Novel symmetry tests in regression models based on Gini mean difference

Abstract

This article proposes two new tests of symmetry based on the Gini mean difference. The symmetry hypothesis of the disturbance in a linear regression model around zero was analyzed using the proposed tests. A Monte Carlo simulation study shows that the tests have good size and power properties for sample sizes as small as 30. The symmetry of the error term in a cross county model of Gini index as a measure of income inequality and consumer price inflation was studied by the proposed tests.