[Update] SP calls for DUO checks to be paused

• Update: DUO algorithm set out of order

MPs Peter Kwint and Renske Leijten (SP) have submitted a long list of written questions (in Dutch) in response to the investigation (in Dutch) which HOP (the news agency to which Delta is affiliated), Investico and NOSop3 published in various media last Wednesday. The SP sees “an alarming picture of bias within DUO’s actions to combat fraud”.

In a letter to the House of Representatives, sent the same day, Education Minister Robbert Dijkgraaf also responds (in Dutch) to the findings, using slightly stronger wording than earlier on camera for broadcaster NOS.

In court cases involving the basic student grant, some 98 percent of students have a migration background. “I find that an especially worrying signal”, Dijkgraaf writes. He says that DUO will evaluate the process. “That is in keeping with the attitude of integrity and learning which characterises DUO in my experience.”

Childcare benefits scandal
But he adds: “At the same time, this does not dispel my questions about the effects of this process – the large representation of students from migrant backgrounds. The lessons of the childcare benefits scandal mean that the government and DUO are more than ever duty-bound to investigate thoroughly every sign or indication of discrimination.” Dijkgraaf pledges to get back on the matter in July.

DUO mainly looks for illogical combinations of addresses

Among other things, the SP would like to know from Dijkgraaf: “Are you willing to publicly disclose this algorithm? If not, why?” DUO was previously only willing to share the general scope and object of this algorithm, and not its precise working.

DUO initially said that origin plays no role in detection, as that information is not included in the algorithm. In its questions, the SP states that this “shows a rather naive view of how it works at best.”

Migration background
Students who live away from home are entitled to a higher basic student grant. Some students therefore register as living with friends or family members, when they are actually still living at home with their parents. DUO mainly looks for illogical combinations of addresses (do parents live closer to where the student is studying than the student themselves), but also examines indicators such as age and educational level, for example.

Part of the problem is that students with a migration background are more likely to live with relatives, and their student rooms can look different from what investigators expect. Also, DUO often relies on dubious local inquiries, asking neighbours: does a student live next door to you?

The news triggered a torrent of reactions, including criticism from people who don’t see what the problem is. Joost Eerdmans of JA21, for example, tweeted that, according to DUO, many students have apparently committed fraud. “And three quarters lost their court case. But the system is flawed?” He announced that he would be asking additional parliamentary questions.

HOP, Bas Belleman
Translation: Taalcentrum-VU