Abstract:
This paper extends the Puga (1999) model by introducing urban frictions. It assumes that the agglomeration of manufacturing in a city imposes a cost on the inhabitants of the agglomerated region. Furthermore, an implicit function methodology is developed to provide a numerical stability function that does not require prior analytical work. Simulations reveal that these numerical stability conditions are consistent with the original Puga (1999) analytical predictions. The central finding is that the extension significantly alters the agglomeration properties of the original Puga framework. In particular, partial agglomeration becomes a stable long run outcome in both with and without migration. Furthermore, the level of sensitivity of the agglomeration to the friction cost market parameters is shown to be different in the both cases. This outlines the need to evaluate the imperfectness of migration when modifying the urban geography as a policy implication.