Crop farm efficiency and climate extremes in Germany

Abstract Using German crop farm data from the EU Farm Accountancy Data Network for 2004–2020, we investigate how weather and climate extremes interact with farm-level inefficiency, accounting for policy. We use a four-component stochastic production frontier model to disentangle farm heterogeneity from persistent and transient inefficiency, and explicitly address economic and econometric endogeneity. We select weather-related determinants of inefficiency using panel random forests. We find that losses associated with one climate extreme day range from 0.15 to 26.31 Euro/hectare. Water-related extremes appear less detrimental than temperature-related extremes. Our results suggest that efficiency can be underestimated if agro-climatic conditions are not acknowledged.