-
Convergence and cluster structures in EU area according to fluctuations in macroeconomic indices
Authors:
Mircea Gligor,
Marcel Ausloos
Abstract:
The cluster analysis methods are used in order to perform a comparative study of 15 EU countries in relation with the fluctuations of some basic macroeconomic indicators. The statistical distances between countries are calculated for various moving time windows, and the time variation of the mean statistical distance is investigated. The decreasing of the mean statistical distance between EU cou…
▽ More
The cluster analysis methods are used in order to perform a comparative study of 15 EU countries in relation with the fluctuations of some basic macroeconomic indicators. The statistical distances between countries are calculated for various moving time windows, and the time variation of the mean statistical distance is investigated. The decreasing of the mean statistical distance between EU countries is reflected in the correlated fluctuations of the basic ME indicators: GDP, GDP/capita, Consumption and Investments. This empirical evidence can be seen as an economic aspect of globalization. The Moving Average Minimal Length Path (MAMLP) algorithm allows to search for a cluster-like structures derived both from the hierarchical organization of countries and from their relative movement inside the hierarchy. It is found that the strongly correlated countries with respect to GDP fluctuations can be partitioned into stable clusters. Some of the highly correlated countries, with respect to GDP fluctuations, display strong correlations also in the Final Consumption Expenditure, while others are strongly correlated in Gross Capital Formation. On the other hand, one notices the similitude of the classifications regarding GDP and Net Exports fluctuations as concerns the squared sum of the correlation coefficients (so called country sensitivity). The final structure proves to be robust against the constant size time window moving over the scanned time interval. The policy implications of the above empirical results concern the economic clusters arising in the presence of Marshallian externalities and the relationships between trade barriers, R&D incentives and growth that must be accounted in elaborating a cluster-promotion policy.
△ Less
Submitted 20 May, 2008;
originally announced May 2008.
-
Cluster Expansion Method for Evolving Weighted Networks Having Vector-like Nodes
Authors:
Marcel Ausloos,
Mircea Gligor
Abstract:
The Cluster Variation Method known in statistical mechanics and condensed matter is revived for weighted bipartite networks. The decomposition of a Hamiltonian through a finite number of components, whence serving to define variable clusters, is recalled. As an illustration the network built from data representing correlations between (4) macro-economic features, i.e. the so called $vector$…
▽ More
The Cluster Variation Method known in statistical mechanics and condensed matter is revived for weighted bipartite networks. The decomposition of a Hamiltonian through a finite number of components, whence serving to define variable clusters, is recalled. As an illustration the network built from data representing correlations between (4) macro-economic features, i.e. the so called $vector$ $components$, of 15 EU countries, as (function) nodes, is discussed. We show that statistical physics principles, like the maximum entropy criterion points to clusters, here in a (4) variable phase space: Gross Domestic Product (GDP), Final Consumption Expenditure (FCE), Gross Capital Formation (GCF) and Net Exports (NEX). It is observed that the $maximum$ entropy corresponds to a cluster which does $not$ explicitly include the GDP but only the other (3) ''axes'', i.e. consumption, investment and trade components. On the other hand, the $minimal$ entropy clustering scheme is obtained from a coupling necessarily including GDP and FCE. The results confirm intuitive economic theory and practice expectations at least as regards geographical connexions. The technique can of course be applied to many other cases in the physics of socio-economy networks.
△ Less
Submitted 28 February, 2008;
originally announced February 2008.
-
Clusters in weighted macroeconomic networks : the EU case. Introducing the overlapping index of GDP/capita fluctuation correlations
Authors:
Mircea Gligor,
Marcel Ausloos
Abstract:
GDP/capita correlations are investigated in various time windows (TW), for the time interval 1990-2005. The target group of countries is the set of 25 EU members, 15 till 2004 plus the 10 countries which joined EU later on. The TW-means of the statistical correlation coefficients are taken as the weights (links) of a fully connected network having the countries as nodes. Thereafter we define and…
▽ More
GDP/capita correlations are investigated in various time windows (TW), for the time interval 1990-2005. The target group of countries is the set of 25 EU members, 15 till 2004 plus the 10 countries which joined EU later on. The TW-means of the statistical correlation coefficients are taken as the weights (links) of a fully connected network having the countries as nodes. Thereafter we define and introduce the overlapping index of weighted network nodes. A cluster structure of EU countries is derived from the statistically relevant eigenvalues and eigenvectors of the adjacency matrix. This may be considered to yield some information about the structure, stability and evolution of the EU country clusters in a macroeconomic sense.
△ Less
Submitted 29 October, 2007;
originally announced October 2007.
-
Cluster structure of EU-15 countries derived from the correlation matrix analysis of macroeconomic index fluctuations
Authors:
M. Gligor,
M. Ausloos
Abstract:
The statistical distances between countries, calculated for various moving average time windows, are mapped into the ultrametric subdominant space as in classical Minimal Spanning Tree methods. The Moving Average Minimal Length Path (MAMLP) algorithm allows a decoupling of fluctuations with respect to the mass center of the system from the movement of the mass center itself. A Hamiltonian repres…
▽ More
The statistical distances between countries, calculated for various moving average time windows, are mapped into the ultrametric subdominant space as in classical Minimal Spanning Tree methods. The Moving Average Minimal Length Path (MAMLP) algorithm allows a decoupling of fluctuations with respect to the mass center of the system from the movement of the mass center itself. A Hamiltonian representation given by a factor graph is used and plays the role of cost function. The present analysis pertains to 11 macroeconomic (ME) indicators, namely the GDP (x1), Final Consumption Expenditure (x2), Gross Capital Formation (x3), Net Exports (x4), Consumer Price Index (y1), Rates of Interest of the Central Banks (y2), Labour Force (z1), Unemployment (z2), GDP/hour worked (z3), GDP/capita (w1) and Gini coefficient (w2). The target group of countries is composed of 15 EU countries, data taken between 1995 and 2004. By two different methods (the Bipartite Factor Graph Analysis and the Correlation Matrix Eigensystem Analysis) it is found that the strongly correlated countries with respect to the macroeconomic indicators fluctuations can be partitioned into stable clusters.
△ Less
Submitted 11 July, 2006;
originally announced July 2006.
-
Statistical distances between countries and cluster structures in EU area according to macroeconomic indices fluctuations
Authors:
Mircea Gligor,
Marcel Ausloos
Abstract:
The paper applies some recent developments of network analysis in order to perform a comparative study of EU countries in relation with the fluctuations of some macroeconomic indicators. The statistical distances between countries, calculated for various moving average time windows, are mapped into the ultrametric subdominant space as in classical Minimal Spanning Tree methods. The novelty consi…
▽ More
The paper applies some recent developments of network analysis in order to perform a comparative study of EU countries in relation with the fluctuations of some macroeconomic indicators. The statistical distances between countries, calculated for various moving average time windows, are mapped into the ultrametric subdominant space as in classical Minimal Spanning Tree methods. The novelty consists in applying the so-called Moving Average Minimal Length Path algorithm, which allows a decoupling of fluctuations with respect to the mass center of the system from the movement of the mass center itself. The present analysis pertains to the Gross Domestic Product and some of its sources for 15 EU countries between 1972 and 2004. It is found that the strongly correlated countries with respect to GDP fluctuations can be partitioned into stable clusters. The final structure proves to be stable against the fluctuations induced by the moving time window over the scanned time interval.
△ Less
Submitted 23 June, 2006;
originally announced June 2006.
