This paper shows that labor market search frictions do not explain fluctuations in the labor wedge per se. However, the introduction of extensive and intensive margin clarifies that measuring the MRS in terms of total hours artificially introduces procyclicality in the MRS. When the MRS is correctly measured in terms of hours per worker, the labor wedge obtained is less variable than the one of the competitive model. Finally, we show that it is possible to measure a strongly procyclical labor wedge when the actual data generating process is a search model that allows for movements in both margins.