With increasingly improved cooking stoves (ICS) that aim to reduce fuelwood consumption by forest-dependent households, more evidence of what drives households to adopt ICS is needed. Using data from a representative sample (N=271) of households in a rural part of eastern Tanzania, we estimated a mixed logit model to take into account the limitations of the standard multinomial logit model and relaxed the restrictive assumption of the conditional logit model.

In this paper, we identify economic implications of the pressure to share resources within a social network. Through a set of field experiments in rural Tanzania we randomly increased the expected harvest of the treatment group by the assignment of an improved and much more productive variety of maize.