Pre-Crime-Fighting: The Ministry for Social Development's interest in Preventive Risk Modelling, as a technique for identifying and rescuing vulnerable children before they grow up to become a burden on society, is strikingly reminiscent of the plot of the science-fiction movie Minority Report. Accusations of racial profiling are inevitable.
HOW LONG WILL IT BE, I wonder, before the Ministry of Social Development (MSD) is accused of racial profiling? Given the statistical techniques currently being developed by the Ministry to identify “vulnerable” clients, such an accusation is practically inevitable.
In collaboration with the University of Auckland, the MSD is perfecting a technique for filtering out all but the worst offenders when it comes to deficient education, poor health, inadequate housing, a history of family violence and/or criminal offending. A filtering process relying upon such variables, however, cannot fail to generate a strong racial bias. In terms of the raw numbers, Pakeha will probably still predominate, but Maori and Pasifika will, almost certainly, find themselves significantly over-represented.
The MSD’s problem is that they cannot avoid using such controversial techniques for identifying vulnerable clients. Downside political risks notwithstanding, they are fundamental to the National Government’s new approach to managing New Zealand’s welfare system. Expressed in its simplest terms, this new approach is about identifying the individuals and families most likely to become a long-term drain of the state’s resources – and making sure that they don’t.
Serious criminal offending, for example, imposes colossal costs upon the state. A person convicted of murder, manslaughter, rape, child abuse, aggravated robbery and/or serious assault can expect to serve anything from 5 to 20 years in prison – at a minimum cost to the taxpayer of $100,000 per year. And that figure does not include the cost of repairing and rehabilitating the victims of criminal offending. The enormous expense of hospitalisation. The loss of productivity associated with the victims’ pain and suffering. All these social costs could be dramatically reduced if the people most likely to impose them could be rescued, early, from themselves.
One of the solutions, according to the MSD, may be found in the statistical technique known as “predictive risk modelling”. According to the Ministry’s own website, a ground-breaking piece of research undertaken by a project team, led by Professor Rhema Vaithianathan of the University of Auckland, has “developed a predictive risk model for children in a cohort who had contact with the benefit system before age two. These children accounted for 83% of all children for whom findings of substantiated maltreatment were recorded by age 5.”
The Ministry further reported that “predictive risk modelling had a fair, approaching good, power in predicting which of the young children having contact with the benefit system would be the subject of substantiated maltreatment by age five. This is similar to the predictive strength of mammograms for detecting breast cancer in the general population.”
Given the well-attested link between childhood abuse and serious criminal offending in later life, the possibilities arising out of Professor Vaithianathan’s and her team’s research are obvious. If predictive risk modelling (PRM) could identify with relative precision which children, in which families, were most likely to suffer abuse, appropriate “wrap-around” intervention by the MSD, the Police, Child Youth and Family, the Department of Corrections and the Department of Courts could ensure that the predicted abuse (and everything likely to flow from it in the future) never happened.
The popular culture reference you’re looking for here is the film Minority Report. Based on the novella by science-fiction writer Philip K. Dick, the movie is set in a futuristic Washington DC, where a special “precrime” squad of police officers use “psychic technology” to arrest and convict murderers before they commit their crime.
Now, it would be quite unfair to suggest that PRM is in any way analogous to “psychic technology”, but it’s undeniable that the former’s widespread use in our social welfare system would give rise to just as many ethical questions as Philip K. Dick’s pre-crime-fighters.
In the section of the University of Auckland study relating to PRMs ethical ramifications, the Project Team drew the MSD’s attention to the dangers of the data arising from its application being misinterpreted:
“It must be acknowledged that some of the data and predictor variables used by the proposed model are highly likely to be misinterpreted by at least some audiences. The decision not to report coefficients in this report, for instance, was based in part upon the belief that the insignificant contribution those factors make to the power of the tool was outweighed by the likelihood of crude and misleading interpretations of that information given existing social prejudices and stereotypes.”
Which brings us back to our original question concerning racial profiling. It would be most surprising if the unwillingness of the Auckland academics to identify all the coefficients used in their predictive algorithms was not, at least in part, related to race as a predictive factor in the maltreatment of children. It would, however, be equally surprising if the prospect of Maori and Pasifika families being targeted for special “precrime” intervention on behalf of their infant offspring was not met with loud, sustained, and entirely justifiable protest.
There is something profoundly disturbing in the very notion that science possesses the power to predict who will – and who will not – inflict harm upon their fellow human-beings. That, somehow, a computer programme can winnow out from tens-of-thousands, the one family in which violence will be done to a child.
Because, even if we could be sure that the child identified through PRM was bound, in every case, to suffer abuse if some form of welfare intervention did not take place, there is another, deeper, question that must be confronted. If individual cases of abuse could be predicted and prevented, what incentive would there be to address the systemic causes of human tragedy?
If Maori and Pasifika appear more often than they should among the perpetrators of child abuse it is only because they appear more often than they should among all the other “coefficients” of dysfunction: illiteracy; the diseases of poverty and overcrowding; the psychological deterioration caused by long periods of unemployment; the mental disintegration associated with drug addiction. These pathologies are the symptoms of class as well as racial oppression. Capitalism and colonialism are “coefficients” too.
Let us leave the final word to another artefact of pop-culture. Perhaps the most surprising of all Elvis Presley’s hits is his extraordinary rendition of Scott Davis’s song, In The Ghetto. To those seeking to transform our social welfare system into something resembling science-fiction, I would strongly recommend Elvis’s poignant retelling of the story of a boy whose fate was sealed at birth: not by the choices he or his mamma made, but by the system that left them with so few:
And as her young man dies
On a cold and grey Chicago morning,
Another little baby child is born
In the Ghetto.
Video courtesy of YouTube
This essay was originally posted on The Daily Blog of Wednesday, 3 June 2015.