Search

This paper presents a new methodology to detect corruption in customs and applies it to Madagascar’s main port. Manipulation of assignment of import declarations to inspectors is identified by measuring deviations from random assignment prescribed by official rules. Deviant declarations are more at risk of tax evasion, yet less likely to be deemed fraudulent by inspectors, who also clear them faster. An intervention in which inspector assignment was delegated to a third party validates the approach, but also triggered a novel manifestation of manipulation that rejuvenated systemic corruption. Tax revenue losses associated with the corruption scheme are approximately 3 percent of total taxes collected and highly concentrated among a select few inspectors and brokers.

This paper examines how providing better information to customs inspectors and monitoring their actions affects tax revenue and fraud detection in Madagascar. First, an instrumental variables strategy is used to show that transaction-specific, third-party valuation advice on a subset of high-risk import declarations increases fraud findings by 21.7 percentage points and tax collection by 5.2 percentage points. Second, a randomized control trial is conducted in which a subset of high-risk declarations is selected to receive detailed risk comments and another subset is explicitly tagged for ex-post monitoring. For declarations not subject to third-party valuation advice, detailed comments increase reporting of fraud by 3.1 percentage points and improve tax yield by 1 percentage point. However, valuation advice and detailed comments have a significantly smaller impact on revenue when potential tax losses and opportunities for graft are large. Monitoring induces inspectors to scan more shipments but does not result in the detection of more fraud or the collection of additional revenue. Better information thus helps curb customs fraud, but its effectiveness appears compromised by corruption.

Over the last few years, customs authorities in many developing countries have introduced modern risk management techniques relying on data mining and statistical scoring techniques. By demonstrating that risk analysis in customs may be a valuable tool to facilitate legal trade and combat fraud more effectively, these techniques have helped improving the performance of customs authorities. However, these risk management techniques may prove to be inefficient in a context of moral hazard and low-performance customs administration. One way to address this weakness is to rely on information gained from discrepancies in bilateral trade statistics. The analysis of discrepancies in bilateral trade statistics (or mirror analysis) is increasingly used to identify high-risk import operations and to estimate revenue losses. By comparing data on fraud recorded by the Gabon customs administration with discrepancies in Gabon’s bilateral trade data, this paper highlights the benefits for a customs administration of a joint analysis of fraud records and mirror trade statistics data, the latter being indicative of the fraud remaining to be detected. Such an analysis helps customs to target ex post audits on risky import declarations unadjusted by the frontline customs officer. Finally, we point that analyzing jointly data on fraud records and mirror trade statistics data may be useful to (i) identify imported products for which the fraud remaining to be detected is large and (ii) monitor the performance of customs inspections.

Les administrations douanières des pays en développement mettent en œuvre depuis plusieurs années des méthodes d'analyse et de gestion du risque à partir des données issues des systèmes informatiques de dédouanement, afin d'orienter les contrôles de première ligne vers les différents circuits de dédouanement. Ces projets ont contribué à promouvoir un changement de culture au sein des administrations concernées en démontrant l'intérêt de recourir aux techniques d'analyse de données. Une des principales faiblesses de ces initiatives est toutefois d'évaluer les risques de fraude uniquement à partir des infractions relevées par les agents des douanes. Dans un contexte d'aléa moral, la portée de l'exercice et sa fiabilité peuvent être limitées, justifiant d'élargir la démarche en recourant à d'autres sources d'informations. Plusieurs études utilisent les écarts observés dans les données miroir du commerce international pour identifier les déclarations frauduleuses et en estimer leurs effets. A partir des données douanières du Gabon et de celles du commerce international, ce document de travail montre les avantages qu'une administration douanière tire simultanément de l'analyse de ses propres données sur les infractions constatées et de l'exploitation des données miroirs qui donnent une idée de la fraude non encore détectée. L'analyse des données miroirs permet d'établir un programme de contrôles a posteriori en ciblant les contrôles sur les opérations à risque qui ont échappé aux contrôles de première ligne. De plus, ces deux sources de données peuvent être combinées pour suivre la performance des contrôles douaniers et identifier les produits sur lesquels l'administration devrait concentrer ses efforts.

To carry out their various missions (collecting revenue, facilitating trade, and ensuring security), many customs administrations have established a risk management unit. In developing countries, however, because of the lack of dedicated human and material resources, intelligence and risk analysis remain insufficiently developed. In view of the lack of resources, this paper proposes a simple methodology aiming at detecting risky import operations. The mirror analysis first helps to identify and target products or sectors with the greatest risk. Based on the examination of customs declarations patterns (data mining), it is possible to identify and target higher risk economic operators (importers and customs brokers). When implemented in Madagascar, this method has helped to reveal probable fraud cases in the present context of customs reform. Estimates suggest that, in 2014, customs fraud reduced non-oil customs revenues (duties and import value-added tax) by at least 30 percent.

Based on a natural experiment that took place in Ecuador in the context of a capital outflow tax, this paper presents suggestive evidence of the emergence of an abnormal transfer pricing behavior. We exploit the fact that some imports were eligible to a tax-credit measure aiming to offset the capital outflow tax. The identification strategy relies on the fact that a foreign multinational enterprise has only an incentive to increase the unit value -transfer price- for profit shifting motivations of products eligible to the tax-credit. Then, comparing the evolution of the unit value of eligible products (the treatment group) with that of non eligible products (the comparison group) before and after the tax-credit measure, we estimate the over-reporting incentive's impact on the reported value. Results show that the declared unit value of eligible imports increased following the introduction of the tax-credit measure. Various tests suggest that this abnormal increase is attributable to a profit-shifting behavior.

Etudes & documents; This paper identifies opportunities for improving the performance of revenue-collectionauthorities. To detect and combat fraud, we argue that revenue-collection authorities should,notably in the absence of reliable third-party information, exploit non-usual sources ofinformation. Specifically, our micro-level study of customs evasion provides evidence thatusing any internal or external available source of information facilitates customsenforcement.Estimates highlight that exploiting historical data and/or relying on aninformation provider - a pre-shipment inspection company - significantly reduces evasion inCameroon. The potential endogeneity of pre-shipment inspections is addressed by usinginstrumental variables. Results are robust to a variety of additional checks.