A rethink on measuring health inequalities using the Gini Coefficient

Objective- We show that a standardized Gini coefficient that takes into account the feasible range of health inequality for a given health attribute is a better instrument than the normal Gini coefficient for quantifying inter-individual health inequality. Methods- The standardized Gini coefficient is equal to the normal Gini coefficient divided by the maximal attainable Gini coefficient, which is computed based on the maximal level of a health attribute an individual could achieve. Both the old and new coefficients are used to estimate the lifespan inequality of 185 countries for year 1990, 2000 and 2006, respectively. The results are then compared both across countries and over time. Findings- Firstly, the standardized Gini coefficient can still be related to the Lorenz curve. Secondly, changes in standardized Gini coefficients can be decomposed into respectively the change in the distribution of health outcomes and the change in the average health outcomes. Thirdly, the standardized Gini coefficient provides richer information and often gives different conclusions regarding health inequality in individual countries as well as country ranking, as compared to the normal Gini coefficient. Conclusion- Accounting for the maximal level of health attribute an individual could achieve is important when measuring health inequality. The proposed standardized Gini coefficient can provide more accurate information regarding the actual level of health inequality in a society than the normal Gini coefficient