Observed Crop-Yield Response Economic and Agro-climatic Factors in Austria - a Spatial Analysis

Abstract

The purpose of this study is to investigate empirically the effects of agro-climatic factors, economic incentives and farmers' land-choice on crop yield responses. We focus on soy bean and maize yields observed in Austrian municipalities over the past 14 years.Weather events, economic factors and technology are the driving forces behind the evolution of annual average yields. In the literature crop yields are frequently derived from crop models or from experimental stations or from reports of farmers (e.g. farm accountancy data network FADN). The data we use are obtained from the statistical services. Our approach therefore differs from studies on crop models or contributions based on data of experimental stations, as we focus on economic outcomes (yield of farmers) and explicitly take economic variables (e.g. market prices) into account.The exact relationship between factors affecting crop yields is not yet well understood due to the complexity of the measurement and definition of weather variables as well as the great variety of technological and managerial factors. Many variables influencing crop yield (weather or soil condition) exhibit a pronounced spatial structure, and therefore yields are spatially correlated as well. We account for this potential bias by applying spatial panel models in order to improve the efficiency of the models explaining variation of yields over time and space.Our results show that the effects of weather conditions are statistically significant and economically sizable. However, we can also show that non-spatial models tend to overestimate the parameter estimates of the respective explanatory variables. We conclude that it is advisable to explicitly control for spatial effects in crop yield response studies which are based on observed crop yields.