Does trust in the extension source improve learning outcomes? Experimental evidence from Southern Africa

Does it matter who trains farmers? We use an artefactual field experiment that includes training on conservation agriculture to test whether trust in the source of extension advice enhances learning outcomes in Malawi, Zambia, and Zimbabwe. We measure trust in public, private, and farmer-to-farmer extension agents using a standard trust game, and link trust to test scores from farmers randomly assigned to participate in training sessions facilitated by each agent. We find higher trust levels in the public extension agents, especially in Zimbabwe. Training increased knowledge outcomes by 7 percentage points (pp), public extension agents increased test scores by 9 pp, and farmer-to-farmer agents reduced scores by 11 pp in the pooled treated sample. At country level, farmer-to-farmer extension agents increased test scores by 19 pp in Zambia and reduced scores by 10 pp in Zimbabwe. Public extension agents increased test scores by 12 pp in Zimbabwe. Trust in the public and farmer-to-farmer extension agents increased test scores by 3 pp. These findings underscore the importance of training and trust in the source of extension in improving knowledge outcomes. Contrary to popular opinion, both public and farmer-to-farmer extensions have a role to play, signifying the need to retool, strengthen, and support both extension systems to enhance service delivery and to build farmer trust. Embracing different information communication technologies (ICTs) and digital advisories could help in this regard.