Automatic Milking Systems in Dairy Farming : A Frontier Eco-Efficiency Approach

Given the mounting importance of today’s agricultural policies to mitigate the environmental pressure from agricultural activities and to bolster the concurrent economic productivity, policymakers advocate that emerging farming technologies might be a viable response to this challenge. We draw on stochastic frontier analysis to examine the effect of automatic milking systems on eco-efficiency in dairy farming systems in Luxembourg. To address endogeneity problems, farmers’ propensity scores of adopting the technology are predicted based on farm accountancy data. In divergence with previous findings from the adoption literature, insights from machine learning models hint that the capital assets of machinery, material expenses, maintenance costs and investment subsidies are the key determinants in the adoption process. Our results indicate that technical and eco-efficiency scores diverge substantially within dairy farming systems. The main findings confirm the positive relationship between technology adoption, technical and eco-efficiency. Sensitivity analysis rules out that the empirical results are affected by the hidden hazard of unobserved heterogeneity