This paper applies a parametric approach to estimate technical and scale (in)efficiencies using input and output data at the level of 850 individual Colombian coffee-farms. Different Stochastic Production Frontier functions are estimated using a two-step procedure that corrects the endogeneity that has been ignored in previous works, leading to more reliable (i.e. unbiased and consistent) results. We conclude that small and medium coffee farmers are technically inefficient and exhibit increasing returns to scale, whereas large coffee farmers are close to being quasi-technically efficient and exhibit decreasing returns to scale. The corrected-for-endogeneity estimation also indicates that small and medium-sized units must prioritise primarily the land factor, whereas large farms should concentrate their efforts on increasing the labour factor. Based on these results, several agricultural and land policy recommendations are made.
Recent trends in agri-food trade and the future in a changing geopolitical environment
Global agri-food trade is undergoing profound structural change, driven by escalating geopolitical tensions, climate-related shocks, and evolving market dynamics. Agri-food...