Nonparametric testing for socio-economic health inequalities: an application to covid-19 deaths

This paper develops a non-parametric test statistic for assessing the equality of the health concentration curve and the 45-degree line, which represents perfect equality. Specifically, the test compares the health concentration curve, C(p,Y), with the line p for p∈(0,1) using an L2 metric. In addition, we propose alternative tests for detecting pro-rich and pro-poor health inequalities based on metrics other than L2. We employ a non-parametric estimator for C(p,Y) and establish the asymptotic properties of the tests, including their size and power. A Monte Carlo simulation study demonstrates the strong finite-sample performance of the tests, with robust size and power across various data-generating processes. Finally, to illustrate the practical relevance of the proposed methods, we apply them to U.S. data to test for inequality in COVID-19 deaths across and within U.S. states, comparing the health concentration curves of these outcomes to the 45-degree line. Our empirical findings reveal varying levels of health inequality across states, with some states showing pro-rich inequality where wealthier individuals experience lower death rates. The results highlight the critical role of socioeconomic disparities in shaping the impact of COVID-19, particularly in states with high poverty rates and lower incomes.