Contingent Inequalities: An Exploration of Health Inequalities in the U.S. and Canada

Speaker: Michael Wolfson

Date: Friday, January 20th from 10 AM – 12 PM

Location: HS 100 (Health Sciences Building, 155 College Street, Toronto ON)

Abstract

There has been considerable debate in the social epidemiology literature as to whether there is an empirical association between income inequality and population health (e.g. as measured by mortality rates) for population groups at various aggregated scales of observation (e.g. from counties to countries).  An intriguing observation is that at the metropolitan level, there is a strong correlation in the US, but none at all in Canada, hence a “contingent correlation”.  On the other hand, there is far less debate about the association at the individual level between socio-economic status (SES) and health; the individual-level SES gradient is ubiquitously observed as an empirical association, and widely accepted as a causal relationship.

In this analysis, we assess alternative theories that could account for the observed contingent correlation between income inequality and measures of population health focusing on Canada and the US.  To this end, we construct an abstract agent-based model (ABM), and explore various possible explanations including whether some mix of neighbourhood income segregation and parent-child transmission of social (dis)advantage can account for the observed patterns where higher income inequality is associated with poorer health.  The THIM (Theoretical Health Inequality Model) ABM incorporates plausible empirically-based but stylized relationships among health status, education, income, mortality rates and neighbourhood sorting / segregation.  THIM generates, among its many outputs, both life expectancies and health-adjusted life expectancies.

Alternative hypotheses are explored via a series of in silico experiments with the model.  THIM is able to reproduce the “contingent correlation” between population health and income inequality observed in Canada and the US.  But perhaps surprisingly, the simulation analysis suggests that higher neighbourhood income segregation alone is insufficient to generate the observed patterns.  Other factors like the structures of urban government appear more important.