Technical efficiency, technological change, and returns to scale in the production of selected food crops in Uganda

Abstract

Although the agriculture sector is an important sector for food security, poverty reduction and overall economic growth in Uganda, the productivity of the food crop sub-sector remains unsatisfactory. While existing studies have mainly been cross-sectional due to lack of panel data, it is imperative that evidence is provided on the temporal performance of the sector across specific major subsectors, highlighting the aspects that have consistently resulted into favourable positive impacts and those that have resulted into poor performance, in order to enhance existing strategies aimed at improving productivity in the sector. The study estimated and investigated three components of productivity; technical efficiency, technological change, and returns to scale in four of the country’s major food crops; maize, beans, banana, and cassava for the period between 20052010. The study also established the determinants of the observed technical efficiency during the same period. Using national panel data for 2005/06 and 2009/10 collected by the Uganda Bureau of Statistics, the study utilized two econometric approaches; a translog stochastic frontier production function model to estimate technical efficiency, technological change and returns to scale, and a robust ordinary least squares regression to establish the determinants of technical efficiency. The results revealed that mean technical efficiency across the four food crop farming households was low during the study period, estimated at 22%, 15%, 15% and 14% for maize, beans, banana and cassava farming households respectively. These results imply that there would still be a possibility to produce 78%, 85%, 85%, 86% more output of maize, beans, banana, and cassava respectively, using the same resources and at the existing technology. Mean technical efficiency declined between the two time periods, and significantly so (at 1% level) for both beans and banana farming households. The key factors that determined technical efficiency across the four farming households were education, extension visits, crop area, location where a household was located, whether in the rural or urban area, and housing index which was composed of a number of features that would indicate the well-being of a household. Although purchased inputs would ordinarily be expected to increase food crop productivity, improvement in productivity among the farming households was propelled more by technical change, resulting from intensified use of both family and hired labour, than technical efficiency. Maize and cassava farming households exhibited increasing returns to scale, implying that expanded use of purchased inputs and crop area would be beneficial to raise their productivity. On the contrary, bean and banana production exhibited decreasing returns to scale, implying that it would neither be worthwhile expanding the use of purchased inputs, nor crop area at the existing technology. In terms of policy, the results underscore the need for government to promote the use of purchased farm inputs through market interventions that will enable input and output prices to motivate investment by maize and cassava farming households, and effort to improve the level of technology to raise returns to scale for beans and banana. Across the four crops, government should pursue a land reform policy that will support farming households to secure and expand food crop area in rural areas, provide market supportive road and physical infrastructure, education and extension support for household heads and spouses, specifically on market dynamics of purchased inputs and food crop output, in order to enhance food crop productivity.