Multivariate Analysis of Structural Economic Indicators for Croatia and EU 27 (Zivadinovic, 2009)

8/10/2019 Multivariate Analysis of Structural Economic Indicators for Croatia and EU 27 (Zivadinovic, 2009)

1/6


2/6

economics restructuring, privatisation andliberalization. These include a sharp fall inindustrial employment and a substantial rise inservice sector employment, but noticeabledifferences with the employment structure of theEU Member States remain. Unemployment hasrisen in all CEE countries to varying extents.Income levels and standard of living have declinedand poverty has spread considerably with avariation between countries and a disproportionaleffect on certain social groups. The spread of sub-national disparities in GDP and unemployment inthe CEECs is smaller than in other EU MemberStates.

According to the previously mentioned research,disparity patterns (at NUTS II level) include thefollowing: GDP per capita in CEE regions is

considerably less than EU average (only Pragueand Bratislava lie above this level), regionalunemployment is relatively low in CEE incomparison to the EU 15 (with noticeable sub-national variation), CEE regions are in generalmore sparsely populated then the EU 15 andagriculture dominates regional employmentstructures in, for example, Romania and Poland tomuch greater extent than in the EU 15. However,the increasing uncertainties regarding the EUabsorption capacity and its future enlargements, aswell as unsorted institutional issues seem not to be

affecting Croatia`s current path towards theaccession [7].Croatia`s small size causes little concern about

the impact it would have on EU institutions,policies and its budget. Therefore it has beenrepeatedly confirmed by EU officials that Croatiawould join the EU as quickly as possible, providedthat it fulfils all the required accession criteriawhich primarily relate to the progress with adoptingand implementing the EU law. In some areas,however, they also include broader political andeconomic reforms [4].

2 Problem FormulationIn this paper the structural indicators of Croatianeconomy (CR) were analysed: GDP per capita(GDPpc), total employment rate (EMPL) andcomparative price levels (PRICE) in comparisonwith those of: Belgium (BE), France (FR), Italy(IT), Greece (GR), Spain (SP), Czech Republic(CZ), Lithuania (LI), Estonia (ES), Latvia (LA),Cyprus (CY), Portugal (PT), Slovenia (SN),

Bulgaria (BU), Hungary (HU), Poland (PL),Romania (RO), Slovakia (SK), Malta (MA),Denmark (DE), Germany (GE), Austria (AU),

United Kingdom (UK), Netherlands (NE), Sweden(SW), Ireland (IR), Finland (FI) and Luxembourg(LU). Using cluster analysis, the main purpose ofthe paper was to explore in which group ofcountries Croatia fits in based on enumeratedstructural economic indicators [1, 2, 3, 5, 6]. Thedata for the analysis were taken from Eurostat website for the year2007.

As an indicator of economic activity, Grossdomestic product (GDP) was chosen. It is definedas the value of all goods and services produced lessthe value of any goods or services used in theircreation. The volume index of GDP per capita inPurchasing Power Standards (PPS) is expressed inrelation to the European Union (EU-27) average setto equal 100. Basic figures are expressed in PPS,i.e. a common currency that eliminates the

differences in price levels between countriesallowing meaningful volume comparisons of GDPbetween countries.

Another structural indicator of interest is totalemployment rate. The employment rate iscalculated by dividing the number of persons aged15 to 64 in employment by the total population ofthe same age group. The indicator is based on theEU Labour Force Survey. The survey covers theentire population living in private households andexcludes those in collective households such as

boarding houses, halls of residence and hospitals.

Employed population consists of those persons whoduring the reference week did any work for pay orprofit for at least one hour, or were not working buthad jobs from which they were temporarily absent.

Comparative price level is the last indicatorchosen for this analysis.Comparative price level isthe ratio between Purchasing power parities (PPPs)and market exchange rate for each country. Theratio is shown in relation to the EU average (EU27= 100).

Although this analysis is accompanied by anumber of constraints that have to be taken into

account when interpreting the results, it is quiteinteresting to know in which groups of EU 27countries was Croatia classified.

3 Problem SolutionThe cluster analysis was applied to classify EU 27countries and Croatia according to the threestructural economic indicators: GDP per capita,total employment rate and comparative price levels.

At first, three previously mentioned variables

were standardized to avoid measurementdifferences.

Proceedings of the 2nd WSEAS International Conference on Multivariate Analysis and its Application in Science and Engineering

ISSN: 1790-5117 135 ISBN: 978-960-474-083-3


3/6


4/6


5/6

Lithuania, Estonia, Latvia, Cyprus and Slovenia.These results confirmed the results of clusteranalysis on 28 countries.

3.4 Non-hierarchical cluster analysis for 13

countriesThe K-means method of non-hierarchical clusteranalysis with the Euclidean distances was providedon standardized variables (SGDPpc, SEMPL andSPRICE). Table 2 shows the classification ofcountries in two clusters given by the Wardsmethod and the K-means method.

Table 2 Classification of countries in two clusters(Wards method and K-means method)

CountryWards

method

K-means

methodBulgaria 1 1Hungary 1 1Croatia 1 1Poland 1 1Romania 1 1Slovakia 1 1Malta 1 1Czech Republic 2 2Lithuania 2 2Estonia 2 2Latvia 2

2Cyprus 2 2

Slovenia 2 2

It can be seen that the K-means method wasresulted in the same structure of the clusters as theclusters given by the Ward's method with thesquared Euclidean distances.

Figure 4 shows the plot of means for the twoclusters obtained by the K-means method.

Fig.4 Plot of means for two clustersPlot of Means for Each Cluster

Cluster 1Cluster 2sgdp sempl sprice

Variables

-2,0

-1,5

-1,0

-0,5

0,0

0,5

1,0

1,5

Figure 4 shows that in the first cluster, in whichCroatia was grouped, all of the three examinedstructural economic indicators are below average.

4 ConclusionThe aim of this paper was to classify Croatia andEU 27 Member States according to structuraleconomic indicators. The cluster analysis was usedto classify those countries into groups which arehomogenous in terms of the three structuraleconomic indicators: GDP per capita, totalemployment rate and comparative price levels.Because of the measurement differences these threevariables were firstly standardized. The hierarchicalcluster analysis and non-hierarchical cluster

analysis were used.Various methods of hierarchical cluster analysiswere first provided to find out the number ofclusters. The best interpretative solution was

provided by the Wards method with the squaredEuclidean distances. The four-cluster solution given

by the Wards method with squared Euclideandistances was chosen. According to the results ofthe Wards method and the three chosen structuraleconomic indicators Croatia was classified alongwith the following EU Member States: Bulgaria,Hungary, Poland, Romania, Slovakia and Malta

A non-hierarchical cluster analysis was thenemployed to improve the results of the four-clustersolution given by the hierarchical cluster analysis.The K-means method was resulted in the similarstructure of the clusters as the clusters given by thehierarchical cluster analysis. On the basis of thethree chosen structural economic indicators and K-means method Croatia was classified into the groupof the following EU 27 Member States: Bulgaria,Hungary, Poland, Romania, Slovakia, Malta andLithuania.

The hierarchical cluster analysis and non-hierarchical cluster analysis were run to classifyCroatia and 12 European countries which joinedduring the last two waves of enlargement. Againthe Croatia was grouped along with Bulgaria,Hungary, Poland, Romania, Slovakia and Malta.

The result of analysis is quite expected, sinceCroatia was grouped with the countries with similareconomic and historical background. However itwas interesting to test that fact methodologically.Further analysis on this particular area will includemore economic structural indicators that might lead

to conclusion in what way will Croatian economybe affected with the accession to European Union.


ISSN: 1790-5117 138 ISBN: 978-960-474-083-3


6/6

References:

[1] Everitt, B. S., Landau, S., Leese, M., ClusterAnalysis, 4th Ed, London, Arnold, 2001

[2] Johnson, R.A., Wichern, D., AppliedMultivariate Statistical Analysis. PearsonInternational Education, 2007

[3] Kurnoga ivadinovi, N., Multivarijatnaklasifikacija upanija Hrvatske, ZbornikEkonomskog fakulteta u Zagrebu, Ekonomskifakultet Zagreb, 2007, pp. 1-15.

[4] Lejour, A., Mervar, A., Verweij, G., TheEconomic Effects on Croatia`s Accession tothe EU, EIZ Working Papers, EIZ-WP-0705,Zagreb, December 2007

[5] Liou, F. M., Ding, C. G., Subgrouping SmallStates Based on Socioeconomic

Characteristics, World Development, Vol. 30,No. 7, 2002, pp. 1289-1306.[6] Soares, J. O., Marques, M. M. L., Monteiro,

C. M. F., A Multivariate Methodology toUncover Regional Disparities: AContribution to Improve European Union andGovernmental Decisions, European Journalof Operational Research, No. 145, 2003, pp.121-135.

[7] Weise, C., The Impact of EU Enlargement onCohesion, DIW Berlin, German Institute forEconomic Research, Oporto, October 2001

[8] http://epp.eurostat.ec.europa.eu[9] http://www.mvpei.hr


ISSN: 1790-5117 139 ISBN: 978-960-474-083-3
http://epp.eurostat.ec.europa.eu/http://epp.eurostat.ec.europa.eu/http://www.mvpei.hr/http://www.mvpei.hr/http://epp.eurostat.ec.europa.eu/

Documents

Multivariate Analysis of Structural Economic Indicators for Croatia and EU 27 (Zivadinovic, 2009)