Motivated by Rajan’s work, we propose that income inequality and systemic banking crises have a nonlinear nexus. In addition to examining the linear “Rajan effect,” we propose a “Kuznets effect” based on an assumption that income inequality has a nonlinear impact on growth, conditional on the stage of economic development, and thus plays a nonlinear role in modeling crises. We test the existence of this nexus using a sample covering 172 economies for a period of fifty years. We confirm that the relationship is U-shaped and identify the threshold level of income inequality that is beneficial for financial stability. Additionally, the U-test and area under the receiver operating characteristic (AUROC) statistics confirm that nonlinearity, and our model has good predictive performance of forecasting. Furthermore, the results of our panel regressions are consistent and are robust to several tests. We then identify the Rajan and Kuznets effects using a two-step test. We conclude that the impact of income inequality on the occurrence of systemic banking crises is U-shaped.

受拉詹工作的启发，我们提出收入不平等和系统性银行危机之间存在非线性关系。除了检验线性“拉詹效应”之外，我们还提出了“库兹涅茨效应”，该效应基于这样的假设：收入不平等对增长具有非线性影响，以经济发展阶段为条件，因此在危机建模中发挥非线性作用。我们使用涵盖 172 个经济体、为期 50 年的样本来检验这种联系的存在性。我们确认这种关系是 U 形的，并确定了有利于金融稳定的收入不平等的阈值水平。此外，U检验和接收器操作特性下的面积（AUROC）统计证实了非线性，并且我们的模型具有良好的预测性能。此外，我们的面板回归结果是一致的，并且对于多项测试来说都是稳健的。然后，我们使用两步测试来识别拉詹和库兹涅茨效应。我们的结论是，收入不平等对系统性银行危机发生的影响呈U形。