The welfare implications of advertising and extension under uncertainty

摘要

This paper specifies an adoption model based upon Bayesian learning and exogenous information generation. Formulae for welfare effects are derived and calibrated using Green Revolution agricultural data. The effects of intervention through the dissemination of new information are then estimated numerically. The simulations indicate that gains to intervention can be substantial. Intervening with slowly adopted marginal technologies is as beneficial as intervening with superior technologies. Taken from Shampine [Am. J. Agric. Econ. 2 (1998).], which examined intervention in the presence of learning externalities, the results suggest that if adoption is slow, and information is the primary constraint, the gains to intervention are generally substantial relative to the costs.