Netherlands childcare benefits fraud automation
Updated: April 2022
The Dutch government has collapsed after a vote of no confidence called by opposition parties stemming from its handling of a childcare benefits scandal ('kinderopvangtoeslagaffaire' in Dutch).
Termed (pdf) an 'unheard-of injustice' by a parliamentary inquiry, the scandal involved national, and racial and ethnic profiling, and led to false accusations of fraud levelled at approximately 26,000 families by the country's Tax and Customs Administration.
These accusations resulted in deep mental stress for many of those involved, financial ruination, and suicides. Over a thousand children had to be taken into foster care.
The more or less fully automated decision-making system used self-learning algorithms to create citizen risk profiles and predict the likelihood of their committing fraud.
Welfare seekers were not informed about the system, and had little scope to complain or appeal decisions. External parties such as researchers, journalists or investigators were not permitted to inspect it.
Fraud and racial profiling
The scheme's shortcomings and failures had been exposed many times before the parliamentary inquiry. Examples of fraud against the system had emerged soon after its launch in 2013, closely followed by cases involving suspected fraud, a number of which proved unfounded.
Substantive allegations of racial and ethnic profiling were levied by broadcaster RTL Nieuws and the Trouw newspaper in September 2018. Dual nationality and low income seen as major risk indicators, as was having a 'non-western appearance'.
In 2020, it was revealed that the tax authorities had kept secret blacklists of potential fraudsters for two decades. Being on the blacklist resulted in a higher risk score in the childcare benefits system.
These allegations led to a damning Dutch Data Protection Authority report which found the system 'unlawful, discriminatory and improper', and a parliamentary inquiry.
The government promised to reimburse each family EUR 30,000, only for the tax authority to demand much of it back in taxes. Eventually, the government cancelled the debts of those families impacted.
Operator: Tax and Customs Administration; Ministry of Social Affairs & Employment
Developer: Tax and Customs Administration; Ministry of Social Affairs & Employment
Sector: Govt - tax; Govt - welfare
Purpose: Detect compliance errors; Detect fraud; Assess & classify risk
Technology: Risk assessment/classification algorithm
Issue: Accuracy/reliability; Bias/discrimination - race, nationality; Privacy
Opacity: Governance; Complaints/appeals; Black box