Three Essays on Wealth and Income Inequality and Population Health in Global and Domestic Contexts Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 19, 2019
  • Brodish, Paul
    • Affiliation: College of Arts and Sciences, Department of Public Policy
  • Essay 1 investigates the contextual effect of community-level wealth inequality on HIV serostatus using DHS data pooled from six sub-Saharan African countries. Multilevel logistic regressions relate the binary dependent variable HIV positive serostatus and two weighted aggregate predictors generated from the DHS Wealth Index. A 1-point increase in the cluster-level Gini coefficient and cluster-level wealth ratio is associated with a 2.35 and 1.3 times increased likelihood of being HIV positive, respectively, controlling for individual-level demographic predictors, with larger effects in males. The association is partially mediated by more extramarital partners. Essay 2 uses multiple cohorts of the National Longitudinal Mortality Study (NLMS) to quantify the absolute income effect on mortality in the United States. Multivariate logistic regressions assess the impact on mortality rate ratios of two hypothetical interventions: lifting everyone living on an equivalized household income at or below the U.S. poverty line in 2000 to the income category just above, and shifting everyone’s income by 10–40% to the mean household income, equivalent to reducing the Gini coefficient by the same percentage. The absolute income effect is in the range of a three to four percent reduction in mortality for a 10% reduction in the Gini coefficient. Larger mortality reductions result from larger reductions in the Gini, but with diminishing returns. Inequalities in estimated mortality rates are reduced by a larger percentage than overall estimated mortality rates under the same counterfactual redistributions. Essay 3 uses multiple NLMS cohorts and multilevel Cox proportional hazards regressions to estimate the contextual effect of state-level income inequality on premature mortality in the United States. It uses six different measures of state income inequality, controls for inflation-adjusted, equivalized family income, and adjusts for eight individual-level socioeconomic and demographic variables, and for state-level percentage black and percentage in poverty. The contextual effect varies markedly by inequality measure, gender, and regression method. Effect sizes are generally in the range of a one to five percent increase in the likelihood of premature death for a one standard deviation increase in income inequality. The contextual effect may cause a sizeable number of premature deaths, especially among males.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Meier, Ben
  • Suchindran, Chirayath
  • Durrance, Christine
  • Perreira, Krista
  • Singh, Kavita
  • Doctor of Philosophy
Degree granting institution
  • University of North Carolina at Chapel Hill Graduate School
Graduation year
  • 2015
Place of publication
  • Chapel Hill, NC
  • There are no restrictions to this item.

This work has no parents.