European Regional Science Association
|
European Regional Science Association
|
|
|
|
|
The abstract for paper number 365:
Guido Pellegrini, Barbara Pacini, Dept. of Statistics, University of Bologna, Bologna, Italy
Proximity and employment rate regional distribution: a non parametric multivariate approach
Quah (1997a,b) proposed a new methodological approach to empirics of economic dynamics based on nonparametric analysis of the complete distribution of growth variables across countries and regions. Quah suggests a two basic methodological tools: the stochastic kernel, a mathematical operator mapping distributions to distributions, and the conditioning scheme, that is analogous to explain, in a regression, a left-hand side variable by some right-hand side conditioning regressors. In models of distribution dynamics, the variable to be explained on the left-hand side is conceptually an entire distribution, rather than some random variable with its own distribution. Conditioning scheme has been applied by Quah and followers only in a univariate context, but the shape of growth variable distribution is affected by several factors: a multivariate approach is needed in order to assess their influence. The aim of this paper is to propose a feasible multivariate approach to the conditioning scheme. The basic idea is to exploit the similarities between a suitable multivariate conditioning scheme and a local nonparametric estimator of a regression function. Therefore we consider different non parametric regression approaches to the estimation of the conditioned distribution. We present also an empirical application on local labour market growth at a very disaggregate territorial level. The conditioning scheme is applied to the distribution of (non agricultural) employment rate in 784 Italian local labour markets (LLMs) in the year 1996: we show how proximity, population density and agricultural employment share affect the shape of the employment rate distribution and explain its bimodality (the “twin peaks” using Quah’s terminology).
References
Quah D. T. (1997a), Empirics for Growth and Distribution: Polarization, Stratification and Convergence Clubs, Journal of Economic Growth, 2, March. Quah D. T. (1997b), Regional cohesion from local isolated actions: II. Conditioning, CEP Working Paper n. 379, December. Wand M.P. and M.C. Jones (1995), Kernel Smoothing, Chapman & Hall, London.
Keywords: Nonparametric multivariate analysis, conditioning, growth empirics.
Unfortunately full paper has not been submitted.