This is a job market seminar organized for selected candidates from the European Job Market.

Information about rival opportunities becomes less valuable as more individuals learn about them. When policy interventions are perceived as rival, secrecy may lower dissemination and uptake. I conducted a lab-in-the-field to test how rivalry affects the pattern of diffusion of information. The lab games began with 100 initial seeders randomly assigned to play one of two games: (a) a rival game in which players received a share of a monetary prize and (b) a non-rival game in which they received a fixed amount. The games followed a snowball structure in which the initial players could invite others in the village to participate in the next round of the game. The initial seeders in the rival game were 18 percentage points less likely to share invitations to the game than those in the non-rival game, which resulted in a lower diffusion into the wider community. I find no effects regarding sharing with people who were less central and therefore had fewer friends with whom to share in the next rounds. In addition to the lab-in-the-field, I conducted experiments to evaluate how the same seeders shared invitations to two policy trainings on two distinct agricultural practices. The practices varied in terms of the level of perceived rivalry as measured by a willingness-to-pay game. The pattern of invitations seen in this experiment indicates a different strategy for dealing with rivalry. Initial seeders did not discriminate in terms of the number of people invited. Instead, seeders invited less-central people to participate in the rival policy compared to the non-rival one people who were less likely to diffuse the knowledge in the network. The results show that when policy interventions are perceived as rival, policy implementers should be careful about how it will affect its diffusion process.