Abstract
A steep decline in coffee prices at the producer level led to considerable pressure for farmers in Costa Rica and producer countries all over the world. One possible reaction was moving to specialty markets, where price pressure was perceived to be lower. We use original survey data from 2002/2003 and 2003/2004 to analyze the factors influencing participation in specialty markets and to estimate separate production functions for specialty and conventional coffee farmers allowing for farm-specific inefficiencies. Applying a sample selection framework, we find significant selection bias in the sub-sample of specialty farmers and evidence for the overestimation of efficiency, if this bias is not adequately controlled for. Among the most important factors that influence farm-specific efficiency levels in the two sub-samples are the availability of additional income activities, experience in coffee cultivation, and membership in cooperatives. Based on the results, we derive policy recommendations to improve farmers’ production performance and ability to cope with the effects of the coffee crisis. These policy measures include the provision of extension services with respect to farm management skills, the creation of income opportunities in rural areas, and the support of farmer-owned cooperatives.
Highlights
► We estimate production frontiers for specialty and conventional coffee farmers. ► Models control for selection bias and analyze farm-specific inefficiency effects. ► Average efficiency is overestimated if selection bias is not adequately controlled. ► Farm-level efficiency determinants are consistent across models. ► Determinants include experience, cooperative membership and off-farm activities.
Keywords
Coffee; Costa Rica; Stochastic frontier analysis; Sample selectivity; Specialty markets; Technological heterogeneity