A recent literature has documented a large secular decline in business dynamism in the US since 1980. In this paper I show that the magnitude of this decline varied dramatically across large and small cities. In particular, the decline was twice as severe in cities that were small in 1980. Recent work by Karahan, Pugsley and Sahin (2019) argues that a large part of the aggregate decline is explained by a slower rate of growth of the labor force. In this paper, I examine the extent to which changing cross-sectional patterns of business dynamism can be explained by differences across cities in labor force growth. To do this I embed a model of firm dynamics into a spatial equilibrium model and calibrate the time profile of city level characteristics (amenities and productivities) needed to generate the cross-sectional variation in labor force growth observed in the data. The calibrated model can rationalize half of the cross sectional heterogeneity in the decline of the firm start up rate.