Resource productivity and environmental degradation in EU-27 countries: context of material footprint.

This study explores the relationship between the resource productivity and environmental degradation in European Union-27 countries. This study tests this relationship in context of high, moderate, and low material footprint sub-samples; these samples are formed utilizing the expectation-maximization machine learning algorithm. Using the panel data set of EU-27 countries from 2000 to 2020, linear and non-linear autoregressive distributed lag (ARDL) are applied for the symmetric and asymmetric evidence and to test environmental Kuznets curve (EKC), linear ARDL with the quadratic function is included. Results of the symmetric relationship find evidence of resource productivity's impact on the environmental degradation. In full sample of EU-27, both symmetric and asymmetric methods show that the short run and long run increase of resource productivity lower the environmental degradation. Only long run asymmetric relationship in high material footprint subsamples supports that the resource productivity controls environmental degradation. Results of moderate material footprint sub-sample are mixed. However, low material footprint countries show that resource productivity in long run controls the environmental degradation in symmetry and only positive resource controls productivity in short run in asymmetric relationship. The reason for mixed results is the quadratic nature of sub-samples. EKC hypothesis is validated in moderate and low material footprint sub-samples. This research has many policy implications.