The changing body is a very important book. It unifies vast literatures and sets of evidence to highlight how the health of human populations in Britain, Europe and America has improved dramatically over the past 250 years. The book is well written and organised. It is also on the whole accessible, though the models and equations in Chapters 2 and 3 will be difficult for non-quantitative scholars (and students) to understand.
Chapter 1 sets out the principal thesis of the book ‘that the health and nutrition of one generation contributes, through mothers and through infant and childhood experience, to the strength, health, and longevity of the next generation; at the same time, increased health and longevity enable the members of that next generation to work harder and longer and to create resources which can then, in their turn, be used to assist the next, and succeeding, generations to prosper’ (p. 2). This is a simplified description of the theory of ‘technophysio evolution’ that Robert W. Fogel and Dora L. Costa first developed.
Chapters 2 and 3 highlight ways of measuring technophysio evolution by describing energy cost accounting, calorie distributions, Waaler curves, and the relationship between calorie availability and labour productivity, famine and population growth. Chapter 3 culminates by attempting to measure the effect of nutrition and health on income per capita. These chapters are the weakest section of the book because many of the key arguments rely upon problematic assumptions and simplified understandings of complex relationships. I will focus on three of these here.
First, in order to construct distributions of calories available to the population, the authors assume that the deciles of wealth, body mass index (BMI), height, basal metabolic rate (BMR) and calories are made up of the same people. Thus, the top decile in wealth contained the tallest, heaviest (in terms of BMI), highest BMR and largest calorie-consuming 10 per cent of the population. However, it is unlikely that these distributions overlapped in this way. For instance, there is no evidence of a strong positive relationship between height and BMI, which would lead the tallest decile to also be the heaviest. These assumptions affect the distribution of BMRs and the calculated physical activity levels (PALs) of each decile, from which the authors draw their inferences on what percentage of the population was able to work. After making all of the adjustments implied above, the PAL level at the top end of the distribution would increase substantially, allowing a larger share of the population to perform strenuous labour. However, the PAL at the low end of the distribution would decrease, suggesting that over 30 per cent of the population did not have enough energy to perform eight hours of light labour tasks per day and 10 per cent barely had enough energy for the survival diet of 1.27 times the BMR. This seems unlikely given mortality rates in Britain in 1800.
In addition, the predicted lognormal calorie distributions are highly sensitive to changes in the mean of the distribution. A 10–20 per cent increase or decrease in the mean calories available per consuming unit could render the distributions presented implausible based on the authors' own logic. Given the potential error in the calorie estimates (see below), it is difficult to trust the results drawn from the distributions. It is possible that up to 20 per cent of the English population did not have enough calories for sustained work effort, but we will need more confidence in the calorie estimates and a less sensitive method for estimating this very important inequality measure before we can be certain.
Second, the authors do not mention the contradictory results in the health economics literature between micro and macro level studies of the influence of health on income. In micro level studies, the effects of health improvements on income, literacy and fertility are strong, but very few macro studies that robustly manage the endogeneity between income and health have been able to establish a significant relationship between improvements in health and increases in income. To be fair, this process is inherently endogenous, and I do not find much of the macro literature convincing. However, the potential endogeneity is not addressed when regressions are run in Chapter 5 and when the authors calculate the effect of increased energy available for work on the annual British growth rate between 1800 and 2000. The income effect would seem especially apt there as Britain purchased more food from abroad over time.
Third, although technology is embedded in ‘technophysio evolution’, the theory does not manage technological change particularly well. For instance, the authors assume that there is a direct relationship between the amount of energy available for human labour and economic output over the past 200 years in Britain. This might be plausible in the pre-industrial era, but as Allen and Wrigley have emphasised, the technology developed in the Industrial Revolution substituted capital and mineralised energy for human labour. Thus, the amount of hard manual labour that a population could perform became less important in determining output than the availability of mineralised energy and the technology to harness it. Further mechanisation and the shift toward a service sector-dominated economy in the twentieth century has further divorced output from the amount of energy available for work. Today, workers in the most productive sectors use fewer calories for work than a beggar would need for begging in 1800.
The meat of the book is in Chapters 4 to 6, which bring together evidence from a wide range of sources to highlight how the health of populations in Britain, Europe and America has changed. The analysis here is incisive and will be extremely helpful to students, non-experts wanting a quick grasp of the field, and to experts seeking a comprehensive analysis of available evidence. The chapters also present many controversial debates (trends in English male heights across the Industrial Revolution and the McKeown debate to name two) with clarity and fairness even if the authors' preferred interpretations remain.
The new energy cost accounting figures for England are especially appreciated and are a significant improvement upon those published in Fogel's Escape from hunger.Footnote 1 The authors should be commended for clearly describing and presenting their calculations in the book so that they can be replicated and checked. While the figures presented seem plausible, several forthcoming papers suggest that they are a lower bound estimate when compared with other estimates.Footnote 2 The authors assume exceptionally high losses from spoilage and distribution of grains, which push their numbers downward relative to others. Readers should also note that there is an arithmetic error in the calculation of wheat output in 1750, which, once corrected, increases the calories available per consuming unit substantially.Footnote 3 Thus, calorie consumption per consuming unit is more or less constant from 1750 to 1850, which makes it difficult to attribute the mortality decline between 1750 and 1837 and the increase in adult male heights from 1770 to 1820 to improving nutrition.
Despite a few problems in the grand interpretation, The changing body is an excellent book that I will reread again and again for insight and inspiration.
