PREDICTING AND EXPLAINING CORRUPTION IN THE MENA REGION: A MACHINE LEARNING APPROACH

Abstract

Corruption is still pervasive and is considered one of the greatest challenges facing modern societies. Many academic studies have attempted to identify and explain the causes and potential consequences of corruption, primarily through theoretical lenses using correlation and regression-based statistical analyses. The present study approaches the phenomenon from the predictive analytics perspective by employing contemporary machine learning techniques to discover the most important corruption perception predictors based on enriched/enhanced nonlinear models with a high level of predictive accuracy. Specifically, within the regression modeling setting employed herein, the Random Forest (an ensemble-type machine learning algorithm) was found to be the most accurate prediction model, followed by Gradient Boosting and CART. Practically, the increased predictive power of machine learning algorithms coupled with a multi-source database revealed the most applicable corruption-related information, contributing to the body of knowledge and generating actionable information for administrators, academics, citizens, and politicians. The variable importance results indicated that trading across borders, time (men in days), getting electricity, and cost (% of warehouse value) are the most influential factors in defining the corruption level of significance.