Martin Dribe, Lund University
Finn Hedefalk, Lund University
The turn of the twentieth century was a period of rapid decline in child mortality and a time of industrialization and urbanization in Sweden. It was also a period of growing social disparities in child mortality. The inequality in child survival was connected to a range of factors, including access to water- and sanitation, housing conditions, infant care, and possibly nutrition. In addition to parental SES, neighborhood conditions likely impacted child health through several of these factors. In this paper, we study the importance of the social neighborhood for infant and child mortality in an industrializing Swedish town (1890-1939). We use individual-level socioeconomic and demographic data from population registers which have been geocoded at the block level. Using a k-nearest neighbor approach we construct dynamic measures of neighborhood conditions from information on social class. In addition, we adjust for physical aspects of the neighborhoods, such as proximity to major industries. We use spatial survival analysis to estimate the association between cumulative social neighborhood variables and the risk of infant and child death, adjusting for parental socioeconomic status. In addition to total mortality, we also analyze cause-specific patterns of broader diagnosis groups, for example, airborne diseases and food- and waterborne diseases.
Keywords: Mortality and Longevity, Neighbourhood/contextual effect analysis, Geo-referenced/geo-coded data, Longitudinal studies