Our experiment was conducted in 72 randomly chosen villages in the potato growing areas of
Hugli and West Medinipur districts. The sample villages under each market catchment area
were randomly assigned to three groups, resulting in 24 villages in each treatment group. In two
groups we conducted two dierent information treatments and one group served as control where
no information was provided. To reduce information spillovers, villages were selected such that
they were at a minimum distance of 8 kilometres from each other.13 In the two treatment groups,