Environmental factors in differential infant and child mortality decline in England and Wales, circa 1895 to 1910.

The purpose of this research was to test the hypothesis that environmental factors made a greater contribution to the differential decline in infant mortality during the early Twentieth Century than did improvements in the standard of 1iving.Infant mortality levels were estimated from retrospective reports of currently married cohabiting women under 45 years of age, who were enumerated in the 1911 Census of England and Vales. They span the period from approximately 1895 to 1910. Material in the published tables are used to create two data sets, one in which the unit of disaggregation was husband's occupation and the other in which it was place of enumeration, for towns, cities and urban districts with fifty thousand or more inhabitants in 1911, and referred to both in the census reports and hereafter as the Great Towns. Two hundred occupations or occupational groupings included sufficient couples to produce robust enough estimates, and as well as their occupation, their class and the extent to which they were urban dwellers was also known. Income level for ninety-five of the occupations was obtained from the Board of Trade Hours and Earnings Survey. For the Great Towns it was possible to obtain directly from the census tables, or otherwise to estimate the type of administrative local authority of the town, its poverty level, a measure of urban development and selective occupational migration. For both major groups it was also possible to control for fertility dec1ine.Infant mortality by father's occupation underwent an average decline of 36% from a peak level of 132 deaths per 1000 births with wide variation about the mean. Analysis of this data set suggested that the removal of poor environmental conditions was particularly important. Highly urban occupations experienced steeper declines from higher peaks than did rural ones. However, income only explained variation in the decline at very high levels although it was also likely to be associated with other possible explanatory variables, such as education. In the Great Towns average decline in infantmorta1ity was 35% from a peak of 146 deaths per 1000 births. In this data set the rate of urbanisation over twenty years which was used as the index for environmental effects accounted for a significant proportion of the differential decline, while measures of poverty, of either level or change explained little additional variation. There was no evidence for concluding that selective occupational migration had distorted the urbanisation effect. There were no strong reasons for concluding from analysis of either data set that fertility decline had an important effect on infant mortality, although it was likely to have been important to the overall decline.