Randomized controlled trials for measuring causal impacts of agricultural extension interventions

Abstract

Extension programs invest substantial resources yet rigorous causal evidence on their impacts remains sparse. This research conducted a systematic review and meta-analysis of 72 RCTs evaluating extension interventions (2005-2023), alongside primary data from 34 evaluations in Tajikistan's Khatlon region between June 2021 and November 2022. Mean effect sizes varied by type: demonstration plots (d=0.41), training programs (d=0.28), advisory visits (d=0.22), and ICT-based (d=0.19). Only 23.5% of trials met all six quality criteria. Sample size calculations were reported in 52.1% and spillover effects measured in only 34.7%.