The Origins of Anthropometric History: A Personal Memoir

I knew nothing of anthropometry—not even the meaning of the word—when, in 1977, Robert Fogel invited me to give a seminar at Harvard. Over lunch after quite a grueling occasion, he asked me if I would be interested in taking part in a project to investigate the long-term decline in mortality in the United States. As he pointed out, the vast majority of migrants to the American colonies and the United States in the eighteenth and nineteenth centuries came from Great Britain and Ireland; it was important, in explaining their subsequent mortality experience, to be able to assess their state of health before they arrived in North America. Their heights and those of the British population as a whole might, he suggested, provide evidence for such an assessment.

I was flattered to be asked to work with one of the leaders of the economic history profession, intrigued by the project—if initially skeptical about the use of height data—and, by the end of a long lunch, enthusiastic about working with Fogel and his collaborator, Stan Engerman, whom I had known for some years. Fogel presented the issues very much within the framework of demographic analysis and the assessment of health status, but it was clear that the project would attack one of the most fundamental topics of American economic history.

Fogel and Engerman had undertaken some work in the Public Record Office in London, collecting height and related information about recruits to the Royal Marines in the eighteenth century. Back in Britain, I decided to begin by investigating the existence of height data in other historical records. I soon found that there were literally millions of individual observations of the heights of recruits to the British army as well as the Royal Marines, dating back in the case of some regiments to the first half of the eighteenth century. Similar records for the Royal Navy had, unfortunately, been destroyed by fire. There were, in addition, records of recruits to the East India Company army, later used by Joel Mokyr and Cormac � Gráda (1996). I considered, but mistakenly decided not to use, information about prisoners; once again, such material later proved to be illuminating (Nicholas and Oxley 1993; Johnson and Nicholas 1995, 1997).

The initial problem was not, therefore, that of identifying useful records but rather that of devising a manageable and efficient sampling method. The task was complicated by the absence of a complete list of army and marine records containing height and age data, by the lack of any description of the recruiting methods that had given rise to the data, and, finally, by my insufficient knowledge of the biological issues that might determine the choice of variables to be recorded; I was puzzled, for example, about whether "chest expansion" should be recorded and decided, erroneously as it turned out, not to record weight on the grounds that, among young men, it was strongly correlated with height. It was immediately clear, therefore, that research into what came to be called anthropometric history would require a blend of skills and knowledge; it was to be, and this accounts for much of its fascination, a truly interdisciplinary field. Luckily, I always found many colleagues, led by James Tanner, very willing to share their knowledge of biology and ever-patient in answering what must have seemed extraordinarily naïve questions. Similar help in dealing with statistical issues came from Harvey Goldstein of the Institute of Education in London, James Trussell of Princeton, and particularly Kenneth Wachter, my collaborator on the Social Science History article (Floud and Wachter 1982). I was also greatly helped by my research assistant, Annabel Gregory, who dealt with the majority of the data collection and contributed in many other ways.

In retrospect, it is surprising that it took some time for me to see the relevance of anthropometric data to issues of changes in living standards. The initial focus of the National Bureau...