Learning and Interdependent Decision Making, Modeling Farmers' Technology Adoption Decisions in Developing Countries

本研究项目的主要目标是开发新的方法来研究采用不确定盈利能力的技术创新。重点将放在学习技术的情况下，无论是从自己过去的经验中，还是从别人的成功中，都是重要的。这一点很重要，因为采用者之间的外部性意味着，传播的速度低于社会最优速度。对这种情况的实证分析也提出了有趣的方法论问题。拟议的研究分为三个部分。首先，农民决策的相互依赖性将被赋予理论基础。结合学习，每个农民的决策问题是一个动态规划问题，并通过应用适当的工具找到均衡采用路径。其次，必须建立反映农民决策中战略相互依存关系的计量经济模型。新技术的采用路径反映了动态离散选择优化。本文将采用一种模拟矩量过程的方法来估计具有邻域效应的动态系统。最后，将使用国际半干旱热带作物研究所的十年面板数据集对开发的理论和计量经济学模型进行测试。这项研究将是在印度半干旱的热带地区采用高产棉花品种。

The primary objective of this research project is to develop new methods for studying the adoption of technological innovations of uncertain profitability. The focus will be on situations where learning about the technology, either from one's own past experience, or from the success of others, is significant. It is important since externalities between adopters imply that diffusion occurs at less than the socially optimal rate. Empirical analysis of such situations also raises interesting methodological questions. The proposed study has three parts. First, the interdependence of farmers' decisions will be given a theoretical foundation. Incorporating learning, the decision problem for each farmer is a dynamic programming problem and equilibrium adoption paths are found by applying the appropriate apparatus. Second, econometric models reflecting strategic interdependence in farmers' decisions must be developed. The path of adoption of a new technology reflects dynamic discrete choice optimization. A simulated methods of moments procedure will be used to estimate a dynamic system with neighborhood effects. Finally, the theoretical and econometric models developed will be tested using a ten year panel data set from the International Crops Research Institute for the SEmi-Arid Tropics. The study will be of the adoption of higher yielding varieties cotton in the semi- arid tropics of India.