Abstract
Besides to be an useful methodology for the detection of distribution dynamics of indicators, stochastic kernel methodology has been generalized to a regression-like rationale (Quah, 1997). The latter allows to determine how a distribution is influenced by a “factor”, through a “conditioning scheme” which is a set of rules stating how the original distribution is altered in order to obtain its “conditioned” version. This paper aims to study the influence of the “spatial factor” on distributions of selected agriculture impact indicators across EU NUTS2. The present work offers an empirical analysis of dynamics of selected indicators of agriculture across NUTS2. Our scope is to give an overlook of EU territorial discontinuities in order to point out time and space features related to “polarization”.The paper describes and explores different space-conditioning scheme (Quah, 1997) and compares their effects over original distributions in highlighting peculiar “local” behaviors of groups of territorial units.