Micro-Spatial Matching and Migrant Health Inequalities in Early Twentieth-Century Madrid

Aitor Garcia Aguirre, Spanish National Research Council
Diego Ramiro, Consejo Superior de Investigaciones Científicas (CSIC)
Michel Oris, Consejo Superior de Investigaciones Científicas (csic)
Stanislao Mazzoni, Consejo Superior de Investigaciones Científicas (CSIC)
Dariya Ordanovich, IEGD-CCHS Spanish National Research Council (CSIC)
Julia de las Obras Loscertales, Spanish National Research Council

This paper examines health disparities between migrants and non-migrants in early twentieth-century Madrid by exploiting newly digitized individual-level data from the municipal register. Building on our previous work digitizing the Madrid padrón, we use a source that identifies residents with high spatial precision, including the street address, building, and, crucially, the floor of residence within the building. This unusually fine-grained residential information allows us to move beyond conventional comparisons based only on broad socioeconomic characteristics or neighborhood averages. Our empirical strategy relies on Propensity Score Matching to construct highly comparable groups of migrants and non-migrants. In addition to matching on standard sociodemographic characteristics such as age, sex, marital status, literacy or education proxies, household composition, and the presence of children, we exploit the micro-spatial structure of the data to compare individuals living in the same building and, where possible, on the same floor. This design substantially strengthens identification by holding constant not only observed social characteristics but also very local environmental and housing conditions that are often unmeasured in historical data. To address concerns related to salmon bias, whereby unhealthy migrants may leave the city and die elsewhere, we focus on sudden death outcomes, which are less likely to be shaped by selective return migration immediately prior to death. By combining rich historical microdata, fine-scale residential matching, and a strategy designed to reduce bias from selective out-migration, the paper offers new evidence on whether migrant status was associated with mortality risk in Madrid’s formative urban transition.

No extended abstract or paper available

 Presented in Session 165. Microdata Methods