NL · Missing: publication date
The Dutch childcare benefits scandal: the tax authority's risk classification model
Image: Amnesty International, illustration from its news release on the report. All rights reserved by the owner.
- Reported by
- Trouw, RTL Nieuws, Amnesty International
- Methods
- FOI / records request
- Parliamentary scrutiny
- Regulator investigation
- Document review
The Dutch tax authority scored childcare benefit applications with a self-learning risk model, and Dutch citizenship was one of its inputs. Tens of thousands of parents were wrongly treated as fraudsters, and migrant and low-income families were hit hardest. Years of reporting by Trouw and RTL Nieuws, pressure from MPs and a data protection investigation established the use of nationality in July 2020. The government resigned in January 2021, and the tax authority was fined €2.75 million. The model was one part of a wider failure.
amnesty.orgThe Dutch childcare benefits scandal: the tax authority's risk classification modelA risk model with self-learning elements, introduced by the Dutch tax authority in 2013, that gave each childcare benefit application a risk score for inaccuracy. Low-risk applications were approved automatically. The highest-scoring applications went to civil servants for manual review, and the applicants' benefit payments were suspended.Netherlands · Reported by Trouw, RTL Nieuws, Amnesty International and others (opens in a new tab)How they investigated
FOI / records request
From summer 2018, Jan Kleinnijenhuis (Trouw) and Pieter Klein (RTL Nieuws) investigated how parents had been labelled fraudsters. A 2019 freedom of information request about one tax authority fraud operation, CAF 11, led to 550 documents in November 2019. The reporters objected to what was withheld and went to court in October 2020.
Parliamentary scrutiny
MPs pressed for internal documents. One, the 'Palmen memo', had concluded long before the scandal that the tax authority acted wrongly; the finance ministry released it in October 2020 after questions from MP Pieter Omtzigt. A parliamentary inquiry reported in December 2020 that the injustice done to parents was unprecedented.
Regulator investigation
The data protection authority began investigating in 2019 and established that 'Dutch citizenship: yes/no' was a parameter in the risk model. A fraud alert about 120 to 150 people with Ghanaian nationality led to an investigation of all 6,047 applicants with that nationality. The regulator said the tax authority had repeatedly misled it.
Document review
Amnesty International did not have the model. It reconstructed how the model worked from the regulator's report and from parliamentary and ministerial documents, put its findings to the government and reflected the government's responses. No code, model or dataset has been published.