This study investigates the technical efficiency of olive farms in Béjaia, Algeria, using primary data from a stratified survey of growers conducted in 2025. A flexible stochastic frontier model, complemented by OLS and Cobb-Douglas specifications as comparative benchmarks, captures the non-linear structure of production technology and isolates managerial inefficiency from environmental noise. Results reveal pronounced heterogeneity, with a strongly bimodal efficiency distribution and an average efficiency level of 0.39. This level of efficiency is relatively low compared to typical estimates reported in Mediterranean orchard systems, suggesting substantial room for productivity improvements. Routine agronomic practices emerge as primary determinant of inefficiency: regular pruning, soil-intervention intensity, and irrigation access significantly reduce distance from the frontier, while tree density contributes positively when embedded in a consistent management regime. Human capital also matters, as older and more educated farmers operate more efficiently. The findings imply that productivity gains in perennial systems depend less on marginal input adjustments and more on the coherence and regularity of orchard management. The findings underline the importance of targeted technical support and improved management practices to enhance farm performance in Mediterranean contexts. The study contributes fresh evidence on efficiency determinants in perennial agriculture and informs ongoing efforts to modernize Algerian olive sector.
