Working Paper: CEPR ID: DP17806
Authors: Patricia Rice; Anthony Venables
Abstract: Spatial variation in the productivity of different sectors is a determinant of sectoral location, with consequences for wages, rents and the cost-of-living in each area. This paper develops an analytical framework which shows how productivity advantage in a highly tradable sector translates into higher nominal wages, rents, and cost of living in an area; in contrast, high physical productivity in non-tradables may result in lower wages, rents and revenue productivity. The theory’s prediction that an area’s bias towards highly tradable activities is positively correlated with its earnings is confirmed by empirical analysis of earnings data for the ITL3 areas of GB. As suggested by the theory, two factors drive this effect. Approximately one-third is a direct result of sectoral composition – on average across GB, tradable sectors pay higher wages. The remaining two-thirds is an equilibrium effect, arising as a productivity advantage in tradables translates into higher local employment and factor prices. While our primary analysis is on recent data, we show that our approach also captures the impact of the structural change that occurred in Britain during the 1970s and 1980s on regional earnings disparities.
Keywords: regional economics; lagging regions; tradability; productivity; disparities
JEL Codes: R11; R12; E62; F60
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| productivity advantage in tradable sectors (F14) | higher nominal wages (J39) |
| productivity advantage in tradable sectors (F14) | higher rents (R21) |
| productivity advantage in tradable sectors (F14) | higher cost of living (J39) |
| higher nominal wages (J39) | increased local employment (J68) |
| increased local employment (J68) | higher factor prices (F16) |
| higher factor prices (F16) | higher cost of living (J39) |
| high physical productivity in nontradable sectors (O49) | lower wages (J31) |
| high physical productivity in nontradable sectors (O49) | lower rents (R21) |
| sparsity bias (C21) | spatial variation in average earnings (R12) |
| sparsity bias (C21) | average earnings (J31) |