Among the various methods adopted to compare health inequality, Makdissi and Yazbeck (J Health Econ 34:84–95, 2014) developed positional stochastic dominance conditions to identify an ordering. To reach a conclusion, their rules require that the (generalized) health concentration curve of the dominant distribution lie above that of the dominated one. However, it is frequently observed in practice that these curves intersect. Our paper proposes new criteria to cope with this problem by allowing a relatively small violation of the condition proposed by Makdissi and Yazbeck (2014). We characterize our conditions by linking them with some ethical constraints of the weight functions. We further use individual data for Côte d’Ivoire and Guinea from the Demographic and Health Survey to demonstrate the usefulness of our newly-proposed method.

在用于比较健康不平等的各种方法中，Makdissi 和 Yazbeck (J Health Econ 34:84–95, 2014) 开发了位置随机优势条件来确定排序。为了得出结论，他们的规则要求主导分布的（广义）健康浓度曲线高于主导分布。然而，在实践中经常观察到这些曲线相交。我们的论文通过允许相对较小的违反 Makdissi 和 Yazbeck (2014) 提出的条件，提出了解决这个问题的新标准。我们通过将我们的条件与权重函数的一些道德约束联系起来来表征我们的条件。我们进一步使用来自人口与健康调查的科特迪瓦和几内亚的个人数据来证明我们新提出的方法的有用性。