Height Patterns during Industrialization

Before attempting to study human health and nutrition in pre-industrial times, it is necessary to determine when the industrial period began. Although some scholars debate whether ‘revolution’ should be appended to ‘industrial’ in describing the process, the changes were revolutionary from the perspective of several centuries of history.

In order to fold heights into the debate over socio-economic performance during industrialization, it is useful to divide the process of change within each country into three chronological parts (early, middle, and late) based on an intuitive rather than a formulistic mix with an emphasis on dates of achieving modern economic growth. Also considered were the extent of mechanization, urbanization, per cent of the labour force involved in agriculture, and the amount of technological change, which have in varying degrees been part of the process of industrialization or modernization in all countries that underwent the experience. Assigning dates and phases is inevitably a matter of judgement, but the contributors to Health and Welfare during Industrialisation were comfortable with three rough categories. In the early industrial period, the transition to a modern industrial economy began.

describing their experiences. Australia, for example, relied extensively on agriculture and mining, while the Netherlands (lacking coal and waterpower) developed banking, shipping, and services.

3.1 Height and economic growth

Table 9.1 depicts average heights and other information during the early, middle, and late stages of industrialization in eight countries whose adult male heights were studied in Steckel and Floud (1997).

In examining the table, one is struck by the diversity of experience across countries in heights, and in life expectancy and the per cent urban. Average heights varied by over 15 cm within each phase, while the per cent urban differed by as much as 30 percentage points in the early phase (the Netherlands versus the United States) and nearly 50 points in the late phase (Japan versus Sweden). While all life expectancies were somewhat compressed in the pre-industrial phase, differing by less than 13 years (Australia versus the United Kingdom), by the late phase Australia had leaped ahead of the pack (65.4 years) while Japan lagged nearly 20 years behind (47 years).

3.2 Height and life expectancy

In arguing for average height as a general welfare measure, it is useful to consider its relationship to a more widely employed measure of health—life expectancy at birth. Understanding of average height as a population's history of net nutrition implies that a positive correlation should exist. Biological deprivation that retards human growth also increases the risk of death, a relationship that is magnified by the synergy between diet and disease (Scrimshaw et al.

Some critics of anthropometric history base their doubts on a selective effect of height on survival. Because studies on individual data have found that short people are less likely to survive (Friedman 1982; Waaler 1984), high mortality erodes the left tail of the height distribution. If this was the only effect of higher mortality, then taller heights would indicate poorer health. But it is not the only effect: survivors also grow less because they endured the biological stress causing higher mortality. The former effect is small because under conditions of increasing biological stress, many tall people also fail to survive. Empirically the second effect far outweighs the former. At the national level, heights and life expectancy are highly and positively correlated in the late twentieth century (Steckel 2001).

Data for estimating the relationship between average height and life expectancy in a historical setting are meagre, but available for all countries in the late industrial periods. Figure 9.1 shows the scatter diagram for the United States, the United Kingdom, the Netherlands, France, Sweden, Germany, Japan, and Australia, which

Table 9.1 Average heights of adult men, life expectancy, and percentage urban by stages of industrialization

Source.

Figure 9.1 Life expectancy and adult male height, late industrial period

Source. Estimated from data in Steckel and Floud (1997: Table 11.2).

incorporates data taken from Table 9.1. The estimated regression line is (/-values in parentheses):

The average height of adult men increased by 0.33 cm for every year by which life expectancy increased. The scatter diagram reveals that Japan is a substantial outlier, and if this country is omitted, the R² rises to 0.79 and the regression coefficient is statistically significant at 0.01. Japan's stature was much too low, given its life expectancy, probably because the time frames of the two measures are different. Conditions in the 1930s and early 1940s, when the young adults of mid-century were growing children, were likely worse than those forming the basis for life expectancy in 1950, a few years following the end of the Second World War. It is also possible that the height—life expectancy relationship is somewhat non-linear at low levels of health.

3.3 Height and urbanization

Historical demographers have long observed the adverse effect of urban areas on health prior to the early twentieth century (see, for example, United Nations 1973: 131—2; Preston and Haines 1991: chs 4—5). Numerous factors explain the connection, including congestion in factories and places of living, which promoted the spread of communicable diseases. Accumulation of waste, impure water supplies, and inflows of people bearing pathogens increased the urban population's exposure to pathogens. High poverty levels and high prices for food (relative to rural areas) also acted to increase mortality rates.

Table 9.1 shows that average male height and per cent urban (towns or cities of 2,500 population or more) varied widely across the countries during the middle phase of industrialization, a period of intense change.

It was not accidental that the Japanese were both the shortest and the most urban. In an era before widespread, effective investments in public health and personal hygiene, the congestion and turnover associated with urban living increased the chances of exposure to pathogens. Other features detrimental to health are often found in cities, such as a large number of poor people who lacked access to food, clothing, and shelter that would have reduced resistance to disease.

The scatter diagram in Figure 9.2 confirms the adverse effect of urbanization on health. The estimated regression equation is (/-values in parentheses):

For every percentage point increase in the degree of urbanization, average male height fell by about 0.15 cm. This magnitude is significant in a practical sense because the transition from a low (say, 20%) level of urbanization to a moderately high level (say 50%) would have decreased average height by 4.5 cm. The notable outlier to the inverse relationship was Australia, which had the tallest population and the second highest level of urbanization. If Australia is dropped from the

Figure 9.2 Per cent urban and average male height

Source. Estimated from data in Steckel and Floud (1997: Table 11.2). regression, the /-value rises to -2.60, R² increases to 0.57, and the regression coefficient increases (in absolute value) by 50%.

What factors explain the exceptional nature of health and urbanization in Australia? One was the relative geographic isolation of the country from major disease currents that affected cities in Europe and in North America.

3.3 Temporal patterns

Additional factors that influenced heights during industrialization, and possibly in the pre-industrial period, can be discerned from study of temporal patterns within countries. Although the various series might be arranged on a continuum in terms of depth and duration of cycles, the patterns are readily grouped into three categories: (a) significant and prolonged declines in health during a large phase of industrialization; (b) a mixture that included a short and modest downturn during industrialization; and (c) sustained, but not necessarily monotonic, gains in stature.

The United States and the United Kingdom fit the first pattern, which is given in Figure 9.3. Americans were very tall by global standards in the early nineteenth century as a result of their rich and varied diets, low population density, and relative equality of wealth. Between 1830 and roughly 1890, however, the average height of American men fell by 4.4 cm, a reversal that was not offset until the 1920s. Consistent with this height decline, life expectancies tabulated from genealogies

Figure 9.3 Average height of soldiers in Britain and of native born American soldiers

Sources. Steckel (forthcoming 2005: figure 12) and Floud, Wachter, and Gregory (1990: Table 4.8). also show deterioration near the middle of the century (Pope 1992). Researchers in the field have suggested numerous possible causal factors for the decline, including the spread of disease affiliated with the development of railroads, canals, and steamboats (for discussions, see Steckel 1995; Komlos 1998). Also mentioned are higher food prices; growing inequality; increasing market integration; the emergence of business cycles that led to malnutrition during contractions, urbanization, and the rise of public schools that exposed children to major diseases. Unfortunately, research has not advanced to the point of assigning plausible weights to these factors, but some progress is being made by examining the relationship between local (county) agricultural, socioeconomic, and demographic conditions and the height of recruits from the county (Haines, Craig, and Weiss 2003).

Although health deterioration of about 9 cm in average height among soldiers also occurred in Britain during the early mid-nineteenth century, the timing is probably more coincidental than emblematic of linkage among similar causal factors across the two countries.⁴ While it is possible that growing trade and commerce spread disease, as in the United States, it is more likely that a major culprit was rapid urbanization and its associated increase in exposure to diseases (Floud and Harris 1997). This conclusion is reached by noting that urban born men were substantially shorter than the rural born, and between the periods of 1800—30 and 1830—70 the share of the British population living in urban areas leaped from 38.7% to 54.1%.

Australia and Wurttemberg illustrate pattern (b) of short, modest cycles shown in Figure 9.4. Both realized gains in health during industrialization, but progress was choppy or otherwise interrupted by relatively brief cycles in height. Adult males reached about 162.8 cm (average of rural and urban) in the province of Wurttemberg

Figure 9.4 Average height of soldiers in Australia and Wurttemberg

Sources. Whitwell, de Souza, and Nicholas (1997: figure 10.3; raw data provided by authors) and interpolated as a weighted average of rural and urban from Twarog (1997: Table 7.29).

for cohorts born on the eve of industrialization, which began in the 1860s (Twarog 1997). A small spurt in average height to 166 cm occurred during the early 1860s which was followed by a decline of 1 cm over the next decade. Recovery but stagnation at 166 cm occurred in the 1880s and 1890s. This temporal pattern was related to the financial crash of 1873 and the subsequent depression, that lasted into the early 1890s. Occupational differences in stature indicate that the professional classes were protected during the early phases of the economic depression, and the loss in health was concentrated among the middle and lower classes. Thus, growing inequality played an important role in Germany's health trends during industrialization.

Two features distinguish the Australian experience: the tall stature (about 172.5 cm) on the eve of modernization that was followed by a large cycle in heights whereby the average height of the mid-1870s was not attained again until the second decade of the twentieth century (Whitwell et al. 1997). The tall stature is undoubtedly related to an inexpensive but diverse diet that was also rich in protein, a phenomenon supporting the view that Australia was a workingman's paradise. Even though the share living in urban areas was relatively high (about 50%), overall population density was low and the country and its major cities were relatively isolated from globalization, which hampered the spread of communicable diseases.

But some troubles occurred even in these relatively idyllic circumstances. The height downturn of 1.8 cm in the 1880s and 1890s was the result of a double whammy. The share living in urban areas was already high (43% in 1881) and then jumped 8 percentage points in the decade following. A sanitary crisis ensued and typhoid fever, which disproportionately affected the young, was epidemic in the cities. Although the pace of urbanization fell considerably during the 1890s, GDP declined and remained relatively low for a decade, thereby dampening any hopes for quick recovery in heights and health.

The Netherlands, France, Sweden, and Japan fall into the last category: sustained increases in stature, interrupted at most by brief, tiny declines or modest stagnation. Prominent in the Dutch experience was a twenty-year pre-industrial height decline of 1.9 cm that began in 1815, which was caused in part by rising food prices and stagnating nominal wages (i.e. a decline in purchasing power). Thereafter, average heights increased more or less continuously into the twentieth century with the exception of small reversals in the early 1840s and in the early 1870s. The former was probably affected by harvest failures (the ‘hungry forties') and the latter was associated with the economic depression of the 1870s (Drukker and Tassenaar 1997).

Similarities and contrasts in the Dutch and the French experiences are shown in Figure 9.5. The French did not have the small and brief cycles of the Dutch before mid-century, but they did experience considerably less height gain thereafter. While the Dutch gained 3.8 cm between birth cohorts of 1850 and 1890, the French increase was slightly less than one-third as much. In France, the modest advance in heights was accompanied by progress in performance measures such as GDP per capita and life expectancy. France also experienced a decline in economic conditions that contributed to slow growth in average heights. In the 1860s, a downturn in real wages was followed by fifteen years of economic stagnation (Weir 1997).

Figure 9.5 Median height of conscripts in the Netherlands and France

Sources: Drukker and Tassenaar (1997: Table 9A.1); personal correspondence from J.W Drukker on 28 February 1999, Weir (1997: Table 5B.1), and van Meerten (1990: 775-6).

Japan opened the industrial era at the end of the nineteenth century with the smallest stature (about 157 cm) of any industrializing country (see Figure 9.6). Hampered by a low protein diet, thereafter progress was slow and significantly correlated with per capita GDP but adversely affected by economic policy that diverted resources to the military (Honda 1997). Its high level of urbanization and modest investments in public health were an obstacle to human health and physical growth. Economic stagnation in the 1920s and the depression of the 1930s (which was rather mild in Japan) brought the modest gains in height to a halt in the mid-1930s.

Sweden realized the most sustained increase in health during the most intense period of industrialization (late nineteenth century). Figure 9.6 shows that average adult male heights rose from 168 to 172.5 cm between 1860 and 1900. The only downturn was the small reversal that occurred during the crop failures of the late 1860s, which had little to do with industrialization (Sandberg and Steckel 1997). Paralleling the growth in stature were declines in childhood mortality rates of roughly 50%. It is notable that Sweden had the least urbanized population among the eight countries studied, and it also benefited from public health measures such as vaccination, and from relatively low food prices created by the spread of potato cultivation and imports of food from America.

3.4 Generalizations

Study of height and mortality patterns in counties, diverse by time period of industrialization and by environmental factors, indicates that a combination

Figure 9.6 Average height of conscripts in Sweden and Japan

Sources. Sandberg and Steckel (1997: Table 4.1) and Shay (1994: Table 10.A6).

of general tendencies and idiosyncratic factors affected health during the industrial revolutions of the nineteenth and early twentieth centuries. In an era when public health policies were often lacking or meagrely enlightened by theories of disease causation, urbanization was a widespread culprit in ill health within countries studied in Europe and in the Pacific, and within the United States. Height was inversely correlated with degree of urbanization across countries, and rising urbanization led to health deterioration, especially in England, Australia, and Japan.

Major business cycles also affected heights and health. France, the Netherlands, Germany, and Australia were victims of major downturns. Changing economic opportunities, in the form of growing inequality, adversely affected heights in Germany and the United States.

Diets were important for health and human growth. Countries with the tallest men (Australia and the United States) had excellent access to a variety of foods, including several rich sources of protein. Food was expensive and the diet was low in protein in the country with the smallest stature (Japan).

Lastly, public health policy (or lack thereof) was also important for heights. Countries that industrialized early, such as the United States and the United Kingdom suffered the most, in part because the adverse effects of trade and population concentrations on health could not be offset by health policies informed by reliable theories of disease causation. Merely arriving late on the scene was no guarantee of protection against the by-products of industrialization, however, as shown by the Japanese case where resources that could have been used for public health and human growth were diverted to the military.

4.