SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information

The welfare of agricultural labourers has recently received renewed interest in both establishing living standards for a baseline group over the long term, and assessing the energy available for increased physical labour in the eighteenth century. Disagreement persists. This article examines a key aspect of agricultural labourers' families' welfare: nutrient consumption. We utilize datasets of the diets of agricultural labourers' households for 1787–96, 1835–46, 1863, 1893, and 1912, to analyse the availability of calories and 11 key nutrients. Self-provisioned foodstuffs are incorporated and adjustments are made for beer consumption. Deficiency is computed against household needs. The results corroborate the general levels of calorie availability identified in agricultural production accounts for the late eighteenth and early nineteenth centuries and assess these as sufficient for productive agricultural labour. However, no improvement is found in the nutritional aspect of household welfare between 1787–96 and 1835–46, thus endorsing pessimistic views of living standards for this group over this time period. Gains were evident in the next half-century, but these improvements were neither consistent nor dramatic and left a large minority of these households with nutrient deficiencies even in the twentieth century.

The welfare of agricultural labourers has recently received renewed interest in both establishing living standards for a baseline group over the long term,1 and assessing the possibility of increased labour supply to fuel the ‘industrious revolution’.2 These works offer differing conclusions. Using detailed data from farm accounts and by reconstructing cost-of-living estimates to reflect the consumption of the agricultural labourer, Clark finds little improvement in the purchasing power of farm wages from 1670 to 1819, followed by substantial growth (44.2 per cent) from 1820 to 1869, although significant regional variation was evident throughout.3 This presents a more optimistic position through the nineteenth century from 1820 than found by Feinstein for all workers4 and, on the basis that agricultural employment share shrinkage was occasioned by migration to the higher paying urban sector, suggests that Feinstein's all-workers real wage series should be raised upwards for this period. A recomputation of building labourers' real wages supports this conclusion.5 Taking a different approach, Allen and Weisdorf compare nominal wages in agriculture with the cost of a constant basic-necessity basket of goods and conclude that farm workers and their families may have been induced to work more days and longer hours through hardship as men's wages fell below the level needed to maintain this basic level of consumption between 1750 and 1818.6 Although the number of days worked needed to satisfy these basic consumption needs decreased by 27 per cent in the years after 1818, labourers decreased their labour supply in line with this decision, suggesting that their labour supply curve was backward bending.7 Agricultural labourers' living standards were essentially static. Indeed, a reworking of the agricultural labourers' real wage series supports this conclusion, finding a distinct pause in real wage growth from the mid-1790s to the early 1830s,8 a pattern very similar to that observed by Feinstein and for rural labourers in Kent.9 This position is somewhat adjusted by Muldrew, who focuses on labourers' diets from 1550 to 1780 and suggests that diets were varied and nutritious, considerably above basic needs, and supplied an energy availability that enabled eight hours' labour per day in 1600 and increased subsequently.10 However, Floud et al. dissent from this view. They consider that calorie availability was too meagre to allow a full day of physical work for large swathes of the population until the mid-nineteenth century.11

The debate over energy availability also coincides with a debate over change in average height. Here Floud et al. observe an increase in height for those army recruits born between 1740 and 1820, followed by a significant decline for those born between 1820 and 1850. Things gradually improved for those born in the second half of the twentieth century.12 Conversely Cinnirella observes a trend decline in heights from the mid-eighteenth to mid-nineteenth century.13 Convict data too support declining terminal height for those born between 1770 and 1815.14 Extracting the heights of those recruits previously employed in agriculture reveals a largely declining trend from 1760 to 1809, some variability until 1824, followed by a rapid decline to the 1840s and 1850s.15 Among other things, height reflects nutritional intake and the movements in height reported here frequently contradict the improvements discerned in real wages and energy availability.

Here we take an alternative approach to examining a key aspect of the welfare of agricultural labourers' families: energy and nutrient availability. The calorie, protein, and sometimes calcium and iron content of labourers' diets for 1780–90 and 1862–3 have been analysed elsewhere.16 These analyses suggest that labourers were typically underfed and suffered dietary deficiency. But here we extend the period of observation by utilizing datasets on agricultural labourers' diets for 1787–96, 1835–46, 1863, 1893, and 1912. We cover a broad range of intakes: calories and 11 key nutrients. While previous work has predominantly considered calorie consumption and the consequent energy available for work, nutritional science has emphasized the role of essential nutrients in maintaining health, ensuring proper physical development, and enabling the acquisition of cognitive and non-cognitive functions, all vital to human development and, relatedly, economic growth. Here we focus on those core nutrients whose availability in foodstuffs can be extracted from food composition tables. We also take account of each household's nutritional requirements in assessing adequacy.

This nutrient analysis allows us to map a crucial aspect of the welfare of farm workers' families and to compare this with alternative estimates. Although nutrition constitutes only one element among many when considering welfare, it has been argued that housing, type of work, and village communities remained largely unchanged through to the mid-nineteenth century, leaving diet, and, relatedly, health and longevity, leisure, and luxury consumption as the variable components.17 Nutrition therefore emerges as central to any assessment of farm labourers' welfare over the long term. Furthermore, it is likely that the recording of food data exhibits greater reliability than much of the income data; all the commentators considered here were desirous of accurately assessing how well the household could survive on prevailing wages and generally went to considerable lengths to obtain good estimates of the food consumed.

Looking at household diet enables us to consider the welfare of the whole family, not just the labourer and his dependents based either solely on the farm worker's wages or by making, necessarily, heroic assumptions about women's and children's work. Welfare is determined at the household level and a window onto it at this level is vital. Of course we are unable to say anything about the division of resources within the household; there is commentary in the various sources that make it clear that meat and protein were largely the preserve of working men, with women and children eating bread, tea, and leftovers.18 But we can say whether there was an adequacy or not within the household even if it was not then shared out to ensure adequacy for all household members, and thus we can observe how the agricultural labourer's household as a unit fared over the course of the nineteenth century.

I

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information

While farm workers' wages can be extracted from a variety of sources, from farm accounts to Parliamentary Papers, information on consumption is more limited. The first surveys of diets of labouring families were those collected by Sir Frederic Morton Eden and the Reverend David Davies in the late eighteenth century.19 These commentators were both keen to understand the position of the rural poor in this period of high bread prices and rapidly rising poor law payments. Eden largely attributed the problems faced by the poor, particularly in the south of the country, to their own wastefulness in desiring to eat wheaten bread and drink sweetened tea in contrast to the more frugal, and nutritious, oatmeal porridge consumed in the north. Davies, on the other hand, considered the loss of employment for women and children, particularly in wool spinning, disastrous and highlighted the impossible plight of a typical family trying to survive on the man's wages alone. Despite their different remits these two authors provided very similar accounts of the incomes, expenditure, and dietaries of the families they observed. In many cases the actual quantities of food bought by the families were recorded. In other cases local price data extracted from within the accounts allow conversion of expenditure into quantities consumed. In total these sources provided 208 observations of the quantities of food consumed by agricultural households in the years 1787–96.20

Of course, these data do not neatly document the exact quantities eaten in all cases. Many of the families had a cottage garden or land on which they could grow and consume potatoes and vegetables, while others kept a pig or even, very occasionally, a cow, and the value of these needs to be incorporated. This is an issue common to many of the household accounts through the nineteenth century. We outline our main arguments in section II, but detail our methods and assumptions in online appendix S1. Additionally, Muldrew has persuasively argued that these accounts seriously understate the amount of beer consumed and that this can vastly change our view of the energy available to these households. In the mid-eighteenth century a ‘well-employed’ family might have drunk 16 gallons of small beer and four gallons of table beer per week, providing an additional 767 calories per day to a child and 1,314 per day to the man.21 In what follows we consider the effect of an increased consumption of beer on our findings.

For the 1830s and 1840s there are relatively few observations of agricultural labouring households' expenditure. Purdy supplies 17 agricultural labourers' budgets with quantities of food consumed for the period 1835–42.22 Most are from Kent and Sussex, but three are recorded for Devon and one each for Suffolk, Cumberland, and Cornwall. Purdy produced these as part of a report to the Statistical Society that compared the state of agricultural labour under free trade in the 1860s with that in the Corn Law era. He extracted the budgets from reports and papers relating to the relief of the poor, some of which were abstracted from unpublished reports made to the poor law commissioners in 1837. Two Parliamentary Papers also furnish additional examples of consumption in agricultural labourers' households. The Poor Law Amendment Act provides 50 budgets for agricultural labourers prior to 1838 in the counties of Leicestershire, Lincolnshire, and Cambridgeshire, and the Reports of the Special Assistant of Poor Law Commissioners on the Employment of Women and Children in Agriculture supply a further three budgets with quantities consumed for Wiltshire in 1843.23 Although doubts have been expressed about the leanings of the Royal Commission set up in 1832 to investigate the poor laws in the wake of the agricultural labourers' revolt of 1830–1 and amidst fears of soaring relief expenditures,24 the household accounts utilized here do not seem to over-represent the poorest. Indeed, they incorporate a high proportion of households who engaged in self-provisioning or kept a pig and many also had earnings from the work of women and children, lack of such work often being identified with household poverty at this time. In total 70 household accounts were collected from these sources for this period.

In 1863 Dr Edward Smith was charged with conducting the first quasi-national survey of nutrition.25 The survey collected information on the diets of indoor workers in various trades around the country and rural labourers in most counties of England, Scotland, and Wales. Smith saw this survey as representative, claiming it to be ‘the first attempt which has been made in any country to ascertain the national dietary’.26 The survey sampled labouring households in well-defined occupational groups, whose characteristics in health, attitude, and employment were representative of their class and thus, when aggregated, of the ‘masses of our labouring people’.27 Smith surveyed 407 agricultural labourers. We have selected a one-in-three sample of the English ones and thus report the nutrition of 127 agricultural labourers' families in 36 counties for 1863. Smith was concerned to ascertain the nutrition of working people across the country. To this end, he collected information on household size, income, and consumption and annotated these with comments on the health and the general state of the family. He meticulously recorded quantities consumed and went to great lengths to ensure people tried to supply averages where purchases were infrequent or seasonal.28 It is clear that he incorporated quantities of food acquired through own-account activities, such as keeping a cottage garden, an allotment, chickens, a pig, and, very occasionally, a cow, and thus no further adjustments for self-provisioning are required for this dataset. Although Smith deliberately selected households that were meant to be healthy, so ensuring a consistent sample of the average working family for his survey,29 it is apparent from the recorded commentary that some households did not satisfy this condition, so we check how nutrient availability correlates with the observed health of the household in section IV below.

In 1893–4 the Royal Commission on Labour was concerned about the condition of the farm worker in the depression of agriculture which prevailed at the time, so in its report it collected and recorded the circumstances of labourers' families in various parts of the country.30 The accounts vary in their completeness. Some detail quantities consumed (30), for others quantities can be reconstructed from expenditure data for the household combined with prices extracted from information on the spending of other households in the same county (27), thus yielding 57 dietaries for this period.31 Again a number of these households had allotments or gardens, and the produce thus obtained requires quantification.

Finally, Rowntree and Kendall conducted a study of the agricultural labourer to ascertain his circumstances given the observed steady migration of population away from rural areas into towns or even abroad.32 They collected 42 household budgets across various counties and provided quantities purchased, weekly menus, the earnings of the family, and a detailed commentary on each household. They clearly went to considerable lengths in their interviews and spent time ascertaining the amounts of food consumed. They separately detailed any food obtained from allotments and gardens, via charitable gifts, or in the way of perquisites. Although only 42 households were interviewed, Rowntree and Kendall were directed to what were deemed typical villages in each area and selected families that had a ‘good reputation for sobriety, thrift and honesty’. They also tried to avoid those with an ‘abnormally large number of children’.33

These sources then provide the data for the nutrient analysis of the agricultural labourer's household from the late eighteenth century to the early twentieth century.

II

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information

A number of the sources record self-provisioning by the sample households and the quantities of food thus obtained need to be ascertained before any nutrient analysis can be performed. Full details of the assumptions we have made are outlined in online appendix S1, but it is necessary to highlight a number of observations.

Table 1 sets out the extent of self-provisioning for each of the surveys we consider. With respect to households having the use of common land or a garden in which to grow vegetables, there is significant variability between the surveys, and it is not immediately clear whether the apparent decrease followed by increase over the long nineteenth century is a real change in the economic circumstances of agricultural households or an artefact of variations in survey design and execution. The figures from Rowntree and Kendall's study in 1912 provide a firm benchmark for the end of our period. For the earlier years we can make comparisons with other estimates. Eden is well known for the household budgets he collected, but the majority of his three volumes on The state of the poor comprises detailed descriptions of the local economy—agriculture, manufacturing, employment, and workhouse operations—of 156 parishes around the country.34 In these accounts he frequently commented on the extent of common land available. Classifying this information on a county basis reveals 20 per cent of the counties in England had no common land, and a further 60 per cent had very few parishes where common land was available. Less than 7 per cent of counties had common land attached to most of their parishes. Thus 35 per cent of the survey households recording a garden or growing their own potatoes would seem reasonable, and concurs with Shaw-Taylor's observation that most labourers would not have had common rights prior to enclosure.35 Parliamentary enclosure removed much of this common land in the late eighteenth and early nineteenth century, and, although the allotment movement gained momentum in the early 1830s, it did little to replace this lost resource initially.36 By 1830 there were only 54 allotment sites in England, most concentrated around Wiltshire.37 This had risen to some 2,000 parishes with allotments by 1848, but this remained less than one-sixth of the parishes in the country. Indeed, one-third of counties had no parishes with allotments in 1841.38 The survey figures fit neatly with these estimates. Allotments became more commonplace as the nineteenth century progressed and this is also reflected in the survey figures. Although Purdy's budgets for 1835–46 record a high incidence of pig keeping, the other surveys record that around one in 12 agricultural households kept a pig. Again, we believe that the figure for 1912 is reliable, given the care with which the survey was executed. For the earlier dates we have few estimates: Shaw Taylor finds a maximum of one-quarter of labourers owning pigs or sheep in unenclosed rural Northamptonshire and considers that pig keeping was rare because, as common land did not produce sufficient resources to fatten a pig, substantial additional inputs were required and these were unaffordable for the labouring family.39 Classification of the responses to the 1834 Rural Queries40 reveals 30 per cent of counties where all rural parishes reported no gardens or pigs kept and a further 40 per cent where only one or two of the respondent parishes said that labourers had these resources.41 Thus only one-quarter of counties had some or most rural parishes where labourers had gardens, grew potatoes, or kept pigs, percentages that would seem to accord with the responses of the survey households. Overall, the recording of self-provisioning would appear to have been done with reasonable accuracy.

Table 1. Self-provisioning activities and beer reported as drunk in survey
SourceEden, State; Davies, CasePurdy, ‘On the earnings’; Poor Law Amendment Act; Reports of the Special Assistant (P.P. 1843, XII)Sixth Report (P.P. 1864, XXVIII)Royal Commission on Labour, Agricultural Labourer (P.P. 1893–4, XXXV)Rowntree and Kendall, How the labourer lives
  1. Note: a A further 16.7% of households record some beer drunk outside the household.

Source: Household accounts; see text.
Year1787–961835–46186318931912
Self-provisioning activities:     
% households recording an allotment, garden, or growing their own potatoes35%17%N/A42%81%
% households reporting keeping a pig8%40%11%9%7%
Beer drinking:     
% households recording beer purchased33.2%1.4%41.7%14.0%14.3%a
Where purchased, household consumption per week, pints.2.3812.06.3712.52.0

It is apparent that few of these households recorded any expenditure on beer, and yet this was deemed to be a staple of the English diet through much of the period. Smith and Rowntree and Kendall particularly tried to capture consumption of beer in their surveys, but both acknowledged that very little entered the household. In some cases men were given beer at work, particularly in the summer months, but Rowntree and Kendall noted a number of households where the man had decided to take this perquisite as cash instead. In other cases, men drank at the ale house. Smith was ‘struck with the small quantity of this fluid which enters the labourers' cottage’.42 Certainly there is reason to believe that men were able to increase their nutrient intake through the consumption of beer, but it appears there was little available for women and children. This contrasts markedly with Muldrew's observation that beer drinking was universal and contributed significant calories and nutrients to the diet of all. In the late seventeenth century he considers maybe two pints per day per capita was consumed.43 Strong beer might provide about 600 kcal per pint, table beer 400 kcal, and small beer 200 kcal.44 On the basis of a budget for a family of six in London in 1734 this could provide at least an extra 5,000 kcal to the household per day.45 Do the labourers' diets here significantly understate beer consumption, or had it reduced by the late eighteenth century? Davies himself claimed that beer was quite out of reach for the families he observed and was only produced on special occasions. Domestic brewing was unlikely to be substituting for purchases elsewhere. Household production of beer was complex and usually only conducted by farmers, the wealthy, and institutions.46 Indeed, only one of the 208 budgets for 1787–96 reported any expenditure on malt and hops. Private production fell precipitously from being over half of the total brewed in 1780–9.47 But from malt returns it is estimated that the total consumption for England and Wales in 1800–4 was 33.9 gallons per capita per annum, approximately ¾ pint per day,48 so it seems probable that there is some understatement of the amount of beer consumed in these household accounts. We thus place little reliability on the fluctuating figures for beer consumption in the household accounts and instead make specific adjustment for this in the nutrient analysis presented below.

The quantities of food available to each household can be converted into nutrients. This utilises the detailed tables on nutritional content of numerous foodstuffs constructed by McCance and Widdowson.49 We were careful to avoid using more recent values where foodstuffs have been fortified; for instance, bread and margarine. Although the tables provide data on the nutritional content of each food once waste has been allowed for, actual nutrition can vary within each type of foodstuff, according to how it is stored and according to how it is prepared and cooked.50 In the case of meat, we do not usually know which cut was eaten or how it would have been cooked or prepared, so we take a composite value of various cuts of beef and mutton, unless otherwise stated. Fruit values are taken in their raw state, but most vegetables would be eaten cooked and we have used cooked nutritional values throughout. Vitamin C, for instance, is destroyed by heat, light, and oxygen, so the nutritional value of cooked vegetables is significantly less than that of raw vegetables. It is probably reasonable to assume that similar storage and cooking conditions will pertain over the time period we are considering, rendering comparisons of relative nutrient availability reliable.

We selected calories and 11 nutrients for analysis: protein, calcium, iron, vitamins A, B6, B12, C, D, thiamine (B1), riboflavin (B2), and niacin (B3).51 These nutrients are a sub-set of those essential and vital to growth, health, and welfare. The function and consequences of deficiency of these nutrients are outlined in table 2. Aspects of nutritional inadequacy were clearly evident in the nineteenth-century population. For instance, iron deficiency was apparent in the anaemia and fatigue reported as endemic among young and pregnant women, often classified as ‘chlorosis’. Iron deficiency also adversely affected work capacity, intellectual performance, and resistance to disease. Scurvy is one of the oldest deficiency diseases recorded and was prevalent among those subjected to extended periods without fresh fruit, vegetables, and meat, such as those on long-distance voyages. By the mid-eighteenth century it was recognized that regular consumption of citrus fruit could prevent this vitamin C deficiency. The reduced incidence of scurvy has also been attributed to the introduction of potatoes, although it continued to make appearances through the nineteenth century and was identified among poor female factory operatives during the Lancashire cotton famine of the early 1860s.52 Deficiency of vitamin D was widespread and manifested itself in rickets, a disease often noted by nineteenth-century commentators. The resultant pelvic bone deformation could also cause complications for women in delivery at childbirth. Some vitamin D is derived from sunlight, and the migration into smoggy cities and the increase in indoor work have been linked to the increased incidence of the disease. Rickets remained a common childhood disability among the poor into the early twentieth century. Vitamin A deficiency too was probably commonplace. The first symptom of deficiency is night blindness, which later leads to complete blindness. It also contributes to maternal mortality, and, among children, low intakes decrease the likelihood of survival of serious illnesses, such as measles. Both these effects were probably evident in nineteenth-century populations. Estimates today suggest widespread vitamin A deficiency among young children worldwide. The 1851–1911 population censuses for the UK report around 0.02 per cent of children under five in England and Wales suffering blindness, less than the 0.1 per cent recorded for the population as a whole, so serious deficiencies were probably limited. Indeed modern recommendations for the prevention of deficiency require a high dose of vitamin A to be administered two to four times a year, so sporadic consumption may have been sufficient. Other vitamin deficiencies were perhaps less evident in nineteenth-century Britain although pellagra, lack of vitamin B3, has been identified among populations in the United States and Italy, and beri-beri, deficiency of vitamin B1, has been widespread in Asia. Kwashiorkor, protein deficiency, has not been specifically mentioned in relation to the British population, but a nineteenth-century disease, ‘marasmus’, was frequently diagnosed in children and is now considered to be related to kwashiorkor. Marasmus was associated with wasting, thinness, and poor muscle development. The poor skin, stunted growth, deformed limbs, bouts of diarrhoea, lack of energy, susceptibility to disease, and maternal complications frequently mentioned in contemporary accounts suggests dietary deficiencies were commonplace among the nineteenth-century British population.

Table 2. Key nutrients, functions, and result of deficiency
NutrientFunctionDeficiency
Source:Department of Health, Dietary reference values, pp. 15–166.
Calories (kcal)Energy/growth; requirements dependent upon age, body mass, and activity 
ProteinGrowth and muscle developmentKwashiorkor, muscle shrinkage, mild mental retardation in children
CalciumBone and teeth developmentBone maintenance, osteoporosis
IronComponent of haemoglobin, myoglobin, and many enzymesAnaemia, poor work capacity and intellectual performance, lowered resistance to infection
Vitamin A (retinol and carotene)Growth and normal development and differentiation of tissuesEye damage, night blindness, reduced ability to fight infection; prolonged deprivation causes death
Thiamine (B1)Metabolism of carbohydrate, fat, and alcohol (related to energy requirements)Beriberi, loss of coordination, rapid heart rate; ultimately confusion, coma, death
Vitamin B6Cofactor for enzymes catalysing amino acidsDeficiency rare. Convulsions
Vitamin B12Needed for enzymes to functionNeurological damage, anaemia
Vitamin CPrevents scurvy, aids wound healing, helps absorption of iron and anti-oxidantsHaemorrhages, bleeding gums, baldness, bone pain; initial symptoms are lethargy, reduced resistance to infection; ultimately fever and death
Vitamin DNeeded for calcium homeostasisRickets in children, skeletal deformity, muscle weakness, osteoporosis in adults
Riboflavin (B2)Essential in oxidative processesLesions of mouth and skin
Niacin (B3)Coenzyme in oxidation and reduction reactionsPellagra—dermatitis, diarrhoea, dementia; fatal if untreated

We consider how best to judge the nutritional adequacy of these diets in section IV, but as an initial consideration of the nutrients available to these agricultural households over time we consider per capita availability (table 3). Notable are what seem to be low levels of per capita calorie consumption. Per capita consumption of other nutrients was also often low in 1835–46 and 1893, periods associated with acute distress for agricultural workers.53 However, the calorie consumption figures fit largely within the range of calorie availability calculated from agricultural production accounts, some 2,100 to 2,977 per day, over the nineteenth century.54 Additionally the households surveyed here all contain children, and family sizes were particularly large in the years when per capita consumption fell below the available levels, 1835–46 and 1912. A more meaningful comparison is to use adult equivalents or consuming units (table 3). A simple computation of adult equivalent calories consumed in the survey households reveals closer correspondence to the similarly evaluated production figures. From the accounts male agricultural workers consumed around 2,500 to 3,400 calories per day. Was this enough to sustain substantial physical labour? Basic subsistence for a man has been estimated at 1,941 calories per day,55 and requirements for a man of average stature and build doing moderate physical work have been computed as 2,816 calories in 1777, 2,894 in 1827, and 2,863 in 1886–93.56 The men in our agricultural households are consuming at least this amount, locating them in the fourth to fifth deciles of the distribution of daily calorie consumption per consuming unit in England in 1800,57 well above the levels consumed by the lowest 20 per cent who could ‘not undertake physically demanding work on a regular basis’.58 Furthermore, as Floud et al. note, even where consumption appears low for men, resources may have been redistributed from women and children within the household to ensure adequate consumption for work by the male breadwinner.59 The levels of calorie intake identified here are considerably above the 2,289 calories per consuming unit per day available in France in 1785 (2,400 in 1806).60 The lower consumption by French workers is reflected in smaller stature and lower productivity than found for the British. By 1863 male agricultural workers were able to consume sufficient calories for a full day of heavy physical work, high consumption that was coincident with the golden age of British agriculture. Did the requirements for male agricultural workers change over the course of the nineteenth century? Towards the end of the nineteenth century mechanization in the form of machine reaping, binding, and threshing of the harvest became commonplace. Hay mowing and baling too were mechanized, but advances such as steam ploughing were more rarely used and many other tasks remained horse drawn or manual. The extent to which these innovations reduced the labour required from the individual compared with reducing the numbers of labourers required at peak harvest times is uncertain, but it is probable that there was a small reduction in the calorie needs for physical work by the agricultural labourer towards the end of the century. Against this the requirement for basal metabolism would have increased with the increase in stature evident from the 1860s onwards. Changes in domestic heating would also have altered energy consumption requirements, but it is unlikely that domestic comfort increased significantly for the majority of agricultural labourers over this period, although there may have been some improvement by 1912. Overall we do not identify any clear trend effect in energy requirements for these agricultural workers over time.

Table 3. Per capita consumption of nutrients per day
SourceEden, State; Davies, CasePurdy, ‘On the earnings’; Poor Law Amendment Act; Reports of the Special Assistant (P.P. 1843, XII)Sixth Report (P.P. 1864, XXVIII)Royal Commission on Labour, Agricultural Labourer (P.P. 1893–4, XXXV)Rowntree and Kendall, How the labourer lives
  1. Notes:

  2. a

    Calculated as adult = 1, child = 0.5.

  3. b

    Taken from Floud et al., Changing body, tab. 4.13, p. 167, years 1800, 1850, 1909–13 (calculated from agricultural production accounts and trade figures).

Source: As for tab. 1.
Year1787–961835–46186318931912
No. of households208701275742
Calories (kcal)2,1271,8622,4192,1531,849
Adult equivalent calories per daya3,0912,5153,3892,8382,800
Calories available per consuming unitb3,249–3,2933,111–3,3633,111–3,363 3,893
Nutrients:     
Protein (g)60.965.7806857
Calcium (g)215131409263245
Iron (mg)7.78.0108.48.3
Vitamin A (μg)1682075585816
Vitamin B6 (mg)1.30.91.50.90.9
Vitamin B12 (μg)0.20.61.20.74.0
Vitamin C (mg)21.08.0342534
Vitamin D (μg)0.10.21.30.20.9
Thiamine (mg)0.90.61.21.30.9
Riboflavin (mg)0.40.30.80.50.6
Niacin (mg)9.51.910.812.68.9
No. in household5.816.205.805.896.69
Total income per week (£)0.4960.5760.7460.8370.850

III

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information

To proceed further in any meaningful way we need to do two things: check the representativeness of our budgets in each time period, and move beyond per capita nutrition to capture the actual nutrition needs of the household given its age and gender composition. An initial check of the representativeness of our household accounts can be made by comparing the wages of these male agricultural workers with those recorded in other sources. In particular we use the observations on wages by county available for a selection of years from 1767 to 1898 as reported by Hunt and Bowley.61 Full details of this comparison are presented in online appendix S2. Once adjustments are made for changes in prices between our budget years and the years of the wage observations and for the extent to which payments-in-kind and perquisites are included in the earnings series, the wages extracted from household accounts are remarkably similar to those reported elsewhere. Only in 1863 and 1893 do they show any divergence, being slightly higher on average (109 per cent) in the former period and slightly lower (93 to 96 per cent) in the latter. We are reassured that the accounts are not sampling unrepresentative agricultural labourers.

From the preceding discussion it is apparent that the data sources do not yield a consistent geographical sample. Although the observations for 1787–96 and 1863 cover at least three-quarters of the counties in England and are spread throughout the north and south and around inland and coastal areas, the observations in other years are more limited in scope. Each still covers the full range of regions but only samples a fifth to a third of all counties.62 Different localities offer different opportunities and rates of pay to the agricultural labourer, and clear distinctions have been established between high and low wage regions of the country.63 To establish the geographical representativeness of our sample we use men's wage by county provided by Hunt for the years 1794–5, 1833–45, 1867–70, and 1898, and the Board of Trade statistics on wages, 1907,64 to classify counties as high, average, or low wage and compare the geographical distribution of our samples (table 4). In 1787–96 and, particularly, 1835–46, there is a tendency towards budgets taken from high wage counties. In 1863 and 1893 more households have been selected in low wage counties, and in 1912 the accounts were fairly evenly distributed between the high and low wage counties. This geographical dispersion implies that we need to be mindful of the regional selection of budgets when we interpret our results. However, any problem that might arise is overcome in the regression analysis where the county distribution of the accounts is controlled for. Generally a high proportion of the sampled counties are those with average wages.

Table 4. Dispersion of household accounts by wage level of county
SourceEden, State; Davies, CasePurdy, ‘On the earnings’; Poor Law Amendment Act; Reports of the Special Assistant (P.P. 1843, XII)Sixth Report (P.P. 1864, XXVIII)Royal Commission on Labour, Agricultural Labourer (P.P. 1893–4, XXXV)Rowntree and Kendall, How the labourer lives
Source: As for tab. 1.
Year1787–961835–46186318931912
Wages fromHunt, ‘Industrialization’Hunt, ‘Industrialization’Hunt, ‘Industrialization’Hunt, ‘Industrialization’Rowntree and Kendall, from 1907 Board of Trade
Wage range used to classify counties:
High wage9s. 3d.–11s. 0d.11s. 9d.–12s. 8d.17s. 0d.–20s. 0d.18s. 7d.–20s. 9d.18s. to 20s. and above
Low wage6s. 8d.–7s. 0d.7s. 10d.–9s. 7d.11s. 6d.–13s. 6d.14s. 5d.–15s. 10d.Less than 16s. to 17s.
Where budget locations fall:     
High wage46/20854/7029/1273/5719/42
%22%77%23%5%55%
Low wage6/2087/7055/12724/5723/42
%3%10%43%42%45%
No. counties observed with wages in average range out of total sample of counties19/303/816/387/150/8

We also consider the effect of the geographical distribution of the budgets on observed calories per capita per day (table 5). Within each time period, the calories consumed show some tendency to rise with the wage level of the counties observed.65 Over time, calorie consumption within each wage level demonstrates the same pattern: a fall in per capita calorie consumption between the 1790s and the 1830s, a rise to 1863, a decline to 1893, and, where we have observations, a further decline to 1912, leaving these families typically no better off than they were at the end of the eighteenth century. This pattern accords with other known facts. The hardship suffered by labouring families in the 1780s and 1790s is widely recognized. War, poor harvests, high bread prices, the loss of common rights, and reduced employment opportunities for women and children left these families in parlous circumstances. Conditions in the 1830s were little better. Continued enclosure and high bread prices led to the Captain Swing Riots in regions of the south of England and, in some places, left labourers so impoverished that allotments and potato grounds were allocated to help alleviate the problems caused by loss of common rights.66 Subsequently, despite contemporaries' concern over the repeal of the Corn Laws, agriculture prospered and during this golden age, farmers and, on the basis of the budgets analysed here, agricultural labourers enjoyed good times in the 1860s. But this was short-lived; large-scale imports of grain, meat, and some dairy products led to severe contraction in the primary sector as evidenced by low wages, reduced rents, and declining employment. Two Royal Commissions were set up to investigate the Depression in Agriculture (1879–82 and 1894–7). Expanded world trade, improved technology, and extensive imperialism continued this trend increase in imports up to 1914, leaving little room for any improvement in the labourers' position and causing concern about the exodus of labour from rural areas.

Table 5. Calories per capita per day from household budgets subdivided by wage level of county (number of observations in parentheses)
 1787–961835–46186318931912
Source: As for tab. 1.
Wage level in county     
Low2,768 (6)1,834 (7)2,153 (55)1,794 (24)1,763 (23)
Average2,074 (156)1,685 (9)2,465 (43)2,423 (30)
High2,224 (46)1,895 (54)2,855 (29)2,334 (3)1,952 (19)
      

At this juncture we should consider whether the household surveys were conducted only at particular times of hardship for agricultural labourers. Comparison of the average real wage for male workers in our households with the average wage extracted for all the years given by Bowley suggests that this is not the case; instead our male earnings lie firmly within the observations available for surrounding years.67

A further check on the representativeness of the accounts can be gleaned by comparing the family sizes of these households by what is known about family size by occupation for this period (table 3). The main information we have is occupational fertility calculated from the 1911 census.68 This shows that completed fertility for agricultural labourers was around 6.75, which was high relative to other occupations (89th percentile in the relative rankings).The households in the surveys analysed here tend to have two to three children fewer than this, but they may not yet have completed fertility, often still having young children at home; older children may have left home; and some children will, sadly, have died. The highest family size comes from Rowntree and Kendall, who we know tried to avoid households with exceptionally large families, which suggests that the others are not unrepresentative, nor do these others vary much over the nineteenth century. The larger size in 1912 may result from reductions in childhood mortality.69

In order to capture the actual nutritional needs of the household we need to ascertain the number of household members that fall into different age groups: under 1 year of age, 1 to 3 years, 4 to 6 years, 7 to 10 years, 11 to 14 years, 15 to 18 years, and over 19 years.70 No account is taken of gender throughout, although needs clearly differ between men and women and differential needs are apparent for children aged 11 and above. Our rationale is partly pragmatic—we often know the numbers and ages of children in the household but not their sex—and partly reasoned—although men and women have differing needs, with women requiring around 75 to 100 per cent of men's requirements for most key nutrients, pregnant and lactating women require considerably more than men for many of these nutrients.71 As many of our families are quite young, the wife may be in one of these states, and the actual demands on these women's bodies and health must remain unknown. It seems reasonable to then treat both sexes as having equivalent requirements. This may lead to some overstatement of household needs, but this overstatement is likely to be consistent across the time period studied.

Unfortunately only the 1912 survey specified the age and sex of each child. For the others, ages of eldest and youngest and numbers of children had to be used to create algorithms by which the numbers of children in each age category could be estimated. Full details of these computations are presented in online appendix S3. Once the numbers in each age category have been ascertained, the nutrient requirement for the household of the selected nutrients can be computed. For this, reference nutritional intakes (RNIs) are used.72 This then establishes the desirable amount of each nutrient for the household, given its composition.73 Note that intakes falling below these levels are not necessarily indicative of nutritional deficiency. Recommended dietary intake or allowances were developed and designed by nutritionists to evaluate food supplies for population groups, and were not intended as a tool for ‘assessing either the adequacy of nutrient intakes or nutritional status’.74 This is because an individual's nutritional status can only be identified by clinical assessment. Nevertheless, in general terms, as Harper has observed:

if the intake of a nutrient is equal to or greater than the RDA, the risk of nutritional inadequacy is remote. If it is less than 50 per cent of the RDA, the risk of inadequacy is high. However, when intake falls between these extremes all that can be said is that the farther intake falls below the RDA the greater is the risk of deficiency.75

Comparison of household availability of each nutrient compared with the computed RNI enables us to establish whether each household had a sufficiency or a deficiency. We express this as a percentage of RNI (see table 6). Given the possibility of measurement error in a number of the computations, a more robust measure of deficiency in the household is required. For all nutrients, following Harper, we take the position that having a deficiency of more than 50 per cent of the RNI indicates that deficiency is likely and classify each household according to whether they suffered a deficiency of the particular nutrient on this criterion. However, it should be noted that the UK RNI recommendations produced in 1991 were based on a mixture of depletion studies (where food is fed to an individual and levels of nutrient monitored via urine or blood analysis until deficiency symptoms are manifested) plus a safety margin, and intake observations in healthy populations. Although the depletion studies are clearly relevant to our classification of deficiency, it should be noted that for some nutrients the safety margin could be as high as 200 per cent.76 Thus while our deficiency variable will pick up households where there is a generally accepted case of insufficiency, it would be wrong then to assume these levels would also give rise to symptoms of deficiency. It may be best to view these as adequacy indicators indicative of a shortfall that is not necessarily observably injurious to health. Deficiencies in the intakes of these key nutrients are liable to be associated with a lowered resistance to infection in the first instance.

Table 6. Proportion of households with less than RNI available
 1787–961835–46186318931912
Source: As for tab. 1.
Calories4771257050
Protein43122
Iron8383419079
Calcium9899659598
Vitamin A100100987248
Vitamin B65080257264
Vitamin B1210094697931
Vitamin C7191617255
Vitamin D1008283
Thiamine4986202129
Riboflavin9799769583
Niacin77100697795

IV

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information

Tables 6 and 7 report the proportion of households with less than RNI or less than 0.5 RNI respectively. Starting with all households where consumption lies below RNI values, not solely those identified as suffering nutritional inadequacy, reveals some clear features. Calorie deficiency follows the same pattern over time as observed for per capita consumption, with the exception of 1912. Here conditions are a little improved after 1893, although below the levels attained in 1863, but not dissimilar to those found in the 1780s and 1790s.77 This pattern is typically followed by all the other nutrients, although to varying degrees, and, for a few nutrients (calcium, thiamine, and niacin) greater occurrence of consumption below RNI levels is observed in 1912 than previously. The majority of households suffered some degree of deficiency in most nutrients in all time periods, with the exception of protein. These agricultural labourers' households almost invariably had sufficient protein in their diets, even though they lacked other key nutrients.

Table 7. Proportion of households with less than 50% RNI available (deficiency)
 1787–961835–46186318931912
  1. Notes: Calorie deficiency is included for completeness but energy availability at levels below RNI would represent an inadequacy relative to needs.

Source: As for tab. 1.
Calories21000
Protein10000
Iron123070
Calcium8093226676
Vitamin A851001004829
Vitamin B61470480
Vitamin B129364395414
Vitamin C578128575
Vitamin D755364
Thiamine613025
Riboflavin7493246636
Niacin49971155

Conditions are considerably improved when the stricter definition of inadequacy is applied, as far fewer households exhibit deficiency of key nutrients. Indeed, as table 7 shows, most have tolerable levels of protein, iron, vitamin B6, thiamine, and niacin, although levels of some of these nutrients remain very low for the eighteenth-century households. But households were generally suffering an inadequacy of calcium and riboflavin for the entire period and had markedly inadequate intakes of vitamin A and vitamin C at various points. The incidence of inadequacy is worse in the 1830s than it is in the 1790s. Considerable improvement is evident by 1863, but this suffers some erosion to 1893; then things largely improved to 1912, with many nutrients showing a similar or better incidence than found in 1863. Thus a general picture emerges for the position of the agricultural labourer. Nutrient intake follows a wavelike pattern about a largely upward trend, but the 1830s represent a low point. We aggregate the patterns of availability of the 12 nutrients into one composite measure to establish a clearer trend later.

Earlier we highlighted the possibility of unaccounted-for nutritional benefits of beer drinking in these labourers' diets. To what extent would beer drinking change our view of nutrition for this group of workers over the nineteenth century?78 Taking a daily consumption of 0.75 pints of beer per person in 1800–4,79 and assuming an average household size of six people, would imply 4.5 pints of beer consumed. This is likely an upper estimate as some beer consumption has already been included in the food calculations, where stated, and some will have been consumed outside the household by the male. Any overstatement of actual beer consumed generated by these assumptions will increase over time as recorded consumption approximately halves.80 Beer contains calories, protein, iron, calcium, riboflavin, niacin, vitamin B6, and vitamin B12. A useful way to consider the impact of an additional 4.5 pints of beer consumed by the household each day is to compute the proportion of a two-adult and four-child household's daily requirement of these nutrients that this consumption would constitute. It would represent 7 per cent of the household's daily calorie requirement, 4 per cent of protein, 0.5 per cent of iron, 6 per cent of calcium, 13 per cent of riboflavin, 11 per cent of niacin, 8 per cent of vitamin B6, and 58 per cent of vitamin B12. Comparison with the levels of deficiency identified in tables 6 and 7 suggests that beer drinking would largely remove vitamin B12 deficiency throughout the period being studied and it would improve the extent of vitamin B6 deficiency, although there are relatively few cases of serious deficiency of this nutrient. This would improve the position of 1787–96 and 1835–46 relative to later years but would probably not change their positions relative to one another. Calorie and protein consumption would be improved by beer drinking, but there are few cases of serious deficiency of these nutrients so there would be little overall impact. The impact on the serious deficiencies observed of calcium, riboflavin, and niacin would be limited, with the majority of households probably still exhibiting intakes of less than 50 per cent RNI, particularly in the earlier years. Overall, including beer drinking in our nutrient assessment has the effect of levelling out the agricultural labouring families' experience over this 125-year time frame and, if anything, indicating even more clearly the plight of those at the end of the nineteenth and beginning of the twentieth century given the parlous conditions we know were suffered by the late eighteenth-century labourer's family and the knowledge that beer drinking declined over the nineteenth century.

Table 7 has established the patterns of nutritional inadequacy for various nutrients. In the following analyses we investigate some of the correlates of poor nutrition. As tables 8 and 9 demonstrate, there is clear evidence that the ability to self-provision through cultivating an allotment or keeping a pig improved nutrition available to the household in the eighteenth and nineteenth centuries. This manifested itself particularly through additional vitamin intakes. It may be less evident in 1912 because few of the surveyed households did not have access to an allotment and maybe those that did not were afforded perquisites or charity instead. Men's earnings sometimes emerge as significantly lower in households where nutritional inadequacy is evident (see table 10a). Overall, though, a significant difference only occurs in nine out of 60 possible cases, and in three cases in the eighteenth century men's earnings are higher in the households with serious deficiencies. Household income levels were more important as they frequently have an impact on whether the household has an adequate diet (12 out of 60 cases in table 10b), but again, perversely, are often higher in households with deficiencies. Overall, therefore, we conclude that earnings and income do not appear to be clear predictors of nutritional inadequacy. As we show in table 10c, there is a more consistent relationship between household size and deficiency. In nearly half the cases deficiency is related to large household size, and the incidence of this is common to all time periods. Having a large number of children in the household made the family vulnerable to inadequate nutrition. Although calorie and protein consumption was unaffected, there were severe consequences for the availability of vitamins and minerals.

Table 8. Proportion of households with less than 50% RNI available (deficiency) if involved in self-provisioning (garden/pig)
 1787–961835–461863a18931912
  1. Note:a Keeps pig only.

Source: As for tab. 1.
Calories03000
Protein00000
Iron43040
Calcium8594145479
Vitamin A60100933329
Vitamin B60300
Vitamin B129241365018
Vitamin C0692103
Vitamin D715768
Thiamine0606
Riboflavin5988144241
Niacin16941406
% households involved in self-provisioning3546114281
Table 9. Significant differences in mean levels of nutrients by whether household is involved in self-provisioning
 1787–961835–46186318931912
  1. Note: * indicates F-statistic significant at 10% level or higher showing positive effect of self-provisioning on mean level of nutrient, *-ve indicates significant negative effect of self-provisioning on mean level of nutrient.

Source: As for tab. 1.
Calories*  * 
Protein  ** 
Iron*  * 
Calcium  ** 
Vitamin A*  * 
Vitamin B6** * 
Vitamin B12*-ve  **-ve
Vitamin C*  * 
Vitamin D     
Thiamine**** 
Riboflavin*****-ve
Niacin** * 
Table 10. (a) Mean values of variables in households with and without deficiencies (<50% RNI): male wage, £ per week; (b) Mean values of variables in households with and without deficiencies (<50% RNI): total household income, £ per week; (c) Mean values of variables in households with and without deficiencies (<50% RNI): number of people in household
 1787–961835–46186318931912
No def.Def.No def.Def.No def.Def.No def.Def.No def.Def.
  1. Note: Only cases where a significant difference arises have been reported (significant at the 10% level or more).

Source: As for tab. 1.
(a)
Calories0.390.29        
Protein          
Iron          
Calcium        0.820.71
Vitamin A0.360.40      0.760.65
Vitamin B6          
Vitamin B12      0.690.62  
Vitamin C0.370.41    0.690.62  
Vitamin D          
Thiamine          
Riboflavin    0.590.54    
Niacin0.370.41        
(b)
Calories          
Protein0.500.73        
Iron          
Calcium    0.720.81    
Vitamin A0.460.50    0.950.73  
Vitamin B6      0.900.76  
Vitamin B12  0.610.56      
Vitamin C0.480.51    0.920.77  
Vitamin D          
Thiamine          
Riboflavin0.460.510.500.58      
Niacin0.470.530.470.58      
(c)
Calories          
Protein          
Iron5.76.56.19.5      
Calcium5.45.94.86.35.66.74.46.75.96.9
Vitamin A4.36.1    5.26.6  
Vitamin B65.76.4    5.46.4  
Vitamin B12  5.36.75.46.55.06.6  
Vitamin C    5.56.7    
Vitamin D4.66.2  5.16.4  6.17.0
Thiamine5.86.9        
Riboflavin5.36.03.86.45.66.43.97.06.17.7
Niacin5.56.11.56.35.67.5    

These results are perhaps surprising. As already noted, household incomes are expected to be a better predictor of nutrition availability than male earnings, as they capture the total monetary resources available to the household, in particular the earnings arising from women's and children's work. These earnings would not necessarily be positively correlated with male earnings. That total income is also an imperfect predictor of nutrition levels requires further explanation. It is well established that expenditure on food increases with income, as does the number of calories purchased. This can be observed for our households in table 5. But our measure of deficiency captures the lack of nutrients relative to the household's need. As the number of children increases the household's need for nutrients increases but income is unlikely to increase to match, at least initially, thus potentially leaving the household deficient. Furthermore, dietary choices made at higher levels of income may not always comprise a more nutritious choice; witness the increased consumption of tea and sugar historically. Children's nutritional needs also differ from those of adults. Children require vitamin D in their diets, while adults do not, and they require vitamin C, iron, and calcium at relatively high levels compared to their relative calorie intake needs. Unless consumption is altered to purchase foodstuffs that reflect the specific needs of children—dairy products, oily fish, and eggs—these needs are unlikely to be met. There is no guarantee that an increase in income will ensure the purchase of the appropriate combination of foodstuffs, and thus numbers and needs of children emerge as stronger predictors of whether nutrient deficiencies are evident.

To disentangle some of these effects we construct a composite nutrition variable and estimate the relationship between this variable and various household characteristics empirically. The composite nutrition variable is constructed by deducting deficiency of each nutrient (<50 per cent RNI) from a maximum value of 12 for each household. This gives each nutrient equal weight in the composite variable. Few households in any period manage to achieve a nutritional level of 12 (table 11), but the mean and median values exhibit the same wave-like pattern around an upward trend over time as found for many of the individual nutrients. The fall in nutrition from 1790 to 1830 is not so noticeable on this measure and the rise into the early years of the twentieth century is more evident.

Table 11. Composite nutrition variable (value 0–12)
 1787–961835–46186318931912
Source: As for tab. 1.
Frequency (%)     
00.5    
1     
22.9    
33.8  1.8 
41.95.7   
525.010.03.13.52.4
615.940.03.912.32.4
716.330.07.122.82.4
817.37.114.214.021.4
99.14.322.010.511.9
106.22.925.212.319.0
110.5 23.610.531.0
120.5 0.812.39.5
Mean6.546.479.228.499.67
Standard deviation2.001.251.592.201.68
Median6.56.09.08.010.0
Mode5.06.010.07.011.0

Smith and Rowntree and Kendall both provide brief commentary on the health of the households they surveyed.81 We classify these qualitative comments into a six-point scale, ranging from 1 = unwell or unhealthy to 6 = very good health. We correlate these with various measures of nutrition. Most notably there is a significant positive correlation between the overall measure of nutrition and health in both years. Mean scores for health were significantly worse where there were calcium, riboflavin, and niacin deficiencies in 1863 and vitamin A, niacin, and thiamine deficiencies in 1912, indicating the impact of nutrients beyond calories and protein on health. The correlation of health with nutrition is supportive of the assertion that we are adequately capturing nutritional status and, correspondingly, an important aspect of household welfare.

To discern nutritional status clearly for agricultural labourers' households over time, we need to control for factors such as income variation within the sample, household size, and differential access to gardens and allotments. We do this using regression analysis for each sample separately. The results given in table 12 confirm the observations made earlier. Only for the early period do men's earnings have a significant impact on nutrition; in some other periods total income is the greater determinant. The number of people in the household has a significant negative effect and access to gardens and perquisites a significant positive effect.

Table 12. Regression analyses, determinants of nutrition variable
 1787–961835–46186318931912
  1. Notes: t-ratios in parentheses. * indicates significance at the 10% level or higher.

  2. a

    The computed nutrition variable uses the significant coefficients generated by the regressions and sums the constant; wage level using Hunt's average male wage for 1787–96; average total household income in each period taken from the budget accounts; assumes a household size of 5; assumes the average wage level in county (= 2); and makes no addition for allotment or perquisites in any time period.

Source: As for tab. 1.
Constant7.478 (13.22)*7.697 (8.00)*10.221 (22.77)*8.205 (14.36)*11.120 (8.85)*
Man's earnings2.562 (2.23)*    
Wage level in county −0.464 (−1.77)*  −0.094 (−0.37)
Total income−0.526 (−0.49)2.641 (1.31)1.309 (2.28)*2.649 (4.56)*1.337 (1.22)
No. in household−0.435 (−6.16)*−0.293 (−4.42)*−0.341 (−5.48)*−0.493 (−6.58)*−0.430 (−2.79)*
Garden/allotment2.496 (11.35)*0.673 (2.55)* 2.206 (6.42)* 
Perquisites    1.226 (2.56)*
Adjusted R20.470.310.180.71 
F45.49*8.88*14.99*44.9*4.58*
Sample size208701275742
Computed representative nutritional variablea6.45.39.58.09.0

We use the coefficients generated from this analysis to construct a representative nutritional variable in each time period which then controls for variations between the samples in household size, income, wage level of county, and availability of gardens and allotments.82 As table 12 shows, the movement of this variable confirms the pattern we identified earlier. After 1835–46 nutritional standards rise considerably, fall back a little to 1893, but almost regain earlier levels by 1912.83 Values for 1787–96 and 1835–46 are considerably below those observed later, with a clear fall between the 1790s and 1830s.84 This view is endorsed when we consider some of the potential biases in the 1830s data when compared with the 1787–96 data. Computation of self-provisioning in 1835–46 was particularly generous.85 The data on household size and composition in the 1835–46 survey did not enable the number of young children to be identified; therefore it is not possible to calculate a requirement for vitamin D for this survey and so no deficiencies for this nutrient resulted. This may artificially raise the aggregate nutrition score.86 Furthermore, deficiencies were identified by calculating household needs against availability. The proxy used to compute household needs in 1835–46 may have understated these needs.87 Finally, beer consumption declined over this period so any mitigating effect of beer drinking on nutrient deficiency would decline between the 1790s and the 1830s. All these factors indicate an upward bias to the observed nutrition levels in 1835–46 and give confidence that there was no actual improvement from 1787–96 and most probably there was a worsening.

V

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information

The analysis of nutritional intake presented here shows that an important element of household welfare showed no discernible improvement between 1787–96 and 1835–46, thus endorsing pessimistic views of living standards improvements over the first four decades of the nineteenth century. Subsequently conditions improved, but gains were not sustained indefinitely—a setback was evident in the late Victorian Great Depression. These results are consistent with what is known about the agricultural labourer's experience over this 125-year period and suggest that nutritional analysis is an appropriate way to capture a key aspect of living standards.

Our analysis corroborates Feinstein's finding and is supportive of Allen and Weisdorf's interpretation of the course of agricultural labourers' living standards during the first half of the nineteenth century. Indeed the decline in nutrition over the first part of the nineteenth century coincides with the decline in heights for the cohorts born after 1820 observed in all series, as well as the downward trend from the 1780s.88 Our results on calorie consumption at the end of the eighteenth century also endorse the general observations of Floud et al., even when we make adjustment for beer consumption. However, the agricultural labourers presented here are not at the bottom of the distribution of food availability and do have sufficient calories to perform a full day of moderate to heavy labour.

All these authors accept that a sustained take-off in wages occurred after the 1840s, but it seems there was no improvement in the material position of agricultural labourers before this time. Indeed, on this evidence, the loss of employment for women and children and of self-provisioned resources through enclosure occasioned an actual decline in consumption standards for these families. Agricultural labourers are used as the benchmark against which to measure the progress of other occupational groups over the industrial revolution; static conditions for farm workers implies gains for the rest. However, we identify a fall in this aspect of living standards for this indicator group and thus contend that the position of all workers remains unknown. Pessimism potentially persists.

Although gains in household nutrition were evident over the next half-century, improvements were neither consistent nor dramatic. The golden age of British agriculture represented the heyday of the agricultural labourer's experience, but even then half of agricultural labourers' households still suffered deficiencies of three or more key nutrients. Subsequently imports and agricultural depression occasioned a falling back of consumption and a raised incidence of nutritional inadequacy. Such deficiencies were likely to be associated with lowered resistance to disease, an increased incidence of morbidity, and, potentially, higher mortality rates.

Footnotes
  1. 1

    Clark, ‘Farm wages’; idem, ‘Long march’.

  2. 2

    Muldrew, Food; Allen and Weisdorf, ‘Was there an “industrious revolution”?’.

  3. 3

    Clark, ‘Farm wages’, tab. 9, p. 496.

  4. 4

    Feinstein, ‘Pessimism perpetuated’.

  5. 5

    Clark, ‘Condition’.

  6. 6

    Allen and Weisdorf, ‘Was there an “industrious revolution”?’.

  7. 7

    Ibid., p. 722.

  8. 8

    Allen, ‘Pessimism preserved’.

  9. 9

    Richardson, ‘Agricultural labourer's standard of living’.

  10. 10

    Muldrew, Food, pp. 29–162.

  11. 11

    Floud, Fogel, Harris, and Hong, Changing body, pp. 72–7, 119–27, 160–8.

  12. 12

    Ibid., pp. 135, 144.

  13. 13

    Cinnirella, ‘Optimists’, pp. 336–7.

  14. 14

    Nicholas and Steckel, ‘Heights’, p. 948; Nicholas and Oxley, ‘Living standards’.

  15. 15

    Cinnirella, ‘Optimists’, pp. 336–7.

  16. 16

    See, for example, Shammas, ‘Eighteenth-century English diet’; Muldrew, Food, pp. 117–62; Oddy, ‘Urban famine’; Barker, Oddy, and Yudkin, ‘Dietary surveys’.

  17. 17

    Clark, ‘Farm wages’, p. 498.

  18. 18

    See, for example, Sixth Report (P.P. 1864, XXVIII); Rowntree and Kendall, How the labourer lives.

  19. 19

    See Muldrew, Food, p. 117; Eden, State; Davies, Case.

  20. 20

    These data were originally extracted and used in Gazeley and Verdon, ‘First poverty-line’.

  21. 21

    Muldrew, Food, pp. 136–7.

  22. 22

    Purdy, ‘On the earnings’, pp. 349–52, 363–9.

  23. 23

    Poor Law Amendment Act (P.P. 1837/8, XXXVIII); Reports of the Special Assistant (P.P. 1843, XII).

  24. 24

    Royal Commission 134:1, cited in Boyer, English poor law, p. 194.

  25. 25

    Sixth Report (P.P. 1864, XXVIII), app. 6.

  26. 26

    Ibid., fo. 220.

  27. 27

    Ibid., fo. 220.

  28. 28

    Ibid., fos. 217–18.

  29. 29

    Ibid., fo. 14, note.

  30. 30

    Royal Commission on Labour, Agricultural Labourer (P.P. 1893–4, XXXV).

  31. 31

    Nutrients per capita in households where quantities were stated and those where quantities were derived using prices from the same geographical area were very similar. Budgets for Cambridgeshire and Somerset where quantities could only be derived using prices from distant locations were omitted from our analysis.

  32. 32

    Rowntree and Kendall, How the labourer lives.

  33. 33

    Ibid., p. 38.

  34. 34

    Eden, State, vols. II and III.

  35. 35

    Shaw-Taylor, ‘Parliamentary enclosure’, p. 658, shows that around 15% of farm labourers had access to common rights in the south and east midlands, a finding that is supported by analysis of anecdotal evidence from contemporary sources; idem, ‘Labourers’.

  36. 36

    In 1700, 29% of England remained open or common; by 1914 this had been reduced to 5%; Allen, ‘Agriculture’, p. 99. Between 1700 and 1850 Acts of Parliament removed the open fields in maybe half of the parishes in England; Shaw-Taylor, ‘Labourers’, p. 95.

  37. 37

    Burchardt, Allotment movement, pp. 36–7.

  38. 38

    Ibid., p. 69.

  39. 39

    Shaw-Taylor, ‘Nature and scale’.

  40. 40

    Rural Queries, appended to the Royal Commission of Inquiry into Administration and Practical Operation of the Poor Laws (P.P. 1834, XXVIII), app. B1.

  41. 41

    See Horrell and Oxley, ‘Bringing home the bacon?’, for further details of this classification.

  42. 42

    Sixth Report, fo. 253.

  43. 43

    Muldrew, Food, p. 65.

  44. 44

    Ibid., pp. 79–80.

  45. 45

    Ibid., pp. 136–40.

  46. 46

    Mathias, Brewing, pp. xxi–xxii.

  47. 47

    Ibid., p. 377, tab. 18.

  48. 48

    Burnett, Plenty and want, pp. 8–9.

  49. 49
  50. 50

    See Barker et al., ‘Dietary surveys’, for a detailed discussion of these aspects of nutrient analysis.

  51. 51

    Vitamin A is calculated as retinol + (carotene/6) as given in Department of Health, Dietary reference values, p. 85.

  52. 52

    Fifth Report of the Medical Officer of the Privy Council (P.P. 1862, XXV), app. 5.

  53. 53

    Vitamin A values appear low for the 1830s and 1863. This results from no recorded consumption of items such as liver, carrots, and dairy products for a number of households in these surveys, but where these items are purchased the availability of vitamin A is within a sensible range when compared with reference intakes.

  54. 54

    Floud et al., Changing body, p. 160, tab. 4.9.

  55. 55

    Allen, ‘Great divergence’, p. 421, tab. 3.

  56. 56

    Floud et al., Changing body, pp. 167–8. Variations over time reflect changes in stature. In subsequent calculations we use the Department of Health, Dietary reference values, p. xix, tab. 1.1, standard of required intake for a 74 kg man doing today's moderate work. This is 2,550 calories per day, a little below the needs suggested here. However, children's needs today are higher than assumed in historical sources and thus compensate for understatement of men's needs when we compute adequacy of nutritional intake for the household.

  57. 57

    Floud et al., Changing body, p. 56, tab. 2.4.

  58. 58

    Ibid., pp. 168–9. These calorie computations do not include any estimate for beer drunk at work or in the alehouse which may have increased energy available for work in the early years studied. These estimates therefore represent lower bounds for actual consumption.

  59. 59

    Ibid., pp. 168–9.

  60. 60

    Ibid., pp. 44, 106.

  61. 61

    Hunt, ‘Industrialization’, pp. 965–6; Bowley, ‘Statistics. I’; idem, ‘Statistics. IV’.

  62. 62

    Regional price variations may also be thought to affect our observations on nutrition. However, our computations largely rely on observed quantities of food purchased and, where these have to be deduced from expenditure, we have only used prices derived locally. In any case, transport improvements were aiding market integration from the 1700s (Bogart, ‘Glorious Revolution’); the integration of England's domestic markets had improved before the industrial revolution (Bateman, ‘Evolution’, p. 466); and regional integration was apparent by 1843 with few differences found in prices paid by poor law authorities (Crafts, ‘Regional price variation’). Indeed, Hunt, ‘Industrialization’, pp. 962–4, concludes that it was not differing prices for necessities that underlay regional differences in agricultural labourers' wages.

  63. 63

    Hunt, ‘Industrialization’.

  64. 64

    1986Ibid., pp. 965–6, app. tab. 6; Rowntree and Kendall, How the labourer lives, frontispiece.

  65. 65

    In 1787–96 observations for the low wage county are taken from Somerset and would appear to be an outlier.

  66. 66

    Burchardt, Allotment movement, pp. 51–69; Moselle, ‘Allotments’; Archer, ‘Nineteenth-century allotment’.

  67. 67

    Real male wage:

    1767–701787–9617951824183318371835–4618501860/11863
    93.387.183.582.3103.891.4108.695.895.799.0
    1867–7218801892189319071912    
    104.0118.7141.2137.3142.2141.4    

    Figures in bold represent real male wage taken from household accounts. Nominal wages were taken from Bowley, ‘Statistics. I’, pp. 706–7, average England and Wales, 1767–70 to 1892; Rowntree and Kendall, How the labourer lives, frontispiece data taken from Board of Trade 1907, total male earnings adjusted for an assumed 27% in payments-in-kind (see online app. S2). The cost of living was taken from Feinstein, ‘Pessimism perpetuated’, pp. 653–4, calculated for 1770–1882, base year 1778–82 = 100; Gazeley, ‘Cost of living’, p. 214, col. 2, ‘Bowley revised’, for 1886–1912. The two series are spliced using Sauerbeck's increase in wholesale price, averaged across both the food index and the overall index, from 1882–86 (Mitchell, British historical statistics, p. 725).

  68. 68

    Szreter, Fertility, app. C, pp. 612–13; ibid., fig. 7.1, p. 312.

  69. 69

    For discussions of this decline, see Harris, ‘Gender, height and mortality’; idem, ‘Gender, health and welfare’.

  70. 70

    These are in line with those cited in Department of Health, Dietary reference values, tabs. 1.3–1.5, pp. xxi–xxiv.

  71. 71

    See comparisons in ibid., reference nutritional intakes, pp. xix–xxiii.

  72. 72

    The 1991 UK RNI values replaced the 1979 recommended daily amounts (RDAs), defined as ‘the average amount of the nutrient which should be provided per head in a group of people if the needs of practically all members of the group are to be met’. In contrast, RNIs are set at a notional two standard deviations above the estimated average requirement (EAR), usually ensuring an amount of a nutrient that is adequate for 97.5% of the population; Department of Health, Dietary reference values, pp. xix–xxiv, 1–3.

  73. 73

    Strenuous physical activity may require more nutrients as well as more energy; Manore, ‘Effect’. But in the absence of alternative measures we retain those detailed above.

  74. 74

    Harper, ‘Evolution’, p. 526.

  75. 75

    Ibid., p. 526.

  76. 76

    Including vitamins B2, B3, B6, and C; Department of Health, Dietary reference values.

  77. 77

    The average deficiency compared with household need for calories, where it existed, was between 15% and 24% in each year.

  78. 78

    But note that consumption of alcohol also has the effect of depleting the body's absorption of vitamins A, C, B12, and D, iron, and calcium so this exercise provides an upper bound estimate of the nutritional benefits of beer drinking.

  79. 79

    See section II.

  80. 80

    The effect on nutrient availability of the inclusion of additional beer will be the same in each time period so will have little impact on the regression analysis performed later.

  81. 81

    Sixth Report; Rowntree and Kendall, How the labourer lives.

  82. 82

    For construction of this variable, see note to tab. 12.

  83. 83

    Indeed Smith comments, ‘It may not be doubted that the condition of the farm labourer and his family has much improved of late years, in his wage, the lower price of provisions, and the work which the extension of trade and communication now offers to him and his family’ Sixth Report (P.P. 1864, XXVIII), fo. 264.

  84. 84

    It may be that giving equal weight to all nutrients is inappropriate in the construction of this composite variable. An alternative, where deficiencies ultimately leading to starvation, lowered resistance to illness and death were weighted 1 (calories, protein, iron, and vitamins A, B1, B3, and C) and the others (calcium, and vitamins B6, B12, B2, and D) were weighted 0.5, generated the same significant results in regression analysis and yielded the same overall pattern for the constructed nutritional variable. Conditions were particularly bad in 1835–46, but the improvement to 1912 shows 1912 having fewer deficiencies than 1863 on this measure. Overall our conclusions are robust to the construction of this variable.

  85. 85

    The division of groceries between tea, sugar, cheese, and butter allowed reasonable amounts of the latter two products to be bought; these are nutritious items, but in reality many households may have opted to buy more sugar, which is not. Moreover, more of the households in 1835–46 kept a pig than earlier.

  86. 86

    The same is true for the construction of the composite nutritional variable in 1893.

  87. 87

    The same proxy was used for 1893 households, but here a number had the actual ages of children stated. Using these for comparison suggested the household nutritional requirements had been slightly understated.

  88. 88

    Cinnirella, ‘Optimists’.

Footnote references

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information
  • Allen, R. C., ‘The great divergence in European wages and prices from the middle ages to the First World War’, Explorations in Economic History, 38 (2001), pp. 411447.
  • Allen, R. C., ‘Agriculture during the industrial revolution, 1700–1850’, in R. Floud and P. Johnson , eds., The Cambridge economic history of modern Britain, I: Industrialisation, 1700–1860 (Cambridge, 2004), pp. 96116.
  • Allen, R. C., ‘Pessimism preserved: real wages in the British industrial revolution’, Oxford University Department of Economics working paper, 314 (2007).
  • Allen, R. C. and Weisdorf, J. L., ‘Was there an “industrious revolution” before the industrial revolution? An empirical exercise for England c. 1300–1830’, Economic History Review, 64 (2011), pp. 715729.
  • Archer, J. E., ‘The nineteenth-century allotment: half an acre and a row’, Economic History Review, L (1997), pp. 2136.
  • Barker, T. C., Oddy, D. J., and Yudkin, J., ‘The dietary surveys of Dr Edward Smith 1862–3: a new assessment’, Dept. of Nutrition, Queen Elizabeth College, University of London, occasional paper no. 1 (1970).
  • Bateman, V. A., ‘The evolution of markets in early modern Europe, 1350–1800: a study of wheat prices’, Economic History Review, 64 (2011), pp. 447471.
  • Bogart, D., ‘Did the Glorious Revolution contribute to the transport revolution? Evidence from investment in roads and rivers’, Economic History Review, 64 (2011), pp. 10731112.
  • Bowley, A. L., ‘The statistics of wages in the United Kingdom during the last hundred years. Part I. Agricultural wages’, Journal of the Royal Statistical Society, 61 (1898), pp. 702722.
  • Bowley, A. L., ‘The statistics of wages in the United Kingdom during the last hundred years. Part IV. Agricultural wages—concluded. Earnings and general averages’, Journal of the Royal Statistical Society, 62 (1899), pp. 555570.
  • Boyer, G. R., An economic history of the English poor law, 1750–1850 (Cambridge, 1990).
  • Burchardt, J., The allotment movement in England, 1793–1873 (Woodbridge, 2002).
  • Burnett, J., Plenty and want: a social history of food in England from 1815 to the present day (3rd edn. 1989).
  • Cinnirella, F., ‘Optimists or pessimists? A reconsideration of nutritional status in Britain, 1740–1865’, European Review of Economic History, 12 (2008), pp. 325354.
  • Clark, G., ‘Farm wages and living standards in the industrial revolution: England, 1670–1869’, Economic History Review, LIV (2001), pp. 477505.
  • Clark, G., ‘The condition of the working class in England 1209–2004’, Journal of Economic History, 58 (2005), pp. 830843.
  • Clark, G., ‘The long march of history: farm wages, population, and economic growth, England 1209–1869’, Economic History Review, 60 (2007), pp. 97135.
  • Crafts, N. F. R., ‘Regional price variation in England in 1843: an aspect of the standard-of-living debate’, Explorations in Economic History, 19 (1982), pp. 5170.
  • Davies, D., The case of labourers in husbandry (1795).
  • Department of Health, Dietary reference values for food energy and nutrients for the United Kingdom (1991).
  • Eden, F. M., The state of the poor: or, an history of the labouring classes in England, 3 vols. (1797).
  • Feinstein, C. H., ‘Pessimism perpetuated: real wages and the standard of living in Britain during and after the industrial revolution’, Journal of Economic History, 58 (1998), pp. 625658.
  • Floud, R., Fogel, R. W., Harris, B., and Hong, S. C., The changing body: health, nutrition, and human development in the western world since 1700 (Cambridge, 2011).
  • Gazeley, I., ‘The cost of living for urban workers in late Victorian and Edwardian Britain’, Economic History Review, XLII (1989), pp. 207221.
  • Gazeley, I. and Verdon, N., ‘The first poverty-line: Davies and Eden and the characteristics of rural poverty in late eighteenth century England’, Explorations in Economic History (forthcoming).
  • Harper, A. E., ‘Evolution of recommended dietary allowances—new directions?’, Annual Review of Nutrition, 7 (1987), pp. 509537.
  • Harris, B., ‘Gender, health and mortality in nineteenth- and twentieth-century Britain: some preliminary reflections’, in J. Komlos and J. Baten , eds., The biological standard of living in comparative perspective (Stuttgart, 1998), pp. 413448.
  • Harris, B., ‘Gender, health and welfare in England and Wales since industrialisation’, Research in Economic History, 26 (2008), pp. 157204.
  • Horrell, S. and Oxley, D., ‘Bringing home the bacon? Regional nutrition, stature, and gender in the industrial revolution’, Economic History Review (2012), doi: 10.1111/j.1468-0289.2011.00642.x.
  • Hunt, E. H., ‘Industrialization and regional inequality: wages in Britain, 1760–1914’, Journal of Economic History, XLVI (1986), pp. 935966.
  • Manore, M. M., ‘Effect of physical activity on thiamine, riboflavin, and vitamin B-6 requirements’, American Journal of Clinical Nutrition, 72, suppl. (2000), pp. 598S606S.
  • Mathias, P., The brewing industry in England, 1700–1830 (Cambridge, 1959).
  • Mitchell, B. R., British historical statistics (Cambridge, 1988).
  • Moselle, B., ‘Allotments, enclosure, and proletarianization in early nineteenth-century southern England’, Economic History Review, XLVIII (1995), pp. 482500.
  • Muldrew, C., Food, energy and the creation of industriousness: work and material culture in agrarian England, 1550–1780 (Cambridge, 2011).
  • Nicholas, S. and Oxley, D., ‘The living standards of women during the industrial revolution, 1795–1820’, Economic History Review, XLVI (1993), pp. 723749.
  • Nicholas, S. and Steckel, R. H., ‘Heights and living standards of English workers during the early years of industrialization, 1770–1815’, Journal of Economic History, 51 (1991), pp. 937957.
  • Oddy, D. J., ‘Urban famine in nineteenth-century Britain: the effect of the Lancashire cotton famine on working-class diet and health’, Economic History Review, XXXVI (1983), pp. 6886.
  • Paul, A. A. and Southgate, D. A. T., McCance and Widdowson's The composition of foods (4th edn. 1978).
  • Purdy, F., ‘On the earnings of agricultural labourers in England and Wales, 1860’, Journal of the Royal Statistical Society of London, 24 (1861), pp. 328373.
  • Richardson, T. L., ‘The agricultural labourer's standard of living in Kent, 1790–1840’, in D. J. Oddy and D. S. Miller , eds., The making of the modern British diet (1976), pp. 103116.
  • Rowntree, B. S. and Kendall, M., How the labourer lives. A study of the rural labour problem (1913).
  • Shammas, C., ‘The eighteenth-century English diet and economic change’, Explorations in Economic History, 21 (1984), pp. 254269.
  • Shaw-Taylor, L., ‘Labourers, cows, common rights and parliamentary enclosure: the evidence of contemporary comment c. 1760–1810’, Past and Present, 171 (2001), pp. 95126.
  • Shaw-Taylor, L., ‘Parliamentary enclosure and the emergence of an English agricultural proletariat’, Journal of Economic History, 61 (2001), pp. 640662.
  • Shaw-Taylor, L., ‘The nature and scale of the cottage economy’ (mimeo, Univ. of Cambridge, 2002) http://www.geog.cam.ac.uk/research/projects/occupations/abstracts/paper15.pdf (accessed on 2 March 2012).
  • Szreter, S., Fertility, class and gender in Britain, 1860–1940 (Cambridge, 1996).

Official publications

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information
  • Poor Law Amendment Act: Returns showing the weekly diet of several independent labourers in the counties of Leicester, Lincoln and Cambridge; and also, a return from several parishes in the counties of Cambridge and Suffolk of the number of pauper marriages and of illegitimate births, in stated periods, before and since the operation of the Poor Law Amendment Act (P.P. 1837/8, XXXVIII).
  • Reports of the Special Assistant of Poor Law Commissioners on the Employment of Women and Children in Agriculture (P.P. 1843, XII).
  • Sixth Report of the Medical Officer of the Privy Council, app. 6 (P.P. 1864, XXVIII).
  • Royal Commission on Labour, The Agricultural Labourer (P.P. 1893–4, XXXV).
  • Fifth Report of the Medical Officer of the Privy Council (P.P. 1862, XXV).
  • Royal Commission of Inquiry into Administration and Practical Operation of the Poor Laws (P.P. 1834, XXVIII).

Supporting Information

  1. Top of page
  2. Abstract
  3. I
  4. II
  5. III
  6. IV
  7. V
  8. Footnote references
  9. Official publications
  10. Supporting Information
FilenameFormatSizeDescription
ehr672-sup-0001-si.doc43K

Appendix S1. Self-provisioning and derived food quantity estimates.

ehr672-sup-0002-si.doc49K

Appendix S2. Comparisons of male wages in the household accounts with those extracted from other sources.

ehr672-sup-0003-si.doc35K

Appendix S3. Computations of numbers of children in each age category.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.