Detecting tax risks in customs declarations regarding the levying and collection of sales tax

Customs

This algorithm helps Customs to select goods for inspection based on risk. It uses declaration data from companies and considers whether or not there are risks of inaccuracies in the declarations for the purpose of determining correct financial measures and levies (including import duties and VAT).

Last change on 10th of December 2024, at 7:53 (CET) | Publication Standard 1.0

• Publication category: Impactful algorithms
• Impact assessment: Field not filled in.
• Status: In use

General information

Theme

Public finance

Begin date

05-2023

Contact information

[e-mail]

Link to publication website

https://belastingnet.belastingdienst.nl/cms/html/boeken/HDU/het_in_het_vrije_verkeer_brengen-het_heffen_en_innen_van_de_invoerrechten.html

Responsible use

Goal and impact

The purpose of this algorithm is to more efficiently select which declarations may have been incorrectly completed. This algorithm focuses on sales tax (VAT) due on import declarations and looks at declarations of goods coming from outside the European Union. It indicates which declarations have a higher risk of fraud or incorrect data. Customs checks companies more or less based on the results. By using this algorithm, declarations are handled more efficiently, resulting in quicker checks and possibly quicker release.

Considerations

Customs processes 700 million declarations every year. With this algorithm (risk profiles), Customs can assess declarations more efficiently and faster than without using this algorithm. So declarations are approved or checked faster. And goods are released faster. Also, the algorithm ensures that 'known and reliable' companies are actually checked less.

Human intervention

If a declaration produces a match with a risk profile, a customs officer reviews the declaration. If necessary, the customs officer asks for more information. The decision to approve a declaration, or instead to start checking, is made by a human being.

Risk management

A risk profile is always developed and checked by at least two customs staff before use. This is the 4-eye principle. And once the profile is in use, a weekly check is made for a profile to see how many declarations produce a match. If necessary, the profile can be adjusted. Periodically, Customs checks per profile whether it is still needed. Customs monitors internal and external complaints and incidents, for example if a profile is insufficiently specific and therefore designates too many declarations as higher risk.

Legal basis

Legal basis for control for the purpose of determining financial measures and levies lies in the Union Customs Code (DWU), the Turnover Tax Act and the General Customs Act. These laws ensure that levies are levied correctly and properly. The use of risk profiles also relies on DWU section 46. This says that customs controls should mainly rely on risk analysis by electronic data processing techniques.

Links to legal bases

• https://wetten.overheid.nl/BWBR0023746/2024-01-01/#Hoofdstuk1 : Algemene Douanewet
• https://eur-lex.europa.eu/legal-content/NL/TXT/PDF/?uri=celex:32013R0952 : Douanewetboek van de Unie
• https://wetten.overheid.nl/BWBR0002629/2014-01-01/1/#Hoofdstuk1: Wet op de Omzetbelasting

Operations

Data

The algorithm uses all data from the goods declaration. For example, type of goods, goods code, weight, number of packages, consignee, declarant, consignor, country of consignment, EORI number, means of transport, etc.

Links to data sources

https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/douane_voor_bedrijven/naslagwerken_en_overige_informatie/andere_onderwerpen/gegevensvereisten/: Gegevensvereisten douaneaangiften (zie Bijlage B Gedelegeerde Verordening Douanewetboek van de Unie + Bijlage VIa van de Algemene Douaneregeling)

Technical design

The algorithm consists of decision rules created in collaboration with content experts. These decision rules provide estimates of which returns have a higher risk of fraud or incorrect data with respect to determination of sales tax due. These decision rules are based on 'if-then-else' combinations. An example: if (if) the information we have on a shipment shows that the goods have an abnormal weight, (then) the declaration is checked manually and the shipment may be selected for extended inspection. If the goods do not have an abnormal weight (else), we do not stop the shipment. Or at least not based on this criterion.

External provider

Internally developed