Preliminary Programme

The role of social class has been a central topic in research on inequalities in early life mortality. Class is a vague proxy for a number of variables that can affect children’s health: nutrition, knowledge about childcare, access to health care and physical environment. On the question of ‘class versus place’, scholars have attempted to disentangle the effect of class from that of unhealthy environments. This paper contributes to this discussion by examining the social and spatial distribution of early life mortality risk in late nineteenth century Copenhagen.
In 1885, Copenhagen was expanding rapidly and the negative health effects of industrialization were at their peak. Housing quality varied dramatically across the city, and I hypothesize that child mortality risk did so as well. The circumstances for investigating this are favorable, as the statistical office of Copenhagen collected and published a wealth of data on housing and living conditions along with the city’s census counts.
My analysis draws on a dataset covering most of the children buried in Copenhagen in 1885, which has been linked to the 1885 census. The individual-level nature of the data allows for analysis of a number of family-level variables. Social class is measured through a combination of HISCLASS and a class scheme specific to Danish society, which was published in 1883. I employ a multilevel modeling approach to distinguish the effect of class and other family-level variables from environmental (neighborhood) effects. (Show less)

The aim of this study is to examine what determined cause-specific infant and child mortality in Amsterdam (1856-1924) to be able to understand how individual, familial and societal factors played a role in determining who died and who survived until age five. In addition, changes over time regarding the impact of specific determinants of child mortality are studied as well. This study is only possible because of unique individual-level cause-of-death registers, which are rare in historical settings. Combined with birth and death certificates a longitudinal dataset that includes the birth of 110.910 children, of whom 27.325 died before age five, is created. Cox proportional hazards models are used to study what determined cause-specific hazards of dying from several categories of causes, such as airborne infectious diseases, waterborne infectious diseases and perinatal causes, to provide new insights in how inequalities in health were shaped. (Show less)

Our understanding of historical environmental health risks are limited by the quality and availability of not only micro-data on individuals based on complete residential histories, but accompanying built and social environmental data that can be used to model such risk across the life course. This project integrates and addresses all of these needs, leveraging a rich spatio-temporal dataset that includes a retrospective cohort of nearly 40,000 individuals across 50 years of data during the early 20th century in the former mining region of Northern Michigan (USA). We link this cohort to built and social environmental data to identify chronic disease risk, and apply this to create a replicable methodological approach for studying these links. This work is a part of a broader project on the influence of the built and social environment on behavior and life-course health.

Our team takes advantage of past work that a) leveraged HGIS analytical tools to uncover hidden patterns in housing inequality and b) identified key methodological shortfalls of other work linking GIS and health. In contrast to research limited by aggregate health data, poor quality built and social environment data, discrepancies in sources and missing data, or limited time horizons, our available data is expansive and detailed and in both temporal and spatial dimensions, offering unparalleled opportunity for historical geospatial inquiry around environmental health issues.

We also explore the ways that various levels of aggregation in the built and social environment influence estimation of disease risk by running sensitivity analyses. These exploratory spatial analyses have afforded us the ability to make recommendations for future research using datasets at a variety of spatio-temporal scales. Our work links potential and actual exposures at a deeper level than would be available for many contemporary datasets, leverages historical geospatial expertise to answer these questions, and uses a novel methodology for matching residential information to create lifetime residential histories. This work will empower future research on the built and social environment focused at the micro-scale by illuminating the potential of existing population micro-datasets and how to integrate them. Other researchers will also have a methodological and spatio-analytical template on which to build future inquiries. (Show less)

For several historians, the decline in mortality rates in sub-Saharan Africa has been one of the major factors contributing to the substantial rise in sub-Saharan population growth in the 20th century (Caldwell, 1985). The extent to which this argument holds has, however, received little attention from quantitatively-oriented historical demographers. Studies quantifying demographic rates primarily focus on the post-1960 period. Recent research using individual-level data has started to shed more light on changes in pre-1960 mortality patterns in some regions of east and southern Africa (see Walters, 2021). Yet, French-speaking West Africa during the colonial era remains understudied (with the exception of Lacombe, 1970), despite the long-lasting legacies of colonialism on demographic regimes (Walters, 2021). In this paper, I will attempt to contribute to closing this gap in West African historical demography, using individual-level data from Senegalese vital registration statistics for the period 1880-1913 to study developments in infant mortality. Limited to the cities along the Senegalese coast, the relatively complete vital event sources on births and mortality contain a sufficient number of certificates to, alongside reconstructing infant mortality rates, examine the role of social class and race. To limit potential biases caused by out-migration, I limit the analysis to deaths occurring in infancy, which is a strong indicator of the overall mortality regime, despite focusing on deaths occurring in the first year (Volk and Atkinson, 2013). By studying four cities of colonial Senegal, Saint Louis (northwest), Dakar, Rufisque (west), and Ziguinchor (southwest), I will try to quantify whether regional, social class, or ethnic differences (European versus African) explain differences in mortality.

To analyze the developments in infant mortality, I focus on certain key years at approximately 10-year intervals. Birth certificates will be selected for 1880, 1891, 1904, and 1912. Employing a life-table analysis, registered deaths of the infants will be linked to the birth certificates to gain an estimation of the infant mortality rate. The years have been selected because they allow for a comparison with data coming from the colonial administrative censuses of 1891 and 1904. Once the risk of dying in the first year of life has been estimated, the data will be analyzed using a Cox proportional hazards regression to investigate which factors contribute to differences and changes in urban mortality. In addition, I will examine whether urban mortality rates were reduced in the early colonial period, as some have claimed occurred in urban Nigeria (Caldwell, 1985).


Caldwell, John C. 1985. “The Social Repercussions of Colonial Rule: Demographic Aspects.” In General History of Africa VII: Africa Under Colonial Domination, 1880-1935, edited by A. Adu Boahen, 458-486. London: Heinemann Educational Books.

Lacombe, Bernard. 1970. Fakao (Sénégal) Dépouillement de Registres Paroissiaux et Enquête Démographique Rétrospective : Méthodologie et résultats. Paris: ORSTROM.

Volk, Anthony A. and Jeremy A. Atkinson. 2013. “Infant and Child Death in the Human Environment of Evolutionary Adaptation.” Evolution and Human Behavior 34, no. 3: 182-192.

Walters, Sarah. 2021. “African Population History: Contributions of Moral Demography. The Journal of African History 62, no. 1: 183-200. (Show less)