Vanishing cities: what does the new economic geography imply about the efficiency of urbanization?

How should the size and number of cities evolve optimally as population grows? Stripped of the constraints of geography itself, the setup of the new economic geography (NEG) implies that de-agglomeration (or de-urbanization) is efficient. The number of cities increases while the size of each decreases on the optimal path until the economy suddenly disperses to tiny towns of stand-alone firms each specializing in a unique good. The cause of this narrow result is the NEG's strong emphasis on intercity trade to satisfy the taste for more goods. For the same aggregate population, a system of smaller cities saves time lost in commuting, has a larger labor supply, and makes more goods than does a system of larger cities. Falling interurban trading costs favor this de-urbanization process. Only if intraurban commuting costs fall sufficiently, can a pattern of growing city sizes be efficient with growing population. Of course, when the number of cities or the geographic space itself is limited or asymmetric, then agglomeration can arise as an artifact of the constraints imposed by geography as demonstrated by numerous NEG models. This reveals that the central agglomerative force in the NEG is space itself and not the underlying economic relations. Copyright 2004, Oxford University Press.