Floud R., Johnson P. The Cambridge Economic History of Modern Britain, Volume 1: Industrialisation, 1700-1860
Подождите немного. Документ загружается.
Household economy 241
households were universally and necessarily larger and more complex
(Laslett 1972a), and that the advent of industrialisation brought a transi-
tion from the extended to the nuclear family (Flandrin 1979).
Polarising the history of the household into pre- and post-industrial
was increasingly out of sync with the revisionist view of the industrial
revolution as involving continuity as well as change. Moreover, both evo-
lutionary and structuralist–functionalist models of the household fell
increasingly foul of growing empirical evidence, which suggested that
therelationship between industrialisation and the household was more
involved. Surprising evidence demonstrated that the English household
wasnot only small and nuclear long before industrialisation but also re-
markably homogeneous across time and space, with perhaps important
implications for understanding England’s precocious economic develop-
ment.
THE SIZE AND COMPOSITION OF
THE PRE-INDUSTRIAL HOUSEHOLD
Some scholars had long resisted the sentiment that the pre-industrial
household was large and complex by pointing to surviving empirical
evidence on household size (Laslett 1972a: 1–13; Wall 1972: 191). The
breakthrough came with the collection and analysis of surviving English
‘listings’ by the Cambridge Group for the History of Population and Social
Structure (Laslett and Harrison 1963; Laslett 1969, 1972b). These docu-
ments enumerated every individual of a particular settlement according
to the household in which he or she belonged, and so enabled the system-
atic quantitative study of household size and composition over time and
across communities of different types. The findings were cataclysmic for
thepresumption that pre-industrial households were large and complex.
Households were small. The majority contained fewer than five persons,
and membership was customarily confined to parents and their unmar-
ried children. Some households contained servants, but there were re-
markably few complex households containing grandparents, parents and
grandchildren.
Tables 9.1 and 9.2 present the estimates of household size and key
quantifiable characteristics of household structure (percentages of house-
holds with kin, lodgers and servants) respectively from the hundred ‘pre-
industrial’ communities originally studied.
The tables also provide comparable figures for England and Wales, and
(following Anderson 1971b and 1972) for Preston and Swansea for 1966,
to show how ‘modern’ the pre-industrial family appeared. True the pre-
industrial family was larger and much more likely to contain servants,
but not much larger, and crucially had only about the same likelihood
of including kin. The large and complex households of western nostalgia
Cambridge Histories Online © Cambridge University Press, 2008
242 Jane Humphries
Table 9.1 Number of persons per household (percentage in various communities)
Household size Mean
household
12345678910+ size
Pre-industrial communities
1564–1821 (N = 100) 61417161512 8535 4.8
England and Wales 1966 15 31 21 18941100 3.0
Preston 1966 18 32 19 15852110 2.9
Preston 1851 1 10 16 17 14 12 10 8 5 8 5.4
Rural Lancashire 1851 3 12 13 12 14 12 11 9 6 9 5.5
York 1851 51516181413 7535 4.7
Nottingham 1851 4.47
Ashford 1851 4.85
Pre-industrial communities
1650–1749 (N = 45) 4.696
Pre-industrial communities
1740–1821 (N = 50) 4.776
Sources: Figures for England and Wales 1966, Preston 1966 and rural Lancashire 1851 from Anderson 1972: table 7.1. Figures for York 1851 from
Armstrong 1972: table 6.1. Figures for Nottingham and Ashford from Armstrong 1972: 211. Figures for pre-industrial communities from Laslett
1972b: table 4.4.
Table 9.2 Percentage of households with kin, lodgers and servants for various communities
Percentage of households with:
kin lodgers servants
Pre-industrial communities 1564–1821 10 <1* 29
England and Wales 1966 10 – 0
Swansea 1966 10–13 <3 <3
Preston 1851 23 23 10
Rural Lancashire 1851 27 10 28
York 1851 22 21 20
Nottingham 1851 17.3 21.8 11.7
Ashford 1851 21 17.5 16.9
Potteries 1861 18 18 9–11
Sources: See Table 9.1. Potteries figures calculated from Dupree 1995: tables 2.2, 2.4 and 2.8b.
* Probably an underestimate, see Anderson 1972: 220.
appeared just that, as far as English history was concerned: figments of a
collective imagination that yearned for a more sociable and less isolated
family life.
The precocious development of wage labour in the English country-
side contributed to the high frequency of small households of two or
three people (for the relatively small size of labourers’ households in the
pre-industrial sample see Laslett 1972b: 154). But generally, pre-industrial
households were not collections of proletarians. The empirical evidence
forthe widespread existence of household production was too ubiquitous
Cambridge Histories Online © Cambridge University Press, 2008
Household economy 243
to be disputed. Why then were households so small and simple? What
forces had produced these modern-looking households in a pre-industrial
environment? The answer required locating the producing activities of
the household in their demographic, economic and cultural context. Con-
ditioning variables were specified as fertility, mortality, expectation of
life, age at marriage, propensity to stay in the parental home after mar-
riage, and relative frequency of widows and eldest sons taking over the
headship of the household after the death of the master. Extended fam-
ilies were rare, produced by freak coincidences within the conditioning
variables, ‘fortuitous outcomes of demographic eventualities and eco-
nomic conveniences, and of particular strong personal attachments as
well’ (Laslett 1972a: 73).
Continuity was assured by the priority given to demographics. These
ripped and tore families into the simple shapes observed. Low life ex-
pectancy for example put an upper limit on the proportion of all fam-
ilies that could include grandparents (Wrigley 1969). But economic and
cultural factors also mattered. Late marriage as well as early death con-
strained the maximum frequency of three-generation households and
residential preference determined what proportion of those who were
demographically able to live with kin actually chose to do so. Economic
conditions, in turn, could bring forward or push back the age at marriage,
and cultural norms could require or not the formation of an independent
household on marriage. Demographic history was to show the importance
of these variables in understanding English population growth (Wrigley
and Schofield 1981; Wrigley et al.1997; and chapter 3 above). In the longer
term,economic changes could also influence mortality and fertility.
The discovery that the pre-industrial household was small and simple
exploded belief in the rise of individualism as a universal explanation of
familial change, and prompted speculation about the role of household
structure in economic development. ‘England had been the first of the
world’s societies to undergo [an industrial revolution], and it seemed quite
possible that her pioneering role might have had something to do with
thesimple structure and small size of English households before ever
industrialisation began’ (Laslett 1972a: 49). Perhaps the history of the
household provides another brick in institutionally driven accounts of
the firstindustrial revolution.
Subsequent research on pre-industrial household size and composition
tended to confirm the initial findings (Schurer 1992; see Wall 1983 for
a comparative European perspective). But the vision of stable continuity
left by the English empirical tradition was misleading. Very different de-
mographic, social and economic forces can produce similar outcomes.
The seemingly constant household could disguise a maelstrom of under-
lying change. Moreover, further scrutiny of the evidence suggested that
farfrom exhibiting continuities both with the past and the present, the
households of the industrial revolution were themselves distinctive.
Cambridge Histories Online © Cambridge University Press, 2008
244 Jane Humphries
THE SIZE AND COMPOSITION OF THE HOUSEHOLD
IN THE INDUSTRIAL REVOLUTION
Even within the original evidence, there were suggestions that the con-
trast was not between pre- and post-industrial households but between
the households of the industrial revolution and those that went before
and came after. Laslett subdivided his original data into two sub-periods,
1650–1749 (forty-five communities) and 1750–1821 (fifty communities), to
suggest that smaller households were more common in 1650–1749, al-
though mean household sizes for the two periods remained close (see
Table 9.1). Ironically it looks as though household size swelled just as
England began to industrialise.
This intriguing suggestion was endorsed by Michael Anderson’s (1971a)
pioneering study of household structure in the industrial revolution.
Anderson calculated household size and the percentages of all house-
holds that contained kin, lodgers and servants, from a 10 per cent sample
taken from the enumerators’ books of the 1851 Census for Preston. The
Lancashire cotton town was viewed as a ‘half-way house between a pre-
dominantly rural pre-industrial England, and the predominantly urban-
industrial/commercial post-capitalist England of the present day’ (1972:
215). Anderson’s findings are shown in Tables 9.1 and 9.2, which also in-
clude summary statistics of household size and structure from a number
of other samples to identify the effects of industrialisation and to check
whether Preston was indeed representative of an environment in tran-
sition. These include: first, a rural sample that Anderson drew with the
object of comparing the family structure of migrants to Preston with the
family types which were found in the villages from which in-migrants
came; second, another sample drawn from the enumerators’ books of
the 1851 Census, but for York, a very different kind of mid-nineteenth-
century town, in which ‘true factory or large-scale production was un-
known’ (Armstrong 1972); third, evidence for Nottingham and Ashford,
nineteenth-century towns which were comparable with Preston and York
respectively (Armstrong 1972); and finally, a sample from the 1861 Census
forthe Potteries, another comparable manufacturing community (Dupree
1995).
Mean household sizes in Preston and rural Lancashire were much
higher than the 1960s benchmarks and somewhat higher than the pre-
industrial mean. This was partly a product of the comparatively high
proportions of households that contained kin in comparison with mod-
ern and pre-industrial communities. Table 9.3 provides some detail on
thekin relationships. Ten per cent of Preston households contained par-
ents and married children, a family composition that was rare in pre-
industrial England and uncommon too in rural Lancashire. Households
that contained other kin were also more numerous in Preston than in the
Cambridge Histories Online © Cambridge University Press, 2008
Household economy 245
Table 9.3 Structure of the families of household heads
Pre-industrial England and Wales Rural Lancashire
Family type communities 1966 (approx.) Preston 1851 1851 Potteries 1861
No related person 17 4 5 5
Married couple only 90 24 10 12 12
Parent(s) and
unmarried children only 49 63 56 65
Parent(s) and married child(ren) 5 9 6
but no other kin
Parent(s) and married child(ren) 10 0 1 0 8
with other kin
Other combinations of kin 4 13 21 10
All (percentage) 100 99 100 100 100
N = 1,533,954 1,240 855 1,432
Sources: See Tables 9.1 and 9.2. Figures for Potteries calculated from Dupree 1995: table 2.2.
pre-industrial communities, though rural Lancashire had an even higher
incidence.
WasPreston’s pattern of co-residence the product of its urban-
industrial environment? Here the comparison with other nineteenth-
century communities is instructive. All the nineteenth-century commu-
nities shared Preston’s relatively high incidence of extended family living
but there were differences in household composition. Oldham, another
nineteenth-century textile town, emulated Preston’s incidence of parents
co-resident with married children (Table 9.3; Foster 1974). But the surpris-
ing popularity of this type of household did not extend to Northampton
or South Shields, where only 5 and 4 per cent of all households respec-
tively included parents and married offspring (Foster 1974). On the other
hand, households that contained other kin were relatively numerous in
Northampton (12 per cent) and South Shields (11 per cent).
Preston also had far larger proportions of households containing
lodgers in comparison with both pre-industrial and 1960s England. House-
holds with lodgers were much less common in rural Lancashire, though
still more frequent than in the pre-industrial or modern communities.
About one-fifth of households in all the industrial towns took in lodgers.
Preston had the highest proportion at 23 per cent, with the other towns
close behind in a cluster.
The proportion of households with servants in Preston was much lower
than in the pre-industrial communities, while they had practically dis-
appeared in modern England. Mid-nineteenth-century rural Lancashire
closely resembled the pre-industrial communities. The relatively low pro-
portions of households with servants in the other industrial towns sug-
gest that servants were disappearing from urban industrial households,
although again Preston stood at the extreme of experience. Moreover, one-
fifth of households in the market towns continued to include servants.
Cambridge Histories Online © Cambridge University Press, 2008
246 Jane Humphries
Subsequent studies have confirmed a relative rise in extended family
living during the mid-Victorian era in Britain and the USA. Ruggles (1987)
analysed twenty-seven separate studies of sixty-eight data sets drawn from
localities and national samples in England and America between 1599 and
1984. Although the majority of households never contained extended kin,
theminority that did approximately doubled between 1750 and the late
nineteenth century, to about 20 per cent of all households (Ruggles 1987:
6). While these findings gave no heart to the old myth that there was a
transition from the extended family to the nuclear family at the time of
theindustrial revolution, they also exploded the new myth that American
and English family structure was always and everywhere overwhelmingly
nuclear. However, the range of experience suggested by the community
studies counsels against taking Preston as a benchmark even for mid-
nineteenth-century urban industrial Britain, let alone the environment
as a whole. Any explanation of household composition must address not
only Preston’s often extreme characteristics but the variation around it
marked out by the other towns.
SELF-INTERESTED INDIVIDUALS AND HOUSEHOLDS
OF MUTUAL ADVANTAGE
The structuralist–functionalists and the English empiricists failed to
translate their macro-models into an account of the human decisions
involved in the formation and dissolution of households and the draw-
ing of their boundaries. The former depicted adaptive behaviour taking
place at the level of society as a whole, while the latter focused on the
measurable characteristics of households detectable in historical records.
More recent approaches such as the New Home Economics (Becker 1965,
1981)and the ‘family strategies’ perspective envisaged the household as
adopting the size, composition and employment structure that was most
advantageous collectively (Tilly and Scott 1978; Hareven 1982). How these
decisions came about and were acted upon remained mysterious. What
was needed to tie the patterns in size and composition to potential social
and economic explanatory variables was a credible account of individ-
ual behaviour with respect to household membership. Without such a
microanalysis of the household, its history had no driving force.
The explanatory vacuum at the micro-level left by the other approaches
wasfilled by Anderson’s instrumentalist utilitarian account of family re-
lationships. In Anderson’s work the ‘economic conveniences’ that Laslett
had acknowledged could influence size and structure emerged in a fully
developed micro-theory of the household based on reciprocal exchange
and mutual advantage.
Co-residence was explained by reference to the advantages and
disadvantages that individuals experienced as a result of household
Cambridge Histories Online © Cambridge University Press, 2008
Household economy 247
membership (Anderson 1972: 226). In particular, large numbers of indi-
viduals from one group of kin (say married children) would only be found
co-residing with another group (say widowed parents) if the present value
of average, lifetime economic advantages of doing so outweighed the eco-
nomic disadvantages. Benefits had to be mutual and calculated on the
understanding that co-residence implied some sharing of resources if re-
quired.
The theoretical framework facilitated an explanation of the distin-
guishing features of the households. The propensity of rural households
to contain kin was easily explained. North Lancashire, unusually for
nineteenth-century rural England, had a high proportion of prosperous
family farms on which kin were employed to mutual advantage (see also
Williams 1963). But on-going economic co-operation could not explain the
high propensity of Preston’s wage-dependent households to harbour kin.
To some extent the tendency to include kin simply represented ‘piling
up’ or ‘huddling’, as individuals sought to spread rents over larger num-
bers in rapidly growing cities. But there was another advantage to co-
residence, for networks of relatives often helped individuals obtain work.
Where kinsmen had small businesses they could provide employment,
and thereby overcome various information problems in hiring and mon-
itoring. Even where households were fully wage dependent, access to em-
ployment could be eased by the information and introductions provided
by parents, uncles, cousins, in-laws and siblings.
What about the higher incidence of married couples co-resident with
parents, unusual seemingly everywhere but Preston and Oldham? Again
theanswer lay in mutual advantage. Members of the working class who
survived to old age faced deprivation unless they could live with married
children, sharing rent and common consumption. Young married couples
could benefit from savings on rent, but the downside of sharing with par-
ents loomed large. As relatives aged they were likely to become liabilities.
Denied co-residence, old people would be saved by the poor law from abso-
lute destitution (Thomson 1991; Smith 1998; Thane, 2000). These consid-
erations kept the proportions of co-resident parents and married children
down in pre-industrial rural areas and most nineteenth-century towns. In
Preston and Oldham, poverty was less intense and there was more room to
manoeuvre (Foster 1974; Anderson 1972). More importantly, the cotton in-
dustry offered employment for married women, thus creating a demand
forsubstitutes to provide childcare and housework. By providing these
services, elderly relatives could reciprocate for bed and board. Everybody
gained. Later work on the Potteries, where poverty was also less biting
and local industry again afforded employment for married women, con-
firmed these links with family structure and composition echoing those
of the textile towns (Dupree 1995; see Table 9.3).
Local opportunities for children’s employment also influenced fam-
ily composition. In Preston children worked in the mills and as
Cambridge Histories Online © Cambridge University Press, 2008
248 Jane Humphries
child-minders, and did not have to leave home in search of jobs, as did
children in pre-industrial England, most rural areas of industrialising
England, and even York. Guaranteed employment, orphaned relatives or
theunderemployed children of kin were attractive additions to Preston
households, thus explaining the large numbers of ‘parentless children’
detected among other kin.
The proportions of households containing servants and lodgers were
also plausibly related to mutual advantage. In Preston the relative absence
of small businesses, the homogeneity of income levels and the existence
of jobs for young people that did not involve co-residence meant that ser-
vants were relatively uncommon, while in Preston and York the relatively
high costs of housing increased the incidence of lodging.
Thus mutual advantage explained not only the distinctive patterns of
co-residence observed in Preston in comparison with pre-industrial house-
holds but also the variations across nineteenth-century industrial and
non-industrial towns. These were related to local differences in housing,
poverty, employment opportunities for women and children, mortality
and migration. For example, the dominance of the nuclear family in the
Potteries and the low incidence of sharing with people outside the nu-
clear family, related to the lower migration into the Potteries and the
more plentiful supply of housing (Dupree 1995).
Moreover, contrary to the conventional wisdom, as the industrial rev-
olution got underway, households acquired new functions as they shed
old ones. The organisation of early industrial labour markets devolved
many functions onto workers’ own families. Relatives, even when them-
selves only employees, often had the power to hire directly. Spinners
hired their own piecers, potters their mould runners, hewers their draw-
ers (Shaw 1903; Smelser 1959, 1967; Humphries 1981). Family teams of
workers extended beyond their well-documented presence in spinning
factories (Smelser 1959; Collier 1964) to many other early industrial work-
places. Working with other family members had advantages. Patriarchal
authority and familial loyalty were adapted to create effective hierarchy
within the labour process, and in dangerous workplaces the presence of
trusted and reliable work-mates increased individual security. Thus, in
thebadly ventilated mines of the industrial revolution, family members
were preferred as co-workers in the belief that they were more likely to
afford assistance in the event of danger (Humphries 1981). The primitive
and contested nature of work evaluation in the early mills and mines,
where workers were often on piece-rates, also put a premium on includ-
ing family members in the work team. Their presence where work was
checked and weighed guarded against cheating by managers and fore-
men and they could be relied upon to struggle for advantages in the
allocation of work and access to equipment (Humphries 1981). Help find-
ing work was particularly important to newcomers to towns and cities,
and strengthened ties with more distant kin, ironically creating a motive
Cambridge Histories Online © Cambridge University Press, 2008
Household economy 249
forextending households in ways which were rare among pre-industrial
producing households.
Today historians have become accustomed to the invasion of eco-
nomics, and rational economic man pops up in the strangest places. But
in 1971, historians accustomed to think of family relations as normative
and affective, balked at his appearance in the households of the indus-
trial revolution. Reference to norms, which freed individuals from inter-
minable computation while ensuring an efficient response, did little to
calm their unease. The reduction of emotional attachments to ‘particu-
lar non-marketable household commodities’ seemed ‘unsubtle’ (Ruggles
1987: 17). The circumstantial nature of the evidence, inferring intention
from outcome, was suspect, prompting Anderson himself to look for sup-
portive accounts of motivation (1972: 231), many of which were how-
ever consistent with more solidaristic interpretations of family relations
(Dupree 1995: 25).
Non-economic explanations of family structure were offered as sup-
plements if not alternatives. The ‘personal attachments’ that Laslett had
grudgingly accorded a role in explaining unusual family forms were
generalised in cultural and emotional explanations of the rise in ex-
tended family living during industrialisation. Urbanisation, mobility, and
widespread economic and social change created unprecedented turmoil
and sense of insecurity, causing many to retreat into the home, the fam-
ily and intense affective relationships (Lasch 1977; Ruggles 1987). Demo-
graphic conditions too had changed in ways that were favourable to ex-
tended families. Increased longevity and earlier marriage, by increasing
theavailability of kin, expanded the proportion of the population that
could choose to live with relatives (Ruggles 1987: 125).
The economistic approach to the household collided with alternative
emotional and cultural interpretations over the extent to which families
provided assistance. It was easy to find examples of help but difficult to
assess whether they were isolated or representative. Care for the elderly
wasanimportant issue. Although the Elizabethan poor law required chil-
dren to support aged parents, poor law records and lists of poorhouse
inhabitants suggest that by the eighteenth century responsibility for the
elderly was shouldered by the parish (Thomson 1991; Thane 2000). On
theother hand, that families only provided help ‘where exchanges were
reciprocal and fairly immediate’ (Anderson 1971: 8) was disputed. Some
historians could find little evidence of a ‘calculative orientation’ among
family members (Roberts 1988: 172; see also Finch 1989). Care or food
was often provided even when poverty held back subsidies. Networks of
non-pecuniary assistance were particularly important for women (Finch
1989; Ross 1993). By subsidising kin and neighbours who were providing
help, the old poor law may have strengthened family and community re-
lations so that they withstood pressure on other occasions. The persistent
miserliness of the old poor law in the north and west (King 2000) left the
Cambridge Histories Online © Cambridge University Press, 2008
250 Jane Humphries
family, for all its drawbacks, as a major provider of assistance in the in-
dustrial districts. The hardening of attitudes in the south and east at the
end of the eighteenth century, and the transition to the new poor law
in the 1830s and 1840s, similarly left the family as an outpost of welfare
provision. The massive retrenchment in poor law expenditures after 1834,
since needs could scarcely have fallen in line, suggests that families filled
some of the gap.
The survival and structure of the family among the wage-dependent
working class owed much to the reciprocal services that members were
able to provide. Changing economic conditions afforded new opportuni-
ties for reciprocal exchange as well as sweeping away old ones, and help
wasperhaps less conditional than earlier thought, with most kin pre-
pared to offer assistance until their own nuclear family was jeopardised
(Humphries 1977; Dupree 1995). But all was not change. Evidence for the
retention of traditional features by an important minority of households
not only supports the revisionist claim that organisational transition was
gradual but also helps to explain why that was so.
CONTINUITY AS WELL AS CHANGE: LIVE-IN
SERVICE AND DOMESTIC ENTERPRISE
Pre-industrial households frequently included live-in servants, who, along
with apprentices, were a significant component of the labour force, rep-
resenting around 15–20 per cent of the adult male population (Stone
1966). Service provided a bridge from childhood to adulthood tailored
forasociety constructed around families rather than individuals (Caunce
1991). With the exception of domestic service, however, living-in has been
regarded as ‘one of the large reptiles of economic history, extraordinar-
ily successful in its time and driven rapidly to extinction when times
changed’ (Kussmaul 1981: 134). The relatively low proportions of house-
holds that contained servants in Preston was held by Anderson to auger
‘the ultimate decline to which this class was destined’ (1971a: 81).
The demise of live-in service is dated from the late eighteenth cen-
tury when economic conditions moved sharply to disfavour it. Yet as late
as the mid-nineteenth century, significant proportions of households in
some communities continued to include servants, trade assistants and
apprentices (see Table 9.2), suggesting that the institution and its fate
are ripe for reconsideration.
Live-in farm servants hired annually and paid largely in kind provided
solutions to some of the problems associated with agricultural labour.
The traditional service contract made it easier to align incentives, and
reduce monitoring and muster costs (Woodward 2000). In return, resi-
dent service cut the costs of travelling to work and insured the worker
against rising rents and food prices. That most live-in farm servants were
Cambridge Histories Online © Cambridge University Press, 2008