Cricket rearing and consumption continues to gain popularity especially among small scale farmers in Kenya. However, low production, high labor-intensive technologies, limited input resources and rearing systems constrain the optimization, consumption and marketing of crickets. Further, little information exists on alternative cheap sources of producing crickets and how such resources in cricket production can be optimally allocated in a fashion that minimizes waste and inefficiency. Therefore, this study sought to determine the economic efficiency of the improvised cricket rearing systems using a Generalized Additive, Stochastic Frontier Model (GAM) approach to assess the efficiency in cricket production. The improvised system comprised of bamboo hideouts, clean scrap blankets for drinking and laying, cut bamboo stem as drinking platter and the plywood-based cages. Twenty-day old Acheta domesticus and Gryllus bimaculatus were separately reared in the cages each with three replications, representing different production farms. 49 crickets were randomly selected from each cage and weighed per week. The maximum likelihood estimates of GAM Stochastic frontier model showed that feed, labor and water were positive and significant at 5% level suggesting their importance and positive influence on cricket output. Similarly, the cost of feed, labor, water and scrap blanket were positive and significant implying that increase in these costs of inputs would increase the total production cost. The mean TE, AE and EE were 85%, 92% and 79%, respectively. Focus should be on strengthening policies that sensitize cricket farmers to use the locally available improvised rearing resources in order to increase productivity, income hence enhance household food security and reduce poverty.
Keywords: Economic Efficiency, Crickets, Generalized Additive, Stochastic Frontier Model