Abstract
This paper tests the expectations-augmented Phillips-curve hypothesis for the 50 states in the US. Unlike previous work both adaptive and rational expectations are incorporated in the modeling of the Phillips-curve relationship. Second, the role of relative regional wages are taken into account. Third, the wage-price controls of 1971–72 and 1972–73 are included in the modeling efforts. The empirical results suggest that the expectations-augmented Phillips-curve model based on adaptive expectations provides better results across the 50 states than the Phillips-curve model based on rational expectations.