Stéphane Gauthier*

It is difficult to say to what extent we can rely on the growth curves in our family health booklets to understand people’s growth patterns in the past. Uncertainty remains because sources that follow substantial numbers of children over a long period before the middle of the 20th century are rare. The proxy method, which consists of calculating averages for each age for different individuals, is considered unreliable because it can hide changes to individual growth patterns. Broad studies leading to surveys of individual anthropometric data were not done until after WWII. For previous centuries, the growth curve that Buffon published is considered the first of its type: it is based on a single individual, the son of Montbeillard, who was measured and weighed every year from his birth in 1759 until 1777. Quetelet’s work a century later on orphans in Brussels, who would not have gone through a growth peak in their adolescence, is still used as a reference for past populations.

In this article, Stéphane Gauthier observes that the sizes reported for the same man in two classic sources of military data, the district recruitment register, and service records, are often different. He mounts a series of arguments to suggest that with the combination of these two sources it is possible to build an individual profile that works for a large proportion of men since the 19th century in France. The recruitment register recorded information gathered at the moment when the men were examined by the review board that decided on their aptitude for military service. This exam took place every year, usually between February and April, and involved all men who had turned 20 in the preceding year. A service record is a sort of individual military identity card. It was most probably drawn up in the recruitment office a few months after the review board’s decision, in the first few days of October, just before the soldier’s induction. In principle, they were created only for those who were deemed fit for service by the review board. However, files for most men born at the end of the 19th century are available because they were re-examined several times during the First World War and often inducted. For these cohorts, the information contained in the recruitment registers and service records provide an almost exhaustive sample. One illustration of the particular case of men born in the Corrèze region in 1887 and examined by the review board in 1908 allows us to estimate a growth of between 0.3 and 0.4 cm over the year following induction. However, in the growth charts of the family health cards booklets, men have already reached their adult height by the age of 20. Gauthier’s thus results confirm theories of reaction norms which state that in the developed economies, men reach a taller adult height and reach it earlier. The example of the men from the Corrèze shows that the distribution of growth in height is far from uniform. The author observes a late catch-up among the smallest: the men who, at the moment of the review board exam are taller than the average height of 163cm for their cohort, do not grow any taller. The late growth of Frenchmen in the past thus reduced the inequality in the final height distribution. It also implies a substantial correction to the current curves when we look at the French men of the last century who were approaching maturity.



Original title of the article: Late height growth from historical individual-level panel data

Published in: PSE working paper n°2021-06

Available at

Credits (picture): Dizfoto – Shutterstock

*PSE Member