Spatial and spatiotemporal models of unemployment rate in Poland

Abstract

The article show interest in modelling of economic processes which are characterized by spatial dependence. An unemployment rate indicating a level of socio-economic development of regions constitutes one of such processes. The correct quantitative description of this process, as one of more important negative socio-economic phenomena, is an important task of econometrics, in particular, of spatial econometrics. The purpose of the article is to provide a spatial analysis of the unemployment rate registered at the end of each year in Polish poviats. The research covers the period of 2004-2009. Two approaches to modelling are used - spatial and spatiotemporal. The neighbourhood matrix based on the common border of poviats is applied to reflect the strength and direction of spatial relations. There are used two exogenous variables: investments and entities of the economy recorded, both per capita. In the spatial as well as in the spatiotemporal approach the main tendencies of the change of the unemployment rate are described with a spatial linear trend. All of the received models have good economic and statistic properties and they could be used for further analysis.