Asymptotic Versus Bootstrap Inference for Inequality Indices of the Cumulative Distribution Function

Abstract

We examine the performance of asymptotic inference as well as bootstrap tests for the Alphabeta and Kobus–Miłoś family of inequality indices for ordered response data. We use Monte Carlo experiments to compare the empirical size and statistical power of asymptotic inference and the Studentized bootstrap test. In a broad variety of settings, both tests are found to have similar rejection probabilities of true null hypotheses, and similar power. Nonetheless, the asymptotic test remains correctly sized in the presence of certain types of severe class imbalances exhibiting very low or very high levels of inequality, whereas the bootstrap test becomes somewhat oversized in these extreme settings.

Inferring inequality: Testing for median-preserving spreads in ordinal data

“…As Davidson and Duclos (2013), we adopt a likelihood ratio statistic combined with bootstrap inference, but we additionally consider both a Z statistic and asymptotic inference. Our work is also related to Abul Naga and Stapenhurst (2015) and Abul Naga et al (2020): while they perform inference on a random variable derived from a particular class of indices consistent with the MPS ordering, we perform inference on the binary outcome given by the partial ordering itself.…”