PhD Candidate, Economics
The Mortality Effects of
Local Boards of Health in England, 1848-70.
(Job Market Paper)
Although world health has improved immeasurably since the nineteenth century—especially among developed countries—there is no clear consensus about the role played by public health interventions in this improvement, and therefore no clear policy prescription for those developing countries in which health outcomes have improved only modestly. In this paper, I explore the Public Health Act of 1848—England’s first attempt at systematic sanitation reform—and in so doing I provide clear quantitative evidence of the effect of public health interventions on mortality in a decidedly developing country. Between 1848 and 1870, the Public Health Act oversaw the adoption of more than 600 local boards of health (which, combined, affected roughly one-quarter of the English population). The Act endowed these boards with the power to tax, borrow, regulate, provide sanitary services, and build, re-build, seize, or otherwise alter local infrastructure. Since the jurisdictions of local boards of health and the jurisdictions of Poor Law unions (i.e. the local geographical units within which mortality statistics were recorded) were not coterminous, the mortality effects of the Public Health Act have been largely unexplored. I introduce a new panel dataset that maps the jurisdictions of local boards of health into the jurisdictions of Poor Law unions. I then leverage variation in both the timing and extent of board adoption across Poor Law unions in order to estimate the cumulative effect of the adoption of a local board of health on mortality 1, 2, 3, and 4 years after adoption. My estimates suggest that the adoption of a union-wide local board of health reduced mortality by 14.2 percent after four years, that approximately 225,000 lives were saved by local boards of health between 1848 and 1870, and that the aggregate English mortality rate was 3.7 percent lower in 1870 than it would have been had the Public Health Act not been passed.
Hardly Worth Chaining Up?
The Effect and Cost-Effectiveness of Welfare Reform
in England, 1857-85.
Between 1601 and 1948, before the modern welfare state, the English Poor Law was a kind of “welfare state in miniature." It provided working-class men, women, and children with a vital, if meagre, guarantee of support in the event that they sustained a negative income shock. It also, according to no few contemporaries, encouraged indolence. In the late 1860s and 1870s, objections to this guarantee of support grew more vocal and more politically formidable. Some localities began to relieve a significantly greater proportion of welfare recipients in workhouses—deliberately disagreeable dwellings intended to disincentivize welfare recipiency. Others did not. Others fell somewhere in-between. This movement was, in effect, a de facto, decentralized welfare reform. In this paper, I use variation in changes in the proportion of welfare recipients relieved in workhouses across English localities between 1865 and 1880 to estimate the effect of the workhouse on the rate of welfare recipiency. First, I introduce a new panel dataset of Poor Law unions—local geographical units that administered the Poor Law, of which there were approximately 600. Second, I employ a difference-in-differences model to estimate whether and to what extent the workhouse affected the rate of welfare recipiency. I find that approximately three in four welfare applicants to whom the workhouse was offered as a “test” refused the offer. I also find that this “workhouse effect” was insufficiently large to meaningfully reduce the costs associated with the Poor Law. Third, I employ a triple-difference model to estimate whether welfare applicants to whom the workhouse was offered were more likely to reject the offer if the workhouses in which they would be relieved were deficient in ventilation, water supply, general sanitation, or diet. I find that they were not.
Wealth and Wealth Redistribution:
Evidence from England, 1881.
How does wealth affect redistribution? Are wealthier people or wealthier places, for instance, more or less likely to favor redistributive policies, such as a welfare system, because of their wealth? In this paper, I isolate quasi-experimental variation in per capita property wealth across English Poor Law unions to estimate the effect of wealth on welfare generosity. Poor Law unions were local geographical units, of which there were approximately 600 in 1881, that administered the English welfare system. They were, for all intents and purposes, autonomous mini-welfare states, in each of which a board of “guardians” determined the level of generosity. They were also agglomerations of parishes—hyper-local geographical units, of which there were approximately 14,000 in 1881. Each parish was entitled to a certain, pre-determined number of guardians to represent it on the board of the Poor Law union to which it belonged. Guardians were only rarely apportioned to parishes in proportion to their population. Some parishes—typically the wealthiest, least populous parishes within a Poor Law union—were over-represented. Others were—typically the poorest, most populous parishes within a Poor Law union—were under-represented. Moreover, the extent of the over-representation of wealthy parishes or of the under-representation of poor parishes varied across Poor Law unions. This is the variation that I exploit. I introduce a new dataset that includes the number of guardians, the population, and the property wealth of each English parish, as well as two distinct measures of the welfare generosity of each English Poor Law union. I then use a measure of the malapportionment of guardians—and, consequently, of the over- or under-representation of wealth on boards of guardians—as an instrument for wealth. This instrument is both strongly correlated with wealth (by construction) and uncorrelated with all other union-level variables that are both available and potentially related to welfare generosity.