Senior Director of Organizational Development and Training, Institute for Health Metrics and Evaluation, University of Washington
Dissertation Title: "Health Inequality: Definition, Measurement and Determinants"
Health inequality is defined as differences in health status across individuals in a population. I present a conceptual framework for the measurement of health inequality, influenced by the long tradition of measurement of income inequality. It is suggested that the ideal for a population would be to achieve equality in individual healthy life expectancy. Healthy life expectancy is the number of years in full health that an individual born today can expect to live. To summarize the distribution of healthy life expectancy a measure of inequality has to be selected. The issues surrounding the selection of a measure are normative. Rather than arbitrarily selecting a measure, an Internet survey of health professionals was conducted to elicit preferences on the normative issues involved in selecting a measure of inequality. The results of this survey are presented and used to select an index of inequality. The inequality measurement framework is then operationalized in the measurement of inequality in child survival in 50 developing countries. The distribution of probability of death across children in each country is estimated using two models: an extended-beta binomial and a random effects logit model. The two models are compared and their results, which do not differ statistically, are shown. Finally, an analysis is undertaken of the effect of key determinants on health inequality, particularly the effect of income inequality and education inequality. Other determinants, such as level of health expenditure, poverty, and urban/rural residence, are also studied. This analysis finds that the determinants most often cited in the literature do not explain the majority of the differences in health inequality across developing countries. The epidemic of HIV/AIDS is shown to have an effect on health inequality, potentially by creating groups of high-risk individuals in populations. It is also suggested that some of the variation across countries could be due to differences in the distribution of health services and resources between urban and rural areas. More data should be collected at the local level, so that the effect of other variables, particularly ones pertaining to health systems, can be analyzed.