Agencies treat OIA requests as threats rather than opportunities for transparency. The maximum time for compliance with a request has become the default, despite the requirement to respond “as soon as reasonably practicable”, and requests for “clarification” reset the clock.
Most troubling is the avoidance of record creation. Officials are pressured not to create records for fear of them being OIA’d, which is a complete inversion of the act’s intent.
A leaked video reported in the Herald on September 22, 2026, showed Education Ministry curriculum advisers being told not to put material in writing for exactly that reason.
All this is the more frustrating because most records are now digital. Proper information management means that large document sets can be narrowed quickly with trained search tools. The grizzle that searching 600 documents will cause delay does not cut it. The digital paradigm should have made compliance easier. The opposite seems to have happened.
Against that background, the Ministry of Justice, which administers the act, has been reviewing its operation. In late 2025 it commissioned Tregaskis Brown to model the system’s costs. The November 2025 report estimated that responding to OIA requests cost the public service more than $183 million in 2024/25 and identified a 394% increase in requests since 2016, partly attributable to changes in reporting and classification.
Further research in 2026 identified the drivers of cost: rising volume and complexity of requests, with publicly available information hard to locate; inconsistent approaches to scoping and triage; fragmented repositories, weak search capacity and poor information-management maturity; and largely manual response processes.
The initial suggestions are unremarkable: more proactive publication, easier access to previously released information, better information-management systems, standardised workflows, self-service for routine information, and improved training. Apart from a veiled reference to charging, these would enhance the objectives of the act.
The August 2026 report is more concerning. It raises possible legislative responses to requests requiring substantial collation, requests considered frivolous or vexatious, and requests involving commercially sensitive information. It looks to Australia, where application fees and an abuse-of-process refusal ground have been proposed. It questions whether the definition of “official information” and the range of agencies covered remain appropriate.
Andrea Vance, writing in the Post, reports that officials want cost-recovery guidelines, citing law firms, finance companies and repossession agents who use the OIA as a free tool for debt recovery and litigation, together with measures to curb broad AI-generated prompts, serial requests and automated submissions. Vance reasonably reads this as a plan for a post-election overhaul.
Two kinds of reform are on the table. The first reduces the cost of providing information. Proactive disclosure, searchable repositories, better records management and self-service portals are not restrictions on access. Rather, they make information easier to obtain. Moving thousands of near-identical requests into a searchable database could cut costs and increase openness at the same time.
The second reform proposal reduces cost by reducing the right to information. Application fees, processing caps, broader refusal powers, expanded vexatious-request provisions, a narrower definition of official information and narrower institutional coverage all alter the balance the act embodies.
The first proposal enhances the act’s promise and sits within its structure.
The second reverses the statutory presumption of information availability and requires a far stronger justification from its proponents.
The ministry places the work under “Transparency and scrutiny of executive action”. The problem is that the exercise is driven predominantly by cost, burden and sustainability.
The danger with this approach is that “sustainability” becomes shorthand for suppressing demand by making access harder. That result would be welcomed by a bureaucracy long practised in obfuscation.
We need to remind ourselves that the central principle is not that government should answer requests as cheaply as possible but that information shall be made available unless there is good reason to withhold it.
The estimated cost of $183 million to administer the scheme is not simply the price of public curiosity.
Much of it is the cost of accountable government. Rising demand reflects the growing complexity and intrusiveness of state activity. And on the Government’s own research, a large share of the cost stems from the state’s own failings such as dispersed records, low information-management maturity, manual processes and inadequate proactive disclosure.
These are distinct problems and should not be conflated. Before restricting the public’s right of access because compliance is expensive, the state should fix the expense it creates for itself.
The same applies to requester behaviour. Serial requests and automated submissions are often a response to delay and evasion. If agencies were more responsive, there would be less need for follow-up requests.
The August 2026 briefing highlights an internal tension. Although proliferating legislative exemptions may undermine perceptions of transparency, yet it nevertheless proposes examining scope and refusal grounds largely because of cost.
Charging would create financial barriers for journalists and watchdogs; one need only count the news stories built on OIA material to see what is at stake.
Narrowing the definition of official information would simply place more State activity under an opaque cloud.
None of this amounts to a return to the Official Secrets Act, but it would be a severely retrograde step, and one that engages the right to seek and receive information under section 14 of the New Zealand Bill of Rights Act 1990.
Some of the proposals, including restrictions on what counts as official information, narrower criteria and charging, would be music to a bureaucrat’s ears.
One wonders whether the consultants’ reports were developed to bolster an existing policy, and why a second report was needed when the first was relatively anodyne.
The reports were released at the start of the electoral cycle and have drawn no comment from the Minister of Justice. The proposals go to the heart of the Fourth Estate’s ability to hold power to account.
An overhaul confined to the state’s own information management would fulfil the promise of the OIA. One that changes the act itself would erode it.
Catch up on the debates that dominated the week by signing up to our Opinion newsletter – a weekly round-up of our best commentary.