Firm Sorting and Agglomeration

Abstract

To account for the uneven distribution of economic activity in space, I propose a theory of the location choices of heterogeneous firms in a variety of sectors across cities. In equilibrium, the distribution of city sizes and the sorting patterns of firms are uniquely determined and affect aggregate TFP and welfare. I estimate the model using French firm-level data and find that nearly half of the productivity advantage of large cities is due to firm sorting, the rest coming from agglomeration economies. I quantify the general equilibrium effects of place-based policies: policies that subsidize smaller cities have negative aggregate effects. (JEL D22, D24, R11, R32)

Market size, occupational self‐selection, sorting, and income inequality

“…We now extend the model to allow for two locations and the sorting of mobile agents across both locations and occupations . Investigating jointly sorting across cities and selection within cities is a challenging task, even in simple models (see Behrens et al., ; Behrens & Robert‐Nicoud, ; Gaubert, ). The reason is that distributions matter in complicated ways.…”

“…The size elasticity of the share of college graduates in U.S. metropolitan areas is, for example, 6.8 percent in the 2000 U.S. Census data. More efficient workers and firms sort into larger cities (Behrens, Duranton, & Robert‐Nicoud, ; Combes, Duranton, & Gobillon, ; Gaubert, ), which partly explains their productivity advantage. Last, city size also matters for the income distribution.…”

Agglomeration and clusters: Tools and insights from coagglomeration patterns

“…Though obvious, this insight has been neglected, mainly because it is hard to model. Analyzing non‐trivial location patterns where multiple industries can sort across multiple locations is an inherently difficult problem (Behrens and Robert‐Nicoud , Gaubert ). The problem is made difficult by the fact that both the sizes and the composition of locations are endogenous, whereas productivity depends on both of these variables.…”

“…Returning to the example of matching, only industries and workers for which matching is more important-or industries that face greater uncertainty and "competitive instability" (Strange et al 2006)-will pay these higher costs and tend to coagglomerate. See Davis and Dingel (2014) and Gaubert (2016) for recent treatments of this problem using the tools of monotone comparative statics and Behrens and Robert-Nicoud (2015) for a survey.…”

“…)—will pay these higher costs and tend to coagglomerate. See Davis and Dingel (2014) and Gaubert () for recent treatments of this problem using the tools of monotone comparative statics and Behrens and Robert‐Nicoud () for a survey.…”