Research notes · regions

Do the best firms make cities productive, or do productive cities attract the best firms?

What the authors found

Gaubert builds a theory of where heterogeneous firms locate across cities and sectors. In equilibrium the distribution of city sizes and the sorting of firms are jointly determined: more productive firms sort into larger cities because they gain more from agglomeration, which in turn makes those cities more productive. The policy implication is uncomfortable — subsidising firms to locate in small cities moves them away from where they are most productive, and the aggregate cost can be large.

What our data adds

The business demography series here counts active enterprises and enterprise births across 1,249 NUTS-3 regions, which is the raw material for exactly this question at European scale. What it cannot supply is the firm-level productivity that does the sorting: it counts firms, not how good they are. So it can show where firms are and where new ones appear, and it is silent on why the good ones are where they are — which is the mechanism the paper is about.

What this cannot show

Enterprise counts treat a shipyard and a hairdresser as one each. Without productivity, size or sector detail there is no way to distinguish sorting from mere density, and both theories the paper adjudicates between predict that firms cluster. This is a note about a dataset that gets you to the doorstep of a question and no further.

The data behind it

Related notes

• Being near other firms helps — but how near is near enough?
• When jobs appear in one place, who actually gets them?
• Why have big cities pulled away from everywhere else since the 1990s?

Was this useful?

It answered my question Partly No

No account, no comment box. One click, counted and nothing else.