Accountability After the Fact
Why independent reviews have become Aotearoa's weakest control
When independent reviews stop reassuring
Independent reviews remain one of Aotearoa’s most trusted instruments of accountability. When systems fail, reviews are commissioned quickly, led by credible figures, carefully scoped and written in serious language. They establish facts, correct misinformation and signal what went wrong. In a high-trust system, that once provided genuine reassurance.
Increasingly, it does not.
This is not because reviews are poorly conducted, nor because the experts involved lack integrity. It is because independent reviews have been asked to perform a function they were never designed to fulfil: to stand in for governance where prevention, enforcement and consequence are weak.
Independent reviews explain failure. They do not prevent its return on its own.
The category error at the centre of accountability
There is a quiet but persistent mistake in Aotearoa governance. Once an independent review is commissioned, especially by a Big 4 consultancy, senior barrister or former judge, it feels that sometimes responsibility is implicitly treated as largely discharged.
In reality, that moment marks the intensification of responsibility.
Commissioning an independent review is not closure; it is diagnosis. And the more credible the diagnosis, the higher the obligation to act. Yet governance attention often relaxes precisely at this point. Findings are accepted. An action plan is released. Assurances are given. Attention moves on.
The error is subtle but profound: analysis is mistaken for control.
Reviews are lagging indicators, not risk architecture
Independent reviews are, by design, lagging indicators. They activate after the harm has occurred. They explain failure once thresholds have already been breached. That is not a flaw; it is their purpose.
The problem is that Aotearoa increasingly treats lagging indicators as if they were preventative measures.
No mature risk system, regardless of industry or area, relies on post-incident explanation to manage live exposure. Prevention sits further upstream: in escalation thresholds, defined trigger events, tracking emerging risk, stop-loss mechanisms and clear ownership of systemic risk. As a risk professional, I find these controls dull, politically inconvenient and largely invisible. They are also decisive.
Where those controls are weak, explanation multiplies.
Mount Maunganui: warning without consequence
The Mount Maunganui landslide illustrates this failure mode with uncomfortable clarity.
In this case, the risk was not unknown. As highlighted in detail by Dr Bryce Edwards, technical advice warned that slopes were primed to fail. Technical advice warned that slopes were primed to fail. The indicators were explicit, documented and time-bound. What did not exist was a mechanism that translated this warning into obligation, a trigger that forced escalation, intervention or transparency before failure occurred.
After the slip, as I expected, media reports of an independent review on the disaster followed. But this tragedy was not epistemic. It was not about insufficient information. It was about preventative governance. Risk has been identified, but it carried no consequence — only inaction.
Where reassurance quietly leaks away: the action plan
Independent reviews almost always conclude with an action or implementation plan. In theory, this is where learning becomes practice. In reality, this is where reassurance most often dissipates.
Action plans sit in an awkward institutional space. They suddenly are no longer independent, but they are rarely binding. Without the reviewing party to carry the momentum and the action plan back in “business-as-usual”, milestones are missed and deadlines are quietly moved back. Commitments are reprioritised without formal acknowledgement and ownership diffuses across agencies or workstreams. Progress is reported through activity to the Boards that have oversight. These are things like policies being drafted, training planned to be delivered and refreshing same frameworks — rather than through direct evidence that risk exposure has actually changed.
From a governance perspective, this is the most fragile link in the accountability chain. The document that matters most is also the one least scrutinised.
The public senses this intuitively. Reviews command attention because they are moments of visibility. Action plans unfold after the attention fades. There is no equivalent ritual confirming that commitments were enforced, controls embedded or risk materially reduced.
Pike River: diagnosis without authority
This is where Pike River matters, not only as tragedy, but as structural warning.
Pike River is often described as a case of warning signs ignored. The deeper lesson is more unsettling: risk was identified, examined and documented. External scrutiny raised concerns about ventilation, methane management and the tension between production pressure and safety. Reviews and audits did not fundamentally misunderstand the risk.
What failed was not diagnosis. It was governance after diagnosis.
Independent scrutiny did not translate into authority strong enough to override operational incentives. Documentation improved. Assurance processes existed. But none of these mechanisms had the power to force cessation, redesign or withdrawal from unsafe conditions.
In that sense, Pike River is not a case where review arrived too late. It is a case where review did not carry consequence.
The system behaved as though risk had been managed because it had been analysed. but risk management without enforcement is simply risk tolerance, expressed in formal language.
Why reassurance keeps thinning
Public unease is not driven by disagreement with findings. It is driven by pattern recognition.
Each new review reassures less because it confirms a deeper truth: the state learns after harm, while prevention remains operational. Over time, trust does not collapse at once; it thins. Cynicism replaces confidence. Populist narratives gain traction, not because people reject expertise, but because expertise appears structurally disconnected from action.
The democratic risk here is quiet but real. A system that can explain failure but cannot anticipate it will eventually lose its authority to govern, even if its intentions remain sound.
What would actually restore reassurance
Reassurance will not be restored by fewer reviews, nor by better-written ones. It will be restored when reviews are repositioned as starting points, not endpoints.
That requires a shift:
from explanation to execution
from activity to control effectiveness
from accepted recommendations to enforced obligation
Preventative controls do not eliminate failure. But they change its frequency, severity and credibility. They demonstrate that warnings carry weight and that known risks cannot sit indefinitely without decision.
In the end, reassurance does not come from being told what went wrong. It comes from seeing that systems are capable of acting before things go wrong.
A state that can explain failure but cannot prevent its recurrence is not accountable — it is merely articulate.
Until that balance is restored, Aotearoa will remain excellent at understanding failure and inconsistent at preventing its return.
Superb insights Angus. I particularly liked this: “The democratic risk here is quiet but real. A system that can explain failure but cannot anticipate it will eventually lose its authority to govern, even if its intentions remain sound.” Those two sentences can be applied far more widely too. For example, the inability to anticipate the potential of inequality to dramatically destabilise a political system. I would argue that being a day late and a dollar short in a New Zealand political context, is cultural. Culture after all simply means, the way things are around here.