This study re-examined the role of geographic scale in measuring income inequality and testing the income inequality hypothesis (IIH) as an explanation of health disparities. We merged Behavioral Risk Factor Surveillance System (BRFSS) 2000 data with income inequality indices constructed at different geographic scales to test the association between income inequality and four different health indicators, i.e., self-assessed health status as a morbidity measure, vaccination against influenza as a measure of use of preventive healthcare, having any kind of health insurance as a measure of access, and obesity as a modifiable health risk factor measure. Multilevel models are used in our regression of the health indicators on measures of income inequalities and control variables. Our analysis suggests that because income inequality is a contextual variable, income inequalities measured at different geographic scales have different interpretations and relate to societal characteristics at different levels. Therefore, a rejection of the IIH at one level does not necessarily negate the possibility that income inequality affects health at another level. Assessment across a variety of scales is needed to have a comprehensive picture of the IIH in any given study. Empirical results also show that whether the IIH holds could depend on the sex group examined and the health indicator used, which implies different mechanisms of IIH exist for different sex groups and health indicators, in addition to the geographic scale. The role of geographic scale should be more rigorously considered in social determinants of health research.
Published by Elsevier Ltd.