Principal investigator: Bruno Benevit
Original title: Watersheds in Child Mortality: The Role of Effective Water and Sewerage Infrastructure, 1880 to 1920
Authors: Marcella Alsan and Claudia Goldin
Location of the Intervention: United States
Sample Size: 2440 municipalities-years
Sector: Health Insurance
Primary Variable of Interest: Infant Mortality Rate
Type of Intervention: Basic sanitation
Methodology: Difference in differences
Summary
Infant mortality rates have decreased significantly over the last 150 years. This improvement is often attributed to technological advances in sanitation and health services, as well as the implementation of public policies. The aim of this article was to analyze the impact of the creation of a water and sanitation district in the Boston area on the infant mortality rate between 1880 and 1920. Using the difference-in-differences-through-multiple-specifications methodology, the authors identified that the joint implementation of both types of infrastructure was fundamental to the reduction in the infant mortality rate observed during that period.
- Policy Problem
Infant mortality was the leading cause of low life expectancy at birth in the United States and much of Europe. In the American state of Massachusetts, infant deaths accounted for 20,4% of all deaths in 1880, while births represented only 2,5% of the total population at the time (Alsan and Goldin, 2019). However, between 1870 and 1930, infant mortality plummeted from about 1 in 5 to 1 in 16 white babies in the United States – in Massachusetts, deaths of children under five years old decreased sevenfold.
This period coincides with the implementation of public water and sanitation systems. Sanitation reduces fecal-oral transmission of pathogens by removing excrement from drinking water sources, decreasing human contact with feces and limiting exposure to gastrointestinal diseases transmitted by flies. Clean water interventions remove impurities, making water safe for consumption and hygiene. These interventions have been considered by several authors as fundamental to explaining the observed downward trend in childhood morbidity, highlighting the importance of this type of public policy.
- Policy Implementation Context
At the end of the 19th century, the Boston metropolitan area was facing a major sanitation crisis. Rapid population growth and rampant industrialization resulted in unsanitary conditions in the region's cities and towns. Sewer systems were primitive and frequently dumped untreated waste directly into Boston's rivers and harbor. This not only polluted water sources but also contributed to outbreaks of waterborne diseases such as cholera and typhoid fever. Furthermore, limited access to clean water exacerbated public health problems, as many communities relied on contaminated water sources.
However, the situation began to change with the implementation of proactive public policies. The creation of the Metropolitan Sanitation District (MSD) and the construction of sewage treatment and water supply systems were crucial steps in improving sanitation conditions. The MSD allowed for efficient coordination between various municipalities to address the sewage problem, while the water treatment systems made the drinking water safe for consumption. Over the following years, infant morbidity showed a significant reduction, indicating the importance of investment in sanitation infrastructure and clean water as vital components of public health and the well-being of urban communities.
- Evaluation Details
The data used in this study were comprised of a set of information databases related to water and sanitation interventions in the Boston metropolitan area. The primary data sources include annual reports from the State Board of Health, the Metropolitan Sanitation Commission, the Metropolitan Water District, and the Metropolitan Water and Sanitation Council. These reports detail the dates on which water and sanitation interventions were completed in different municipalities within the region, providing records of public sanitation service implementations over time. In addition, the data encompass demographic and vitality information at the municipal level, collected from the Massachusetts Secretary of State's annual vital statistics reports.
The primary sample comprised a panel of 60 municipalities between 1880 and 1920, consisting of 2440 municipality-years. This sample included all municipalities within the immediate area of the Boston Harbor basin, as well as municipalities outside the immediate area of Boston that were incorporated cities by 1895. Additionally, two coastal municipalities (Ipswich and Weymouth) and five with the largest populations in 1880 (Attleboro, Clinton, Milford, Natick, and Peabody) were considered.
- Method
To assess the impact of sanitation policies on health, outcome variables related to infant mortality were considered. The main variable of interest was the infant mortality rate (logarithmic), as well as mortality rates for other ages: under one, under two, and over five years. Thus, the differences in observed characteristics between the treated and control municipalities were initially evaluated.
As the main empirical strategy, the Difference-in-Differences (DID) method was employed, assuming the condition of parallel trends in the trajectory of health outcome variables between the treated and control groups in the absence of intervention by water supply and sanitation infrastructure policies. Several model specifications were presented, considering different sets of variables, fixed effects (time and municipal) and temporal trends, as well as count and linear models at the (non-logarithmic) level. Standard errors were grouped at the municipal level throughout the analysis.
As a robustness strategy, the authors analyzed possible co-occurring effects. Analyses were conducted to examine potential effects on population composition resulting from infrastructure improvements, which could abruptly affect infant mortality rates due to the immigration of wealthier individuals. Furthermore, heterogeneous effects were verified considering the time of year (spring-summer and autumn-winter), population composition (municipal population growth rate, proportion of British foreigners, and proportion of Irish foreigners). The effects of the interventions were also estimated by category of disease causing death and on the non-infant mortality rate. Finally, the study presents the results with samples considering different groups of treated and control municipalities.
- Main results
Initial results demonstrated similarity in most characteristics between the municipalities in both the treated and control groups. Overall, few changes were observed in mortality rates in the municipalities over the 9 years preceding the intervention.
The results of the main analysis indicated that the combined introduction of water supply and sewage systems led to a significant 33,6% reduction in the infant mortality rate in the municipalities that underwent the intervention over the 41 years observed. When analyzed separately, the results indicated significant effects of smaller magnitude resulting from the implementation of sewage systems, while the water supply system did not show statistical significance. These results were consistent considering various model specifications.
Regarding the heterogeneous effects analysis, the results indicated that, while there was a significant reduction in infant mortality from gastrointestinal diseases (associated with improvements in basic sanitation), a less significant effect was identified in the reduction of deaths from respiratory diseases and non-significant effects on infant mortality from tuberculosis and on the non-infant mortality rate. The infant mortality rate showed significant effects when considering the period from April to September (spring-summer), and non-significant effects between October and March (autumn-winter), results consistent with the greater vulnerability of children to gastrointestinal diseases, which are more common in warmer periods.
In terms of relative gains related to socioeconomic conditions, the results indicate that municipalities with a higher proportion of Irish immigrants (who were more vulnerable at the time) experienced more significant reductions in infant mortality rates compared to those with a higher proportion of British immigrants. Additionally, municipalities with lower population growth showed significant reductions in infant mortality rates, unlike half of the municipalities with higher population growth rates. The results remained significant considering different compositions of municipalities in the sample, indicating the robustness of the findings.
- Lessons in Public Policy
This article conducted an analysis of the importance of basic sanitation infrastructure, focusing on water and sewage interventions, in reducing infant mortality in the Boston metropolitan area between 1880 and 1920. The results revealed that the implementation of clean water systems and effective sanitation played a key role in improving child health, accounting for approximately one-third of the reduction in infant mortality over 41 years.
The results of this study have significant implications for the formulation of public policies related to child health and basic sanitation infrastructure. They highlight the importance of investments in clean water and sanitation systems as effective measures in reducing infant mortality. Furthermore, they emphasize the need for comprehensive and coordinated approaches in the planning and implementation of sanitation infrastructure, especially in urban areas. These findings can guide future policy decisions and investment strategies aimed at improving children's health in urban contexts, both in the United States and in other parts of the world where similar challenges persist.
References
ALSAN, M.; GOLDIN, C. Watersheds in Child Mortality: The Role of Effective Water and Sewerage Infrastructure, 1880–1920. Journal of Political Economy, v. 127, no. 2, p. 586–638, apr. 2019.
