Abstract: We discuss the ability of standard estimates of the correlation of wages and employment to measure the relative strength of aggregate demand and supply shocks, given that the choice of time period, deflator, and explanatory variables inherently biases the estimated cyclical coefficients toward identifying labor supply or demand. We determine that a closer look at the standard wage/labor correlation shows that it can neither provide information on the relative strength of supply and demand shocks, nor give an indication of the response of wages to aggregate demand shocks. Following this, we test the predictions of a neo-Keynesian model for the correlation of employment and wages using restrictions generated by the model to identify movements along or shifts in labor demand. Our results are consistent with the theory of nominal wage rigidity and we find no reason to reject the neo-Keynesian model based on the correlation of wages and employment.

Full paper (130 KB PDF) | Full paper (218 KB Postscript)