Information sharing between UHT and aid organisations

Benefits Department

The algorithm helps other support organisations to obtain the correct information from the Childcare Allowance Recovery Agency. This enables these organisations to provide the right support to parents who have (or have not) been affected by the problems with the childcare allowance.

Last change on 22nd of July 2026, at 9:09 (CET) | Publication Standard 1.0

• Publication category: Impactful algorithms
• Impact assessment: DPIA
• Status: In use

General information

Theme

Social Security

Begin date

2021-12

Contact information

[e-mail]

Link to publication website

https://herstel.toeslagen.nl/hulp-bij-schulden/gegevens-delen/ 

Responsible use

Goal and impact

The Benefits Recovery Implementation Organisation (UHT) of the Benefits Service is committed to helping those affected by the problems with the childcare allowance as effectively as possible. We are rectifying past errors and doing our utmost to address the financial consequences of those errors.

In addition to the help provided by UHT, those affected can also receive support from other (government) organisations, such as local authorities and Sociale Banken Nederland (SBN). To ensure efficient cooperation with these organisations, it is important that UHT coordinates and collaborates effectively with them. To this end, it is essential that the correct and necessary information is shared with these organisations. This involves sharing not only the necessary information about victims but also about non-victims, for example to support local authorities in discontinuing general support services for these parents.

To speed up the process and improve support for victims, UHT is automating certain tasks using algorithms. The algorithm quickly selects the relevant information held by UHT on their behalf, so that parents receive the help they need more quickly and effectively. In this way, victims are assisted as efficiently as possible and receive the right support. If the information indicates that someone has not been affected, further support is discontinued.

Considerations

Using this algorithm offers the following benefits:

· Speed and efficiency: victims receive the right support more quickly; for those who are not victims, support is withdrawn more quickly.

· Consistency: the approach ensures that the provision of support to those affected and the discontinuation of support for those not affected are monitored in the same way.

· Accuracy: the determination of the appropriate support is carried out with great precision

· Traceability and reproducibility: every step is documented

· Flexibility: existing processes can be easily adapted to new rules

When weighing up the use of the algorithm, the following potential drawbacks were taken into account:

· Privacy risks: the use of the algorithm may result in sensitive information being shared with various organisations. For this reason, additional checks are carried out to ensure that only the correct information is shared.

· Dependence on good data: the algorithm only works properly if the data is accurate and complete. For this reason, the data is subject to additional checks and the results are sometimes verified manually.

· Limited human judgement: the algorithm is less able to take exceptional or unique situations into account. In exceptional cases, a member of staff reviews the results. There are additional manual checks and a clear escalation process.

Human intervention

· In exceptional circumstances, such as where bank details are not known to UHT/Toeslagen, the algorithm flags this to the staff member.

· In addition, random checks are carried out: a random group of parents is selected for a manual check to verify that the details are correct and that the algorithm selects the correct parent.

· Those affected must give their consent for UHT to share their data with the local authority.

· If they have any questions, members of the public can call the UHT service team or other support organisations, such as local authorities or service teams.

· The support organisations can provide assistance with individual cases and maintain personal contact with the affected parents.

Risk management

· Risk management and safeguards

The Benefits Service has drawn up conditions for the development of algorithms, in the form of a quality framework. This sets out rules and agreements that must be followed when developing the algorithm. The conditions originate from the Central Government Audit Service and are binding. The algorithm has business rules based on legislation and regulations. These business rules are tested and maintained to ensure continued compliance with legislation, regulations and policy objectives. Where changes are made to an algorithm, the Benefits Service assesses whether the quality requirements are still being met.

· Equality, non-discrimination and privacy

The use of data is assessed against the General Data Protection Regulation (GDPR). The GDPR stipulates that we may not use more data than is necessary to achieve a specific purpose. This is known as data minimisation. The Benefits Service examines which data is necessary and may be used.

· Profiling

The use of data is assessed by the Benefits Service to prevent unlawful profiling from taking place.

· Citizens’ rights

Citizens have the right to view the information collected about them by the Benefits Service. They may also submit a request to have this information amended.

Legal basis

Act on the Recovery of Child Benefit Allowances

For further information, see: https://wetten.overheid.nl and wetten.nl – Regulations – Benefits Recovery Scheme Act

Links to legal bases

Government Legislation:  https://wetten.overheid.nl

Elaboration on impact assessments

Privacy and the GDPR

The use of data must be assessed against the General Data Protection Regulation (GDPR). By assessing personal data, any privacy risks can be identified and appropriate measures taken.

The GDPR stipulates that no more data may be used than is necessary. This is known as data minimisation. The Tax and Customs Administration regularly reviews whether the data used is still necessary and may therefore be used.

The GDPR compliance document has been drawn up. GDPR-related risks have been identified and recommendations have been followed up with appropriate measures.

Equality and non-discrimination

The algorithm is assessed in line with applicable non-discrimination principles relating to direct and indirect discrimination. By processing as little personal data as possible, the risk of direct discrimination is reduced. Staff involved in the development and management of the algorithms receive training on data protection and bias.

Impact assessment

Data Protection Impact Assessment (DPIA)

Operations

Data

A dataset is being compiled containing all those affected who are entitled to compensation and assistance. This is based on the register of parents who have been identified as affected by the issues surrounding the childcare allowance. Relevant information Relevant information will be shared with, amongst others, local authorities and Sociale Banken Nederland so that they can offer further assistance. The information shared will depend on the support organisation concerned and may include, amongst other things:

· A list of citizen service numbers of affected parents who have received at least 30,000 euros (Catshuis scheme)

· Partners of those affected

· Bank details for the CJIB (Central Judicial Collection Agency)

· Decision letters, including the date of the letter

· Information on whether there is an ongoing appeal, an award or rejection decision from the Actual Damage Commission (CWS) and/or a signed settlement agreement (VSO)

· Check whether a letter of acceptance has been sent indicating whether a parent is eligible for admission to the Stichting Gelijkwaardig Herstel (SGH)

· Information on whether compensation has been paid for hardship, bias or wilful misconduct/gross negligence (OG/S)

· Check whether someone is a duplicate applicant

The exact information shared varies from organisation to organisation. In addition, data is also provided on non-victims, for example to local authorities, in order to discontinue general support for this group where applicable.

Technical design

The algorithms are designed to determine what information is important for other aid organisations to receive in order to help those affected by the childcare allowance issues. They carefully analyse the available data to select only the relevant information. This is important because it means that no unnecessary or irrelevant information is shared with other organisations supporting those affected.

At UHT, the various types of available data are examined. Based on specific criteria, the exact data is selected using algorithms. These criteria help to identify precisely the information needed to assist those affected.

Once the algorithms have selected the correct information, the data relevant to the aid organisation in question is shared. Support organisations include, for example, local authorities or Sociale Banken Nederland. Only information that is necessary for the organisations in question is shared. This ensures that all organisations involved have the correct and necessary data to carry out their tasks efficiently and effectively, so that those affected can quickly receive the help they need.

All these steps are carried out with due regard for parents’ privacy, in accordance with the rules of the General Data Protection Regulation (GDPR). This means, amongst other things, that we only use the data we genuinely need to achieve our objective.