Adoption of settlement decision

Dutch Tax and Customs Administration

The purpose of the algorithm is to support the Tax Administration in offsetting an amount a citizen or business is owed from the Tax Administration against their outstanding tax debts.

Last change on 28th of October 2024, at 10:43 (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

Field not filled in.

Contact information

[e-mail]

Link to publication website

https://over-ons.belastingdienst.nl/onderwerpen/omgaan-met-gegevens/algoritmeregister/

Link to source registration

https://over-ons.belastingdienst.nl/onderwerpen/omgaan-met-gegevens/algoritmeregister/vaststellen-verrekenbeslissing/

Responsible use

Goal and impact

The purpose of the algorithm is to support the Tax Administration in offsetting an amount a citizen or business owes the Tax Administration against their outstanding tax debts. To do so, the algorithm is provided with, among other things, information about the credit and all outstanding debts of the citizen or business concerned.

It is a rule-based algorithm. The decision rules are prepared by content experts based on laws and regulations, accumulated expertise and implementation policies.

The algorithm indicates when and how a credit can/may be offset against debts and implements it whenever possible.

The algorithm is not self-learning. This means that it does not evolve during its use.

Considerations

The algorithm helps in the automated settlement of balances against outstanding debts of citizens and businesses. By deploying the algorithm, these offsets can be processed more efficiently and faster, giving citizens and businesses faster clarity.

Human intervention

Human intervention occurs when human supervision of decision-making is substantial and carried out by someone authorised and competent to change the decision.

In the operation of this algorithm, there is no human intervention. The algorithm makes decisions about offsetting a credit against one or more outstanding debts.

Risk management

No special personal data are used in the algorithm and there are no indications that the outcome of the algorithm model contains ethical bias (undesirable biases for specific groups).

The decision rules and algorithm are tested against non-discrimination laws. The Tax Administration aims to process as little personal data as possible. Employees involved in developing and managing the algorithms receive training on data protection and bias. To prevent neutral features from leading to unequal treatment, the Tax Administration conducts research on effects of the algorithms. The Tax Administration aims to (continue to) assess all algorithms for bias.

The General Administrative Law Act requires the government's actions to be transparent and lawful (guarantee character). When applying and developing algorithms, the Tax and Customs Administration acts in accordance with the general principles of good administration, such as the principles of equality, due care, justification, proportionality, legal certainty and trust.

Conditions, a policy framework, have been drawn up by the Tax Administration for the development of algorithms. This contains rules and agreements that were followed when developing the algorithm.

Legal basis

The collection and use of the data described above is regulated in the:

• Invorderingswet 1990 (including Section 24) and the Invorderingswet 1990 Implementation Decree based thereon and the Invorderingswet 1990 Implementation Regulations).
• General provisions Citizen Service Number Act.
• General Income Support Act.
• Collection Guidelines 2008.
• The Instructions on Collection and Tax Bailiffs (IIB).

The further details and choices in interpreting the above-mentioned legislation for drawing up the decision rules for the automated settlement process are laid down in policy rules.

Links to legal bases

• Invorderingswet 1990: https://wetten.overheid.nl/BWBR0004770/
• Uitvoeringsbesluit Invorderingswet 1990: https://wetten.overheid.nl/BWBR0004772/
• Uitvoeringsregeling Invorderingswet 1990: https://wetten.overheid.nl/BWBR0004766/
• Wet algemene bepalingen Burgerservicenummer: https://wetten.overheid.nl/BWBR0022428/
• Algemene wet inkomensafhankelijke regelingen: https://wetten.overheid.nl/BWBR0018472/
• Leidraad Invordering 2008: https://wetten.overheid.nl/BWBR0024096/
• Instructie Invordering en Belastingdeurwaarders (IIB): https://download.belastingdienst.nl/belastingdienst/docs/instructie_invordering_en_belastingdeurwaarders-v1_1_ov0211z63fd.pdf

Elaboration on impact assessments

The use of data is reviewed against the General Data Protection Regulation (AVG). This review against the AVG will highlight potential privacy risks and allow appropriate measures to be taken. The AVG prescribes that no more data should be used than necessary. This is called data minimisation. The Tax Authority regularly examines whether the data used are still necessary and may therefore be used.

Operations

Data

The algorithm uses data collected under various tax laws. The use is permitted by law. As required by the AVG, no more data is used than necessary.

• Debt and payment data by BSN
• Schemes and different treatments
• Official decisions

Links to data sources

• Schuld- en betalingsgevens per BSN: Belastingdienst
• Regelingen en afwijkende behandelingen : Belastingdienst
• Ambtelijke beslissingen: Belastingdienst

Technical design

Within the algorithm, a distinction is made between a regular netting or an ex officio netting request. A regular set-off is started and completed fully automatically. An ex-officio set-off request is often started by a debt collection employee on request of a citizen or company, where the employee can indicate to which debts and/or in which order the credit should be set off.

The algorithm goes through the following steps:

1. First, the algorithm assesses whether the credit may actually be offset;
2. Then the debts are sorted according to the type of tax and how old the debt is;
3. Next, the algorithm assesses which debts may actually be offset.

External provider

The algorithm was developed by the Inland Revenue and is also maintained internally.