This CABI Working Paper uses quasi-experimental data to test for changes in orange producer input costs, yield and profitability attributable to plant clinics operating in the Teso region of Uganda. Methods used include producer fixed effects difference in difference regression; propensity score matching with difference in difference; and inverse propensity weighted producer fixed effects difference in difference estimation of impact. The tests find some evidence of lower yields and revenue for farmers visiting the plant clinic, and some suggestions of increased expenditures on labour. Potential explanatory factors for the decline in yield and revenue are discussed.

The paper also presents data gathered on a number of contextual questions pertinent for the continued operation and improvement of the plant clinic programme. Finally, the paper discusses challenges for this impact evaluation and makes recommendations to improve future impact evaluations in the areas of evaluation design and survey design, and suggests ways to ensure better plant clinic functioning.