Improving agricultural productivity in Uganda remains a major policy objective given the key role of agriculture in the economy. In this study we evaluate the impact of access to extension services on farm productivity. We use comprehensive baseline survey data collected for monitoring and evaluation of the Agricultural Technology and Agribusiness Advisory Services (ATAAS) project. Applying the ivtreatreg Stata command, and probit 2-stage least squares (2SLS) model that addresses the selection and endogenous bias, we found that access to extension services does not significantly improve the crop productivity of farmers. The finding is consistent with similar studies that control for selection and endogenous bias when estimating treatment effects. We argue that the insignificance of extension contact on productivity when selection and endogenous effects are addressed may reflect the inefficiency of the current extension services in improving farmers’ productivity. In conclusion, the study shows that increasing extension impact on farm productivity will require efforts to improve the quality of extension services that directly translate into productivity effects.
