The price adjustment hazard function - the probability of a good’s price changing as a function of its price misalignment - enables the examination of the relationship between price stickiness and monetary non-neutrality without specifying a micro-founded model, as discussed by Caballero and Engel (1993a, 2007). Using the micro data underlying the U.S. Consumer Price Index going back to the 1970s, we estimate the hazard function relying on empirical patterns from high and low inflation periods. We find that the relation between inflation and higher moments of the price change distribution is particularly informative for the shape of the hazard function. Our estimated hazard function is relatively flat with positive values at zero. It implies weak price selection and a high degree of monetary non-neutrality: about 60% of the degree implied by the Calvo model, and much higher than what menu cost models imply. In addition, our estimated function is asymmetric: price increases are considerably more likely to occur than price decreases of the same magnitude.

正如 Caballero 和 Engel 所讨论的，价格调整风险函数——一种商品价格随其价格偏差而变化的概率——能够在不指定微观模型的情况下检查价格粘性和货币非中性之间的关系（ 1993a，2007）。我们使用 1970 年代作为美国消费者价格指数基础的微观数据，根据高通胀和低通胀时期的经验模式估计风险函数。我们发现通货膨胀与价格变化分布的较高时刻之间的关系对风险函数的形状特别有用。我们估计的风险函数相对平坦，正值为零。它意味着较弱的价格选择和高度的货币非中性：大约是 Calvo 模型所暗示的程度的 60%，远高于菜单成本模型所暗示的程度。此外，我们的估计函数是不对称的：价格上涨比同等幅度的价格下跌更有可能发生。