Use of extended ESQ estimaters with residuals in investigation of stationariness of time series

The aim of this study is to discuss the use of the Residuals Augmented Least Squares method proposed by Im (1996) in time series stationarity tests. These tests will also be compared with conventional stationarity tests. The method of Residuals Augmented Least Squares is an alternative estimation method used when one of the assumptions of the Least Squares method, that the errors are normally distributed, is not met. The application study to compare the conventional and residuals-extended unit root tests is analysed with the example of GINI coefficient, which is an indicator of income inequality. For 62 countries, a data set covering different time dimensions is used. In this study, the process of each country over time is analysed with different unit root tests with and without taking into account the non-normal distribution information and the results are presented in a report. The results of the study show that in countries where the coefficient of income inequality is stationary, exogenous shocks have temporary effects and a return to equilibrium is achieved. These countries are more resistant to economic and political fluctuations. On the other hand, income inequality is more sensitive to exogenous shocks in countries with unit-rooted series and these shocks may have permanent effects in the long run.