Validating approximate slope homogeneity in large panels
大面板中的近似斜率同质性检验
In this paper, we introduce new inference methods for slope homogeneity in large regression panels. While most existing tests are developed for the hypothesis of slope homogeneity (equality of all individual slopes), we propose to test the more realistic relaxation of approximate slope homogeneity (similarity of all slopes). We present new test statistics for dense and sparse alternatives to approximate homogeneity. In the dense setting, the main focus of this paper, we…