Estimable equilibrium models of locational sorting and their role in development economics

Geography plays a prominent role in many problems in development economics - directly in analyses of the spatial distribution of important variables like poverty and productivity, and indirectly through the role of local spillovers in economic growth. Empirical work on these topics is complicated by the fact that the behavioral consequences of such spillovers cannot be distinguished from those of unobservable local attributes using only the observed location decisions of individuals or firms. This problem can be solved with an instrumental variables strategy derived from the internal logic of a structural model of residential sorting. We show practically how the strategy is implemented, provide intuition for the instruments and econometric identification, demonstrate how traditional techniques overstate agglomeration externalities, and use the model to value changes in spillovers from urban centers. © Oxford University Press 2005; all rights reserved.