The study uses randomisation to estimate the impact of interventions to improve extension programmes in ten provinces in Cambodia.
Many farmers in low- and middle-income countries are either unaware of new and improved cultivation practices or they are unable to employ these practices optimally. This situation has led governments and other agencies to invest considerable resources in agricultural extension programmes. PADEE and ASPIRE are two such programmes operated by the Cambodian Ministry of Agriculture, Forestry, and Fisheries (MAFF). This study evaluates innovative models that use incentives and information and communication technology (ICT) to deliver extension services through these programmes.
- Can ICT methods strengthen the effectiveness of traditional extension in terms of farmer learning?
- Do ICT methods enhance the effectiveness of traditional extension on farming outcomes?
- How cost-effective are ICT-related additions at increasing farmer’s learning outcomes?
- Can financial incentives for extension workers improve the effectiveness of extension programmes in developing countries?
- By reducing costs of information delivery and increasing the critical mass of informed farmers, can ICTs improve farmers’ knowledge and adoption of agricultural practices?
The study measures the impact of innovative models for delivering extension services to farmers. Through the PADEE programme, MAFF provided dedicated extension workers to work with farmers’ groups, and also developed software that allowed extension workers to provide personalised recommendations on seeds, fertiliser, and pest control. Some PADEE extension workers also received monetary incentives based on how effectively they disseminated information to their client farmers (measured using quizzes administered to farmers). The ASPIRE programme also involved extension workers working with farmers’ groups, but in addition, some farmers also received automated voice messages to their mobile phones with information on recommended farming practices.
Theory of Change
The motivating belief behind the PADEE and ASPIRE interventions is that innovative approaches to extension will lead to increased awareness and knowledge among farmers of better agricultural techniques like seeds or inputs. Enhanced knowledge will, in turn, encourage more farmers to adopt these technologies and improve their practices, resulting in an increase in agricultural productivity and household welfare.
The study implements a multi-arm randomised controlled trial along with a mixed-methods approach. For the PADEE intervention, in one treatment group extension workers received tablets loaded with specialised software (ePADEE) to allow them to provide targeted advice to farmers. In the second treatment group, the extension workers received the tablets and were also eligible for monetary incentives based on farmers in their area performed on periodic knowledge tests. Both these treatment arms were compared to a control group that received the basic PADEE intervention, which provides extension workers but without the tablets and specialised software.
For the ASPIRE intervention, there were three treatment conditions in addition to the control condition. The first treatment condition sent mobile voice messages to member of farmers’ groups once, while the second condition sent the messages a second time at the time they needed to take the action recommended in the message. In the third condition, multiple messages were also sent to farmers outside the farmers’ groups to test whether network effects could help promote adoption.
The evaluation found that the ePADEE software aided extension workers in conveying rice farming recommendations to farmers, some of which were implemented. Although the evaluation found no effects of ePADEE on downstream outcomes like rice production and productivity, the programme increased adoption of recommended practices with respect to seed and fertiliser. The incentives to extension workers had no discernible effects.
The ASPIRE phone calls were popular with farmers, who frequently listened to the entire messages and felt the messages were helpful for increasing their production. These effects were stronger when farmers received the calls multiple times. However, household surveys did not reveal any improvements in productivity as a result of the calls.
The authors conducted a cost analysis and estimated that the cost per farmer of the ePADEE software intervention was about US$14 per farmer, over and above the costs of the basic PADEE model. The marginal cost of sending voice messages to ASPIRE was US$2.39 per farmer.
Implications for implementers
Although the evaluation did not detect significant effects of either ICT solution on agricultural production, the authors recommend that MAFF continue with the voice messages, given their lower costs compared to ePADEE and their popularity among farmers.
Implications for policy and practice
In addition to recommending that the voice calls continue, the authors also note there are several potential avenues for enlisting the private sector (e.g. mobile service providers) to deliver similar interventions. For example, given that farmers show demand for extension information, mobile service companies may wish to provide access to call-in extension services to maintain their customer bases.
Implications for further research
Given the relatively short-term nature of this evaluation, it would be valuable for future research to explore whether more sustained exposure to innovative extension models can spur increases in outcomes like productivity, income, and food security. Future research could also explore the viability and effectiveness of relying on private-sector technology companies to deliver extension services through ICTs.