Using a randomized experiment, we assess the impact of a training session where information on a more efficient cooking stove (ICS) is provided along with the opportunity to purchase it at market price. We find positive direct and spillover effects of being invited to the session on ICS ownership in 6 to 9 month following our treatment.

We also find a positive impact on usage, measured both through self-reporting and with monitoring devices, along the extensive and intensive margin. Our training session appears to have no impact on the knowledge about ICS and any measure of welfare. Within our session, we furthermore randomly assign participants to receive information on a peer’s actual purchase or previous ownership and find that women are more likely to adopt the product if the information they receive is on a peer who purchased the product and whose opinion is respected. This allows us to investigate the various mechanisms of social interaction potentially at play and provide evidence supporting imitation effects, rather than social learning or constraint interaction.