Public health, nutrition and the decline of mortality: the McKeown thesis

revisited

Bernard Harris

Division of Sociology and Social Policy

University of Southampton

Highfield

Southampton SO17 1BJ.

1

Public health, nutrition and the decline of mortality: the McKeown thesis

revisited

By BERNARD HARRIS.∗

SUMMARY. The medical writer, Thomas McKeown, can justifiably claim to have

been one of the most influential figures in the development of the social history of

medicine during the third quarter of the twentieth century. Between 1955 and his

death in 1988, he published a stream of articles and books in which he outlined his

ideas about the reasons for the decline of mortality and the ‘modern rise of

population’ in Britain and other countries from the early-eighteenth century onwards.

Although McKeown’s main aim was to deflate the claims made by the proponents of

therapeutic medicine, his publications have sparked a long and protracted debate

about the respective roles of improvements in sanitation and nutrition in the process

of mortality decline, with particular emphasis in recent years on the impact of

sanitary reform in the second half of the nineteenth century. This paper attempts to

place the debate over the ‘McKeown thesis’ in a more long-term context, by looking

at the determinants of mortality change in England and Wales throughout the whole

of the period between circa 1750 and 1914, and pays particular attention to the role

of nutrition. It offers a qualified defence of the McKeown hypothesis, and argues that

nutrition needs to be regarded as one of a battery of factors, often interacting, which

played a key role in Britain’s mortality transition.

Keywords. Public health, sanitation, diet, nutrition, mortality, living standards, real

wages, housing, state intervention, ‘McKeown thesis’.

2

Between 1700 and 1911-15, the crude death rate in England and Wales declined

from 27.9 deaths per thousand living to 14.4 and average life expectancy at birth

increased from 37.1 (in 1701) to 53.5 (in 1910-12).1 These figures reflect a major

improvement in the life-chances of the British population over the course of the

period, and therefore it is hardly surprising that the ‘McKeown thesis’, which attempts

to account for the decline of mortality and the ‘modern rise of population’, should

have played such an important part, not only in debates about economic, social and

medical history, but also in the fields of population studies and historical

epidemiology. However, in recent years opinion has undoubtedly moved against

McKeown and his coauthors, and it has even been suggested that the time may

have come ‘to draw a line under the McKeown interpretation and simply

acknowledge that its greatest strength has proved to be its enduring ability to

stimulate debate’.2 If this is right, then this paper may, in some respects, seem

somewhat ill-timed.

Although McKeown examined a wide range of factors in his efforts to account

for the decline of mortality, he attached the greatest importance to the improvement

of nutrition, and this has led many of his critics to complain that he failed to take

sufficient account of such factors as changes in the virulence of infectious

organisms, improvements in personal and domestic hygiene, medical intervention

and, perhaps most importantly, the beneficial effects of the sanitary revolution of the

second half of the nineteenth century. However, as Hionidou has recently reminded

us, epidemiologists have continued to emphasize the importance of the role which

nutrition can play in combating disease, during famines and at other times.3 This

suggests that the real issue is not whether nutrition is capable of influencing trends in

3

mortality, but whether changes in nutrition did have an influence on the pattern of

mortality in England and Wales during the eighteenth and nineteenth centuries.

Although it is difficult to offer any categorical answers to this question, this paper will

suggest that it would be wrong to exclude the role of nutritional change altogether,

and that nutrition should be regarded as one of a battery of factors, often interacting,

which played a key role in Britain’s mortality transition.

The synergistic relationship between nutrition, infection and mortality

During the course of the last century, there have been many advances in our

knowledge and understanding of the essential requirements for human nutrition.

Seebohm Rowntree confined his analysis of essential nutrients to fats,

carbohydrates and proteins, but during the interwar period, nutrition researchers

became increasingly interested in the role played by vitamins and minerals in the

aetiology of what subsequently became known as ‘deficiency diseases’.4 After the

Second World War, this research led to the identification of what is now known as

‘protein-energy malnutrition’ or ‘protein-calorie malnutrition’, which Tomkins has

defined as a condition in which ‘deficiencies of major body nutrients, resulting from a

diet which is generally inadequate in energy and protein, are frequently accompanied

by deficiencies of micronutrients’.5

Nutrition researchers have also drawn an important distinction between the

concept of nutrition, or diet, and nutritional status. The concept of nutrition refers to

the amount and quality of the food consumed by each individual, whereas nutritional

status refers to the balance between the food consumed by each individual and the

4

claims made upon it. As Dasgupta and Ray have argued, the most important

category of claims are those required for the maintenance of the body’s basic

functions, such as temperature control, the circulation of the blood, and breathing,

but human beings also require additional quantities of food in order to enable them to

perform work and ward off the effects of disease. Individuals who perform large

amounts of energy-intensive work require more food than individuals engaged in

more sedentary occupations, whilst individuals who are subjected to repeated bouts

of infection require more food than individuals whose environments are largely

disease-free. The amounts of food required by individuals also vary according to the

size of their bodies. Children require less food than adults, and the average woman

requires less food for the maintenance of her normal functions than the average

man.6

Although it is generally accepted that nutrition and infection are interrelated, it

can often be difficult to distinguish between the two. This is partly related to the fact

that those individuals who are most likely to suffer from defective diets are also the

individuals who are most likely to be living in insanitary and overcrowded

environments, with the result that those individuals who are most likely to show signs

of malnutrition are also the individuals who are most susceptible to outbreaks of

infectious disease.7 However, it is also important to recognize that the effects of

malnutrition and infection can also be mutually reinforcing. As Scrimshaw and

SanGiovanni have argued, infection can have an adverse effect on nutrition,

because it leads to the suppression of appetite and inhibits the body’s ability to

digest those nutrients which are consumed, and malnourished individuals are less

likely to recover from an infection once it has been contracted.8 As a result, it can

5

often be difficult to determine which factor plays the greatest role in determining the

outcome of an infectious disease.

However, despite these problems, it seems clear that there is a widespread, if

not universal, consensus on the part of the epidemiological community that

nutritional deficiencies do play an important part in the development of a wide range

of infectious diseases. Scrimshaw, Taylor and Gordon argued that even though

nutritional factors tended to have little effect on the outcome of an infection in cases

where natural resistance was either high or low in relation to the virulence of the

infection, nutritional factors were important in situations where there was an

equilibrium between the natural or constitutional resistance of the host and the

virulence of the infective agent, and this conclusion has been reinforced by

subsequent research.9 In 1991, Lunn concluded that ‘whatever the cause of the

deterioration in nutritional status, it has become generally accepted that malnutrition

predisposes an individual to infectious diseases’ and that ‘when illness does strike, it

is likely to be more severe, prolonged, and carries an increased risk of death or

permanent damage’.10

Although it seems clear that nutrition can play an important role in the

development of a number of infectious diseases, the nature of this role may not

always be straightforward. Aaby argued that nutritional factors played a very limited

role in the outcome of measles epidemics, and that the progress of the disease was

more likely to be related to the extent of overcrowding and the intensity of the

infective dose,11 but this interpretation has not been accepted uncritically. In a wide-

ranging review, Tomkins argued that even though cases of mild to moderate

malnutrition had little effect on the initial stages of measles infection, they did have

6

an effect on the outcome of post-viral complications, and severe cases of

malnutrition had an effect on all stages of the disease.12

During the last two decades, epidemiologists have devoted considerable effort

to understanding the ways in which nutritional deficiencies can affect a range of

different conditions. In 1982, a multidisciplinary group of historians, demographers,

economists, food scientists and nutritionists concluded that nutritional status had

relatively little effect on the development of such diseases as plague, typhoid,

tetanus, smallpox or malaria, and that it only exerted a ‘variable’ impact on the

outcomes of diphtheria, influenza, syphilis and typhus.13 However, they believed that

it had a strong effect on the outcomes of such diseases as cholera, measles,

leprosy, tuberculosis, and both bacterial and viral respiratory infections. From the

point of view of the later sections of this paper, it may be worth noting that this list

includes a number of conditions, including tuberculosis, which made a major

contribution to the decline of mortality in Victorian Britain.14

One of the main areas of difficulty in interpreting these arguments concerns

the question of whether or not there is a ‘threshold’ at which levels of malnutrition

become sufficiently severe to affect disease outcomes, or whether all levels of ‘sub-

optimal’ nutrition can have adverse consequences. The majority of observers

appear to believe that the relationship between nutrition and infection is not

continuous and that, to quote Scrimshaw and SanGiovanni, nutrient deficiencies are

only likely to affect disease outcomes if they are ‘sufficiently severe’.15 However,

Lunn suggested that ‘some deficiencies, particularly those in the cell-mediated

immune system, do appear to be affected at an early stage of undernutrition’, and

even though it is not yet clear how severe such changes need to be before immune

status is affected, ‘it seems probable that where reductions in the immune

7

components described are found, there will be an overall increased susceptibility to

infectious disease’.16

Nutrition and the decline of mortality before 1820

During the last twenty years, there has been considerable debate over the precise

pattern of mortality change in England (or England and Wales) since the beginning

of the eighteenth century. Razzell suggested that mortality may have declined

substantially during the first half of the eighteenth century, but this suggestion was

rejected by Wrigley, Davies, Oeppen and Schofield on the grounds that it depended

largely on the behaviour of adult mortality, and failed to take sufficient account of

countervailing trends in mortality at younger ages.17 Although Wrigley and his

coauthors now accept that mortality change exerted a stronger influence on

eighteenth-century population trends than they had previously thought, their latest

findings reinforce their earlier view that the secular decline in English mortality began

during the second half of the eighteenth century and continued into the first two

decades of the nineteenth century. There was then an arrest of progress between

the 1820s and 1860s, followed by a resumption of mortality decline from the late-

1860s onwards.

In addition to these disputes over the course of mortality decline, there has

also been a vigorous debate over its causes. In a series of publications, McKeown

and his coauthors denied that either sanitary improvement or therapeutic intervention

(in the form of smallpox inoculation) had any significant impact on mortality before

1800,18 but these conclusions have been strongly challenged by a number of

8

historians. Razzell and Mercer have argued that both inoculation and vaccination

led to a sharp reduction in the incidence of smallpox mortality,19 and Dobson has

shown that the draining of marshlands in Essex, Kent and Sussex helped to reduce

mortality from malaria in the south-eastern corner of England between 1670 and

1800.20 Jones and Falkus have drawn attention to the various environmental

improvements which took place in the market towns of southern England, and both

Porter and Landers have demonstrated the importance of environmental

improvement in London during the second half of the eighteenth century.21

However, this research does not mean that the influence of nutritional factors should

be discounted altogether. Even though McKeown and his colleagues have often

been accused of basing their arguments on inference rather than direct evidence,

there is now a growing body of evidence which does suggest that nutritional factors

may also have played a part in reducing mortality.

As we have already seen, there is now a growing debate among nutritionists

as to whether it is possible to identify a ‘threshold’ below which individuals can be

said to be sufficiently malnourished for the degree of malnutrition to result in

increased susceptibility to infectious disease, and this debate has important

implications for our efforts to examine the relationship between malnutrition and

mortality among populations in the past. As Floud, Wachter and Gregory have

shown, eighteenth-century Britons were both shorter and lighter than their modern

equivalents,22 but this does not mean that their food requirements were necessarily

lower. One of the most important factors which needs to be considered is the extent

to which individuals were able to digest the nutrients they consumed. Dasgupta and

Ray showed that individuals who are subjected to repeated bouts of diarrhoeal

infection and whose diets contain large proportions of dietary fibre are only able to

9

digest approximately eighty per cent of the nutrients they consume, whilst

conventional estimates of food adequacy tend to assume that around 95 per cent of

nutrients will be digested.23 This implies that, just as conventional estimates of

dietary need are likely to underestimate the food needs of deprived populations in

the modern world, they are also likely to underestimate the food needs of

populations in the past.

In addition to considering the adequacy of the standards used to establish

dietary standards for individuals, it is also important to consider the distribution of

food within populations. Livi-Bacci has argued that ‘a population which could rely on

a normal consumption of 2000 calories per head would have been, in centuries past,

an adequately-fed population, at least from the point of view of energy’,24 but this

view takes little account of inequalities in the distribution of food within either

households or populations.25 It is also difficult to reconcile Livi-Bacci’s view that

nutritional deprivation was not an important factor in the high mortality of pre-

industrial societies with his account of the role played by nutrition in the improvement

of mortality in Britain after 1850. If one assumes that there was little or no

relationship between nutrition and mortality in the eighteenth century, it is difficult to

see how ‘improved nutrition certainly did play a positive role in increasing life

expectancy’ during the second half of the nineteenth century,26 unless one is also

prepared to argue that there was a significant reduction in the level of food

consumption, or an increase in dietary needs, between the two dates.

In view of these arguments, it seems clear that a substantial proportion of the

British population was inadequately fed during the first half of the eighteenth century,

but it does not automatically follow that the subsequent reduction of mortality was

caused by an improvement in the population’s diet. In the years following the

10

publication of McKeown’s original findings, both Wrigley and Schofield and Livi-

Bacci argued that it was extremely unlikely that dietary standards would have

improved during the second half of the eighteenth century because the average

value of real wages was falling,27 but it is important to recognize the extent to which

these conclusions are dependent on the accuracy and reliability of the price and

wage indices used to calculate real wages. In 1998, Feinstein published new

estimates of the value of real wages (or, to be more precise, real earnings) in

England between 1770 and 1870 which went a long way towards reconciling some

of the apparent contradictions between real wages and mortality. Previous accounts

had suggested that real wages rose substantially during the first half of the

eighteenth century, fell back between 1750 and circa 1810, and then rose sharply

from the 1820s onwards.28 In contrast, Feinstein’s figures suggested that real

earnings rose by 12.5 per cent in Great Britain between 1770/2 and 1818/22, and by

23.1 per cent between 1818/22 and 1848/52 (see Figure 1).

Figure 1 about here

In view of these disagreements over the trend in real wages, it is not

surprising that other historians have looked elsewhere for direct evidence of

nutritional standards. Towards the end of the eighteenth century, Davies and Eden

presented information about the consumption patterns of 213 labourers’ households,

and these data have been analysed by modern historians to yield new estimates of

the nutritional value of the diets available to labourers’ families in the preindustrial

period. In 1995, Clark, Huberman and Lindert estimated that the average value of

the diet consumed by the typical Davies-Eden household was equivalent to 1508

11

calories per person per day and 27.9 grammes of protein per person per day,

whilst Shammas, using a restricted sample of 22 Davies-Eden households,

estimated that mean daily consumption was equal to 1734 calories per person in the

south of England, and 2352 calories per person in the north.29 It is not entirely clear

why these estimates should differ, and neither set of figures includes any allowance

for the food which was provided for individuals at work or which they were able to

produce for themselves, but it still seems clear that the overall figures are very low.

In her 1984 article, Shammas concluded that ‘the south’s 2100-2500 [calories per

day per adult equivalent] would not seem to provide the energy for hard labour or

growth in children and the north’s 2800-3200 was, at best, barely adequate’.30

One of the greatest challenges posed by the analysis of dietary information in

the past in the difficulty of knowing exactly how the groups represented in any one

survey should be compared with the population as a whole. As we have already

seen, Clark, Huberman and Lindert believed that the value of the diets consumed by

the Davies-Eden households was very low, but they also believed that these

households were drawn from the very poorest sections of the population. By

contrast, Shammas does not appear to have regarded the families in her sample as

being especially poor, and Fogel has argued that they were in fact close to the

median for the population as a whole.31 In 1989, Fogel used the information

provided by Shammas to estimate the distribution of food consumption among the

population as a whole in 1790. He estimated that the average value of the diets

consumed by the poorest decile of the population was equivalent to 1545 calories

per adult equivalent per day, whilst the bottom twenty per cent of the population still

consumed less than 2000 calories per adult equivalent per day. By contrast, the

12

richest decile of the population consumed a diet with an average daily value of

more than 4000 calories per adult equivalent (or ‘consuming unit’).32

In addition to estimating the level of food consumption at the end of the

eighteenth century, it is also important to obtain some impression of the likely trends

in consumption over time. Shammas argued that the nutritional value of the average

diet may have declined because the proportion of bought foods increased and the

the consumption of milk and dairy products declined.33 By contrast, Fogel has

estimated that the nutritional quality of the average diet increased between 1750 and

1800 (because of an increase in the proportion of calories derived from meat

products), and that the average calorific value rose by just under seven per cent

between 1700 and 1800, and by a further 5.6 per cent between 1800 and 1850.

Nevertheless, even in 1850, the average British consumer only consumed

approximately 75 per cent of the calories consumed by his or her successors at the

end of the 1980s.34

As the above paragraphs have demonstrated, it is difficult to reach any

categorical conclusions regarding the main trends in dietary standards before 1800,

but it seems reasonably clear that the average level of nutrition was low and that a

substantial proportion of the population subsisted on diets which fell significantly

below the standards recommended by modern nutritional experts. The main area of

uncertainty concerns the question of change. As we have already seen, there are

some grounds for believing that the quality of the average diet may have declined,

but other evidence suggests that the proportion of meat may have increased and

that overall energy levels may also have risen, even though the extent of any

increase was not very great, and the value of the food consumed remained very low.

Thus, even if the decline of mortality between 1780 and 1820 was related to

13

improvements in nutrition, the limited nature of these improvements may help to

explain why mortality rates only fell at the rate they did.

Although there is a considerable body of evidence to show that the poorer

members of the population were clearly undernourished by modern standards, it is

rather more difficult to argue that the high mortality rates experienced by the better-

off were related to malnutrition. As we can see from Figure 3, there was very little

difference between the life expectancy of aristocratic infants and that of the

population at large before the early years of the eighteenth century, and even during

the eighteenth and nineteenth centuries, members of the aristocracy experienced life

expectancies which fell well below modern standards. These facts have led Razzell

and Livi-Bacci to question the whole basis of the argument that the decline in

mortality as a whole can be related to improvements in nutrition.35

Figure 2 about here

In The modern rise of population, McKeown suggested that even though the

aristocracy were not themselves malnourished, they were affected by the presence

of malnutrition among the population at large. As we have already seen, modern

epidemiologists believe that the presence of severe malnutrition can render a

population more susceptible to infection, as well as impeding their prospects of

recovery, and McKeown argued that the health of the aristocracy was undermined

because the poor nutritional status of the general population increased the incidence

of infection in the population as a whole. However, he was unable to explain why the

improvement in nutrition which was central to his explanation for the decline of

aristocratic mortality should only have led, in the first instance, to an increase in the

14

average life expectancy of the better off, without any immediate or equivalent

impact on the life expectancy of the poor.36

In view of these problems, it is difficult to argue that the decline in aristocratic

mortality was initiated by a general improvement in nutrition, but this does not mean

that nutrition was unimportant overall. As we have already seen, Scrimshaw,

Gordon and Taylor suggested that nutritional factors were only likely to exert a

decisive influence on the outcome of an infectious disease in situations where there

was an epidemiological balance between the natural resistance of the host and the

virulence of the infective organism, and this may help to explain why nutritional

factors do not appear to have played a leading role in the determination of mortality

before circa 1750. 37 Kunitz and Engerman argued that during the sixteenth and

seventeenth centuries, the major causes of premature death were epidemic or

pandemic diseases, but these either became less important (in the case of plague)

or more endemic (in the case of smallpox) during the course of the eighteenth

century, and this allowed social factors, such as differences in hygiene, domestic

arrangements and nutritional status, to become more important as the century

progressed.38 Consequently, the real significance of the ‘peerage paradox’ is not

that it demonstrates the irrelevance of nutritional explanations for mortality decline,

but, rather, that it demonstrates the need for a more sophisticated understanding of

the relationship between these factors and the changing nature of the disease

environment.

15

Mortality change c. 1820-1850

Although there are good grounds for believing that nutritional factors were involved in

the decline of mortality before 1820, it is rather more difficult to argue that the

cessation of progress which followed this period was caused by a deterioration in

nutritional standards between circa 1820 and 1850 (or even 1870), and there are

even grounds for believing that nutritional standards may even have risen. However,

the most obvious explanation for the cessation of progress between 1820 and 1850

is that the beneficial effects of any improvement in nutrition were undermined by the

impact of urbanization on the nature of the disease environment in which a growing

proportion of the population lived.

During the last century, few debates in economic history have attracted more

attention, or generated more controversy, than the debate over the ‘standard of

living’ during the industrial revolution.39 In 1816, the government statistician, John

Rickman chastised his friend, Robert Southey, for entertaining the belief that ‘[the]

state [of the poor] has grown worse and worse of late’,40 and in 1926 Sir John

Clapham pointed out that, according to the latest estimates, ‘the purchasing power of

wages in general – not, of course, everyone’s wages’ was definitely increasing

between 1820 and 1850.41 However, in 1930 the socialist historian, J. L. Hammond

retorted that even though the industrial revolution may have made life easier and

more comfortable for thousands of men and women, ‘the ugliness of the new life,

with its growing slums, its lack of beautiful buildings, its destruction of nature and its

disregard of man’s deeper needs, affected not this or that class of worker only, but

the entire working class population’.42

16

During the following decades, a new generation of economic historians

tried to challenge Clapham on his own territory. In 1957, Eric Hobsbawm criticized

the evidence which Clapham had used to measure real wages, and contrasted this

with the available evidence on food consumption. He argued that there were

declines in the per capita consumption of wheat, milk and cheese, and that the

consumption of meat also declined, principally because the number of cattle and

sheep slaughtered at London’s Smithfield Market failed to keep pace with the growth

of the capital’s population.43 However, many of these claims were strongly rebutted

by Max Hartwell in 1961. In particular, he argued that there was no firm evidence for

the view that grain consumption declined, and he challenged Hobsbawm’s

arguments about meat consumption by demonstrating that it was impossible to

generalize from the experience of Smithfield market at a time when London’s other

meat markets were growing much more rapidly.44

The difficulties associated with the measurement of trends in the consumption

of domestically-produced goods have prompted a number of writers to focus rather

more attention on the consumption of imported items, such as tea and sugar.

Burnett argued that the consumption of sugar fell between 1811 and 1821, and did

not begin to rise significantly before the mid-1840s, whilst the consumption of tea

(which he regarded as a much truer indicator of standards) only began to increase in

the 1830s.45 Mokyr used similar information to estimate changes in real wages

overall. He concluded that the average level of real wages remained largely

unchanged in the three decades following the end of the Napoleonic Wars, and only

began to improve substantially from the mid-1840s onwards.46

These findings highlight the need for caution in assessing changes in real

wages, but most economic historians have continued to argue that the average level

17

of real wages did increase during the first half of the nineteenth century, and the

main areas of contention concern the question of when they began to increase and

by how much. As we have already seen, many commentators, including Wrigley and

Schofield, Lindert and Williamson, Crafts and Schwarz, believed that real wages

stagnated or fell during the second half of the eighteenth century, before rising

sharply from the early-1800s onwards,47 but Feinstein’s more recent calculations

suggest that these authors may have been wrong to suppose that real earnings were

falling in the earlier period, and that they may have exaggerated the extent to which

they increased in the later period. In his article, Feinstein tended to focus most

attention on the movement of real earnings after 1820, and it was for this reason that

his findings appeared to lend most support to the ‘pessimist’ case. However, it is

worth noting that even his revised figures implied that real wages rose by 0.7 per

cent per year between 1818/22 and 1848/52, and by 23.15 per cent over the whole

period.48

Although it is widely accepted that real wages did increase, on average, over

the course of this period, it does not necessarily follow that aggregate levels of food

consumption should have risen as a consequence. Clark, Huberman and Lindert

attributed part of the gap between the available estimates of food consumption and

their own estimates of the increase in the level of real wages to the fact that urban

and industrial occupations tended to be less energy-intensive than traditional rural

occupations, and to the possibility that urban workers ‘chose’ to consume a less

nutritious diet.49 However, as Feinstein pointed out, much of the ‘food puzzle’

identified by Clark, Huberman and Lindert was based on their own, possibly inflated,

estimates of the growth in real earnings. If the growth in real earnings was, as he

18

believed, significantly lower than they argued, then the food puzzle, as such,

would largely disappear.50

In addition to examining questions of taste and income elasticity, economic

historians have also focused attention on the question of relative prices. Komlos has

suggested that the quality of the average diet may have deteriorated because

increases in the relative price of meat encouraged consumers to transfer expenditure

from meat and dairy products to grains, but it is not entirely clear how far the

available data support his view.51 He appears to have based his interpretation on

Horrell’s study of home demand in Britain during the industrial revolution, but her

data suggest that the proportion of expenditure devoted to meat and dairy products

rose by seventeen per cent between 1801 and 1841, and Feinstein’s data suggest

that the share of total expenditure devoted to meat and dairy products remained

unchanged between 1828/32 and 1858/62.52

In view of these findings, it is unlikely that we can attribute a deterioration in

the quality of the working-class diet to changes in the relative price of meat and dairy

products, but other writers have suggested that the demand for food may have been

affected by changes in the price of industrial goods.53 It is also worth noting that

even though food prices were falling between 1820 and 1850, the cost of housing

increased, and it is possible that workers may have attempted to recoup some of

their additional housing costs by economising on the purchase of food.54 However,

even though the proportion of working-class expenditure devoted to food fell

between 1828/32 and 1858/62, the extent of this reduction was outweighed by the

overall increase in the real value of working-class earnings.55 Consequently, even

though working-class families devoted a smaller proportion of their overall budget to

the purchase of food, the increase in the real value of their earnings meant that that

19

the amount of food which they were able to purchase was more likely to have

increased than decreased.56

As a result of these calculations, it seems unlikely that the amount of food

consumed by the wage-earning population of Britain declined between 1820 and

1850, and there are clear grounds for believing that it ought to have increased, even

though the extent of this increase was probably small, and may not necessarily have

been reflected in the aggregate accounts. This conclusion is reinforced by an

examination of the most recently-published analyses of eighteenth- and nineteenth-

century household budgets, but these also need to be treated with care. During the

1990s, both Oddy and Clark, Huberman and Lindert estimated the average

nutritional value of the food consumed by working-class families at different points in

time between the end of the eighteenth century and the middle of the nineteenth

century, but there are significant differences between the results which they reported

and those reported by Shammas, even though all their findings were derived from

substantially the same sources.57 Nevertheless, all three sets of results still suggest

that there were significant improvements in both the quantity and quality of the diets

consumed by poor English workers, and their families, during this period (see Table

1).58

Table 1 about here

These findings make it unlikely that the cessation of mortality decline after

1820 can be attributed, to any significant degree, to a deterioration in the level of

nutrition, but it may be attributable to a change in the nature of the disease

environment in which a growing proportion of the population lived. During the period

20

between 1800 and 1850, several commentators drew attention to the high death

rates, associated with urban areas, and these fears were compounded by the pace

at which the size of the urban population was increasing.59 As Wohl has pointed out,

the number of towns and cities containing more than 100,000 inhabitants increased

from one to eight between 1801 and 1851, and by the end of the period the

population of London had risen to more than one million.60 During the period as a

whole, the proportion of the population residing in towns containing more than

10,000 inhabitants rose from 24 per cent to 44 per cent, and the proportion of the

population residing in towns with more than 100,000 inhabitants rose from eleven

per cent to 25 per cent.61 Between 1801 and 1851, the population of Manchester

increased from 75,000 people to 303,000, whilst the population of Liverpool rose

from 82,000 to 376,000. The population of Bradford rose from 13,000 to 104,000.62

During the last two decades, a number of authors have argued that urban

conditions were not only less healthy than rural conditions, but that conditions within

towns may have grown worse. Huck showed that there was a significant increase in

the level of infant mortality in nine urban parishes in central and northern England

between 1813/18 and 1831/6, and evidence from four of these parishes (Walsall,

West Bromwich, Wigan and Ashton) suggested that this deterioration was unlikely to

have been reversed, and may even have continued, into the 1840s.63 This evidence

is supported by Floud, Wachter and Gregory’s account of changes in the average

height of British army recruits who were born in different parts of Britain between the

1820s and 1850s. They found that there was a substantial deterioration in the

average heights of men born in London and other urban centres, and that this was at

least partly responsible for the decline in the average height of the male population

as a whole during this period.64

21

Although there is strong evidence of a deterioration in infant mortality and in

stature, it has proved rather more difficult to establish whether or not there was any

deterioration in mortality rates as a whole within urban areas. In 1985, Woods used

data for 1861 and 1911 to estimate the main trends in mortality in four different types

of area between 1811 and 1911. He suggested that average life expectancy at birth

rose in all parts of Britain during the first half of the nineteenth century, including rural

areas, small towns, large provincial towns and the capital, and that the absence of

any clear improvement in national mortality rates was a direct result of changes in

the proportions of the population inhabiting different types of districts65 but Szreter

and Mooney have argued that average life expectancy at birth declined sharply in

the largest provincial cities during the 1830s, and only began to improve consistently

from the 1870s.66 However, as Woods himself has pointed out, Szreter and

Mooney’s figures for the early part of the nineteenth century were largely (though not

entirely) dependent on the evidence for Glasgow, and even if one accepts the

reliability of the Glasgow data, it is still open to question how far they can be used to

estimate the main trends in mortality in cities as diverse as Bradford, Newcastle,

Sheffield, Bristol, Leeds, Birmingham, Manchester and Liverpool.67

In view of the obvious dangers associated with these figures, it is clearly

advisable to treat any estimates regarding the pattern of mortality change in different

parts of the country in the early part of the nineteenth century with a high degree of

caution, but some points are beginning to become a little more clear. Although

Woods has questioned the validity of Szreter and Mooney’s overall conclusions, he

has also revised his own estimates of changes in life expectancy in the different

types of area, and he now believes that average life expectancy may have declined

after all in both large and small towns between the 1830s and 1850s.68 However,

22

whilst this represents an important concession to supporters of a more pessimistic

interpretation of the impact of urbanization, it also raises other questions about the

pattern of mortality change in the country as a whole. As we can see from Table 2,

the new estimates suggest that the average level of life expectancy at birth in towns

containing between 10 and 100,000 inhabitants fell by 0.3 years during the 1840s,

but this figure may not have been true of all these areas. In their paper, Szreter and

Mooney emphasized the fact that it was not the size of population per se which

mattered, but the rate at which the population was increasing, and this suggests that

even though many smaller and medium-sized towns experienced a deterioration in

their standards of public health, others may have witnessed some improvement, and

this impression is reinforced by the fact that the average level of life expectancy in

rural areas also appears to have increased.69

Woods’ figures also raise questions about the pattern of mortality changes in

the country as a whole. Szreter and Mooney argued that there was a catastrophic

fall in life expectancy in the largest provincial cities during the 1830s, and even

though their figures suggested that the average level of life expectancy rose by four

years during the 1850s, they nevertheless concluded that there was no substantial

increase in life expectancy before the 1870s. By contrast, Woods’ figures point to a

somewhat smaller reduction in average life expectancy in the 1830s and 1840s,

followed by a small but not insignificant improvement from the 1850s onwards.

These figures suggest that the improvement in urban life expectancy cannot simply

be attributed to sanitary intervention, since even Szreter concedes that there is little

evidence of effective sanitary intervention in these areas before 1870, and Bell and

Millward have argued that sanitary intervention is unlikely to have made much

difference to mortality rates before the 1890s or even the early-1900s.70

23

Table 2 about here

These calculations suggest that even though urbanization was undoubtedly

the most important influence on Britain’s mortality rate during this period, it was not

the only influence. As we have already seen, the evidence of real wages,

expenditure patterns and food consumption suggests that dietary standards were

also improving, and this may help to explain why mortality rates continued to improve

within certain types of area, even though they failed to improve in the country as a

whole. This interpretation is reinforced by Floud, Wachter and Gregory’s analysis of

changes in the average height of army recruits, which began to improve with the

birth cohorts of the 1850s. Although they recognized that diet was only one of the

factors which was likely to influence stature, they nevertheless concluded that ‘the

height data make the link between nutrition (although in a wider sense) and mortality

which McKeown could only infer’.71

Floud, Wachter and Gregory’s findings also raise questions about the timing

of the factors which determine both childhood and adult mortality. In 1934, Kermack,

McKendrick and McKinlay highlighted ‘certain regularities in the vital statistics of

Britain’,72 in which each cohort or generation appeared to experience a consistently

lower level of mortality throughout the life course than its predecessor, and this has

led many subsequent researchers to devote much more attention to the impact of

cohort-specific factors on mortality and to the ways in which nutritional and

environmental influences in early life can influence mortality at later ages.73 These

investigations certainly raise important new questions about the pattern of mortality

change in the second and third quarters of the nineteenth century.74

24

Nutrition and the decline of mortality 1850-1914

Although the McKeown hypothesis was concerned with the whole of the period from

1700 onwards, much of the fiercest debate has focused on the critical period

between 1850 and 1914. In 1962, McKeown and Record examined changes in the

main causes of death and concluded that ‘it seems unquestionable that the decline

of mortality between 1851/60 and 1891/1900 was attributable almost exclusively to a

reduction in the frequency of death from infectious disease’.75 In 1976, McKeown

argued that 33.30 per cent of the decline in mortality between 1848/54 and 1901 was

caused by a decline in the incidence of deaths from water- and food-borne diseases,

whilst 43.63 per cent of the decline was caused by a decline in the frequency of

deaths from airborne diseases. He argued that a substantial proportion of the

decline in the incidence of deaths from water- and food-borne diseases was caused

by improvements in hygiene, but that there was little evidence to show that

environmental improvements had any substantial effect on exposure to airborne

diseases before the start of the twentieth century. He therefore concluded that the

decline in the death rate associated with these conditions (respiratory tuberculosis;

bronchitis, pneumonia and influenza; whooping cough; measles; scarlet fever and

diphtheria; smallpox; and infections of the ear, pharynx and larynx) was most likely to

have been caused by an improvement in the population’s capacity to resist infection,

which was itself the result of improvements in nutrition.76

During the last two decades, several historians have attempted to cast doubt

on the form of McKeown’s argument and on the conclusions he drew from it. One of

25

the most intriguing features of the argument concerns the treatment of the decline

in the number of deaths attributable to scarlet fever. In 1962, McKeown and Record

argued that ‘scarlet fever was responsible for about nineteen per cent of the

reduction of mortality during the second half of the nineteenth century’ and that ‘there

is no reason to differ from the general opinion that [the decline in the number of

deaths attributed to scarlet fever] … resulted from a change in the nature of the

disease’,77 but in 1976 McKeown included scarlet fever and diphtheria in the

category of airborne diseases whose decline he attributed to an improvement in

nutrition. If the decline in the number of deaths attributed to scarlet fever and

diphtheria had been excluded from this category, then the overall contribution of the

remaining airborne diseases to the overall decline in mortality would have fallen from

43.63 per cent to 31.20 per cent – i.e. less than the figure associated with the decline

in the incidence of mortality from water- and food-borne diseases (see Table 3

below).

Despite the obvious importance of scarlet fever to the construction of

McKeown’s argument, his critics have often tended to focus rather more attention on

the question of tuberculosis. In 1988, Szreter argued that McKeown had

exaggerated the extent of the contribution made by pulmonary tuberculosis to the

overall decline in mortality, and that this had led him to underestimate the

significance of improvements in the incidence of mortality from water- and food-

borne diseases.78 However Szreter’s own interpretation of the tuberculosis statistics

has also been challenged, and he subsequently conceded that ‘the recorded fall in

respiratory tuberculosis was probably genuine, [even though] it also confirms the full

extent of the contradictory rise in bronchitis/pneumonia/’flu’.79 This conclusion has

since been reinforced by Woods’ analysis of the cause-of-death data in the

26

Registrar-General’s Decennial Supplements for the period 1861/70 to 1891/1900.

Although Woods’ figures cover a slightly different period to those employed by

McKeown, they are broadly consistent with his original interpretation (see Table 3).

Table 3 about here

One of the most interesting attempts to account for the decline of mortality in

the latter part of the nineteenth century in recent years has come from attempts to

relate the history of mortality to that of morbidity. Riley has argued that one of the

reasons for the decline of mortality was that improvements in medical attendance

had enabled sick people to manage their diseases (and, especially, respiratory and

organ diseases) in such a way as to enable them to survive for longer and postpone

mortality.80 However, much of the increase in the duration of sickness episodes on

which Riley based his case may have been caused by increases in the average age,

and changes in the age distribution, of the population who submitted sickness

claims.81 It is difficult to know how far this may have been true of the friendly society

members whose health Riley investigated, but it certainly appears to provide the

most likely explanation for changes in the pattern of morbidity experienced by

members of the Hampshire Friendly Society, in the south of England, over the same

period.82

Although Riley has raised important new questions about the effects of

medical attendance and the relationship between mortality and morbidity, his work

offers a much less sustained critique of the McKeown thesis than Woods’ exhaustive

analysis of the main changes in causes of death in each of Britain’s 614 registration

districts between the 1860s and the 1890s. Woods’ most striking finding was that

27

even though the decline in the death rate from pulmonary tuberculosis accounted

for more than one-third of the total decline in mortality during this period, it appeared

to decline at much the same rate in all parts of the country. He therefore concluded

that this decline was unlikely to have been caused by changes in either diet or

environmental conditions, and that the most likely explanation was a change in

virulence of the infective organism.83 However, this argument has not been

accepted uncritically. Landers thought it was ‘rather unlikely’ that ‘tuberculosis

underwent a spontaneous reduction in virulence … given the apparent evolutionary

age and stability of the tubercle bacillus’, and Szreter noted that Woods’ discussion

of the main trends in tuberculosis was ‘uncharacteristically lacking’ in references to

the most recent medical research.84

In view of these criticisms, it is worth looking more closely at recent studies of

the recrudescence of tuberculosis in Britain since the beginning of the 1980s.

Spence et al. found that there was ‘a strong relation … between notification rates of

tuberculosis and poverty in the last decades of the twentieth century’ and Bhatti et al.

argued that ‘the evidence favours a major role for socio-economic factors that affect

all residents in accounting for the increase in tuberculosis in poor districts’. Although

these authors recognized that poverty was ‘multifaceted’, they also argued that it

was most likely to influence levels of tuberculosis through its association with poor

diet and overcrowding.85 Spence et al. concluded that ‘where social deprivation is

rife, particularly in areas of poor housing, high unemployment, and low incomes, an

increased awareness of tuberculosis as a possible diagnosis is necessary. Far from

diminishing, tuberculosis both worldwide and in the United Kingdom is increasing.

Poverty may be a factor causing this increase’.86

28

During the last three decades, a number of researchers have examined the

relationship between housing standards and tuberculosis mortality in the latter part of

the nineteenth century. Pooley and Pooley found that housing density accounted for

only seventeen per cent of the variance in crude mortality rates in Manchester

between 1871 and 1875,87 and Vögele argued that there was surprisingly little

relationship between housing and population density and either all-cause mortality or

tuberculosis mortality in the country as a whole.88 Moreoever, both Vögele and

McKeown believed that there was comparatively little evidence of any substantial

improvement in housing conditions before the start of the twentieth century.89

However, other historians have challenged this view. Daunton suggested that the

introduction of wallpaper and linoleum in working-class homes during the second half

of the nineteenth century ‘possibly contributed to the reduction in tuberculosis’,90 and

Cronjé argued that there was a considerable amount of evidence to show that

‘housing improved considerably in the second half of the nineteenth century, with a

rise in the quality of new building, a reduction in excessive overcrowding, and

increased government regulation bringing better sanitation and water supply’.91

In view of the conflicting nature of these statements, it is worth looking more

closely at the extent of housing change during this period. As many historians have

pointed out, a very large number of working-class families continued to inhabit ill-

ventilated and overcrowded accommodation at the end of the nineteenth century,

and this may well have contributed to the persistence of high rates of tuberculosis

mortality at the end of the century.92 However, this does not mean that there had

been no improvement in housing conditions before this point. As Daunton has

shown, the substantial increase in the value of real wages after 1850 enabled a

significant proportion of working-class families to pay higher rents in return for

29

improved accommodation, and this was reinforced by the introduction of new

building bye-laws which helped to establish higher standards for the construction of

new houses from the 1870s onwards.93 Consequently, even though it may be

difficult to establish a precise relationship between housing improvement and

mortality decline, there are strong circumstantial grounds for believing that these

improvements did make an important contribution to the decline of tuberculosis (and,

possibly, other diseases) after 1850.

Although many historians might be prepared to accept the view that housing

conditions did play a part in the decline of tuberculosis mortality, the relationship

between diet and mortality has proved even more contentious. Hardy has claimed

that ‘recent research … reinforces the conclusion that, for a significant proportion of

Britain’s population, rising real incomes had little direct impact on improving

nutritional standards’, but the evidence for this is far from clear.94 The statement is

apparently based on the third edition of Burnett’s classic study of Plenty and Want,

but Burnett himself argued that the period between 1873 and 1896 ‘brought

increased purchasing power … and a bigger margin which could go towards

providing a better and more varied diet’ and that ‘improvements are observable in

the general standard of the working-class diet’ during the last quarter of the

nineteenth century.95

In view of the apparently contradictory nature of these statements, it is clearly

worth looking more closely at some of the available evidence on food consumption

during this period. Oddy estimated that the calorific value of the food consumed

each day by the average working-class consumer fell by 360 calories between the

1860s and the 1890s, but he also argued that quality of the diet improved, and that

this was reflected in the reduced consumption of bread and an increase in the

30

consumption of meat and sugar.96 Oddy’s emphasis on the quality of the average

diet was supported by Dewey’s analysis of changes in food consumption between

1880 and 1954. He argued that ‘well before 1914, consumption had begun to shift

away from “inferior” foods (cereals, potatoes) towards “superior” ones (mainly meat

and dairy produce). This process reflected mainly the rise in average real income

which became apparent from about 1850, but it was also related to the greater

availability of certain foods and in some, but not all, cases, a lowering of the price of

individual commodities’.97

Although this paper has been particularly concerned with the impact of

nutritional factors on the decline of mortality, it is also important to consider the role

played by sanitary improvements, especially after 1870. As we have already seen,

McKeown believed that up to one-third of the total decline in mortality between

1848/54 and 1901 was associated with a decline in the incidence of mortality from

water- and food-borne diseases, and that a substantial part of this decline could be

explained by improvements in the sanitary condition of the urban environment.98

Szreter believed that it was no accident that the onset of more rapid mortality decline

after 1870 should have coincided with an increase in the level of local public health

activity, and Hardy has argued that ‘the problems of nineteenth century cities called

for a systematic and professional administrative response, which was provided with

the establishment of sanitary authorities and, more especially, of Medical Officers of

Health in London in 1855 and elsewhere in 1872…. It was [the Medical Officers of

Health] … who spearheaded the Victorian struggle against infectious disease, and it

was their initiative and their labours which led to the eventual eradication of the

epidemic streets’.99

31

In view of the importance of arguments about the effectiveness of social

intervention, it is worth examining this question in a little more detail. In his recent

book, Robert Woods argued that fourteen of the 614 registration districts in England

and Wales (containing thirteen per cent of the total population) were responsible for

25 per cent of the national decline in the number of deaths from diarrhoea and

typhus, but these two diseases were responsible for more than 28 per cent of the

overall decline of mortality in the country as a whole. These findings demonstrated

the importance of sanitary intervention and showed that ‘if sufficient progress could

be made in sanitary reform in even a small number of the most populous places its

effects might outweigh more substantial advances in places with smaller populations,

such as the small towns’.100 However, even though Woods was anxious to

emphasize the importance of sanitary reform, especially when compared to the

effects of any possible nutritional improvements, he also warned against placing

excessive reliance on what he called ‘simple conclusions’.101 One of the most

important problems for supporters of a ‘public health’ explanation for mortality

change is the persistence of high rates of infant mortality before 1900. Even though

the aggregate rate of mortality from diarrhoeal diseases fell by 23.25 per cent

between 1861/70 and 1891/1900, the rate of diarrhoeal mortality among infants only

began to decline after the turn of the century.102

Although it would clearly be wrong to underestimate the importance of

sanitary intervention, it is also important to ask why its impact on mortality should not

have been greater. In 1998, Frances Bell and Robert Millward pointed out that most

of the initial increase in sanitary expenditure was concerned with the provision of

better water supplies, but such measures were often counterproductive because the

authorities failed to make any equivalent investment in sewerage facilities.103 They

32

argued that it was only in the 1890s that many local authorities began to make

significant investments in the full range of sanitary measures and that, consequently,

it was only after this date that the ‘sanitary revolution’ really began to take hold. As a

result, although many local authorities began to invest more heavily in their sanitary

infrastructure from the 1870s onwards, it was not until the first decade of the

twentieth century that these efforts began to exercise a decisive effect on the decline

of mortality in the country as a whole.104

Conclusions

In some respects, this paper may be said to have raised more questions than

answers. During the last three decades, McKeown’s critics have often accused him

of failing to provide any direct evidence to justify the weight he attached to nutritional

factors in his interpretation of the onset of mortality decline. This paper has shown

that such evidence may exist, but it remains difficult to attach precise quantitative

estimates to the impact of different determinants at different points in time.

Although the paper has, therefore, offered a qualified defence of McKeown’s

work, it is also important to acknowledge some particular weaknesses. One of the

most important problems with McKeown’s work is his failure to distinguish between

the concepts of nutrition (or diet) and nutritional status. In his work, he argued that a

large part of the decline of mortality between 1700 and 1914 must be attributable to

improvements in diet because it could not be explained by changes in any of the

other ‘obvious’ causes, but this ignores the extent to which the adequacy of a

person’s diet may also be affected by changes in their epidemiological environment

33

and in the nature and amount of the work they are expected to perform. It is for

this reason that many students of this question have tended to prefer the term

‘nutritional status’. Even though some critics may feel that this concept tends to blur

the boundaries between dietary issues and epidemiological issues, it also reflects a

much better understanding of the synergistic relationship between nutrition and

infection on which so many aspects of human health depend.105

The distinction between nutrition and nutritional status is particularly important

when one considers the way in which McKeown attempted to draw a clear line

between the different types of disease and the factors which may (or may not) have

contributed to their development. Whilst a number of writers, including Szreter, have

drawn attention to the inconsistencies which existed between the diseases which

McKeown listed in the same categories (such as, for example, the differences

between the trends in respiratory tuberculosis and bronchitis), it is also important to

recognize the synergies which may exist between diseases in different categories.

For example, in his work McKeown drew a clear distinction between water- and food-

borne diseases such as diarrhoea, which were amenable to the effects of sanitary

intervention, and airborne diseases, such as respiratory tuberculosis, whose decline

could only have come about through improvements in nutrition, but this ignores the

extent to which reductions in the incidence of water- and food-borne diseases may

have contributed to the decline in nutrition-based diseases by improving the

population’s capacity to absorb essential nutrients.106

It is also arguable that McKeown failed to pay sufficient attention to the way in

which urbanization affected the nature of the disease environment within which

people lived. As we have seen, there was a substantial increase in the proportion of

the population which was exposed to urban influences in the first half of the

34

nineteenth century, and it is highly likely that the urban environment itself also

deteriorated, even if this was not true of the entire country. Consequently, even

though there are grounds for believing that the general standard of nutrition may

have improved, it is difficult to deny that the expansion of urban conditions either

created or exacerbated a public health problem which could only be addressed

properly by effective sanitary intervention.

Although McKeown emphasized the importance of improvements in the

‘standard of living’ in the second half of the nineteenth century, it is arguable that he

interpreted the consequences of this improvement rather too narrowly. In The

modern rise of population, he tended to argue that improvements in mortality could

only come about because of a change in the biological relationship between

organism and host, or because of therapeutic improvements, or as a result of

sanitary intervention, or because of improvements in the standard of nutrition, and it

is therefore arguable that he failed to pay sufficient attention to the ways in which

increases in the value of real wages also enabled people to purchase better housing.

This factor has sometimes appeared to be obscured by an emphasis on the role of

housing legislation (though the impact of this should not be neglected), and it seems

particularly relevant to the analysis of changes in the incidence of mortality from

tuberculosis.

In his famous critique of McKeown, Szreter claimed that one of the major

weaknesses of the ‘McKeown thesis’ was the extent to which it appeared to offer

support to ‘the extreme laissez-faire position that health and welfare gains may be

generated most effectively as a by-product of economic growth’,107 but this criticism

is surely misplaced. Even if it could be shown that McKeown had intended to offer

his analysis as part of an argument for the promotion of free-market economics, the

35

fact remains that the improvements which did occur in Britain’s health were not

that great, as the recruiting officers who surveyed the physical state of army

volunteers discovered to their cost in 1914.108 It would therefore be wrong to

assume that the slow and tortuous path which Britain took towards better health in

the eighteenth and nineteenth centuries represents a model to which other countries

should now aspire at the start of the twenty-first century.109

Acknowledgements

This paper was originally prepared for a conference to mark the life and work of

Thomas McKeown at the Postgraduate Medical Centre, Queen Elizabeth Hospital,

University of Birmingham, on 21 September 2002, and has also been presented in

seminars at the Universities of Southampton and Chicago. I should like to thank

participants in these meetings for their helpful comments. I am also grateful to Javier

Birchenall, Claudia Edwards, Robert Fogel, Andrew Hinde, Leonard Schwarz, Simon

Szreter, Robert Woods and Tony Wrigley for their comments on written versions of

the paper.

36

Table 1. Average daily consumption of calories and proteins per person, 1787-1863.

1787-93 1796 1787-96S 1787-96N 1787-96 1837-41 1841 1863

Shammasa kcalories - - 1,734 2,352 1,874 - - -

Oddyb kcalories 1,990 2,170 - - 2,028 - 2,300 2,600

protein (g) 49.0 62.0 - - 51.8 - 62.0 66.0

CHLc kcalories - - - - 1,508 1,974 - 2,395

protein (g) - - - - 27.9 32.1 - 37.7

Notes.

a. Shammas’ figures were derived from Davies’ study of the diets of rural families between 1787 and 1793, andfrom Eden’s study of the diets of urban and rural families in 1796. She based her results on the analysis of 7northern families (1787-96N) and 15 southern families (1787-96S).

b. Oddy’s figures for the eighteenth century were derived from the same Davies and Eden studies asShammas’ figures. However, he based his findings on the analysis of 32 families from the Eden study(mainly northern labourers) and 119 families from the Davies study (mainly southern families). His figuresfor 1837-41 were based on William Neild’s study of the diets of Manchester and Dukinfield and ‘several othercontemporary budgets’. The figures for 1863 are based on the information contained in Edward Smith’ssurvey for the Medical Department of the Privy Council.

c. Clark, Huberman and Lindert based their figures for 1787-96 on 195 households in the Davies-Edensurveys. Their analysis of the 1841 data was based on the figures supplied by Neild for Manchester andDukinfield, Purdy (for agricultural labourers) and two Parliamentary commissions on mining and child labour.Their figures for 1863 were also derived from Edward Smith’s surveys.

Sources: C. Shammas, ‘The Eighteenth-Century English Diet and Economic Change’, Explorations in EconomicHistory, 21 (1984), 254-69, pp. 256-7; G. Clark, M. Huberman and P. Lindert, ‘A British Food Puzzle, 1770-1850’,Economic History Review, 48 (1995), 215-37, pp. 222-3; D. Oddy, ‘Food, Drink and Nutrition’, in F.M.L.Thompson, ed., The Cambridge Social History of Britain 1750-1950, 3 vols., vol. 2 (Cambridge, 1990), 251-78,pp. 269-74.

37

Tab

le2.

Ave

rage

life

expe

ctan

cyat

birt

hin

diffe

rent

part

sof

Eng

land

and

Wal

es,1

801-

1900

.

Woo

ds(1

985)

1801

1811

1821

1831

1841

1851

1861

1871

1881

1891

Lond

on-

3031

3233

3537

3941

44

Ove

r10

0,00

0-

3030

3132

3435

3739

41

10-1

00,0

00-

3233

3536

3840

4244

46

Rur

al-

4142

4343

4545

4647

49

Eng

land

and

Wal

es-

3839

4040

4141

4244

46

Woo

ds(2

000)

1801

-10

1811

-20

1821

-30

1831

-40

1841

-50

1851

-60

1861

-70

1871

-80

1881

-90

1891

-190

0

Lond

on35

.036

.036

.936

.936

.738

.037

.740

.442

.643

.7

Ove

r10

0,00

032

.032

.532

.732

.632

.032

.333

.036

.639

.039

.6

10-1

00,0

0034

.235

.336

.236

.336

.037

.238

.041

.444

.044

.8

Rur

al42

.243

.343

.543

.544

.045

.546

.547

.751

.053

.5

Eng

land

and

Wal

es40

.341

.141

.140

.740

.441

.141

.243

.045

.346

.1

Woo

ds(1

985)

1801

1811

1821

1831

1841

1851

1861

1871

1881

1891

Larg

epr

ovin

cial

tow

ns-

3030

3132

3435

3739

41

Oth

erce

ntre

s-

3939

4040

4243

4445

47

Eng

land

and

Wal

es-

3839

4040

4141

4244

46

Szr

eter

&M

oone

y(1

998)

1800

s18

10s

1820

s18

30s

1840

s18

50s

1860

s18

70s

1880

s18

90s

Larg

epr

ovin

cial

tow

ns-

n/a

3529

3034

3438

4042

Oth

erce

ntre

s-

n/a

4142

4242

4244

4747

Eng

land

and

Wal

es40

4141

4141

4141

4345

46

Woo

ds(2

000)

1801

-10

1811

-20

1821

-30

1831

-40

1841

-50

1851

-60

1861

-70

1871

-80

1881

-90

1891

-190

0

Larg

epr

ovin

cial

tow

ns32

.032

.532

.732

.632

.032

.333

.036

.639

.039

.6

Oth

erce

ntre

s40

.441

.441

.641

.441

.342

.442

.944

.647

.248

.6

Eng

land

and

Wal

es40

.341

.141

.140

.740

.441

.141

.243

.045

.346

.1

Sou

rces

:R.W

oods

,‘T

heE

ffect

sof

Pop

ulat

ion

Red

istr

ibut

ion

onth

eLe

velo

fMor

talit

yin

Nin

teen

th-C

entu

ryE

ngla

ndan

dW

ales

’,Jo

urna

lofE

cono

mic

His

tory

,45,

645-

51,p

p.64

8,65

0;S

.Szr

eter

and

G.M

oone

y,U

rban

izat

ion,

Mor

talit

yan

dth

eS

tand

ard

ofLi

ving

Deb

ate:

New

Est

imat

esof

the

Exp

ecta

tion

ofLi

feat

Birt

hin

Nin

teen

th-C

entu

ryB

ritis

hC

ities

’,E

cono

mic

His

tory

Rev

iew

,51,

84-1

12,p

.104

;R.W

oods

,The

Dem

ogra

phy

ofV

icto

rian

Eng

land

and

Wal

es(C

ambr

idge

,200

0),p

p.36

2,36

9.

38

Tab

le3.

The

cont

ribut

ion

mad

eby

diffe

rent

dise

ases

toth

ede

clin

eof

mor

talit

yin

Eng

land

and

Wal

esdu

ring

the

seco

ndha

lfof

the

nine

teen

thce

ntur

y.

McK

eow

n&

Rec

ord

(196

2)M

cKeo

wn

(197

6)W

oods

(200

0)

1851

/60

1891

/190

0C

ontr

ibut

ion

toch

ange

1848

/54

1901

Con

trib

utio

nto

chan

ge18

61-7

018

91-1

900

Con

trib

utio

nto

chan

geD

eath

sp

erm

illio

nD

eath

sp

erm

illio

n%

Dea

ths

per

mill

ion

Dea

ths

per

mill

ion

%S

tand

ardi

zed

deat

hsD

eath

s%

Pht

hisi

s/R

espi

rato

rytu

berc

ulos

is2,

772

1,41

843

.89

2,90

11,

268

33.3

477

7,35

042

6,22

435

.19

Bro

nchi

tis,p

neum

onia

and

influ

enza

--

2,23

92,

747

-10.

37-

--

Dis

ease

sof

the

lung

/res

pira

tory

syst

em-

--

--

-97

1,69

61,

044,

719

-7.3

2

Who

opin

gco

ugh

433

363

2.27

423

312

2.27

140,

748

115,

670

2.51

Mea

sles

357

398

-1.3

334

227

81.

3111

9,47

112

6,84

1-0

.74

Sca

rletf

ever

779

152

20.3

2-

--

272,

437

48,2

9022

.46

Dip

ther

ia99

254

-5.0

2-

--

52,3

1980

,671

-2.8

4

Sca

rletf

ever

&di

pht

heria

--

-1,

016

407

12.4

3-

--

Sm

allp

ox20

213

6.13

263

105.

1746

,713

4,05

84.

27

Infe

ctio

nsof

the

ear,

phar

ynx

and

lary

nx-

-75

100

-0.5

1-

--

Cho

lera

,dia

rrho

ea,d

ysen

tery

990

715

8.91

1,81

91,

232

11.9

829

4,64

322

6,14

36.

86

Non

-res

pira

tory

tube

rcul

osis

--

753

544

4.27

--

-

Tab

esm

esen

teric

aan

dot

her

tub

ercu

lous

and

scro

fulo

usdi

seas

es70

660

33.

34-

--

121,

864

189,

782

-6.8

1

Typ

hus

891

184

22.9

2-

--

268,

467

55,9

9621

.29

Typ

hoid

and

typ

hus

--

990

155

17.0

5-

--

Can

cer

--

307

844

-10.

9611

9,41

323

2,17

8-1

1.30

Vio

lenc

e-

-76

164

02.

4728

3,48

420

2,36

38.

13

Oth

ers

(McK

eow

n&

Rec

ord)

13,8

9014

,024

-4.3

4

Oth

ers

(McK

eow

n)-

-9,

967

8,42

131

.56

--

-

Oth

ers

(Woo

ds)

--

--

-4,

076,

386

2,89

5,52

328

.30

Tot

al21

,209

18,1

2410

0.00

21,8

5616

,958

100.

006,

573,

295

5,57

5,43

510

0.00

Not

e.A

ccor

ding

toW

oods

(The

Dem

ogra

phy

ofV

icto

rian

Eng

land

and

Wal

es,p

.351

),th

eto

taln

umbe

rof

deat

hsin

1891

-190

0w

as5,

575,

375.

How

ever

,the

figur

esin

his

colu

mn

sum

to5,

575,

435,

and

this

figur

eha

sbe

enus

edto

calc

ulat

eth

efig

ures

show

nin

this

tabl

e.

Sou

rces

:T

.McK

eow

nan

dR

.G.R

ecor

d,‘R

easo

nsfo

rth

eD

eclin

eof

Mor

talit

yin

Eng

land

and

Wal

esdu

ring

the

Nin

etee

nth

Cen

tury

’,P

opul

atio

nS

tudi

es,1

6(1

962)

,94-

122,

p.10

4;T

.McK

eow

n,T

heM

oder

nR

ise

ofP

opul

atio

n(L

ondo

n,19

76),

pp.5

4-62

;R.W

oods

,The

Dem

ogra

phy

ofV

icto

rian

Eng

land

and

Wal

es(C

ambr

idge

,200

0),p

p.35

0-1.

39

Figure 1. Real wages, 1770/2-1848/52

60.00

70.00

80.00

90.00

100.00

110.00

120.00

130.00

140.00

1770-1772

1773-1777

1778-1782

1783-1787

1788-1792

1793-1797

1798-1802

1803-1807

1808-1812

1813-1817

1818-1822

1823-1827

1828-1832

1833-1837

1838-1842

1843-1847

1848-1852

period

real

wag

es(1

778/

82=

100)

Phelps Brown & Hopkins Feinstein

Sources: E.A. Wrigley and R.S. Schofield, The Population History of England 1541-1871: a Reconstruction (Cambridge, 1981), pp. 642-4; C.H. Feinstein, ‘PessimismPerpetuated: Real Wages and the Standard of Living in Britain During and After theIndustrial Revolution’, Journal of Economic History, 58 (1998), 625-58, p. 648 (realearnings adjusted for unemployment in Great Britain).

40

Figure2. Averagelifeexpectancyat birth, 1541/46-1866/71

15

25

35

45

55

65

1541-1546

1566-1571

1591-1596

1616-1621

1641-1646

1666-1671

1691-1696

1716-1721

1741-1746

1766-1771

1791-1796

1816-1821

1841-1846

1866-1871

period

Life

expe

ctan

cyat

birt

h(y

ears

)

1550-1574

1575-1599

1600-1625

1625-1649

1650-1674

1675-1699

1700-1724

1725-1749

1750-1774

1775-1799

1800-1824

1825-1849

1850-1874

England(WDOS) Ducal families(Hollingsworth)

Sources: T.H. Hollingsworth, ‘The Demography of the British Peerage’, PopulationStudies, 18 (Supplement, 1966), pp. 56-7; E.A. Wrigley, R. Davies, J. Oeppen andR. Schofield, English Population History from Family Reconstitution 1580-1837(Cambridge, 1997), p. 614.

41

∗ Division of Sociology and Social Policy, School of Social Science, University

of Southampton, Highfield, Southampton SO17 1BJ. Email:

bjh2@soton.ac.uk.

1 The figures for 1700 and 1701 are for England only; figures for 1991-5 and

1993-5 are for England and Wales. For sources, see E. A. Wrigley and R. S.

Schofield, The Population History of England 1541-1871(Cambridge, 1981),

pp. 230, 533; R. Fitzpatrick and T. Chandola, ‘Health’, in A. H. Halsey and J.

Webb, eds., Twentieth-century British social trends (Basingstoke, 2000), 94-

127, pp. 95, 98.

2 R. Woods, The Demography of Victorian England and Wales (Cambridge,

2000), p. 359.

3 V. Hionidou, ‘Why do People Die in Famines? Evidence from three Island

Populations’, Population Studies, 56 (2002), 65-80.

4 B. Harris, ‘Seebohm Rowntree and the Measurement of Poverty, 1899-1951’,

in J. Bradshaw and R. Sainsbury, eds., Getting the Measure of Poverty: the

Early Legacy of Seebohm Rowntree (Aldershot, 2000), 60-84, pp. 73-5.

5 A. M. Tomkins, ‘Protein-energy Malnutrition and Risk of Infection’,

Proceedings of the Nutrition Society, 45 (1986), 289-304, p. 289.

6 P. Dasgupta and D. Ray, ‘Adapting to Undernourishment: the Biological

Evidence and its Implications’, in J. Drèze and A. Sen, eds., The Political

Economy of Hunger. Volume 1. Entitlement and Well-being (Oxford, 1990),

42

191-246, pp. 213-5; see also P. Pellett, ‘Energy and Protein Metabolism’, in

K. Kiple and K. C. Ornelas, eds., The Cambridge World History of Food, 2

vols., vol. 1 (Cambridge, 2000), 888-913, pp. 894-6; J. Kim, ‘Nutrition and the

Decline of Mortality’, in Kiple and Ornelas, eds., Cambridge World History of

Food, vol. 2, 1381-9, pp. 1384-6.

7 P. Eveleth, ‘Population Differences in Growth: Environmental and Genetic

Factors’, in F. Falkner and J. Tanner, eds., Human Growth, 3 vols., vol. 3

(London, 1979), 373-94, p. 388.

8 N. Scrimshaw and J. P. SanGiovanni, ‘Synergism of Nutrition, Infection and

Immunity: an Overview’, Supplement to the American Journal of Clinical

Nutrition, 66 (1997), 464S-477S, p. 464S.

9 N. Scrimshaw, C. E. Taylor and J. E. Gordon, Interactions of nutrition and

infection (Geneva, 1968), pp. 144-5; Scrimshaw and SanGiovanni,

‘Synergisms’, pp. 466S-467S; N. Scrimshaw, ‘Infection and Nutrition:

Synergistic Interactions’, in Kiple and Ornelas, eds., Cambridge World History

of Food, vol. 2, 1397-1411, p. 1398; Tomkins, ‘Protein-Energy Malnutrition’.

10 P. G. Lunn, ‘Nutrition, Immunity and Infection’, in R. Schofield , D. Reher and

A. Bideau, eds., The Decline of Mortality in Europe (Oxford, 1991), 131-45, p.

136; see also Scrimshaw and SanGiovanni, ‘Synergisms’, p. 472S.

11 See e.g. P. Aaby, ‘Malnourished or Overinfected? An Analysis of the

Determinants of Acute Measles Mortality’, Danish Medical Bulletin, 36 (1989),

93-113; idem., ‘Lessons for the Past: Third World Evidence and the

43

Reinterpretation of Developed World Mortality Declines’, Health Transition

Review, 2 (Supplementary Issue, 1992), 155-83.

12 Tomkins, ‘Protein-Energy Malnutrition’, pp. 295-7. Although Aaby’s work has

been cited by a number of historians, the views of other writers have not

always received similar attention. See e.g. J. Landers, ‘Introduction’, Health

Transition Review, 2 (Supplementary Issue, 1992), 1-28, pp. 21-2; Woods,

Demography of Victorian England and Wales, pp. 322-3; W. Muraskin,

‘Nutrition and Mortality Decline: Another View’, in Kiple and Ornelas, eds.,

Cambridge World History of Food, vol. 2, 1389-97, pp. 1391-2. A. Hardy, The

Epidemic Streets: Infectious Disease and the Rise of Preventive Medicine,

1856-1900 (Oxford, 1993), pp. 34, 283, also examines Aaby’s work, but

compares it with the more nutritionally-oriented approach associated with

David Morley.

13 Although the Bellagio conferees argued that nutritional factors played a very

small role in the aetiology of smallpox, Duncan, Scott and Duncan have

claimed that nutritional factors were responsible for the timing of smallpox

epidemics in Penrith between 1630 and 1800. See Bellagio Conference, ‘The

Relationship of Nutrition, Disease and Social Conditions: A Graphical

Presentation’, in R. Rotberg and T. K. Rabb, eds., Hunger and History: the

Impact of Changing Food Production and Consumption Patterns on Society

(Cambridge, 1985), 305-8; S. R. Duncan, S. Scott and C. J. Duncan, ‘The

dynamics of smallpox epidemics in Britain, 1550-1800’, Demography, 30

44

(1993), 405-23, p. 420; idem., ‘Smallpox epidemics in cities in Britain’, Journal

of Interdisciplinary History, 25 (1994), 255-71.

14 Woods, The demography of Victorian England and Wales, pp. 350-1.

15 Scrimshaw and SanGiovanni, ‘Synergism’, p. 464S.

16 Lunn, ‘Nutrition, Immunity and Infection’, p. 145.

17 P. Razzell, ‘The Growth of Population in Eighteenth-Century England: a

Critical Reappraisal’, in P. Razzell, Essays in English Population History

(Hampstead, 1994), 173-206, pp. 185-95; E. A. Wrigley, R. Davies, J. Oeppen

and R. S. Schofield, English Population History from Family Reconstitution

1580-1837 (Cambridge, 1997), pp. 282-4; but see also P. Razzell, ‘The

Conundrum of Eighteenth-Century English Population Growth’, Social History

of Medicine, 11 (1998), 469-500, pp. 485-500.

18 T. McKeown and R. G. Brown, ‘Medical Evidence Related to English

Population Changes in the Eighteenth Century’, Population Studies, 9 (1955),

119-41, p. 126; T. McKeown, R. G. Brown and R. G. Record, ‘An

Interpretation of the Modern Rise of Population in Europe’, Population

Studies, 26 (1972), 345-82, pp. 346, 351; T. McKeown, The Modern Rise of

Population (London, 1976), pp. 124-7; idem., ‘Food, Infection and Population’,

in Rotberg and Rabb, eds., Hunger and History, 29-49, pp. 31-2, 45-8.

19 P. Razzell, ‘Population Change in Eighteenth-Century England: a

Reinterpretation’, Economic History Review, 18 (1965), 312-32; idem., The

Conquest of Smallpox: the Impact of Inoculation on Smallpox Mortality in

45

Eighteenth-Century Britain (Sussex, 1977), pp. 140-58; A. Mercer, ‘Smallpox

and epidemiological-demographic change in Europe: the Role of Vaccination’,

Population Studies, 39 (1985), 287-307; idem., Disease, Mortality and

Population in Transition: Epidemiological-Demographic Change in England

since the Eighteenth Century as Part of a Global Phenomenon, (Leicester,

1990), pp. 46-73.

20 M. Dobson, Contours of Death and Disease in Early-Modern England

(Cambridge, 1997).

21 E. L. Jones and M. E. Falkus, ‘Urban Improvement and the English Economy

in the Seventeenth and Eighteenth Centuries’, in P. Borsay, ed., The

Eighteenth-Century Town: A Reader in English Urban History 1688-1820

(London & New York, 1990), 116-58; R. Porter, ‘Cleaning Up the Great Wen:

Public Health in Eighteenth-Century London’, in W. F. Bynum and R. Porter,

eds., Living and Dying in London (Medical History, Supplement no. 11,

London, 1991), 61-75; J. Landers, Death and the Metropolis: Studies in the

Demographic History of London 1670-1830 (Cambridge, 1993).

22 R. Floud, K. Wachter and A. Gregory, Height, Health and History: Nutritional

Status in the United Kingdom 1750-1980 (Cambridge, 1990), pp. 134-86; R.

Floud, ‘Height, Weight and Body Mass of the British Population since 1820’,

NBER Working Paper Series on Historical Factors in Long-Run Growth,

Working Paper no. 108, p. 35; R. Fogel, ‘Economic Growth, Population

Theory and Physiology: the Bearing of Long-Term Processes on the Making

of Economic Policy’, American Economic Review, 84 (1994), 369-95, p. 372.

46

23 Dasgupta and Ray, ‘Adapting to Undernourishment’, pp. 215-6. I am grateful

to Robert Fogel for directing my attention to this article.

24 M. Livi-Bacci, Population and Nutrition: An Essay on European Demographic

History, Cambridge, 1991), p. 27.

25 Fogel, ‘Economic Growth, Population Theory and Physiology’, pp. 373-4.

26 M. Livi-Bacci, The Population of Europe: A History (Oxford, 2000), p. 143.

27 Wrigley and Schofield, Population History, pp. 351-5; Livi-Bacci, Population

and Nutrition, pp. 99-107.

28 See e.g. P. Lindert and J. Williamson, ‘English Workers’ Living Standards

During the Industrial Revolution: A New Look’, Economic History Review, 36

(1983), 1-25; idem., ‘English Workers’ Real Wages: Reply to Crafts’, Journal

of Economic History,45 (1985), 145-53; N. Crafts, ‘English Workers’ Living

Standards During the Industrial Revolution: Some Remaining Problems’,

Journal of Economic History, 45 (1985), 139-44; L. D. Schwarz, ‘The Standard

of Living in the Long Run: London 1700-1860’, Economic History Review, 43

(1985), 24-41; idem., ‘Trends in Real Wages, 1750-90: A Reply to Hunt and

Botham’, Economic History Review, 43 (1990), 90-8; R. Floud and B. Harris,

‘Health, Height and Welfare: Britain 1700-1980’, in R. Steckel and R. Floud,

eds., Health and Welfare since Industrialization (Chicago, 1997), 91-126, p.

95.

29 C. Shammas, ‘The Eighteenth-Century English Diet and Economic Change’,

Explorations in Economic History, 21 (1984), 254-69, pp. 256-8; idem., The

47

Preindustrial Consumer in England and America (Oxford, 1990), p. 134; G.

Clark, M. Huberman and P. Lindert, ‘A British Food Puzzle, 1770-1850’,

Economic History Review, 48 (1995), 215-37, pp. 222-3

30 Shammas, ‘The Eighteenth-Century English Diet’, p. 257. According to Clark,

Huberman and Lindert, Shammas based her findings on the analysis of 28

household budgets in the Davies-Eden surveys. However, Shammas herself

only refers to 22 budgets. See Clark, Huberman and Lindert, ‘A British Food

Puzzle’, p. 222 (note 16); Shammas, ‘The Eighteenth-Century English Diet’, p.

256 (note 3); idem., The Preindustrial Consumer, p. 134.

31 Shammas, ‘The Eighteenth-Century English Diet’, pp. 255-6; idem., The

Preindustrial Consumer, p. 134; Clark, Huberman and Lindert, ‘A British Food

Puzzle’, p. 222.

32 R. Fogel, ‘Second Thoughts on the European Escape from Hunger: Famines,

Price Elasticities, Entitlements, Chronic Malnutrition and Mortality Rates’,

NBER Working Paper Series on Historical Factors in Long-Run Growth,

Working Paper no. 1 (1989), p. 40; reprinted in S. R. Osmani, ed., Nutrition

and Poverty (Oxford, 1992), 243-86, p. 269; R. Fogel, ‘New Sources and New

Techniques for the Study of Secular Trends in Nutritional Status, Health,

Mortality and the Process of Aging’, Historical Methods, 26 (1993), 5-33, p.

12; idem., ‘Economic Growth, Population Theory and Physiology’, pp. 373-4.

In their article, Clark, Huberman and Lindert defended their findings by

claiming that their estimate of the calorific value of the diets consumed by the

families in the Davies-Eden surveys was comparable to the figure which

48

Robert Fogel estimated for the bottom decile of the whole population.

However, since Fogel’s estimate was calculated on the basis that the Davies-

Eden households were actually quite close to the median of the overall

distribution, it is difficult to see how the two sets of figures can be reconciled.

See Clark, Huberman and Lindert, ‘A British Food Puzzle’, p. 222 (note 16);

Fogel, ‘Second thoughts’ (1989), pp. 37-43; idem., ‘Second Thoughts’ (1992),

pp. 268-71.

33 Shammas, ‘The Eighteenth-Century English Diet’, p. 262; idem., The

Preindusrial Consumer, pp. 145-7.

34 R. Fogel, The Escape from Hunger and Premature Death 1700-2100: Europe,

America and the Third World, (Cambridge, in press), pp. 10-11. I am grateful

to Robert Fogel for permission to cite material from this book, prior to its

publication. In the current paper, page references refer to the manuscript of

Fogel’s text.

35 P. Razzell, ‘“An Interpretation of the Modern Rise of Population in Europe” – A

Critique’, Population Studies, 28 (1974), 5-17, pp. 6-7; idem., ‘The Growth of

Population in Eighteenth-Century England’, pp. 152-3; Livi-Bacci, Population

and Nutrition, pp. 63-7; idem., The population of Europe, pp. 56-7; Mercer,

Disease, mortality and population in transition, p. 37.

36 McKeown, The Modern Rise of Population, pp. 139-42.

37 Scrimshaw, Gordon and Taylor, Interactions of Nutrition and Infection, pp.

144-5.

49

38 S. Kunitz and S. Engerman, ‘The Ranks of Death: Secular Trends in Income

and Mortality’, Health Transition Review, 2 (Supplementary Issue, 1992), 29-

46, p. 33.

39 P. Mathias, ‘Preface’, in A. J. Taylor, ed., The Standard of Living in Britain in

the Industrial Revolution (London, 1975), vii-ix, p. vii.

40 Quoted in M. D. George, England in Transition: Life and Work in the

Eighteenth Century (Harmondsworth, 1953), p. 76.

41 J. Clapham, An Economic History of Modern Britain (Cambridge, 1926), p. 7.

42 J. L. Hammond, ‘The Industrial Revolution and Discontent’, Economic History

Review, 2 (1930), 215-28, p. 225.

43 E. Hobsbawm, ‘The British Standard of Living, 1790-1850’, in Taylor, ed., The

Standard of Living, 58-92, pp. 75-81.

44 R. M. Hartwell, ‘The Rising Standard of Living in England, 1800-50’, in Taylor,

ed., The Standard of Living, 93-123, pp. 108-17.

45 J. Burnett, Plenty and Want: a Social History of Diet in England from 1815 to

the Present Day (2nd edition, London, 1983), pp. 24-6.

46 J. Mokyr, ‘Is There Still Life in the Pessimist Case? Consumption during the

Industrial Revolution, 1790-1850’, Journal of Economic History, 48 (1988), 69-

92, p. 83.

47 Wrigley and Schofield, Population History, pp. 638-41; Lindert and

Williamson, ‘English Workers’ Living Standards’; idem., ‘English Workers’

50

Real Wages’; Crafts, ‘English Workers’ Living Standards’; Schwarz, ‘The

Standard of Living’; idem., ‘Trends in Real Wages’.

48 C. H. Feinstein, ‘Pessimism Perpetuated: Real Wages and the Standard of

Living in Britain During and After the Industrial Revolution’, Journal of

Economic History, 58 (1998), 625-58, p. 648.

49 Clark, Huberman and Lindert, ‘A British Food Puzzle’, pp. 234-5.

50 Feinstein, ‘Pessimism Perpetuated’, p. 652.

51 Cf. J. Komlos, ‘Shrinking in a Growing Economy? The Mystery of Physical

Stature During the Industrial Revolution’, Journal of Economic History, 58

(1998), 779-802, p. 786.

52 S. Horrell, ‘Home Demand and British Industrialization’, Journal of Economic

History, 56 (1996), 561-604, p. 592; Feinstein, ‘Pessimism Perpetuated’, p.

635.

53 M. Haines, ‘Shrinking People – Growing Incomes. Can Development be

Hazardous to Your Health? Historical Evidence for the United States,

England and the Netherlands in the Nineteenth Century’. Paper presented to

the First International Conference on Economics and Human Biology,

University of Tübingen, 11-14 July, 2002, p. 11; Clark, Huberman and Lindert,

‘A British Food Puzzle’, pp. 233-4.

54 Feinstein, ‘Pessimism Perpetuated’, p. 640.

55 Ibid., pp. 648, 653.

51

56 According to Feinstein, the proportion of total working-class expenditure which

was devoted to food fell from 65 per cent to 61 per cent between 1828/32 and

1858/62. However, the real value of working-class earnings (after allowing for

changes in the rate of unemployment) rose by 23.4 per cent over the same

period (Feinstein, ‘Pessimism Perpetuated’, pp. 640, 648, 653).

57 D. Oddy, ‘Food, Drink and Nutrition’, in F. M. L. Thompson, ed., The

Cambridge Social History of Britain 1750-1950, 3 vols., vol. 2 (Cambridge,

1990), 251-78; Clark, Huberman and Lindert, ‘A British Food Puzzle’.

58 The view that dietary standards improved during this period is also supported

by Roger Scola’s analysis of the food supply of Manchester between 1770

and 1870. See R. Scola, Feeding the Victorian City: the Food Supply of

Manchester 1770-1870 (Manchester, 1992), pp. 261-2.

59 A. Wohl, Endangered Lives: Public Health in Victorian Britain (London, 1984),

pp. 4-6.

60 Ibid., p. 4.

61 F. M. L. Thompson, ‘Town and City’, in Thompson, ed., Cambridge Social

History of Britain, vol. 1, 1-86, p. 8.

62 Wohl, Endangered Lives, p. 4.

63 P. Huck, ‘Infant Mortality and Living Standards of English Workers During the

Industrial Revolution’, Journal of Economic History, 55 (1995), 528-50, pp.

534-5. Residents of Walsall, West Bromwich, Sedgeley and Handsworth

52

might have been somewhat surprised by Huck’s assertion that these midlands

parishes were situated ‘in the industrial north of England’.

64 Floud, Wachter and Gregory, Height, Health and History, pp. 205-6, 326;

Floud and Harris, ‘Health, Height and Welfare’, p. 105.

65 R. Woods, ‘The Effects of Population Redistribution on the Level of Mortality

in Nineteenth-Century England and Wales’, Journal of Economic History, 45

(1985), 645-51.

66 S. Szreter, ‘Economic Growth, Disruption, Deprivation, Disease and Death:

On the Importance of the Politics of Public Health for Development’,

Population and Development Review, 23 (1997), 693-728, p. 700; S. Szreter

and G. Mooney, ‘Urbanization, Mortality and the Standard of Living Debate;

New Estimates of the Expectation of Life at Birth in Nineteenth-Century British

Cities’, Economic History Review, 51 (1998), 84-112, p. 104.

67 Woods, Demography of Victorian England and Wales, pp. 357-8, 368.

68 Ibid., pp. 360-80.

69 Szreter and Mooney, ‘Urbanization, Mortality and the Standard of Living

Debate’, pp. 89, 107.

70 Szreter, ‘Economic Growth, Disruption, Deprivation, Disease and Death, p.

26; F. Bell and R. Millward, ‘Public health expenditures and mortality in

England and Wales 1870-1914’, Continuity and Change, 13 (1998), pp. 242-3.

71 Floud, Wachter and Gregory, Height, Health and History, p. 314.

53

72 W. O. Kermack, A. G. McKendrick and P. L. McKinlay, ‘Death Rates in Great

Britain and Sweden: Some General Regularities and Their Significance’,

Lancet, 1 (1934), 698-703, p. 698.

73 G. Davey Smith, D. Gunnell and Y. Ben-Shlomo, ‘Life-Course Approaches to

Socio-Economic Differentials in Cause-Specific Adult Mortality’, in D. Leon

and G. Walt, eds., Poverty, Inequality and Health: An International

Perspective (Oxford, 2001), 88-124; D. Leon, ‘Common Threads: Underlying

Components of Inequalities in Mortality Between and Within Countries’, in

Leon and Walt, eds., Poverty, Inequality and Health, 58-87.

74 B. Harris, ‘“The Child is Father to the Man”. The Relationship between Child

Health and Adult Mortality in the Nineteenth and Twentieth Centuries’,

International Journal of Epidemiology, 30 (2001), 688-96.

75 T. McKeown and R. G. Record, ‘Reasons for the Decline of Mortality in

England and Wales during the Nineteenth Century’, Population Studies, 16,

(1962), 94-122, p. 103.

76 McKeown and Record, ‘Reasons for the Decline of Mortality’, pp. 109-22;

McKeown, The Modern Rise of Population, pp. 117-23.

77 McKeown and Record, ‘Reasons for the Decline of Mortality’, p. 117; see also

McKeown, The Modern Rise of Population, p. 82; S. Szreter, ‘The Importance

of Social Intervention in Britain’s Mortality Decline, c. 1850-1914: A

Reintrepretation of the Role of Public Health’, Social History of Medicine, 1

(1988), 1-37, p. 12; Hardy, Epidemic Streets, p. 66; Woods, Demography of

Victorian England and Wales, p. 323.

54

78 Szreter, ‘The Importance of Social Intervention’, pp. 11-17.

79 S. Guha, ‘The Importance of Social Intervention in England’s Mortality decline:

The Evidence Reviewed’, Social History of Medicine, 7 (1994), 89-113, pp.

96-100; S. Szreter, ‘Mortality in England in the Eighteenth and the Nineteenth

Centuries: A Reply to Sumit Guha’, Social History of Medicine, 7 (1994), 269-

82, pp. 274-8.

80 J. Riley, Sick, not Dead: The Health of British Workingmen During the

Mortality Decline (Baltimore, 1997), p. 197; idem., ‘Reply to Bernard Harris,

“Morbidity and Mortality During the Health Transition: A Comment on James

C. Riley”’, Social History of Medicine, 12 (1999), 133-7, p. 137.

81 See e.g. H. Emery, ‘Review of James C. Riley, Sick, not dead: the health of

British workingmen during the mortality decline’. EH.Net, H-Net Reviews, July

1998. URL: http://www.h-

net.msu.edu/reviews/showrev.cgi?path=17775899999635.

82 C. Edwards, M. Gorsky, B. Harris and P. R. A. Hinde, ‘Sickness, Insurance

and Health: Assessing Trends in Morbidity through Friendly Society Records’,

Annales de Démographie Historique, 1 (2003), 131-67.

83 R. Woods and N. Shelton, An Atlas of Victorian Mortality (Liverpool, 1997),

pp. 143-4; Woods, The Demography of Victorian England and Wales, pp. 340,

359.

55

84 J. Landers, ‘Review Article’, Continuity and Change, 15 (2000), 466-8, p. 468;

S. Szreter, ‘Review of Robert Woods, The Demography of Victorian England

and Wales’, Social History of Medicine, 14 (2001), 562-3, p. 563.

85 D. P. S. Spence, J. Hotchkiss, C. S. D. Williams and P. D. O. Davies,

‘Tuberculosis and Poverty’, British Medical Journal, 307 (1993), 759-61, p.

760; N. Bhatti, M. R. Law, J. K. Morris, R. Halliday and J. Moore-Gillon ,

‘Increasing Incidence of Tuberculosis in England and Wales: A Study of the

Likely Causes’, British Medical Journal, 310 (1995), 967-9, p. 969; see also K.

Tocque, M. A. Bellis, N. J. Beeching, Q. Syed, T. Remmington and P. D. O.

Davies, ‘A Case-Control Study of Lifestyle Risk Factors Associated with

Tuberculosis in Liverpool, North-West England’, European Respiratory

Journal, 18 (2001), 959-64.

86 Spence, Hotchkiss, Williams and Davies, ‘Tuberculosis and Poverty’, pp. 760-

1.

87 M. E. Pooley and C. G. Pooley, ‘Health, Society and Environment in

Nineteenth-Century Manchester’, in R. Woods and J. Woodward, eds., Urban

Disease and Mortality in Nineteenth-Century England (London & New York,

1984), 148-75, pp. 171-2.

88 J. Vögele, Urban Mortality Change in England and Germany, 1870-1913

(Liverpool, 1998), p. 145.

89 McKeown, The Modern Rise of Population, p. 118; Vögele, Urban Mortality

Change, p. 145.

56

90 M. Daunton, ‘Health and Housing in Victorian London’, Medical History,

Supplement no. 11 (1991), 126-44, p. 143.

91 G. Cronjé, ‘Tuberculosis and Mortality Decline in England and Wales, 1851-

1910’, in Woods and Woodward, eds., Urban Disease and Mortality, 79-101,

p. 99.

92 See e.g. Wohl, Endangered Lives, pp. 304-6; J. Burnett, ‘Housing and the

Decline of Mortality’, in Schofield, Reher and Bideau, eds., The Decline of

Mortality in Europe, 158-76, p. 174.

93 M. Daunton, House and Home in the Victorian City: Working-Class Housing

1850-1914 (London, 1983), p. 7; idem., ‘Health and Housing in Victorian

London’, p. 130.

94 A. Hardy, Health and Medicine in Britain since 1860 (Basingstoke, 2001), p.

38.

95 J. Burnett, Plenty and Want: A Social History of Diet in England from 1815 to

the Present Day (3rd edition, London, 1989), pp. 108, 176.

96 Oddy, ‘Food, Drink and Nutrition’, pp. 267-75.

97 P. Dewey, ‘Nutrition and Living Standards in Wartime Britain’, in R. Wall and

J. Winter, eds., The Upheaval of War: Family, Work and Welfare in Europe

1914-18 (Cambridge, 1988), 197-220, p. 215.

98 McKeown, The Modern Rise of Population, p. 121.

99 Szreter, ‘The Importance of Social Intervention’, pp. 22-6; Hardy, Epidemic

Streets, pp. 293-4.

57

100 Woods, The Demography of Victorian England and Wales, pp. 350-6.

101 Ibid., p. 358.

102 Ibid., pp. 350, 358.

103 J. Hassan, ‘The Growth and Impact of the British Water Industry in the

Nineteenth Century’, Economic History Review, 38 (1985), 531-47, pp. 543-4;

Bell and Millward, ‘Public Health Expenditures and Mortality’, pp. 237-40.

104 Bell and Millward, ‘Public Health Expenditures and Mortality’, pp. 242-3; see

also R. Millward and F. Bell, ‘Economic Factors in the Decline of Mortality in

Nineteenth-Century Britain’, European Review of Economic History, 2 (1998),

263-88.

105 See also Kim, ‘Nutrition and the Decline of Mortality’, pp. 1384-6.

106 S. Preston and E. van de Walle, ‘Urban French Mortality in the Nineteenth

Century’, Population Studies, 32 (1978), 275-97, pp. 281-2; Tomkins, ‘Protein-

Energy Malnutrition’, pp. 297-8; Lunn, ‘Nutrition, Immunity and Infection’, p.

135.

107 Szreter, ‘The Importance of Social Intervention’, p. 37.

108 See also Woods and Shelton, An Atlas of Victorian Mortality, p. 145.

109 This, it seems to me, is one of the main lessons to be drawn from James

Riley’s recent review of Rising life expectancy: a global history (Cambridge,

2001).