Determinants of Corruption in the Balkan Countries: An Empirical Analysis

Abstract

Purpose: Corruption represents one of the most serious problems faced by developing countries, particularly those with a lack of democratic potential. The aim of this paper is to examine, through the application of a dynamic statistical model, the long-term and short-term effects of economic freedom, democracy, and the level of economic development on the level of corruption in the case of Balkan countries.

Мethodology: The research is based on panel data for 10 Balkan countries. The observed dependent variable is the World Bank’s Control of Corruption indicator, while the independent variables include the Economic Freedom of the World index published by the Fraser Institute, the Electoral Democracy Index published by the V-Dem Institute, and the level of economic development measured by GDP per capita from the World Bank database. The analysis employs an ARDL model, which enables separate examination of the long-term and short-term effects of the independent variables on the dependent variable.

Findings: A statistically significant impact of electoral democracy, economic freedom, and the level of economic development on the level of corruption in the long run has been confirmed. Only economic freedom also has an effect in the short run, with the direction of the effect remaining unchanged.

Originality/value: This is the first study of these determinants of corruption conducted on a Balkan sample, and, to the author’s knowledge, it is only the second paper in the overall literature to be based on the application of the ARDL model.

Practical implications: Research on the determinants of corruption always has practical relevance in the fight against corruption. Confirmation that corruption is a consequence of a lack of democracy and economic freedom in the Balkans provides guidance for the necessary reforms of the economic and political system.

Limitations: A practical limitation of the study relates to the relatively small number of observed variables, which is a consequence of the choice of the statistical model. Because including a large number of variables would lead to the problem of model overidentification.