Going Beyond Analysis of Internal Data to Support Customs Modernization: A Case Study in Gabon

Abstract

Customs administrations in developing countries increasingly use risk-based techniques relying on data mining and statistical scoring. By demonstrating the value of using data analysis techniques to orient frontline controls so as to facilitate legal trade and combat fraud more effectively, these projects have helped promote a cultural change in these organizations. However, these risk management techniques may prove to be ineffective in assessing fraud risks based only on frauds detected by customs inspectors. In a context of moral hazard and low-performing customs administration, one way to address this weakness is to expand the approach by relying on other sources of information such as discrepancies in bilateral trade statistics. Several studies use these statistical discrepancies (mirror data) to identify fraudulent declarations and estimate their effects. By comparing Gabon's import customs data with discrepancies in its bilateral trade data, this paper stresses the usefulness of simultaneously analyzing customs fraud records and mirror trade statistics data. Such an analysis helps quantifying undetected fraud and therefore constitutes a valuable tool to target ex post audits. Then, based on the combination of these databases, the paper defines indicators to monitor the performance of customs controls.