of the elderly in the United States in the nineteenth century living with or next door to their children, as calculated by Ruggles.

The ability of kin to function as support could, however, be compromised by population mobility. There are two issues to consider: what proportion of relatives did live locally and whether residential proximity has lost some of its significance in present-day society. Even in the past, relatives, particularly those more distantly related, were very likely to be living not only in another household but in another parish, as further evidence from Long Melford makes clear. Micro-simulation suggests that in seventeenth-century England about 7 in 10 males aged 3–7 could expect to have a living grandparent, yet of those resident in Long Melford, in 1684 only 16 per cent of boys in early childhood had a grandparent who lived locally.Footnote ⁷⁶ For nearly 80 per cent of these boys who had a grandparent, that grandparent lived elsewhere. In early adulthood (between the ages of 23 and 27), two-thirds of the men who had at least one uncle or aunt had none who resided in Long Melford. On the other hand, their parents and siblings were much more likely to be present in Long Melford: 80 per cent of these young men who still had a parent living and 82 per cent of those with a surviving sibling. As is well known, most migration rates were particularly high during adolescence and early adulthood and this too had consequences for the kinship network, with only apparently random changes in the proportions of siblings living locally after the age of 35.Footnote ⁷⁷ The proportion of men who had cousins identified as resident in Long Melford also ceased to decline once the men reached their mid-twenties.Footnote ⁷⁸

Another study indicates that there has been surprisingly little change between the late eighteenth and the late twentieth century in the proportion of married children living at no great distance from their parents.Footnote ⁷⁹ For example, nearly half of all married sons lived within 5 miles (8 kilometres) of their parents who resided in the Bedfordshire parish of Cardington in 1782. Two-thirds of married sons whose parents lived in the Nottinghamshire parish of Caunton in 1846 had their own homes no more than 5 miles away. For comparison, a study of kinship patterns in five towns in 1983 indicated that a third of married sons were living within 5 kilometres (3 miles) of their parents.Footnote ⁸⁰ From the point of view of elderly parents, the important criterion would have been the proximity of the nearest child. In Cardington in 1782 a third had a child resident in the same parish; in the five-town study in 1983, 36 per cent of elderly parents lived within 1 kilometre of one of their children.Footnote ⁸¹

The fact that the proportion of adult children living close to their parents has not changed over the past two centuries is made more interesting by the fact that improved means of transport and the development of new means of communication have made it much easier to keep in contact even when people were not living in the same locality. In the past residential proximity was critical if practical help was needed, whether with house repair or personal care, and was also important for casual social contacts. In a study of residence patterns in Wolverhampton in 1948 J. H. Sheldon selected five minutes' distance to define ‘living near’, on the basis that this represented the maximum distance that would allow a relative to transport a hot meal without the need for re-heating.Footnote ⁸² Admittedly, in some cases residential proximity will still be important today, particularly if practical assistance or care is needed on a medium- to long-term basis, but social contacts can now be maintained regardless of the distance between residences – although whether they are as satisfying as meeting in person is doubtful.

One further issue that needs to be considered is whether family support was available when it was most needed, for example following widowhood or in old age. The populations from the eighteenth and nineteenth centuries that have yielded the information on how close parents lived to their married children are unfortunately too small to permit effective analysis of variations in rates of residential proximity according to whether the parents were married or widowed. However, there is evidence from late-nineteenth and early-twentieth-century England that indicates that the widowed were more likely to have their children living with them than was the case for married couples of the same age. The differences, however, were relatively slight and widows over the age of 75 were much less likely than were widows in their late 50s or 60s to share a household with an adult son and were also slightly less likely to have an adult daughter present.Footnote ⁸³ What these results indicate, therefore, is that whatever additional support was needed after widowhood or in old age it was not provided via higher rates of co-residence with an adult child.

6. THE STRENGTH OF KINSHIP TIES AND THEIR LIMITATIONS

The evidence presented above has illustrated a number of the different aspects of kinship patterns in English society in the past. The discussion has included some of the ways members of particular populations, as well as historians and sociologists, have interpreted the presence of different types of kin in certain households; the extent of the economic support provided to their parents by children of working age living in the parental household; the importance of assistance, whether in the form of cash, kind or work, provided by children and other relatives who had households of their own; and the impact of demography on the ability of kin to render effective assistance.

Some of the evidence is admittedly patchy, its geographical, temporal and social compass being limited in ways that hinder effective assessment of the significance of the role of kin in English society. Nevertheless a number of conclusions can be drawn. One of the most striking aspects is that a variety of evidence suggests that kin were of limited importance. There is, in the first place, the fact that kin constituted such a small component of English households, much smaller than servants for example.Footnote ⁸⁴ Secondly, migration limited the effectiveness of kin who moved elsewhere to render everyday assistance. Such considerations may perhaps explain why, as noted above at the beginning of this article, the authorities responsible for the 1851 Census of Great Britain emphasized the importance of co-residence regardless of whether the persons who lived together were related or whether they had close relatives such as children residing elsewhere. Thirdly, it has been possible to show that even those children of working age who lived with their parents often withheld a sizeable proportion of their earnings for their personal expenditure. Fourthly, it is clear that children or other relatives were only one of a number of sources of support that were potentially available in the event of need. Combinations of different sources of support, including personal earnings, savings, property, relatives and the Poor Law, were usual rather than exceptional. Only a fifth of the elderly recorded as ‘assisted by relatives’ in Booth's study of the elderly in England and Wales in 1892 were not also supported in some other way, either from their own labour or savings or by the Poor Law or by charity.Footnote ⁸⁵ Yet support from the Poor Law did not preclude assistance being provided by adult children and it might even be encouraged.Footnote ⁸⁶ Finally, it is clear that the existence of close kin, such as the potential donor's own children, limited the extent of the assistance that could be provided to more distant kin just as under inheritance law in England closer kin excluded the more remote, as pointed out by Lloyd Bonfield in this issue.Footnote ⁸⁷

On the other hand, there is also evidence of very significant support from family members in some circumstances: by children living with their widowed mothers, for example, or in some cases from children and other relatives living elsewhere.Footnote ⁸⁸ In times of crisis, such as when there was a shortage of affordable housing after the two World Wars, households could expand as parents found a place in their home for a married child. It was in these years that households in England were most likely to contain relatives beyond the core group of parent and unmarried child.Footnote ⁸⁹ It is also apparent that the tasks undertaken by kin could be diverse. Some household members were present in anticipation of inheritance, others for work-related reasons and many to perform a caring or service role within the household or to receive care.Footnote ⁹⁰ This article has placed stress on the role of kin as providers of welfare, but this reflects the fact that much of the evidence presented above concerns the lives of the poor. For other social groups, farmers or entrepreneurs for example, work and inheritance considerations were of greater importance. It is also important to recognize that acceptance of assistance, if not provided in recognition of past service, could impose a moral obligation on the recipient to help the donor in the future, should the occasion arise. Michael Anderson interpreted the motivations for co-residence of relatives as calculative, in that all the parties involved could see the advantages that they would obtain – either immediately or in due course – from such an association.Footnote ⁹¹ Rees also thought that some assistance provided to neighbouring farmers in Llanfihangel yng Ngwynfa was intended to place them under an obligation to reciprocate.Footnote ⁹² Yet much assistance was influenced by custom, by what was expected and sometimes without any realistic prospect of its being reciprocated.Footnote ⁹³ Moreover, as both Anderson and Rees noted, even calculative behaviour can generate its own norms.

Of course, although it is possible to distinguish these different aspects of kinship, in many cases the roles overlapped. As pointed out above, heirs worked for their parents while awaiting their inheritance. Sons who were to marry off the farm and establish their own households also contributed their labour to the parental economy in anticipation of assistance from the parents in establishing their own economies or as payment for previous assistance. An elderly mother housed by her daughter helped with housework. A small kin component within the household does not necessarily mean, therefore, that kinship was unimportant.

There is also some evidence that suggests that the provision of assistance was seen as the responsibility of close kin, with more distant kin or neighbours intervening only in the absence of close kin. For example, analysis of Flora Thompson's classic account of her early childhood as the daughter of a stonemason in the 1880s in Juniper Hill on the Oxfordshire–Northamptonshire border revealed that neighbours would offer to help at a confinement or with housework, cooking or washing, but only if there was no daughter or other female relative available. Flora Thompson had a good deal to say about neighbours and considerably less about kin; nevertheless, in major respects relatives were more important than neighbours.Footnote ⁹⁴ In this poor community, only relatives ever ate with the family and only relatives explicitly intervened in the life of the family by chastising the children.

What is much more difficult to assess is the extent to which the role of kin in the social structure of England may have evolved in the long term. Micro-simulation has proved a very useful tool in showing how changes in fertility, nuptiality and mortality can transform the composition of the kin group in both the medium and the long term – as Murphy demonstrates in this issue. The technique has, for example, established that by the early nineteenth century a widow in England had more close kin, and potentially more effective kin, because her children and grandchildren were somewhat older at the onset of her widowhood than had been the case for widows in the second half of the seventeenth century. For men born in the mid-sixteenth and mid-seventeenth centuries, micro-simulation indicates a decline in the numbers of children and grandchildren alive when a man reached his mid-60s. Admittedly, one weakness with the CAMSIM simulation is that currently it is based on a composite measure of the demographic history of England, but it would be possible to introduce variants to take account of differences between social groups in age at marriage, fertility and mortality and thereby to determine whether they differed in the numbers, relationships and age distribution of their kin. Micro-simulation could also be exploited to explore how kinship patterns might be reshaped as a result of a major demographic crisis involving lower fertility, higher mortality and increases in remarriage rates. However, even refined, micro-simulation can only reveal those kin who were potentially available for interaction; there can be no certainty that these interactions actually occurred.

On the other hand, evidence has been produced above which shows that the extent and type of interactions with kin could vary considerably, depending on the nature of the local economy. Kinship ties within and between households were very important in the pastoral farming communities in Wales and Northern England studied respectively by Rees and Williams in the 1940s and 1950s, and so too were ties with neighbours.Footnote ⁹⁵ Contacts with kin appear to have been much less frequent in Terling, Essex in the seventeenth century.Footnote ⁹⁶ There may well, therefore, have been marked regional differences in patterns of interaction with kin in the English past and even intra-regional differences where there were distinctive local economies,Footnote ⁹⁷ just as there were major differences in farm economies in terms of farm size, whether pastoral or arable farming and the extent of reliance on wage labour. As farming practices evolved, for example with shifts to or from arable farming and with the mechanization of farming practices which previously had been labour-intensive, so the role of kin in such communities is likely to have been transformed, increasing with a shift to a pastoral economy and reducing with the introduction of machinery. The nature of interactions with kin will also have changed with industrialization and urbanization in the nineteenth century, probably lessening in intensity due to more varied work and housing environments, although kin were important contacts for new migrants, and potential migrants, to cities.Footnote ⁹⁸