Spatial and Non-spatial Approaches to Agricultural Convergence in the EU

In the European Union, real convergence has a political and financial importance and, in this context, a renewed interest is deserved to the contribution of agriculture. The theoretical debate on the issue has evolved over time also supported by the empirical literature and the improvement in the investigation techniques.

The prevailing contributions on economic growth are referred to the one-sector economic growth model by Solow according to which, due to diminishing marginal returns of input factors in a production function with constant return to scale, economies converge towards a dynamic long-run steady state, a trend only driven by the rate of technological progress.

On the basis of this perspective Barro and Sala-i-Martin have elaborated the concept of absolute b-convergence and Mankiw et al. that of conditional b-convergence that have given rise to a wide body of empirical investigations aimed at detecting the neoclassical hypothesis of convergence by OLS and panel models, addressing, more recently, the issues related to unit root tests in panel data.

In addition, the theoretical literature has focused on spatial effects. As knowledge is entirely disembodied and, technically, understood as pure public goods, in the neoclassical perspective distance does not play a role in the convergence process. This assumption has been strongly criticised by the theoretical and empirical literature. The economic geography theory usually assumes that knowledge is a regional public good with limited spatial range and plays the same role as classical spatial interaction related to population, capital and material. From this assumption there is the possibility of different paths of regional growth and of coexistence of divergent and convergent groups of territorial units. This theoretical perspective explains the agglomeration process and the geographic spillover in the interaction between geography and growth. To address the issue, new techniques of analysis have been introduced. They can be divided into univariate and multivariate spatial data statistics.

In the light of these considerations, the paper provides a critical review of the spatial and non spatial techniques to convergence, within the evolution of the theoretical debate, and the application of the most important spatial and non spatial panel approaches to a sample EU regions in order to emphasise their specific contribution to the understanding of the process of agricultural convergence and their limitations particularly from a policy perspective.