On Thursday evening, outgoing minister Dijkgraaf (D66) entered into a debate with the House about the mistakes made by the Education Executive Service (DUO) in its checks on fraud with the basic grant for those living away from home.
DUO’s fraud hunters mainly targeted students with a non-western migration background, research by the Hoger Onderwijs Persbureau, Investico and NOSop3 showed last year. Further research by consultancy firm PwC confirmed this. There had been signals earlier that checks were not going well, but nothing was done about it.
Innocent students
The cabinet has apologised for ‘indirect discrimination’. During the debate, many political parties said they appreciated this, but they doubted whether it went far enough. “Does the cabinet also recognise that innocent students have been affected by this?”, asked MP Sandra Beckerman (SP). “Will the cabinet think about a compensation for students who are really harmed by this policy?”
The same asked MP Doğukan Ergin of DENK. He foresaw lawsuits from victims (‘That will also cost a lot of money’) and thought the minister would be better off coming up with a compensation scheme himself.
Ergin also inquired what was so ‘indirect’ about DUO’s discrimination. After all, officials had long known that students with non-Dutch-sounding surnames were more likely to get inspectors visiting them. They had also already been warned against bias. “This is rock-hard institutional discrimination,” Ergin said.
Slept through the alarm clock
Dijkgraaf reiterated his apologies and shrugged. “We actually slept through the alarm clock. We didn’t hear it.”
At the same time, he didn’t commit to anything. PwC did not look at individual cases. “Based on existing investigations, it can not be determined whether innocent students were affected,” the minister said. So he won’t promise financial compensation . “I really understand that question,” Dijkgraaf said. “But I am going to talk to students first.”
Burden of proof
One of the problems with the inspections is that DUO only has to make the abuse of the basic grant for living away from home ‘plausible’, which was sometimes done on the basis of very superficial conversations with neighbours. Once suspected, the burden of proof lies with the student: he or she must refute the suspicion.
Some parties want to put that ‘burden of proof’ back on DUO. After all, people are innocent until proven guilty, write SP, DENK and GroenLinks-PvdA in a motion on the subject.
But Dijkgraaf wants to think about it longer. “I understand the idea of reversing the burden of proof,” he told the House, “but the motion commits me to a follow-up step in a certain direction and I think that is still too early.”
Flames from the roof
Still, there is a chance that the motion will get a majority, as NSC seems to support it. “The flames are coming out of the roof and the building is ablaze,” MP Aant-Jelle Soepboer said of DUO’s inspections. “Models and systems in the wrong hands are disastrous. That is also apparent now.”
He therefore is in favour of reversing the burden of proof. If DUO only has to make it ‘plausible’ that someone is fraudulent, then the same should be true the other way round: as far as he is concerned, the student only has to make it ‘plausible’ that he or she lives at the right address after all. “Then it is equal again”, Soepboer mused. And then DUO should eventually come up with better evidence.
Targeted control
Some other voices were also heard. The VVD understood all the criticism, but did see a role for the use of ‘good’ risk profiles. “How do we ensure that targeted control of fraud remains possible in the future, without discriminating against people?”, MP Claire Martens-America wanted to know from the minister.
Dijkgraaf did not yet know, he said. Martens-America insisted: is it by definition ‘indirect discrimination’ if one population group is found to be overrepresented in the checks? He could not just say that either. The checks, he said, should focus on ‘the group of people who are abusive’.
'Why apologise?'
The PVV stood differently in the debate. MP Reinder Blaauw found it strange that individual cases were not looked at. “We consider that vital information after all. In how many cases is there actually an unjustified finding of fraud?” Without that information, you cannot say at all that there was actual discrimination, he argued.
‘Indirect’ discrimination was a concept unfamiliar to Blaauw. “If there really had been direct discrimination, or if there had been unjustified accusations of fraud against students, we would disapprove of that and an apology from the minister would have been in order.”
But Dijkgraaf did not change his mind on this. “The argument that you can’t be discriminatory because you don’t actually use migration background as an input of your process is a naive attitude towards this kind of control mechanism,” he said in the debate.
Indirect discrimination occurs when seemingly neutral requirements or criteria hit certain groups harder. For example, DUO looked at who lives close to their parents’ house; as it turns out, students of non-Western origin are more likely to do that, so they were targeted more often.
Fines
Since 2012, almost 25 thousand home visits have been made and almost ten thousand times the inspectors believed fraud was involved. Basic grants worth 13.2 million euros were recovered and 7.4 million euros in fines were handed out.
Currently, the discriminatory algorithm has been disabled and DUO uses ‘random sampling’. 356 students have since been checked. Of these, 244 lived at the specified address, the minister said. In the case of 52, DUO could not establish it. Of ten students, DUO believes there was abuse of the basic bursary for those living away from home.
