We theoretically study a general family of economic geography models that features endogenous agglomeration. In many-region settings, the spatial scale---global or local---of the dispersion force(s) in a model plays a key role in determining the resulting endogenous spatial patterns and comparative statics. A global dispersion force accrues from competition between different locations and leads to the formation of multiple economic clusters, or cities. A local dispersion force is caused by crowding effects within each location and induces the flattening of each city. By distinguishing local and global dispersion forces, we can reduce a wide variety of extant models into only three prototypical classes that are qualitatively different in implications. Our framework adds consistent interpretations to the empirical literature and also provides general predictions on treatment effects in structural economic geography models.
↧