Measurement of Mortality

Abstract

Although there had been anticipations by Ibn Kaldun in the fourteenth century, G. Botero in 1589, Francis Bacon in 1612, and others, the first great study of mortality statistics was made by John Graunt (1620–1674) with his observations on the bills of mortality of the city of London in Graunt (1662), Hull (1899). The bills had commenced in 1517 or 1519 at the order of Henry VIII as a means of warning against increases in the incidence of plague; such use of them has been described by Mullett (1956). The bills were compiled weekly by the parishes and sent on to the Company of Parish Clerks, who combined them into a total for each year ending with the Thursday before Christmas. Causes of death were given under some 60 headings, but neither age nor sex was given. There were separate tabulations of the numbers of males and of females, christened or dying. Since plague was endemic in London and, in many years, the dominant cause of mortality, Graunt confined his main discussion to the deaths occurring in the 20 years, 1629 to 1636 and 1647 to 1658, during which England was comparatively free from the plague. He noted that it was immaterial to his purpose whether a man aged 70 years was recorded as dying from cough or from age. However, he would have liked to know the exact age in years of children who died. As this information was not available to him, he found that of 229,250 people dying in 20 years, 71,124 died from thrush, convulsions, rickets, “teeth,” and worms or as “abortives,” “chrysomes,” infants “liver-grown,” and “overlaid.” He guessed that about half of 12,210 deaths due to smallpox, swine pox, measles, and worms without convulsions were also of children under 6 years of age. He concluded that 36% of all births (quick conceptions) resulted in death before the age of 6 years. Later statistics, when they became available in the early part of the nineteenth century, showed that Graunt’s guess was probably a very good one. In general, Graunt (1662) demonstrated the regularity of certain vital phenomena, such as births, deaths, and marriages, the excess of male over female births and the approximately equal numbers of the sexes in the adult population, the high rate of mortality in the earliest years of life, and an urban death rate that exceeded the rural death rate. Graunt carefully examined the reliability of the data and enquired, for example, whether the apparent changes over the years were due to transfer of deaths from one disease rubric to another. His work can still be read with interest because of the great insight displayed in his criticism of the sources and the caution with which he put forward new hypotheses (see Glass, 1963, 1964; Greenwood, 1948b; and Lancaster, 1962b).