A number of authors have suggested that economies face a long-run inflation-unemployment trade-off due to downward nominal-wage rigidity. This theory has implications for the nature of the short-run Phillips curve when wage inflation is low. Akerlof, Dickens and Perry have developed an empirical model in which a variable (S) designed to capture the effect of downward nominal-wage rigidity is constructed as part of the estimation of the short-run Phillips curve. Adding this variable dramatically improves the dynamic out-of-sample inflation forecasts of the curve in both the United States and Canada.

In this paper we perform a variety of tests using both real and constructed data to address whether the addition of S truly does provide a better estimate of the short-run Phillips curve, and whether this constitutes evidence that downward nominal-wage rigidity increases the natural rate of unemployment in times of low wage inflation. Our main conclusion is that the performance of the S-enhanced Phillips curve in dynamic simulations is independent of whether downward nominal-wage rigidity is an important feature of the macroeconomy.