In May 2025, New Zealand’s Office of the Auditor-General published an inquiry into Oranga Tamariki’s procurement and contract management. The agency had cut $60 million from contracted social services in a single year, recovered funds it considered unspent by providers, and run the entire process outside open or competitive procurement. The Auditor-General’s conclusion: ‘the effects of decisions on children and their families are still not known.’

This is procurement logic applied to social services commissioning. The logic worked exactly as it was designed to. That is the problem.

The exhibit

Oranga Tamariki contracts roughly NZ$500 million per year in social services through external providers. In 2024/25, the agency decided to reduce contracted services spending by $60 million and simultaneously recover funding it considered unspent. The OAG found the decisions were not made through open, transparent, or competitive procurement. There was no system for managing conflicts of interest. Te Tiriti partners were not considered. Decisions were ‘not adequately informed by evidence or an understanding of how decisions would affect children and their families.’ And outcomes for children ‘are still not known.’

The question the agency asked was: where can we reduce contracted spend? The question the agency did not ask was: what are children currently receiving through these contracts, and what happens to them when the funding changes?

These are different questions. Procurement frameworks are broad enough to engage with the second — total cost analysis, demand mapping, social value criteria, consequence-weighted decisions. But the capability tends to concentrate in sourcing. Ask a sourcing specialist whether contracted spend can be reduced and the answer is yes. The question of what happens to the people receiving those services does not register in that frame, not because the tools are absent, but because nobody trained to cut costs was asked to think about what the costs were attached to.

Why the framework produced this result

‘Value for money’ in contracted social services is not the same calculation as value for money in a goods or services market. In a standard commercial context, spend reduction produces a financial benefit with assessable trade-offs: you pay less, you receive less, the gap is visible. In social services commissioning, the unit of value is not a deliverable. It is whether a child received the support their assessed needs required. Spend reduction does not generate information about that question. A $60 million cut in contracted services is a reduction in some quantum of service to some number of children whose identities and needs were not enumerated in the decision.

The OAG’s finding that outcomes are ‘still not known’ is not a monitoring gap. It is the direct consequence of the analytical frame used to make the decision. If the question asked is financial, the answer returned is financial. The children appear as a line item, not as the subject of the analysis.

The incentive chain that nobody examined

Recovering ‘unspent’ funds from providers creates a specific and predictable incentive for every provider in the system: do not underspend, because underspend is reclassified as an overpayment. The rational response is to spend to contract regardless of whether the spending reflects service need, or to suppress reporting of underspend to avoid clawback in the next period.

The 12% budget reduction applied across contracted social services creates a supply-side shock that procurement frameworks have no standard tools to measure. Social services providers are not interchangeable on short notice. Many are Maori and community providers whose relationships with the families they serve are not transferable to alternative suppliers. Excluding te Tiriti partners from the process, as the OAG found, did not simply miss a consultation step. It removed the providers most likely to understand what the service demand actually looked like.

Contract management systems track payments against milestones. They do not track whether the child who was receiving a service in March is still receiving it in September, or whether their circumstances changed when the contract did. The incentive effects of the procurement decisions were real and traceable. They were simply not part of the analysis.

What the same decisions look like, made by someone who read the source

The decision facing Oranga Tamariki in 2024 was a commissioning decision, not a procurement decision. The right question was: given the children who need social services, what is the minimum adequate level of contracted provision, which providers have the capability and community relationships to deliver it, and what does responsible reduction look like if reduction is unavoidable?

Procurement tools can support that analysis. They cannot substitute for it. An evidence-informed reduction would have mapped service demand against contracted capacity, identified which contracts could absorb cuts without service disruption, maintained competitive processes to preserve provider confidence in the system, and built monitoring capable of detecting gaps after the changes took effect. The OAG found none of these. The absence is not incidental. It follows directly from treating a commissioning decision as a procurement problem.

What the standard response misses

The procurement industry response to findings like this is predictable: clearer contract KPIs, improved commissioning frameworks, consultation with affected providers, stronger governance. These are process corrections. They do not address the underlying problem, which is that value-for-money logic applied to social services commissioning as though it were a standard market consistently produces decisions whose consequences cannot be assessed.

The OAG did not find that Oranga Tamariki had poor procurement governance in the abstract. It found that procurement decisions affecting an unknown number of children were made without asking what would happen to those children. Adding a contract management plan to that process would have produced better-documented decisions with the same informational blind spot.

Procurement professionals who want to work in social services commissioning need to accept that the frameworks they carry were built for different problems. If you cannot say what a $60 million cut did to the children it affected, you have not demonstrated value for money. You have demonstrated that value for money was the wrong measure.

The Auditor-General said it plainly. The effects on children are still not known. That sentence is not an indictment of record-keeping. It is an indictment of the question the agency thought it was answering.

Source: Office of the Auditor-General New Zealand, Oranga Tamariki: Inquiry into Procurement and Contract Management, May 2025, https://oag.parliament.nz/2025/oranga-tamariki