2. Agglomeration and Human Capital Externalities: empirical evidence
In our model, the interaction between high-skilled workers in a region increases the
productivity of each worker by means of a knowledge diffusion process. Regions with a
higher average level of human capital are therefore more attractive for high-skilled workers.
The skill premium is endogenous and increasing in the regional quota of high-skilled
workers. One of the implications of the model is the existence of a wage and productivity
differential between the core and peripheral regions.
Economic theory gives several explanations on why nominal wages may be different
across regions in the short run, one of the main reasons being the existence of asymmetrical
regional shocks and, overtime, divergences of regional business cycles. In the long run, after
the necessary adjustments in local labour markets have taken place, regional nominal wages
should, in principle, converge.
One possible explanation for the observed lack of convergence in the long run is
linked to the existence of differences in the regional “endowment” of amenities. Local
amenities, such as good climate, favorable physical morphology of the area, air and water
quality, may affect regional wages through two main channels. First, through a positive (or
negative) effect on consumer utility. For example if consumer values the warm and sunny
weather of Southern Italy, they will require a wage premium to live and work in rainy and
cloudy climate of Northern Italy. As a result, ceteris paribus, we expect workers of similar
characteristics willing to accept lower salaries in region with pleasant weather conditions.
Secondly, local non-exclusive amenities may also have a direct influence on labour
productivity. If the effect is positive we will expect higher wages. The attractiveness of the
area, due to a positive regional wage differential, will induce migration, and therefore these
productivity differences will be capitalised into differentials in land rents.