Alternate formats: the government doesn’t know the cost

Callum McMenamin — 30 April 2026

I asked the government how much each alternate format type costs. Here’s what I found.

Alternate formats are translations or adaptations of information for disabled people. They include:

• audio
• Braille
• large print
• NZSL
• Easy Read.

For a while, I’ve felt the alternate formats system in government isn’t working well. So I decided to do a bit of digging.

I wanted to understand how government agencies are allocating spending on these alternate format translations. So, I asked Whaikaha | Ministry of Disabled People for the following information:

Seems like a simple enough request, right?

Well, Whaikaha refused to answer the request. If you’re interested, here’s the OIA response.

To be fair, Whaikaha did consult with me quickly to inform me my first request was likely to be unsuccessful — this is good practice — agencies should work with requesters quickly to help refine requests.

A centralised service with no financial data

For some context, Whaikaha runs a centralised Alternate Formats Service, which is supposed to centrally co-ordinate government agency requests for translation, where they are prioritised then connected to alternate formats providers.

But, Whaikaha argues that while they centrally co-ordinate the alternate formats process, they don’t have any kind of oversight of alternate format expenditure across the government — basically, they haven’t been recording financial data:

This was rather frustrating. I found it strange that Whaikaha runs a centralised conversion process for government, but has literally zero financial data on the system. This means it’s impossible to research or understand government spending on alternate formats.

A simpler question — also refused

So I decided to refine my request to a question I felt was easier to answer, and Whaikaha could surely complete:

I felt this gave Whaikaha a lot of flexibility on how they could answer my request. If I got average figures on cost for each type of format, I can then extrapolate government-wide costs from other data.

But, this above request was also refused by Whaikaha:

Was the refusal justified?

So, supposedly, we, the public, are not allowed to know the cost of each alternate format type. The public simply isn’t allowed to know.

Whaikaha argues, that if the public knew how much alternate formats cost it would damage the commercial position of the organisations that produce alternate formats.

Whaikaha did not provide any justification for how the commercial position of alternate format providers would be damaged — they just asserted this as fact.

The Ombudsman provides guidance for agencies on how OIA Section 9(2)(b)(ii) should be administered by agencies. The Ombudsman guidance advises:

A bigger issue is that before refusing an OIA request under s. 9(2)(b)(ii), agencies must weigh the refusal against the public interest in releasing the information. In this instance, Whaikaha's refusal to answer my OIA contained no evidence the public interest test was applied in this case.

There is obvious and strong public interest here: disabled people, advocates, journalists, and politicians need to understand whether information is being made accessible. You can’t do that without the numbers. We need to understand whether the alternate format provider market is working, whether the centralised model is functional, and whether funding allocations are adequate or appropriate.

What did I discover? A collapse in translation numbers

Whaikaha did provide me with some useful data, even though they could not answer my core request. I do think this is helpful, and good practice.

What it revealed, is that alternate format translations have decreased significantly over the past few years.

Note that these figures only cover translations that went through Whaikaha's Alternate Formats Service. Whaikaha notes that some agencies bypass this service, so the data isn’t necessarily capturing all translations across the government.

That’s a 62% drop over three years, though the earlier figures are flagged as incomplete — so treat that number with some caution. Even the most recent year-on-year comparison, FY24/25 to FY25/26, shows a 38% drop. The trend is consistent: fewer translations are going through the central service, which might suggest that disabled people are getting less accessible government information.

This is why collecting this kind of data, and releasing it to the public, is so important. Before I ran my investigation, this information was not available. It’s crucial for the public to know.

Whaikaha also provided translation numbers for their own agency alone:

What this reveals, is that Whaikaha's own translations account for 44.6% of all translations through the Alternate Formats Service in FY 2025/26.

I think we can conclude with some certainty, that alternate format translations through the central service have collapsed over the past few years. To be more certain, that would require Whaikaha to actually properly collect and store this data in the first place.

The real scandal is not that the numbers are bad. It’s that the system is arranged so we can’t even know if the numbers are bad.

Can we guess at the alt format cost?

Whaikaha did provide the total alternate format cost for its own agency, and the total of translations performed. It did not break down this cost for each format.

For FY25/26, Whaikaha spent $214,436.13 across 268 translations — roughly $800 per translation on average.

To me, this average cost does seem reasonable — however such high-level averages hide a lot of nuance. We don’t know how expensive Easy Read or NZSL is, compared to large print, for instance. I was particularly curious to know how expensive audio and large print formats are, given these formats can be 100% automated (like they are for this blog post you’re reading!)

Institutional memory loss

My OIA for this research was impacted by institutional memory loss.

The Alternate Formats Service used to be run by the Ministry of Social Development, and it was then shifted into Whaikaha. When this transition occurred, institutional memory was lost, indicating poor or incomplete data management processes at either MSD, Whaikaha, or both:

What should happen next?

Whaikaha should:

• collect core statistics for the Alternate Formats Service
• show per-agency translation rates
• publish those statistics proactively.

I have a track record of persuading public service systems into being transparent on their accessibility — I think this is the next system that should be opened up. Accessibility cannot be done in the shadows: progress only eventuates if it’s public.

Conclusion

So what did I learn in this process?

1. The public isn’t allowed to know how much individual alternate formats cost
2. Whaikaha has poor quality data for its Alternate Formats Service
3. The shift from MSD to Whaikaha exposed institutional memory loss
4. Alternate Formats Service translation rates have significantly collapsed over the past few years, if we trust the data they provided.

This experience has reaffirmed my belief that the OIA is not fit-for-purpose. In the past few years, I’ve run a few government investigations using the OIA as my primary investigative tool. Getting good information out of the government is often difficult, bureaucratic, and you get OIA requests refused under dubious grounds. I’ve spoken with reporters who have had exactly the same experiences as me.

I think the OIA is a core pillar of New Zealand’s democracy. The OIA is a crucial law that allows the public to know what the government knows. It lets us speak truth to power. It lets us protect the rights and freedoms we enjoy today.

Many disability rights advocates rely on the OIA to give us crucial information we need to inform our advocacy work. When agencies start refusing, it diminishes our ability to be an informed advocate, as we’re refused access to the truth.

I hope we see reform some day, that substantially reduces wait times, and refusal rates of OIA requests. I think that will help to strengthen our democracy, and it’d give advocates the crucial data they need to do their job effectively.

— Callum