Social learning and parameter uncertainty in irreversible investments, and, Partial maximum likelihood estimation of a spatial Probit model

The first paper discusses the social learning and parameter uncertainty in irreversible investments. The adoption of new technology usually involves irreversible investments where the future payoff is uncertain. In addition, investors often have to contend with a limited understanding of the technology itself, which can be modeled as uncertainty regarding the parameters of the stochastic process describing the future payoff. It is hypothesize that social learning (having previous adopters in the farmer's social network) increases the probability of the farmer adopting the new technology. This is posited based on theory: social learning would reduce parameter uncertainty, and thus the overall level of risk facing the farmer-investor, and thus induce investment. The paper tests this hypothesis using Chinese farm household data on adoption of greenhouses. The latter are of the "intermediate technology" type, made of clay walls, a plastic-sheet roof, and a straw mat roll-out awning for cold nights. The empirical findings of this paper support the hypothesis. It is also found that market volatility discourages adoption.;The second paper analyzes a spatial Probit model for cross sectional dependent data in a binary choice context. Observations are divided by pairwise groups and bivariate normal distributions are specified within each group. Partial maximum likelihood estimators are introduced and they are shown to be consistent and asymptotically normal under some regularity conditions. Consistent covariance matrix estimators are also provided. Finally, a simulation study shows the advantages of the new estimation procedure in this setting. The proposed partial maximum likelihood estimators are shown to be more efficient than that of generalized method of moments counterparts...