Insight into regional similarities of Serbian districts using appropriate statistical methods

Abstract

The territory of the Republic of Serbia is highly heterogeneous due to large differences in climatic, geo-morphological and socio-economic factors. In Serbia, there are big regional differences, and addressing the problem of balanced regional development and its financing is the question that will be actualized with the opening of negotiations for accession to the European Union. The objective of this paper is to present the methodology for the classification of administration districts of the Republic of Serbia based on their selected characteristics in order to perceive the similarities between regions in a way to provide the basis for the regional development policy. The classification of 25 Serbian districts was performed using the factor and cluster analysis, based on the eighteen selected socioeconomic, demographic and agricultural indicators. The factor analysis resulted in the extraction of factors which represent the main dimensions of the achieved development. The cluster analysis was performed based on these dimensions. The resulting clusters are homogenous and consist of districts with similar characteristics. The paper demonstrates that applying of appropriate statistical methods can provide an analytical tool for the separation of districts into groups with similar characteristics, which represent more appropriate framework for the creation and implementation of the regional development policy measures.

Područje Republike Srbije je izuzetno heterogeno, kako zbog vrlo različitih klimatskih i geo-morfoloških tako i ekonomsko-socijalnih faktora. U Srbiji su prisutne velike regionalne razlike, a rešavanje problema ravnomernog regionalnog razvoja i njegovog finansiranja su pitanja koja će biti aktuelizovana sa otpočinjanjem pregovora za pridruživanje Evropskoj uniji. Cilj rada je da prikaže metodologiju klasifikacije upravnih okruga Repu­blike Srbije na osnovu odabranih obeležja, kako bi se sagledala regionalna sličnost i dala osnova politici regionalnog razvoja. Klasifikacija 25 okruga Republike Srbije je izvršena primenom faktorske i klaster metode, a zasniva se na osamnaest odabranih društvenih, demografskih i poljoprivrednih pokazatelja. Korišćenjem faktorske analize dobijaju se faktori koji predstavljaju glavne dimenzije razvoja promatranogprostora. Klaster analizom se, na osnovu zastupljenosti svake od dobijenih dimenzija u svakom okrugu, isti svrstavaju u grupe (klastere). Tako dobijeni klasteri su homogeni i sastavljeni su od okruga sličnih karakteristika. Rad pokazuje da se korišćenjem odgovarajućih statističkih metoda mogu analitički izdvojiti grupe okruga sličnih obeležja, koje predstavljaju prikladniji objekt za primenu mera regionalne politike.