VoxEU Column Health Economics

How have Europeans grown so tall?

The last century has seen dramatic improvements in the health of Europeans. Young adult males are about 11 centimetres taller than their counterparts were a century ago. This column examines and explains the remarkable long-run trends in the average height of Europeans.

With so much of the focus in healthcare research on ageing populations (see for example Breyer et al. 2011 on this site), we might be forgiven for taking our eyes off the younger generations. But a glance in their direction can reveal some striking trends.

Young people are tall these days. Whereas a six foot (183cm) man was once viewed as distinctly tall, this height is now only a shade above average in some countries. Less well known is the fact that this process has been going on in Western Europe for five generations. New data reveals that the average height of young adult males has increased by about 11 centimetres (more than 4 inches) in the last century. This is a remarkable and unprecedented spurt in human stature.

Long-run trends in average height can be calculated either from looking at the heights of different cohorts at a point in time or, further back in the past, looking at the heights of military recruits in different years (hence we look only at males). The new dataset examined here uses a combination of these methods to derive new long-run series for men aged about 21. The data for 15 European countries are plotted in Figure 1. It shows that for the countries of middle and northern Europe height increased from around 168cm for those born in the early 1870s to around 179cm for the cohort of the late 1970s. Southern Europeans were significantly shorter and have remained so.

Figure 1. Adult male height by birth cohort, 1856-60 to 1976-80

Source: Hatton and Bray (2010), Table 1.

The literature on human height identifies two proximate determinants of height: nutrition and disease. While the effect of the former is obvious, the latter works through infections that reduce the body’s ability to convert food into bone and tissue during infancy and childhood. Of course the differences between individuals in height are largely genetic. But the trend in average height is largely determined by the socioeconomic conditions that influence it through the two channels of nutrition and disease.

Looking more closely at the underlying trends in height reveals an interesting puzzle Table 1 shows that for northern and middle Europe, the period of fastest increase in height was 1911-15 to 1951-55. This is somewhat surprising as it was a period of relatively slow growth that embraced two world wars and the Great Depression. By contrast the period from the 1950s, an era that saw improvements in medical technology, the advent of national health services, and above all faster growth in living standards, was accompanied by a more modest trend in average height. For southern Europe (which here includes France) the trend in height increased across each of these major periods.

Table 1. Trends in height (cm per decade)

 

1871-75 to

1976-80

1871-75 to

1911-15

1911-15 to

1951-55

1951-55 to

1976-80

North

1.24

0.84

1.43

0.99

Middle

1.13

0.80

1.36

1.02

South

1.12

0.68

1.05

1.70

Note: Growth rates estimated by fitting linear trends, with fixed country effects. Grouping of countries as follows: North: Denmark, Finland, Netherlands, Norway, Sweden; Middle: Austria, Belgium, Germany, Great Britain, Ireland; South: France, Italy, Greece, Portugal, Spain.

So what can account for these distinctive trends? In a recent CEPR Discussion Paper (Hatton 2011), I explore the key socioeconomic determinants of average height. Some economists think of the relationship between living standards and heath in terms of a “health production function”. This is depicted in Figure 2 as a concave relationship between height and per capita income.

Figure 2. Health production function

Two implications follow from this. First, think for the moment of movements along HPF1 and associate the income levels with different points in time (such that Y1 is 1871-75, Y2 is 1911-15, Y3 is 1951-55 and Y4 is 1976-80). It is possible to explain why the upward trend declined in the last period (H3 to H4) as the relationship becomes flatter. But it is more difficult to explain why it accelerated from the First World War (i.e comparing growth from H1 to H2 and from H2 to H3). A second implication is that, because of the concavity of the health production function a fall in inequality should lead to an increase in average height.

So what actually happened? Econometric analysis reveals that per capita income helps to explain some of the increase in height, but the fall in inequality does not. And the overall upward trends are only partly explained. The results suggest that the twentieth century experience has been largely an upward shift of the health production function, rather than a movement along it. This is shown in Figure 1 as a shift from HPF1 to HPF2. So what can account for this? One possibility is falling family size, as parents sought to reduce the quantity and increase the “quality” of their children. Another is parents’ education, working though better knowledge of nutrition and hygiene and perhaps also higher aspirations for their children.

But the most important factor shifting up the HPF is the improvement in the disease environment. This is reflected in the steep downward trend in infant mortality, something that has been associated, above all, with sanitary reforms and improved housing. As Table 4 shows, when infant mortality is used as a proxy for the disease environment, its steep fall after 1911-15 accounts for much of the acceleration in height, with more modest contributions from falling family size and rising education.

Table 2. Contributions to increase in height by period and region

Region

 

Log GDP per capita

Infant mortality

 

Family size

 

Parental education

 

1871-75 to 1911-15

North

0.53

1.48

-0.04

0.34

Middle

0.48

0.09

0.06

0.49

South

0.42

0.83

0.03

0.26

 

1911-15 to 1951-55

North

0.80

2.58

0.57

0.44

Middle

0.38

2.97

0.43

0.28

South

0.50

2.41

0.21

0.16

 

1951-5 to 1976-80

North

0.83

0.81

0.12

0.31

Middle

0.98

1.46

0.07

0.15

South

1.21

1.80

0.06

0.66

 

1871-75 to 1976-80

North

2.16

4.87

0.65

0.87

Middle

1.89

4.52

0.57

0.90

South

2.08

5.28

0.29

1.17

 

Of course infant mortality is itself determined by deeper causes – causes that include trends in income, in family size, and in parental education. Indeed, if infant mortality is omitted from the analysis then the contribution of these factors increases. It is also possible to identify positive effects of public expenditure on social services and the advent of universal health systems. In addition the evidence suggests that improvements in the urban environment contributed to the upward trend in height.

But after taking account of these effects, the most sobering result is that, in the absence of a direct proxy for the disease environment (infant mortality), only about half of the century-long increase in height can be explained. It seems likely that advances in medical knowledge and improvements in sanitary infrastructure account for much of the missing explanation. Perhaps most important is the way in which parents converted better understanding of hygiene and nutrition into practices that radically improved the well-being of their children.

References

Breyer, Friedrich, Joan Costa-i-Font, Stefan Felder (2011), “Does ageing really affect health expenditures? If so, why?”, VoxEU.org, 14 May.

Hatton, TJ (2011), “How have Europeans Grown so Tall?”, CEPR Discussion Paper 8490.

Hatton, TJ and E Bray (2010), “Long Run Trends in the Heights of European Men, 19th and 20th Centuries”, Economics and Human Biology, 8:405-413. 

16,798 Reads