Explaining declines in US rural mortality, 1910–1933: The role of county health departments

https://doi.org/10.1016/j.eeh.2018.08.003Get rights and content

Abstract

This study assesses the mortality decline associated with one of the first and most extensive rural public health efforts in the United States. Between 1908 and 1933, county governments enacted county health departments (CHDs) that provided sanitation and child-oriented health services. Using an event study design, I exploit variation in the location and timing of CHD arrival to measure the health effects. I find that CHD entry led to a decline in infant mortality, but provided little advantage to overall population health. For infants, CHDs prevented two deaths per 1000 births, which accounts for 8–10% of the period-specific mortality decline. The effect is most substantial in rural-only counties where infant mortality declines by roughly three deaths per 1000 births.

Introduction

At the beginning of the twentieth century, urban-focused public health in the United States neglected the nearly 60% of the population residing in rural areas. While urban centers were considered magnets of disease and filth, both urban and rural infants faced a one in ten chance of dying before their first birthday.1 To combat the rural health problem, the United States Public Health Service (USPHS) and related organizations encouraged county governments to establish full-time public health departments over 1908–1933. In this study, I question whether the services offered by county-level health departments (CHDs) impacted the local mortality conditions in targeted counties. Reports surrounding the history of CHDs detail the advantages of the program, but the causal inference is limited. For instance, in Tennessee CHD counties over 1927–1929 ”the recorded death rate from diphtheria was about 20% lower and that from typhoid fever about 40% lower” (Lumsden, 1920 pp. 2630). Due to the limited inference from these direct comparisons, a more robust program evaluation is warranted.

I reconsider the effect of the national program using an event-study design that exploits variation in when and where CHDs were adopted. This event-study tracks mortality in each year before and after the CHD arrived.2 To estimate the event study, I digitize a novel dataset of health department administrative records from 1908 to 1933. I combine these records with county-level mortality data, census data, and taxation data.3 Using the event-study design, I find that CHDs persistently reduced local infant mortality but had no consistent impact on overall population mortality.4 Unlike urban public health investments, CHDs were only beneficial for infants and cannot explain the overall mortality transition in rural areas. The estimates suggest that health investments persistently reduced infant mortality by two deaths per 1000 births. Based on the period decline in infant mortality from 88 deaths per 1000 births to 62 deaths per 1000 births, CHD access accounts for eight percent of the rural infant mortality decline over the span of the rollout.5 The magnitude of this effect is similar but slightly lower than the impact of the Sheppard–Towner Act found in Moehling and Thomasson (2014).6 By contrast, the effect is small relative to urban sanitation investment, which (Alsan and Goldin, 2015) attribute with 44% of the decline in infant mortality between 1880 and 1915.

Next, I consider the effect on subpopulations and find that the mortality reductions are highest in worse-off communities. CHDs disproportionately benefited nonwhites, rural-only counties, and counties with lower wealth and limited access to private physicians.7 Ideally, I would also test which diseases CHDs were nationally effective against. Unfortunately, a major limitation of this study is that lack of county-level by-cause mortality data during the rollout. As an alternative, I test the disease-specific effects with by-cause morbidity and mortality data from Michigan, Alabama, and North Carolina. I find that CHDs are associated with declines in cases of diphtheria, typhoid, and smallpox, as well as lower maternal mortality. These findings indicate that CHDs may have targeted infant mortality with treatments, vaccines, or hygiene information.8

This study contributes to the literature by being one of the first to focus on a rural-oriented public health program during the early twentieth century. A closely related study, (Higgs, 1973), documents the rural mortality decline leading up to this study (1890–1915) and argues that standard of living improvements produced the decline rather than rural public health programs. This study picks up where (Higgs, 1973) left off and is one of the first to focus on rural public health in the early twentieth century. While rural health has been sparsely studied, there is an extensive literature covering state and city-level efforts. The city-level literature has shown that urban public health investments led to significant declines in adult and infant mortality (Troesken, 1999, Troesken, 2001, Haines, 2001, Cutler, Miller, 2005, and Alsan and Goldin, 2015). Prior work has also documented the benefit of more unified public health efforts across several states (Olmstead, Rhode, 2004, Bleakley, 2010, and Moehling and Thomasson, 2014). Two closely related studies, Moehling and Thomasson (2014) and Bleakley (2007), have considered similar funding initiatives in related contexts. Moehling and Thomasson (2014) examines the Sheppard–Towner Act and finds decreased state-level infant mortality in targeted areas. Bleakley (2007) considers the efforts of the Rockefeller Sanitation Commission’s participation (RSC) in hookworm eradication.

The remainder of this study proceeds as follows. Section 2 covers the history of the health investment. Section 3 presents the event-study design and then Section 4 describes the administrative, mortality, and tax data used throughout the study. Section 5 displays the event-study results. Section 6 considers the mechanisms behind the effect. Section 7 concludes.

Section snippets

History of funding

The country’s first CHD was internally funded and opened outside of Louisville in Jefferson County, Kentucky. Beginning in 1911, the USPHS piloted a second department in Yakima County, Washington to contain an outbreak of typhoid fever. The Yakima department improved irrigation and water sanitation throughout the county, and subsequently reported a steep decline in typhoid-specific mortality. After that success, the USPHS took the CHD movement to the South and advanced water sanitation to

Event-study specification

To measure whether health departments were effective at improving rural mortality conditions, I exploit variation in the timing and the location of CHD entry. I track mortality before and after health department arrival using a flexible event-study design. The event study accounts for changes in mortality that could have spurred county governments to open a CHD.22

Healthinvestment data

For this study, I digitized the History of County Health Organizations (Ferrell et al., 1936), which tracks the rollout of CHDs from 1908 until 1933. The USPHS compiled the document as a central financial record of the rural health movement. This data source includes detailed annual data regarding the employed staff and the budget by source. The document maintains that it is the complete record of CHDs in the United States.28

Main results

Fig. 6 plots the coefficients from the event-study dummy variables from Eq. (1). Panel A shows the overall mortality rate in green and Panel B shows the infant mortality rate in blue. Each plotted line represents a weighted least squares estimation of the county-level mortality rate, with the overall rate weighted by the population and the infant mortality weighted by the number of births. The plotted points connected by solid lines represent the estimated coefficients on the county-level

Intensive margin effects

The main finding of this paper suggests that CHDs produced a robust and significant decline in infant mortality by two deaths per 1000 births. The channels by which CHDs improved infant mortality remain unclear from the baseline results. To illuminate how health departments improved infant outcomes, I modify my approach to a difference-in-differences estimation and test how infant mortality is affected by altering the intensity of CHD efforts. To test which CHD efforts were most productive, I

Conclusion

This study tracks one of the first and largest rural health programs in US history. The rollout of county health departments was expansive and included roughly 725 counties throughout 38 states. Despite the size and scope of the CHD program, there has been little effort to study the causal impact of the program. The present study is the first to investigate the health effects of CHD access and is one of the first papers to explore rural public health in the early twentieth century. With the

References (35)

  • R. Higgs

    Mortality in rural america, 1870–1920: estimates and conjectures

    Explor. Econ. Hist.

    (1973)
  • U.S.V. Statistics Mortality Statistics: Vital Statistics of the United States: 1890–1938 United States Bureau of the...
  • J.J. Wallis

    A history of the Property Tax in America

    (2001)
  • M. Alsan et al.

    Watersheds in Infant Mortality: The Role of Effective Water and Sewerage Infrastructure, 1880 to 1915

    Technical Report

    (2015)
  • M. Bailey et al.

    County-Level Natality and Mortality Data, 1915–2007 ICPSR 36603

    (2017)
  • M.J. Bailey et al.

    The war on poverty’s experiment in public medicine: community health centers and the mortality of older americans

    Am. Econ. Rev.

    (2015)
  • E.L. Bishop et al.

    Public health organization

    White House Conference on Child Health and Protection. Section II: Public Health Service and Administration. Committee on public health organization.

    (1932)
  • H. Bleakley

    Disease and development: evidence from hookworm eradication in the american south

    Q. J. Econ.

    (2007)
  • H. Bleakley

    Malaria eradication in the americas: a retrospective analysis of childhood exposure

    Am. Econ. J. Appl. Econ.

    (2010)
  • C. Bureau

    Wealth, Debt, and Taxation: 1912

    (1915)
  • C. Bureau

    Wealth, Public Debt, and Taxation: 1922

    (1924)
  • C. Bureau

    Financial Statistics of State and Local Governments: 1932

    (1935)
  • M.P. Center et al.

    IPUMS Restricted Complete Count Data: Version 1.0 [Machine-readable database]

    (2013)
  • C.V. Chapin

    Municipal Sanitation in the United States

    (1900)
  • C.V. Chapin

    A Report on State Public Health Work: Based on a Survey of State Boards of Health

    (1916)
  • T. Clasen et al.

    Interventions to improve water quality for preventing diarrhoea: systematic review and meta-analysis

    BMJ

    (2007)
  • D. Cutler et al.

    The role of public health improvements in health advances: the twentieth-century united states

    Demography

    (2005)
  • Cited by (21)

    • Public health departments and the mortality transition in Latin America: Evidence from Puerto Rico

      2023, Journal of Development Economics
      Citation Excerpt :

      Finally, this paper follows research on historical public health in the United States (Alsan and Goldin, 2019; Anderson et al., 2019, 2020, 2021; Clay et al., 2019; Ferrie and Troesken, 2008; Kitchens, 2013; Komisarow, 2017; Olmstead and Rhode, 2004) and Puerto Rico (Amador, 2015), including investments made by the Rockefeller Foundation (Bleakley, 2007), as well as in Cuba (Díaz-Briquets, 1981; McGuire and Frankel, 2005), another former Spanish colony in the Caribbean which was intimately linked to the United States before the communist revolution. This paper is most closely related to Hoehn-Velasco (2018) and Hoehn-Velasco and Wrigley-Field (2022), which demonstrate that local health departments led to a modest reduction in mortality in the rural, but not urban, United States in the early 20th century. The results of this paper are largely in line with Hoehn-Velasco (2018), as county health departments and public health units both significantly reduced infant mortality, while the effect on general mortality was less pronounced.

    • The association between educational attainment and longevity using individual-level data from the 1940 census

      2022, Journal of Health Economics
      Citation Excerpt :

      We then average these measures across the years when an individual was 6-17 years old. We collected data from various years of the Financial Characteristics of Cities (Ager et al., 2020; Curran, 1979; Hoehn-Velasco, 2018; Hoehn‐Velasco and Wrigley‐Field, 2022; Tabellini, 2020); the publication reported data for cities with population larger than 30,000 from 1905 (with 154 cities) to 1930 (with 310 cities). We focused on per capita sanitation expenses and per capita health expenses.

    • Quantifying the impact of aid to dependent children: An epidemiological framework<sup>⁎</sup>

      2020, Explorations in Economic History
      Citation Excerpt :

      Moehling and Thomasson (2014) also estimate that during the 1920s, the Sheppard−Towner Act accounted for 21 percent of the decline in infant mortality. Hoehn-Velasco (2018) finds that county health departments in the period 1910−1933 lowered infant mortality by 8 percent to 10 percent. For comparison on other health developments in U.S. history, Alsan and Goldin (2019) find that sanitation investments between 1880 and 1915 accounted for nearly 50 percent of the decline in infant mortality.

    View all citing articles on Scopus

    I am grateful to Claudia Olivetti, S Anukriti, and Robert Margo for their gracious feedback on this project. I thank Jacob Penglase, all the participants at the Harvard Economic History Lunch, the Boston University Economic History Reading Group, the 2016 NBER DAE Summer Institute, the 2016 EHA Annual Meeting, the 2017 WEAI Graduate Student Workshop, the Boston College Applied Micro Workshop, and the Boston College Dissertation Workshop for their valuable suggestions and comments.

    View full text