Document Type : ORIGINAL ARTICLE
Authors
1 Master of Economic Development and Planning, University of Tabriz
2 Professor, Department of Economic Sciences, Faculty of Economics and Management, University of Tabriz, Tabriz, Iran
3 Associate Professor, Department of Economic Sciences, Faculty of Economics and Management, University of Tabriz, Tabriz, Iran
Abstract
Public debt in Iran’s economy has become one of the central challenges of fiscal policymaking. This study aims to identify the determinants of Iran’s public debt using a Bayesian regression approach and machine-learning algorithms. Twenty-seven explanatory variables, grouped into six categories—international, monetary, demographic, public sector, political, and other variables—were entered into the model for the period 1979–2024 (1358–1402). The methodology relies on two Bayesian algorithms, BACE and BAS, while the random forest algorithm is used to assess the robustness of the results. The findings of BACE and BAS indicate Iran’s public debt is a multifactorial phenomenon and that no single variable can fully explain it. Based on posterior inclusion probability (PIP), total employment, export value, oil revenues, the free-market exchange rate, the capital account balance, and population show the strongest evidence. By contrast, political and institutional variables such as the democracy index, the corruption perceptions index, and the Gini coefficient display weaker independent and persistent effects in the presence of other macroeconomic variables. The random forest results confirm these findings, showing that employment, oil revenues, export value, and the free-market exchange rate make the largest contributions to predicting Iran’s public debt. Public debt in Iran is shaped primarily by complex interactions among the external sector, the labor market, oil revenues, monetary conditions, and demographic pressures. Accordingly, when designing public debt management strategies, policymakers should adopt multidimensional strategies and manage the multifaceted interactions among variables.
Keywords
Main Subjects
)