It was just a passing comment that struck me, from a Māori woman, stating what her mother had told her when she became a mother. She was talking about what is now Oranga Tamariki (OT). “Don’t give them a reason – don’t given them any reason to start looking at your parenting. Make sure everything appears perfect and don’t, whatever you do, give them any cause to start in on your family”. What is the level of cultural penetration of a child protection agency in the lives of families when the generational advice, along with feeding and sleeping and nappy-changing, includes how to protect yourself from state intervention? But this level of penetration does not apply to all families, everywhere. Our research shows that if you live in the most deprived 10% of neighborhoods in this country, your chance of having a family group conference held about your family is 35 times greater than if you live in the least deprived, and you have ten times the chance of having your child placed in fostercare (Keddell, Davie & Barson, 2019).
So if you live in the most deprived communities in the country, you are likely to have more than a passing understanding of the operations of the state in everyday life, and in the least deprived, you probably don’t even know what an ‘FGC’ is. This immense difference should concern us because it shows the hugely disproportionate chances of child protection intervention for those in poorer areas. As Māori and Pasifika whanau are over-represented in highly deprived areas, this difference affects those communities more than others. Such large disparities suggest that children’s experiences of childhood are strongly affected by where you happen to be born – in effect a ‘postcode lottery’. Whether you consider ‘intervention’ to be good or bad, we can agree that where such inequalities exist, we should be questioning what could or should be done to prevent it. It’s not inevitable.
What exactly does the rate of intervention represent? Is it the harm of a higher chance of abuse and neglect, or the harm of the state intervening in family life? This question leads us to a much larger conversation about the causes of contact with the child protection system, and the nature of intervention. On the one hand, there is the ‘risk’ argument – that poorer neighborhoods have truly higher rates of child abuse and neglect so therefore need much greater emphasis on community-based prevention services and social protections to reduce that risk. In this perspective, reducing the intervention rates of OT means increasing the range of protections for families before they end up at OT’s door.
What evidence is there for this argument? The rates of hospitalisations for physical abuse (a more objective measure than notifications to the child protection system) do have a social gradient (Duncanson et al., 2018). Theories of abuse generally emphasise the ‘stress’ thesis: that when people are subject to the stresses of poverty, and everything that comes with it, the chances of abusive behaviour and more generally harmful situations for children really do increase. On the other hand, there is the ‘bias’ argument – that people in poorer neighborhoods are subject to at least three kinds of bias – exposure bias, direct bias and institutionalised bias. The first means that people in deprived neighborhoods are over-exposed to people likely to refer them to OT compared to those in more affluent neighborhoods, for example, if there is domestic violence the police are more likely to be called than if it was in a more affluent home, prompting a mandatory notification to OT. More deprived areas are more likely to have a school social worker and have a range of other health and social services that have strong notification policies. People in more deprived areas are less likely to be able to avoid exposure either – for example, less able to afford legal representation to get out of the child protection system quick–smart if they are notified. Another source of evidence is the fact that community studies of abuse incidence are generally much higher than the rate of those notified to child protection services, but not so strongly correlated with poverty and ethnicity.
Direct bias means people in more deprived areas may be more likely to be perceived as ‘risky’ due to both classist and racist assumptions, affecting both notifiers and OT staff. There is also evidence for this type of bias, for example our study on bias found that Māori families were likely to be perceived as higher risk than Pākeha families in exactly the same situation, have more decisions made about them, and be perceived as ‘in need of care and protection’ (Keddell & Hyslop, 2019). It’s likely both exposure bias and direct bias play their part to increase the rates of system contact for people in deprived neighborhoods, in a way that exacerbates any real difference in incidence.
But institutionalised bias can operate in less direct ways too. Risk assessments emphasise previous contact with the child protection system, and police records of domestic violence. As both are more likely to have happened if you are poor and Māori or Pacific, the emphasis on those factors in assessments in turn ‘ratchets’ some families further into the system, while others are not subject to this multiplier effect (consider that much domestic violence and child abuse among more affluent and Pākeha populations is never reported). So there can be a racialised and classist outcome due to these institutionalised factors, even in cases where there is no direct bias.
And then there is the complexity. Beneath broad patterns of inequalities in life chances of system contact, you get outliers. There are some places in Aotearoa New Zealand where there is high deprivation, but much lower chances of coming into care than in other equally deprived areas. This was especially true if the larger region is less deprived. In our decision variability study, we found that site offices and even individual practitioners could differ a lot in their ways of assessing risk, what they understood risk to ‘look like’ and their general attitude towards the inevitable balancing of family preservation, with the need to protect children. Some sites have a strong commitment, for example, to supporting families to keep children if possible, while others less so. Some have more or different services available to them out in the community than others, also shaping their practice decisions. It’s easier to close a case if you know there is an NGO service available to work with the family. So location, site office and individual differences can buck broad trends, shaping practice in unexpected ways.
Policy change also affects the direction of practice and the distribution of resources. The reforms of 2015 embraced a ‘child protectionist’ emphasis on early removal, with little acknowledgement of social context factors. It was inevitable that this general policy orientation, with no real attention to prevention, would result in even more children in highly deprived areas coming into care. It was a further swing away from a ‘child welfare’ orientation, which has a focus on supporting families to retain children in their care and tends to embrace a broader public health type perspective. The protectionist agenda can legitimise very heavy handed actions. But this is not just a problem for OT. The child protectionist emphasis also obscures the contributory factors to OT contact, of both risk and bias. For example, we have not (yet) seen any significant actions to reduce the stresses of low income and poor housing, or increase in the funding of preventive services in the community.
So there are many complex intersecting reasons as to why there is greater intervention in more deprived and disproportionately Māori communities. These relate to risk, bias, site cultures/resources, individual perceptions and policy direction. Some of it relates to OT practices directly but there is much more to the picture – reflecting the intersecting nature of OT with other services and universal social protections.
Whatever the reasons, where a child gets born should not be such a strong determinant of contact with the state child protection service. Efforts to reduce both risk and bias are needed to even begin to address the clearly unequal outcomes. ‘Don’t give them a reason’ shouldn’t be part of received intergenerational wisdom. It is time to move past emotive rhetoric and for the complex inequalities embedded within the child protection system to be acknowledged, understood and addressed.
Keddell, E., Davie, G & Barson, D. (2019). “Child protection inequalities in Aotearoa New Zealand: Social gradient and the ‘inverse intervention law’.” Children and Youth Services Review.
Keddell, E. and I. Hyslop (2019). “Ethnic inequalities in child welfare: The role of practitioner risk perceptions.” Child & Family Social Work 0(0): 1-12.