Generalizing the Stochastic Approach to Price Indexes
证明传统价格指数公式和时间虚拟特征回归方法都能一致估计同一广义定价模型中的增长参数，并指出随机方法无法帮助选择指数公式，而时间虚拟法是被低估的优良方法。
I show that conventional price index formulas and time‐dummy hedonic regression techniques all consistently estimate growth parameters in the same generalized model of product pricing. I then use that result to make two points. First, the “stochastic approach” is not a helpful tool for choosing price index formulas, because in its complete form it can justify any of them. Second, the literature uses flawed arguments for replacing time‐dummy hedonic regression with hedonic imputation. The time‐dummy method is an excellent, underrated option.