Dataset for: Group-based ICT Extension Support Systems: Evidence from Rural Peru (“Go Digital Peru”)

We investigated how to make group-based ICT extension work more effective and efficient, to understand whether different types of result-based incentives improve focal points’ effort and whether this effort translates into increased farmers' participation, adoption of improved agricultural practices, and ultimately crop yields. Between September 2024 and August 2025 (main potato season), we organized 1,242 farmers in groups in 102 communities in the highlands of the Provinces of Sanchez Carrión, and Pataz (La Libertad region, northern Peru) and assigned them to three treatment arms: (a) individual incentives for focal points, (b) social incentives for the community, and (c) no incentives. Three extension officers from the implementing partners (Asociacion Pataz) were the technical advisors of the platform, supported by CIP experts. We periodically submitted flyers with information about potato production best practices through the WhatsApp platform and responded to questions by the groups. We conducted baseline and endline surveys and four rounds of telephone tests to measure farmers knowledge gains. To understand potential spillover effects, we identified 142 additional farmers who lived in the same communities but were not part of the WhatsApp platform. This database presents a comprehensive evaluation of the Social and Individual incentives. The sections are organized to provide a logical progression from the validation of the experimental design to the assessment of specific agronomic and economic impacts. We begin by outlining the study’s structural foundations, detailing the cluster-based randomization process and the resulting sample properties. Following this, we assess the internal validity of the design through a rigorous Evaluation of Attrition, confirming that participant loss between baseline and endline was non-differential and introduced no systematic bias. We then present the Baseline Balance Analysis, which verifies the comparability of treatment arms across key demographic, economic, and connectivity indicators prior to the intervention. Once the validity of the design is established, the report defines the Estimation of Causal Impact, outlining the econometric specifications, including the Intention-to-Treat (ITT) strategy and covariate adjustments used to ensure robust estimates.