Dynamic Discrete Choice Estimation of Agricultural Land Use

Résumé

I develop a new framework for analyzing land use change with dynamically optimizing landowners. My empirical approach allows for unobservable heterogeneity and avoids the burden of explicitly modeling the evolution of market-level state variables like input and output prices. Using a rich new data set on land use in the United States, I estimate a relatively large long-run cropland-price elasticity of 0.3. Compared to static estimates using the same data, my dynamic estimates suggest that biofuels production leads to dramatically more land use change and substantially smaller price increases in the long run.