Shimer (2005) and Hall (2005) have documented the failure of standard labor market search models to match business cycle fluctuations in employment and unemployment. They argue that it is likely that wages are not adjusted as regularly as suggested by the model, which would explain why employment is more volatile than the model predicts. We explore whether this explanation is consistent with the data. The main insight is that the relevant wage data for the search model are not aggregate wages, but wages of newly hired workers. Our results show that wages for those workers are much more volatile than aggregate wages and respond one-for-one to changes in labor productivity. Thus, we find no evidence for wage rigidity.