Exploring Social Disorganization: A Multilevel Analysis of Child Mortality in Metropolitan São Paulo, Brazil

Kuniko Chijiwa, University of Florida
John D. Reitzel, Illinois State University

In this study we investigate the effects of local environmental characteristics as indicators of collective well-being or collective disadvantage in predicting child mortality within a social disorganization framework. Child mortality is employed as an indicator of an individual’s ultimate health status in Metropolitan São Paulo, the highly unequal and rapidly changing metropolitan Brazilian society. We use multilevel negative binomial regression model to predict individual mother’s number of children lost, which are nested in the geographic distribution of higher-level (aggregate) structural factors. Our investigation focuses on how structural variables that are indicative of ineffectual community organization, which earlier studies on health and communities found to be important, contribute in predicting child life expectancy across geographic domains. We consider how the effects of individual-level deprivations on child mortality could be mitigated or enhanced by community-level characteristics, while also developing a framework for understanding the mechanisms that produce healthier communities.