The regional adjustment model is used to analyze changes in population and employment across the American metropolitan landscape between 1990 and 2015. Estimates are made for the effects of natural and human-created amenities on population change and the effects of wages, self-employment, patents, economic specialization, and age composition on employment change. Short-run impacts, estimated by linear regression, allow identification of a 2 by 2 “growth operator” matrix; long-run impacts are estimated by powering this matrix. In the early years of the 25-year study period, employment numbers largely drove population change, but, once the direction of causality reversed, population numbers largely drove employment change. Clearly, the balance between the overall effects of population and employment can shift over long periods of time.