Wealth and Child Mortality in the Nineteenth-Century United States: Evidence from Three Panels of American Couples, 1850-1880

With only a few exceptions, the historical study of individual-level correlates of child mortality in the United States has been limited to the period surrounding the turn of the twentieth century, when children ever born and children surviving data collected by the 1900 and 1910 censuses allow indirect estimation of child mortality. The recent release of linked census data, such as the IPUMS MLP datasets, allows a different type of indirect estimation over a longer period. By following couples across subsequent decennial
censuses, it is possible to infer child mortality by measuring whether couples’ own children in the first census were still present in the second census. We focus our analysis on children aged 1–3 in the first of two linked censuses, who were less likely to be undercounted by the census than infants, and unlikely to be living apart from their parents in the second census. We estimate child mortality over the intervening decade and use OLS regression to correlate that mortality to the residence location and socioeconomic characteristics of their parents’ households. We limit our analysis to three panel datasets for married couples linked between the 1850–60, 1860–70, and 1870–80 censuses, when real estate and personal estate wealth data were collected. Our results indicate a significant negative relationship between wealth and child mortality across all regions of the United States and over the entire period examined.