Farr, William (1807–1883)

Abstract

William Farr, born in Kenley, Shropshire on 30 November 1807, died in London on 14 April 1883, was a statistician in the General Register Office who had been appointed in 1840 as ‘compiler of abstracts’ and was 2 years later made Statistical Superintendent, a post he held until his retirement in 1880. He pioneered the quantitative study of morbidity and mortality and in the process became one of Victorian England’s most prominent figures in the public health and reform movements (Cullen 1975). He made major contributions in the fields of data collection, being largely responsible for the introduction of a cause of death classification which was linked with his derivation of the ‘zymotic’ theory of epidemic disease (Eyler 1979; Pelling 1978). As an Assistant Census Commissioner for each of the censuses of 1851, 1861 and 1871, he was largely responsible for the development of reliable procedures for the recording of occupations (McDowall 1983). He is, however, best known as a statistical analyst, for in 1843 he constructed the first English Life Table based on deaths in 1841 linked to the census of that year. At the same time he established the formula for deriving from a rate of mortality by age m the probability of survival p at the initial age. In 1850 and 1864 Farr produced his second and third English Life Tables, the last mentioned being used as the actuarial basis for the life insurance scheme set up by the Post Office for its employees. Farr in his work on occupational mortality was the first to make extensive use of the standard mortality rate, allowing comparisons of the mortality of different groups by means of a summary statistic which took account of differences in the age structure of the groups being compared. A recurring theme in his work was the identification of variation in mortality in different urban areas of the country. Such differential mortality was viewed as an index of human welfare. For example, in 1850 one-tenth of the registration districts, those he named ‘healthy districts’, had average mortality rates not exceeding 17 per 1,000, a rate he thought indicative of the ‘natural’ mortality which, when exceeded, would indicate those deaths attributable to unnatural and preventable diseases. An underlying aim in much of his work was to discover statistical laws or numerical expressions of regularities such as he proposed in the laws of recovery and death in smallpox, the elevation law for cholera mortality in London (Lewes 1983) and the law of the relation between population density and mortality. He was also an early contributor to human-capital theory (Kiker 1968) arguing, in particular, that the economic value of men varied with age as well as social class, and this he used as powerful publicity for urban reform by drawing attention to the financial losses that followed from diseases that were the causes of death and illness in society at large.