Joint Discussion Paper Series in Economics 16-2017
Based on data from a representative survey among more than 2200 households, this paper empirically examines the agreement to the German energy transition in total and to six single energy policy measures, which are components of this challenging national policy approach. Our micro-econometric analysis with uni- and multivariate binary and ordered probit models reveals that both economic calculus and personal and social values are relevant for this agreement. An expected future electricity price increase due to the energy transition (which especially incorporates the price expectations in the hypothetical case that the measures of the energy transition are withdrawn) is significantly negatively correlated with the agreement to the two core measures of the energy transition, namely the nuclear phase-out and the financial support of the expansion of renewable energies. While other economic variables like income and energy expenditures also have some significant effects, our estimation results especially reveal that political identification and other personal values are at least equally relevant. For example, an overall left-green orientation is significantly positively correlated with the agreement to the energy transition in total and especially to the aforementioned core measures of the energy transition, which are also significantly positively affected by strong environmental values. Our econometric analysis suggests that studies that only include economic variables or only include personal and social values in order to explain the acceptance of policy measures can lead to biased estimation results and thus distorted conclusions. For policy makers our study additionally identifies important skeptical population groups which might be addressed in order to increase the acceptance of climate and energy policy measures like the German energy transition.
climate and energy policy measures energy transition energy cost expectations political identification environmental values uni- and multivariate binary and ordered probit models