Working Paper: NBER ID: w4781
Authors: Robert Dekle; Jonathan Eaton
Abstract: We use Japanese prefectural wage and land price data to estimate the magnitude of agglomeration effects in manufacturing and finance. We also examine the range of agglomeration effects by estimating the extent to which they diminish with distance, using a specification that encompasses the polar cases of purely local agglomeration economies, on the one hand, and national increasing returns to scale, on the other. We find that agglomeration effects are slightly stronger in financial services than in manufacturing, and that they diminish substantially with distance in either sector. Our estimates indicate that agglomeration effects can explain about 5.6 per cent of the growth in Japanese output per worker in manufacturing and about 8.9 per cent of the growth in output per worker in financial services during 1976-1988. Our estimates imply that, while the average elasticity of productivity with respect to agglomeration is between 10 and 15 per cent, agglomeration economies in the largest prefectures are nearly exhausted.
Keywords: Agglomeration; Productivity; Urban Economics; Land Prices
JEL Codes: R11; R12; O18
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Agglomeration (R11) | Productivity Growth (O49) |
| Agglomeration (R11) | Elasticity of Productivity (D24) |
| Distance (R12) | Agglomeration Economies (R11) |
| Agglomeration (R11) | Output per Worker (J54) |
| Agglomeration (R11) | Local Activity (R11) |
| Agglomeration (R11) | Nationwide Activity (R59) |