MULTIVARIJATNA ANALIZA POKAZATELJA DRUSTVENO-EKONOMSKOG RAZVOJA U ODABRANIM EUROPSKIM ZEMLJAMA/MULTIVARIATE ANALYSIS OF SOCIO-ECONOMIC DEVELOPMENT INDICATORS IN SELECTED EUROPEAN COUNTRIES

Headnote

SAZETAK: U ovome radu provedene su faktorska i klaster analiza pokazatelja odrZivog drustveno-ekonomskog razvoja u odabranim europskim zemljama. Faktorska analiza je pokazala da se iz odabranih devet pokazatelja mogu izlueiti dva faktora od kojih je jedan visoko koreliran s nezaposlenoscu, a drugi s BDP-om po stanovniku, izdacima za istraZivanje i razvoj i stopom zaposlenosti, eime je potvrdena prva istraZivaeka hipoteza. Klaster analiza Wardovom metodom i K-means metodom pokazala je da se promatrane zemlje mogu klasificirati u tri klastera prema BDP-u, izdacima za istraZivanje i razvoj i stopi zaposlenosti. Medutim, spomenute metode pojedine zemlje svrstavaju u medusobno razlieite klastere. U oba slueaja klasteru najslabije razvijenih zemalja pripadaju zemlje kandidatkinje: Hrvatska i Turska, dio zemalja koje su pristupile Europskoj uniji nakon 2004. godine, te dvije elanice EU-15: Italija i Spanjolska. Italiju i Spanjolsku u razvoju su prestigle sljedeee zemlje koje su pristupile Europskoj uniji 2004. godine, a koje se na1aze u drugom klasteru po razvijenosti: Slovenija, Estonija, Ceska i Cipar.

KLJUCNE RIJECI: Faktorska analiza glavnih komponenata, varimax raw rotacija faktora, klaster analiza, K-means metoda, Wardova metoda, pokazatelji drustveno-ekonomskog razvoja.

ABSTRACT: Factor analysis and cluster analysis of sustainable socio-economic development indicators in the European Union member countries and some candidate countries are conducted in this paper. The main subject of this paper is conducting the. The factor analysis showed that two factors can be extracted out of nine selected indicators: one which is highly correlated to unemployment, and the other to the GDP per capita, research and development expenditures and employment rate, what confirms first research hypothesis. Implementation of the cluster analysis using the Ward method and the K-means method showed that he observed countries can be classified into three clusters according to the GDP, research and development expenditures and employment rate. However, mentioned methods classify some countries in different clusters. In both cases, the cluster of least developed countries comprises candidate countries: Croatia and Turkey, part of the countries that joined the EU after 2004, as well as two EU-15 members: Italy and Spain. Italy and Spain are outrun by countries of second cluster that joined the EU in 2004: Slovenia, Estonia, Czech Republic and Cyprus.

KEY WORDS: Principal components factor analysis, varimax raw factor rotation, cluster analysis, K-means method, Ward method, socio-economic development indicators.

...