Uncertainty, Competition, and the Adoption of New Technology
构建了一个贝叶斯学习模型，分析企业在不确定新技术价值且面临竞争时，如何通过序贯信息收集决定是否采纳新技术，并刻画了最优策略的阈值特征。
Faced with the decision of whether or not to adopt a new technology whose economic value cannot be gauged with certainty, the manager of the firm may elect to decrease the uncertainty by sequentially gathering information (at a unit cost of c > 0), updating his prior beliefs in a Bayesian manner. Uncertainty regarding the actions of a competitor, however, cannot be reduced. Our model allows a manager to account for potential competition, either substitute or complementary, by inclusion of strategic considerations modelled in a game theoretic setting. We show that the firm's best response mapping satisfies a dynamic programming recursion (even when the one period reward function is unbounded) providing management with a familiar tool to solve its problem. Best response behavior is characterized by a monotone sequence of pairs of threshold numbers which give rise to a “cone-shaped” continuation region. The continuation region is shown to shift up (down) with increases in the expected level of substitute (complementary) competition making the firm less (more) likely to adopt the innovation. Finally, the existence of a Nash equilibrium in cone-shaped strategies is established.