Poverty, Religious Differences, and Child Mortality in the Early Twentieth Century: The Case of Dublin

Abstract

Across many cities in the early twentieth century, one in five children died before their fifth birthday. There is much we do not know about how infant and child mortality was reduced or why it declined at different rates across populations. This article investigates mortality using data from 13,247 families in Dublin City in the 1900s with a novel approach that incorporates geographic information systems, spatially derived predictors, and multilevel modeling. In the early twentieth century, Dublin had one of the highest early-age mortality rates in the British Empire. Whereas experts attributed the death of young children to the unhygienic behaviors of indigenous Roman Catholics, others made claims of a social injustice rooted in economic inequality and the indifference of public authorities toward the health of the lower classes. This article finds that high Catholic mortality was mainly driven by poverty and the conditions engendered by residential segregation. Low mortality rates among Dublin's small Jewish population are not easily explained by location or economic characteristics.

在二十世纪早期的诸多城市中, 每五个孩童就有一个在五岁前死亡。我们对于婴儿与孩童的死亡率如何降低, 以及在各人口中为何以不同的速率降低所知甚少。本文运用都柏林市在 1900 年代的 13,247 个家庭的数据, 并採用纳入地理信息系统、空间推倒的预测以及多层级模式化之创新方法来探讨死亡率。在二十世纪初期, 都柏林在大英帝国中, 是早夭率最高的地方之一。有些专家将年轻孩童的死亡归咎于当地罗马天主教徒不卫生的行为, 另有人宣称此为植基于经济不平等的社会不公, 以及公共单位对下层阶级的健康置之不闻所导致。本文发现, 天主教徒的高死亡率, 主要是由贫穷和居住隔离引发的状况所导致。都柏林的少数犹太人口的低死亡率, 则不易由地点或经济特徵进行解释。

En numerosas ciudades de principios del siglo XX moría uno de cada cinco niños antes de cumplir los cinco años. No es mucho lo que sabemos sobre cómo pudo reducirse la mortalidad de bebés y niños, o por qué aquella declinó con tasas diferentes a través de las poblaciones. Este artículo investiga la mortalidad con el uso de datos de 13.247 familias de la ciudad de Dublín en los años 1900, mediante un enfoque novedoso que incorpora sistemas de información geográfica, predictores derivados espacialmente y modelado de nivel múltiple. A principio del siglo XX, Dublín tenía una de las más altas tasas de mortalidad del Imperio Británico para las edades tempranas. Mientras los expertos atribuían la muerte de niños de poca edad a los comportamientos antihigiénicos practicados por los nativos de religión católica romana, otros creían en la responsabilidad de una injusticia social arraigada en la desigualdad económica y en la indiferencia de las autoridades públicas hacia la salud de las clases más bajas. Este artículo encuentra que la alta mortalidad entre católicos estuvo determinada principalmente por la pobreza y las condiciones generadas por la segregación residencial. Las bajas tasas de mortalidad entre la pequeña población judía de Dublín no se explican fácilmente por localización o características económicas.

Key Words:

• child mortality
• historical GIS
• multilevel modeling
• segregation and health
• spatial demography

关键词：:

• 儿童死亡率
• 历史
• 多层级模式化
• 隔离与健康
• 空间人口学。

Palabras clave:

• mortalidad infantil
• SIG históricos
• modelado de nivel múltiple
• segregación y salud
• demografía espacial

Acknowledgments

This article has benefited from comments and encouragement from Leah Boustan, Cormac Ó Gráda, Kelvyn Jones, David Rigby, Jamie Goodwin-White, Michael Storper, Roger Waldinger, Gerald Mills, Helga Leitner, Ashley Fent, Alice Reid, Sara Hughes, Mia Bennett, Peter Catron, John Sullivan, Susan Watkins, Eric Shepherd, Adrian Connor, Geert Molenberghs, and Marc Swyngedouw. Alan Fernihough, Tine Ningal, and Catriona Crowe very generously shared data with me and three anonymous reviewers provided excellent feedback and suggestions.

Funding

This research was supported by the Denis Phelan Scholarship, the NUI Travelling Studentship, the Population Specialty Group of the American Association of Geographers, and a Roter Research Travel Grant from the Center of Jewish Studies at UCLA.

Notes

1. African Americans and other minority populations in the United States have infant mortality rates up to double that of whites.

2. The World Health Organization reports present-day under-five mortality rates of seventy-six per thousand for low-income countries and seven per thousand for high-income countries.

3. The term holocaust, as used here, predates World War I and was not a reference to the experience of the Jewish people during World War II. In this context, it was used to describe a more general political negligence toward public health.

4. Thornton and Olson (2011) are an important exception to this.

5. Each census form was completed and signed by the head of household.

6. A separate analysis in the online Appendix indicates that classifying couples differently would not change these results. See http://media.wix.com/ugd/8b202c_e78a03f5f6794f419151320f054dfdd6.pdf.

7. The HISCAM index refers to the relative position within the stratification structure. Values approaching 100 correspond to the most prestigious occupations and values approaching 0 correspond to the lowest. Occupations on the lower end of this scale include miners and general laborers, whereas doctors, chemists, dentists, and professors have HISCAM scores of around seventy and up.

8. Ó Gráda (2004, 2006) previously found low mortality among Dublin's Jewish community.

9. This approach closely follows the procedure developed and discussed in Appendix A of Garrett et al. (2001) and Garrett, Reid, and Szreter (2012).

10. These categories of low, medium, and high refer to cuts at scores of thirty-six, forty-five, and fifty-three, respectively.

11. In the online Appendix, I show that streets with higher levels of mortality were less likely to be located on historic maps.

12. Grimshaw and Cameron (1888) commissioned the Royal Engineers' Department to map these typhoid fatalities.

13. This surface was estimated using an exact interpolation method (ordinary kriging) with a spherical model variogram with a search distance of 300 m. See point map in online Appendix for comparison: http://media.wix.com/ugd/8b202c_e78a03f5f6794f419151320f054dfdd6.pdf.

Additional information

Notes on contributors

Dylan Shane Connor

DYLAN SHANE CONNOR is a PhD Candidate in the Department of Geography, University of California, Los Angeles, Los Angeles, CA 90095. E-mail: [email protected]. His research interests include urban and regional development and inequality, spatial demography, and economic history.