Pro-environmental behavior and its determinants: evidence for Mato Grosso do Sul

This article investigates the importance of individual characteristics, including a proxy for testosterone levels (using the biological marker 2D:4D ratio as proxy), to explain students’ pro-environmental behavior. We employ machine learning modeling from artificial neural networks. We elaborate an environmental score based on 13 questions related to the environmental impact generated by the individuals. We analyzed which profile variables are most important to explain pro-environmental behavior. The results are in line with the literature, which proves the importance of the 2D:4D ratio as a predictor of pro-environmental behavior (testosterone levels), gender, and age. These results suggest evidence for elaborating public policies to preserve and reduce environmental degradation based on individuals’ characteristics and biological markers.