Empirical Likelihood-Based Inference for Poverty Measures with Relative Poverty Lines
提出一种基于经验似然的推断方法，用于比较共享同一相对贫困线（收入中位数的一定比例）的两个子群体的贫困程度，并通过蒙特卡洛模拟和加拿大家庭收入数据验证。
In this article, we propose an empirical likelihood-based method of inference for decomposable poverty measures utilizing poverty lines which are some fraction of the median of the underlying income distribution. Specifically, we focus on making poverty comparisons between two subgroups of the population which share the same poverty line. Our proposed method is assessed using a Monte Carlo simulation and is applied to some Canadian household income data.