DP2191 How Regions Converge
|Author(s):||Francesco Caselli, Wilbur John Coleman II|
|Publication Date:||July 1999|
|Keyword(s):||Regional Convergence, Skill Acquisition, Structural Transformation|
|JEL(s):||O14, O18, O41|
|Programme Areas:||International Macroeconomics|
|Link to this Page:||cepr.org/active/publications/discussion_papers/dp.php?dpno=2191|
We present a joint study of the US structural transformation (the decline of agriculture as the dominating sector) and regional convergence (of Southern to Northern average wages). We find that empirically most of the regional convergence is attributable to the structural transformation: the nation-wide convergence of agricultural wages to non-agricultural wages, and the faster rate of transition of the Southern labor force from agricultural to non-agricultural jobs. Similar results describe the Mid-West's catch up to the North-East (but not the relative experience of the West). To explain these observations, we construct a model in which the South (Mid-West) has a comparative advantage in producing unskilled-labor intensive agricultural goods. Thus, it starts with a disproportionate share of the unskilled labor force and lower per capita incomes. Over time, declining education/training costs induce an increasing proportion of the labor force to move out of the (unskilled) agricultural sector and into the (skilled) non-agricultural sector. The decline in the agricultural labor force leads to an increase in relative agricultural wages. Both effects benefit the South (Mid-West) disproportionately since it has more agricultural workers. With the addition of a less-than-unit income-elasticity of demand for farm goods and faster technological progress in farming than outside of farming this model successfully matches the quantitative features of the U.S. structural transformation and regional convergence, as well as several other stylized facts on U.S. economic growth in the last century. The model does not rely on frictions on inter-regional factor mobility, since in our empirical work we find this channel to be less important than the compositional effects the model emphasizes.