Confidence in the Phillips Curve (PC) as predictor of inflation developments along the business cycle has been shaken by recent “inflation puzzles” in advanced countries, such as the “missing disinflation” in the aftermath of the Great Recession and the “missing inflation” in the years of recovery, to which the Euro-Zone “excess deflation” during the post-crisis depression may be added. This paper proposes a newly specified Phillips Curve model, in which expected inflation, instead of being treated as an exogenous explanatory variable of actual inflation, is endogenized. The idea is simply that if the PC is used to foresee inflation, then its expectational component should in some way be the result of agents using the PC itself. As a consequence, the truly independent explanatory variables of inflation turn out to be the output gaps and the related forecast errors by agents, with notable empirical consequences. The model is tested with the Euro-Zone data 1999–2019 showing that it may provide a consistent explanation of the “inflation puzzles” by disentangling the structural component from the expectational effects of the PC.

对菲利普斯曲线 (PC) 作为商业周期通胀发展预测指标的信心已被发达国家近期的“通胀难题”所动摇，例如大萧条之后的“通货紧缩缺失”和“通胀缺失”在经济复苏的年代，可能会加上欧元区在危机后萧条期间的“过度通货紧缩”。本文提出了一种新指定的菲利普斯曲线模型，其中预期通胀不是被视为实际通胀的外生解释变量，而是内生化的。这个想法很简单，如果 PC 用于预测通货膨胀，那么它的预期成分在某种程度上应该是代理使用 PC 本身的结果。作为结果，通货膨胀的真正独立解释变量是产出缺口和代理人的相关预测误差，具有显着的经验后果。该模型用 1999-2019 年的欧元区数据进行了测试，表明它可以通过将结构性成分与 PC 的预期效应分开，为“通胀难题”提供一致的解释。