Modelling Spatio-Temporal Dynamics in Multi-Output Stochastic Frontiers for the European Agribusiness Industry

Abstract This paper introduces a maximum likelihood estimation approach for multi-output stochastic frontier models with simultaneous effects, cross-equations and temporal and spatial components to analyse the aggregate production of agricultural-related industries in European OECD countries in the period 1996–2019. The result is a comprehensive empirical assessment of input, inefficiency and shocks-related spillovers between two of the main sectors in the agribusiness industry, i.e. agriculture and food and beverage manufacturing. Our findings reveal the existence of positive spillovers in the short term from both efficiency and innovation shocks, as well as input variations, which modify into competitive pressures in the long run. Insights from this study allow policymakers to evaluate how the productive performance of each sector in the agribusiness industry influences the production output of the other, both within and between countries. Additionally, it allows for the inspection of various transmission mechanisms and contagion phenomena, aiding in the design of international support plans for the entire agribusiness industry.