How helpful are mirror statistics for customs reform? Lessons from a decade of operational use

Abstract

From a tool for trade economists, mirror data, which consist of comparing export and import data reporting by exporting and importing countries, have been increasingly used to detect potential fraud in developing countries. This dramatic change can be explained by the simplicity and low cost of using open worldwide databases. Even though mirror data discrepancies do not guarantee fraud records, this approach contributes to improved risk analysis (on origins and tariff lines, as well as brokers/importers, locations and inspectors), estimates of potential revenue losses, and new dialogue between customs administrations, importers, brokers and, sometimes, political authorities. The use of mirror statistics in developing countries is, therefore, promising and should be used increasingly for customs reforms. However, this statistical tool is even more efficient with other reforms, such as human resources reforms.