Revealed preference land use models in Andalusia: integrating commercial and environmental variables

We investigate transitions among crop, grass and forest land uses in Andalusia (Spain) through an econometric model that uses individual land use decisions observed in the Corine Land Cover database as dependent variable, and net income, subsidies and biophysical and environmental factors as explanatory variables. Using multinomial logit models (fixed and random parameters), we identify both monetary and environmental and biophysical variables as significant in explaining land use decisions, but model performance significantly improves with the latter variables. We also identify the role of inertia but its effect is low as compared to the other variables. Based on the estimated model, we simulate the hypothetical variation of some explanatory variables in four scenarios up to 2022 and 2038. We find no large changes in land uses in most of the scenarios except for the one with an important reduction of subsidies and a decrease in protected areas. Climate change variables have a moderate impact as the ranges analyzed are narrow. The regional application shows the feasibility of a European wide application.