Title: Productivity analysis of pond aquaculture in northern Malawi: Application of the stochastic frontier production function.

In this study, a stochastic frontier production function was applied to estimate both the productivity (technical efficiency) scores and determinants of low productivity (inefficiency) for 96 fish farms in northern Malawi. The study used the Cobb-Douglas model, in which efficiency estimates showed that fish farmers in the region were more technically inefficient, with mean efficiency scores of 29%. The empirical results show that fingerlings and commercial feed are the most significant inputs that affect production. On average, fish production in northern Malawi is carried out under decreasing returns to scale, meaning that an increase in inputs will more than proportionately decrease output. Except for club membership, all inefficiency determinants are negative, although none of the coefficients are significant. It is, therefore, suggested that policy variables such as feed and fingerlings are important determinants of productivity for farms to reduce the existing 71% yield gap. Thus, the government must ensure that the provision of support services to fish farmers, such as the availability of high-quality input suppliers (feed and fingerlings), are readily available, accessible, and affordable.