Productive Efficiency of Specialty and Conventional Coffee Farmers in Costa Rica: Accounting for the use of Different Technologies and Self-Selection

As a result of considerable oversupply of green coffee on the international market, world coffee prices dropped to their lowest levels in 30 years giving rise to the most severe crisis experienced by the coffee sector. In many countries, coffee prices did not cover the average cost of production causing widespread financial and social hardships among producers. Economic losses and the lack of viable income alternatives forced many farmers to abandon their coffee plantations and migrate to urban areas in search of employment. Overall, the effects of the crisis pose serious threats to the prospects for sustainable rural development. In the face of this situation, policymakers and development practitioners have shown their willingness to assist farmers in improving their production performance and thus their ability to cope with the crisis. To avoid wasting scarce resources, policy actions must be carefully tailored to the needs of farmers. On this account, the paper presents a diligent investigation of the factors that determine farmers' technical efficiency in coffee production. As inefficiency in production results in a failure to maximize profits at the farm level, increases in productive efficiency enhance the competitiveness of coffee production and could help farmers to confront the adverse economic conditions caused by the coffee crisis. An empirical evaluation of the factors determining efficiency is critical to identify the constraints faced by farmers in the specific situation and to derive adequate policy measures. The empirical analysis is based on a sample of 216 households that were randomly chosen from within two of the main coffee regions in Costa Rica. A standardized questionnaire was administered to households to collect data on coffee production as well as on the socio-economic characteristics of household members. The information collected partly includes recall data covering the production periods 2003/04 and 2002/03. Given its favorable natural conditions, Costa Rica has a competitive advantage with respect to the production of high-quality coffee. In the face of the crisis, the country has put emphasis on exploring that potential, motivating farmers to adjust their production to the requirements of specialty markets. The specialty segment, including high-quality and sustainable coffee, has been attributed major importance in providing farmers with a sustainable alternative to conventional coffee markets. The household sample used for the empirical analysis includes both farmers producing in the specialty segment as well as in the conventional segment. In order to analyze farmers' production performance, a stochastic frontier model is estimated and the effects of a range of farm-specific variables on technical efficiency are determined simultaneously. Given that farmers in the sample use different sets of technologies, two different production frontiers are estimated for farmers cultivating conventional and specialty coffee, respectively. As these sub-samples are unlikely to represent unbiased representations of the population, the ignorance of self-selection bias will produce inconsistent estimates. Following Heckman (1979) and Lee (1978), an inverse Milli's ratio is included in the models to control for self-selection bias. The inverse Milli's ratio is derived from a pooled probit model, which is used to estimate the probability that a farmer chooses to grow specialty coffee. Whereas previous studies have estimated technical efficiency separately for two sub-samples of farmers using different production technologies, the present work explicitly controls for the selection bias that can result from splitting the original sample and that has been neglected in former studies. The probit model indicates that the probability of participation in the specialty segment increases with the farmer's experience in coffee cultivation, education and access to specialized extension service as well as with the size of the farmland. Furthermore, membership in a coffee cooperative has a positive impact on participation. The inverse Milli's ratio that is then included in the stochastic frontier analysis proves to be significant indicating that self-selection is present. Average output of specialty coffee farmers is larger than it would be if all farmers were cultivating specialty coffee. Likewise, average output of conventional farmers is smaller than it would be if all farmers were using the respective technology. This underscores the need for taking possible selectivity bias into account. Controlling for selection bias, the two frontier models identify several factors that determine efficiency levels of farmers. In both models, the effect of other income-generating activities on efficiency is positive, which is likely to be a result of better access to liquidity and information of those farmers who have additional income sources. In the case of specialty farmers, efficiency increases if farmers keep book of their activities and expenditures. Furthermore, efficiency decreases with farm size. This might be interpreted as an advantage of small-scale farms in the cultivation of specialty coffee, although it has to be kept in mind that overall, smaller farmers are less likely to participate in specialty coffee cultivation. In the case of conventional farmers, model results reveal that membership in cooperatives significantly contributes to the achievement of technical efficiency at the farm level. The results of the empirical analysis stress the importance of actively involving small-scale farmers, who are otherwise easily excluded from new market developments. The study suggests that once they are able to overcome the barriers that prevent their participation, small-scale farmers can successfully compete in the specialty segment. Further policy recommendations derived from the empirical results include the provision of extension services, the support of coffee cooperatives, and the diversification of the rural economy.