The IMF Conditionality and Corruption

• In Event: IMF, Conditionality, and Intellectual Authority

Abstract

Do International Monetary Fund (IMF) programs reduce corruption? Although this question has once again recently come to the limelight with the IMF’s withholding of funds from Ukraine due to corruption concerns in 2019, there is still very little systematic analysis on the impact of IMF loans on borrowing countries’ corruption levels, with anecdotal evidence suggesting that IMF programs may actually increase corruption. In this paper we argue that the extent and effectiveness of IMF anti-corruption measures are affected by: a) the recipient country’s political proximity to the major shareholders of the Fund, particularly the Unites States, b) the type of corruption in the country (centralized versus decentralized); and c) administrative turnover in the country. The paper also shows that even when IMF programs improve measurements of corruption, they may actually worsen the public perception of corruption in the borrowing country.

In order to test our theory in a large-N setting, we use a disaggregated approach to IMF programs and employ regression analysis on a novel dataset that codes the anti-corruption measures in all IMF programs from 1980 to 2015.

The paper contributes to the IMF literature by using a disaggregated approach to IMF conditionality and corruption.

Authors

• Merih Angin, Koç University

• Ayse Kaya, Swarthmore College

• Saliha Metinsoy, University of Groningen