The Role of Economic and Political Features in Classification of Countries-in-Transition by Human Development Index

• Main Authors: Jani KINNUNEN, Armenia ANDRONICEANU, Irina GEORGESCU
• Format: Article
• Language: English
• Published: Inforec Association 2019-01-01
• Series: Informatică economică
• Subjects: artificial neural network; democracy; governance; human development index; logistic regression multiple correspondence analysis; market economy
• Online Access: http://revistaie.ase.ro/content/92/03%20-%20kinnunen,%20georgescu.pdf

Typical classification research of the United Nations’ Human Development Index, HDI, has focused on its direct underlying sub-indices, i.e., Gross National Income, GNI, Education and Health. In this paper, economic and political systems within which the elements of HDI are created are under study. We use Bertelsmann Foundation’s qualitative data from period 2008-2016 on 124 countries-in-transition including features of market economy, democracy and governance. The purpose is to identify the most important economic and political features predicting the level of HDI and to compare the classification performances of the applied models: an artificial neural network, ANN, and a logistic regression. The multi-method approach is complemented by multiple correspondence analysis for descriptive analysis purposes. The main results and original contributions include proving the effectiveness of the ANN over the logistic regression and showing that the higher levels and specific individual factors of marked economy, governance and democracy, the higher the HDI.