Working Paper: CEPR ID: DP4602
Authors: Richard Baldwin; Toshihiro Okubo
Abstract: A Melitz-style model of monopolistic competition with heterogeneous firms is integrated into a simple NEG model to show that the standard assumption of identical firms is neither necessary nor innocuous. We show that re-locating to the big region is most attractive for the most productivity firms; this implies interesting results for empirical work and policy analysis. A ?selection effect? means standard empirical measures overestimate agglomeration economies. A ?sorting effect? means that a regional policy induces the highest productivity firms to move to the core while the lowest productivity firms to move to the periphery. We also show that heterogeneity dampens the home market effect.
Keywords: economic geography; estimation of agglomeration economies; heterogeneous firms; home market effect
JEL Codes: H32; P16
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| firm productivity (D22) | location choices (R32) |
| relocation to larger region (R23) | selection effect (C24) |
| higher productivity of firms (D21) | overestimate of agglomeration economies (R12) |
| regional policies aimed at increasing industrial presence (L52) | attract less productive firms (R32) |
| sorting effect (C69) | concentration of high productivity firms in core regions (R32) |
| sorting effect (C69) | settlement of low productivity firms in periphery (R32) |
| firm heterogeneity (D21) | dampening of home market effect (R21) |