A mishandled allegation of abuse has revealed shortcomings in how Oranga Tamariki uses one of its most powerful tools for keeping children safe.
The Child Protection Protocol is meant to ensure effective joint action by Oranga Tamariki (OT) and police to investigate and safeguard a child where there is a risk of immediate serious harm, or serious abuse is identified or suspected.
But in the case of a Northland boy who accused his mother's partner of abusing him, before retracting the allegation soon after, social workers failed to follow protocol in more than a dozen different ways - including not following up on the boy's "ongoing safety" and not looking into his retraction.
"Many aspects of it were not completed to a high standard," OT said in a summary of a review into the case, released to RNZ after a two-month wait.
The man accused says because the investigation was mishandled, it was inconclusive, and so was unsafe for the boy as well as for the blended whānau.
OT did not acknowledge it had social workers who were not following the protocol, until after it was forced to do an internal review in response to a complaint from the accused to its chief executive's advisory panel and the Ombudsman.
The review found "a lack of scrutiny and oversight".
When it went looking further, the agency told RNZ it found "inconsistencies across the country".
The flaws in Child Protection Protocol (CPP) practice have not been made public until now.
In the Northland case, investigators found:
- the Whangarei OT office did not have a good grasp of the protocol
- no updated assessment of the boy retracting the allegation
- "unusual" use of evidential interviews, with little preparation
- "very unusual" use of an extended forensic assessment
- lack of consultation, supervision and meetings, and delays signing-off reports
- OT did not tell police it had closed the case
The Oranga Tamariki investigation ran from July to October 2018; the review came out last August.
The agency refused to release a copy of the 10-page review to RNZ citing privacy reasons; instead, it eventually released a six-paragraph summary. Read the summary here.
The summary - which does not identify Northland - said that despite the protocol practice needing to be stronger, social workers "maintained a focus on the wellbeing and best interests of the child".
The review itself does not say that.
A copy RNZ obtained said there was "limited" social work done around the evidential interview, and no detailed assessment plan of the boy's circumstances, especially necessary given the retraction he made.
"After the initial CPP consultation, there was no further assessment recorded relating to [the boy's] ongoing safety, where he was living after the school holidays and who was living in the household."
A joint visit to the home by OT and police was ticked off as not being required, with no record of any discussion, even though agencies did not know if the accused was living at the house.
The case was close in October 2018 without an updated assessment that explored the boy's retraction of the allegation.
The boy's safety, and any further investigation, was then "left to the police".
Police kept the file open for a year, interviewed the accused, then closed it with no further action, the review said.
The man said it had been a real struggle to find out the outcome of the review he sparked.
It shows the problems with the decade-old protocol extended further afield than Northland, to patchy training and an auditing vacuum.
"We are not aware of any regular quality assurance activity which happens within OT around CPP compliance, either regionally or nationally" despite that being required, the review said.
"We are aware training about CPP process and the joint obligations is variable across the country ... we also understand that no nationally-led joint training has occurred for some time."
No professional development training module existed, even though documents referred to it.
After RNZ raised this with OT, it said: ''We have identified inconsistencies across the country and made improvements."
The Northland case influenced a national review last year - a routine review, OT said, that led to "signficant" changes, including the start of national training next month.
Also for the first time in July it said it would establish "regular mechanisms" to assess regional and national compliance.
"This will include considering the role for any independent assessment or assurance processes," the agency said in a statement.
The agency had also now appointed protocol contact people at each office, who must meet police at least once a month to review all CPP cases.
Police said mechanisms existed to monitor CPP activities, and they had regular local joint training and meetings with social workers.
"It is important we get this right," police said.
The accused's complaint against a social worker was dismissed by the Social Workers Registration Board, saying the "issue appears to be systematic".
The revised protocol updates references to the law; amends the process of launching a CPP case; improves guidance to support child victims and whānau; and adds sections on cumulative harm and a stronger multiagency approach.