Examines earnings of synthetic cohorts contained in the Current Population Surveys (CPS) for each year from 1968 to 1975. The CPS data are useful to test for the importance of local labor markets because individuals can be assigned to markets by calendar year and area of residence. This study finds that contrary to the established conviction, wage rates are sensitive to business cycles. This sensitivity appears, holding industry constant, to be skill neutral so that business cycle movements apparently explain little of the recent time series behavior of relative wages of college graduates. Furthermore, the paper has identified some important movements in wage rates among areas. These wage differentials seem more sensitive to employment deviations and trends than were initially expected.