A validation forecast using robust estimators into environmental application

Investigating the effect of energy consumption and economic growth on carbon dioxide emissions has received much effort due to the global environmental issues. Multiple methods are used to explain that relationship, but findings are conflicting, that might due to inaccurate chosen statistical methods and the frequent presence of outliers in the data. The main objective of this study is to shed light on obtaining the best model in detecting that relationship using various robust estimators (M, Median, S and MM-estimator) against OLS in the presence of different types of outliers in the panel data, then, models are evaluated by using train-test forecasting approach. The panel data include 29 countries, divided into two groups based on the economic level, 17 developed countries versus 12 developing countries from 1960 to 2008. The main findings support that the robust estimators have better properties than the OLS estimator when the dataset has outliers. The M-estimator is the best robust estimator which could fit the data in the presence of different types of outliers. The energy consumption and economic growth have negative and positive relationship with CO2 emissions, respectively. Moreover, developing countries affect CO2 emissions more than the developed countries. In conclusion, energy consumption contributes to higher environmental degradation particularly in CO2 emissions, thus it is recommended to policy makers to consider it in their policies for a better future.