A Minimax Regret Approach to Decision Making Under Uncertainty
提出一种最小化遗憾方法用于不确定性下的最优要素需求，无需指定工具变量或可能状态数量，可估计事后生产冲击并进行统计推断，通过实证应用展示其效果。
Abstract We propose a minimax regret approach to optimal factor demand under uncertainty. Regret is the deviation of any given decision from the optimal decision based on a specified set of possible scenarios for the uncertain variables. This approach does not require the specification of instrumental variables to control for unobserved states of nature, and also does not require specification of the number of possible states in advance. Importantly, ex post production shocks can be estimated using our approach, and full statistical inferences can be obtained. Econometric techniques are based on Bayesian analysis using Markov Chain Monte Carlo techniques. A substantive empirical application is provided to illustrate the new approach.