Improving efficiency assessments using additive data envelopment analysis models: an application to contrasting dairy farming systems

Authors

• Andreas Diomedes Soteriades Scotland's Rural College, Future Farming Systems Group, Edinburgh
• Philippe Faverdin INRA, Domaine de la Prise, Saint-Gilles INRA-Agrocampus Ouest PEGASE, Physiologie, Environnement et Génétique pour l'Animal et les Systèmes d'Elevage, Rennes
• Margaret March Scotland's Rural College, Future Farming Systems Group, Dumfries
• Alistair William Stott Scotland's Rural College, Future Farming Systems Group, Edinburgh

DOI: https://doi.org/10.23986/afsci.49446

Keywords:

efficiency indicators, slacks-based measure (SBM), undesirable outputs, savings potentials, slack shares, experimental dairy farm data

Abstract

Applying holistic indicators to assess dairy farm efficiency is essential for sustainable milk production. Data Envelopment Analysis (DEA) has been instrumental for the calculation of such indicators. However, ‘additive’ DEA models have been rarely used in dairy research. This study presented an additive model known as slacks-based measure (SBM) of efficiency and its advantages over DEA models used in most past dairy studies. First, SBM incorporates undesirable outputs as actual outputs of the production process. Second, it identifies the main production factors causing inefficiency. Third, these factors can be ‘priced’ to estimate the cost of inefficiency. The value of SBM for efficiency analyses was demonstrated with a comparison of four contrasting dairy management systems in terms of technical and environmental efficiency. These systems were part of a multiple-year breeding and feeding systems experiment (two genetic lines: select vs. control; and two feeding strategies: high forage vs. low forage, where the latter involved a higher proportion of concentrated feeds) where detailed data were collected to strict protocols. The select genetic herd was more technically and environmentally efficient than the control herd, regardless of feeding strategy. However, the efficiency performance of the select herd was more volatile from year to year than that of the control herd. Overall, technical and environmental efficiency were strongly and positively correlated, suggesting that when technically efficient, the four systems were also efficient in terms of undesirable output reduction. Detailed data such as those used in this study are increasingly becoming available for commercial herds through precision farming. Therefore, the methods presented in this study are growing in importance.