There is a growing body of literature that analyzes the Phillips curve model in the context of subnational labor markets. To date the results of these analyses have been mixed. This paper estimates a tiered version of a modified Phillips curve for local labor markets in the midwest region of the U.S. The purpose of the model is to test the significance of a wage-unemployment trade-off at the metropolitan level, and to determine the significance of a downward wage transmission effect through the tiers of an urban hierarchy. Using the method of ordinary least squares, the results provide no evidence of a significant Phillips curve type of relationship between wages and unemployment at the local level. However, the wage transmission effect from the upper order cities of the hierarchy to the lower order cities appears to be a significant determinant of the rate of change of wages in metropolitan areas.

No abstract

No abstract

No abstract

a regression-equation structure have been violated. Hence, R-H's using the single-equation approach to the expectations hypothesis does not yield entirely meaningful results. ItIThis Comment has shown analytically that R-H's single-equation approach to the expectations hypothesis for wage determination in tier i is clearly inappropriate: their single-equation approach yields biased and inconsistent estimators.Accordingly, it is suggested here that it is desirable for R-H to adopt, in lieu of single-equation regressions estimated by ordinary least squares (OLS), an alternative empirical approach, such as properly specified, multi-equation models (in W.-and P..) to be estimated by two-stage, least squares (2SLS), or perhaps evenl~hree-slt~ge, least-squares (3SLS). Such a change would be necessary in order to account for feedback between the rate of change of prices and that of money wages in tier i. Moreover, given the public policy implications of the Phillips curve, this is a change which R-H should not ignore. , we estimated a model to test the significance of a wage-unemployment trade-off at the metropolitan level, and to determine the significance of a downward wage transmission effect through the tiers of an urban hierarchy. In his comment on our article Richard J. Cebula makes two criticisms of our work. First, he argues that we do not justify the hypothesized positive relationship between the rate of inflation (P) and the rate of change of wages (W) which is the dependent variable in our model. Second, he argues that our "model, because of the way in which it is specified and estimated, unequivocally suffers from a simultaneity bias and hence yields