Demographic, Mortality, and Intergenerational Effects of Family Support Programs

This project looks at how an intervention in early childhood affects health, reproduction, and mortality in later life and across three generations. The focal outcomes are fertility (births) and mortality (age at death). Understanding how childhood conditions affect such outcomes decades later – and for these individuals’ descendants – is highly relevant to present-day questions about the early-life origins of health and longevity and the long-run effects of supporting children in poor families. But examining the intergenerational effects of an intervention in childhood is extremely difficult: while reductions in support for poor families have been shown to have negative short-run effects on health, we often lack the data to study impacts in the longer run and across generations. Prospective studies, if they could be done ethically, are not feasible because of the long timeline and excessive costs. However, using historical census microdata from England and Wales, this proposal seeks to exploit the drastic reduction of “poor relief” due to the Poor Law Amendment Act of 1834 to understand the impact on fertility and mortality across multiple generations. In addition to the availability of fertility and mortality outcomes across three generations, the data to be used in this project are also unique in providing adequate sample sizes and sufficient variability in the independent variable—changes in family support—to produce valid estimates. Specifically, we will collect novel data on parish-level poor relief spending and district-level fertility and mortality, and use a difference-in-differences design to study effects on these health and demographic outcomes in later life and across three generations. Individual-level measures of fertility are also observable in the 1861 and 1891 censuses of England and Wales, and by linking males across these years, we can also provide more detailed estimates of mortality impacts. Thus, we will be able to study the impacts on individuals who were children at the time and on two generations of their descendants. As a preliminary validation of our project, we used coarser publicly available versions of these data and a simplistic and imprecise record-linking strategy (a “naïve” record-linking strategy) to examine our outcomes of interest across these 6 decades. In the very short run, we find evidence of increased mortality in counties where poor relief declined the most. For longer term outcomes, surviving cohorts of children from counties where poor relief declined the most have more offspring as adults. Consistent with greater fertility being a sign of poverty due to the tradeoff between having more children and investing more in their human capital, we find that fewer of their female children – the granddaughters of those exposed to poor relief declines as children – were in school. This proposal therefore focuses on collecting and gaining access to more detailed data and refining the record-linking strategy used in our preliminary study. Our aim is to more accurately and precisely estimate long-run and intergenerational health and longevity impacts and improve our understanding of factors that mediate these effects at critical periods for the health and well-being of children and multiple generations of their descendants.