Acta Universitatis Lodziensis. Folia Oeconomica (May 2014)

RANKING THE STANDARDS OF LIVING IN DISTRICTS IN POLAND BETWEEN 2003 AND 2012 INCLUDING SPATIAL CORRELATION

  • Marek Sobolewski,
  • Aldona Migała-Warchoł,
  • Grzegorz Mentel

Journal volume & issue
Vol. 6, no. 308

Abstract

Read online

In this paper, we present the ranking of living standards in Poland in a cross-section of counties as constructed by using selected methods of linear ordering. The analysis was of a dynamic character and the study involved data from the years 2003 to 2012. The positions of the counties in the annual section rankings were distinguished. Additionally, we assessed the trends of the synthetic measure of the standard of living for each country between 2003-2012. An important issue when creating rankings was determining the spatial extent of the study. For example, when examining the standard of living in the counties of the Podkarpackie province, should not the counties in other provinces be taken into account as a reference point? In the case of multivariate analysis, the interaction of objects can indeed be changed by an extended context analysis. Therefore, in this paper, comparisons made of the rankings compliance obtained in the analysis were narrowed down to a single province with the results of the analyses being carried out in the broader context of all the counties in Poland. As with today’s data processing, it is not important whether the population of a few dozen or a few hundred objects is analysed. The conclusion from the conducted considerations is that one should always strive for the widest possible context for the research. We then propose a modification of the linear ordering method that takes the spatial relationships between the districts into account. Two rankings were presented: one where the neighbourhood matrix between the counties was applied and one where the length of the shared border was considered. In the discussion of the results, we highlighted the fact that spatial relationships should be determined separately for each diagnostic variable. In this way, the directions for further research were determined.

Keywords