What will Warkworth to Te Hana cost us? And other burning questions

Last week, late on Thursday afternoon, Minister of Transport Chris Bishop announced that  the government had signed the contract for a Public Private Partnership for Warkworth to Te Hana.

This answers our long-running question: are they seriously going to go ahead with this?

It also leaves many more questions, which are only just beginning to be asked more widely, and answered in any kind of detail.

What will Warkworth to Te Hana cost us?

The announcement put the sticker price for the project at $3.649 billion. However, we do not yet know the total cost for the project, as there has been no mention of what the annual PPP payments will be.

As the minister said:

“A PPP is like getting a mortgage to buy a house. Rather than paying the full cost upfront, it is spread over a longer period and repaid over time.”

And, just as with a mortgage  – and as we have seen with Transmission Gully, where the final bill is likely around triple the initial named price – the total cost to the public over the lifetime of a PPP will be much higher than the initial number. So, every journalist reporting on this project needs to ask what those annual repayments will be.

One number we do know: the government has provided a Crown Loan to NZTA of $1.6 billion, which leaves around $2 billion that will be financed by the contract winners.

This means if Warkworth to Te Hana tracks similarly to Transmission Gully, we could be looking at three times that $2 billion over the total contract of the PPP.

In other words, the public could be stumping up anywhere up to $7.6 billion for this road over the 25-year term. But we don’t know exactly, as the government has said nothing, and NZTA has redacted any information about the PPP financing numbers.

What’s the claim about “savings”?

In Thursday’s announcement, the Minister said the project “represents good value for taxpayers”, pointing to the “$3.9 billion Public Sector Comparator approved by Cabinet in March last year.” That’s an estimate of what the price would have been “under traditional public sector procurement”, i.e. if the government were to cover financing for all costs itself –  a highly theoretical situation in any case, as the intention has always been to procure through a PPP.

So, now with the PPP price tag of $3.649 billion, the government is claiming a “savings” of $251 million.

But the thing is, because they’ve signed a PPP committing us to decades of repayments, they will almost certainly be spending billions more to achieve that claimed “savings” of millions.

It’s like purchasing a Bentley on a payment plan, then crowing about getting a wee discount on the car stereo. Not a great look for a government that’s telling everyone else to cut costs and tighten their belts.

What new information do we now have about the costs?

Funding questions aside, one of our other long-running questions was: what makes this project such a top priority that this government wants to pour all of our collective transport funding for new projects into it, for the foreseeable?

NZTA has now released a bunch of information about the implementation plan, including an Investment Case dated January 2025. Throughout, key figures and numbers have been redacted – mainly relating to anything to do with what the PPP might actually cost. For example:

The lines in the table above were redacted under section 9(2)(j) of the Official Information Act, which is normally used to:

enable a Minister of the Crown or any public service agency or organisation holding the information to carry on, without prejudice or disadvantage, negotiations (including commercial and industrial negotiations); or

But considering the contract has now been signed, these redactions hardly make a lick of sense?

How has the project met the stated thresholds for greenlighting?

The next thing to catch the eye in the January 2025 Investment Case is that NZTA says that two of the trigger points for proceeding with the this project have been met – one being safety and related road closures; the other being traffic volumes:

These triggers for implementation have been met, with the forecast Dome Valley DSI savings (down to zero) already exceeded in the first 18 months and the number of unplanned closures well above 100% higher than 2018 levels. Whilst current traffic volumes are not in excess of 25,000 AADT, forecasts indicate this being reached in the long term.

Let’s start with safety. Wait, so the Dome Valley safety improvements to SH1 were supposed to reduce Deaths and Serious Injuries down to zero?

Isn’t it interesting how Vision Zero can suddenly re-enter the chat, when you least expect it.

In fact, those safety fixes – carried out under the previous government between 2019 – 2022 – have dramatically improved safety, and are saving lives, as you can see on the graph below. All for the bargain price of around $90 million (less than half what’s been spent on property acquisitions for Warkworth to Te Hana, at $190 million).

However, apparently because some serious crashes are still occurring along the Dome Valley stretch, that’s a “trigger” to spend 40x as much on the safety fixes to carve 26km of an entirely duplicate new four-lane highway through the landscape.

NZTA also says the claimed trigger for reliability has been hit (i.e. how many times the road is unexpectedly closed). This is surely due largely to the impacts of extreme weather, which particularly impacts the Brynderwyns where there are very real, urgent and costly problems with frequent slips and closures. So, another question:

Why then proceed with a brand new project that, by encouraging more driving, will increase climate-changing emissions, thus increasing the likelihood and impact of said extreme weather events? Especially when the brand new four-lane expressway from Warkworth to Te Hana doesn’t even touch the Brynderwyns?

Indeed, before the 2023 election, National explicitly called out proactive resilience improvements as crucial for this part of the road network. Spending (a minimum of) $3.649 billion on an entirely new stretch of road that stops short of the truly vulnerable section is a very costly and roundabout way to (not) solve those reliability issues.

Lastly, NZTA couldn’t claim to have met the traffic volume threshold of 25,000 vehicles per day. That’s because this won’t happen until 2050, even assuming a heroic growth in traffic volumes.

So by every measure, it still makes no sense to urgently commit to spending all this money on Warkworth to Te Hana as currently designed.

And yet NZTA and the government have chosen to do so.

Running costs are the least of it

And yet it should be noted that this stretch of four-lane expressway will continue to cost us (the nation as a whole) a minimum $15 million a year to maintain and operate, forever.

Meanwhile, the old SH1 through Dome Valley will most likely revert to Auckland Council, meaning Aucklanders will shoulder (i.e. pay for) maintenance and renewal and deal with all the issues that extreme weather damage will bring to that toll-free alternative route, in perpetuity:

No substantive upgrades to the existing SH1 are proposed between Warkworth and Te Hana due to the recently completed Dome Valley safety improvements, however given recent resilience challenges in this part of the network, additional maintenance and repairs have been allowed for in the Do Minimum costings (at $60M, and we note that the recent works following the Auckland Anniversary weekend floods was in the order of $20M).

So what about those Costs and Benefits?

The 2025 document also reveals the Benefit-Cost Ratio. Remember, this is the top secret number that NZTA and the Minister of Transport had argued could not be released, lest it put into question the government’s ability to commit to the project.

How bad could it be?

1.4

As in, for every dollar invested, this road will return a dollar forty. Not much more than break-even.

That’s 1.4 excluding Wider Economic Benefits (WEBs); rising to 1.6 if you include WEBs.

What’s striking about this calculation is that the BCR has now doubled since the 2019 business case, where it was 0.7 – even though the project has doubled in cost in the meantime.

How is this possible? It’s down to a change of calculus. A new method lowers the discount rate applied to benefits over years. We “value” more immediate benefits, and reduce their value over time through a discount rate. And in short, NZTA has changed the way it measures projects, so that more “benefits” can be claimed as time passes by reducing their “value” less over time.

This was intended to better capture the longterm benefits of investment; it’s not clear that’s what’s actually happening, nor that the full range of benefits and disbenefits are in the picture (induced demand, anyone?).

Regardless, to be effective, this method really needs to apply to all projects and options on the table… which this wasn’t.

We can also see what the BCR might look like with higher discount rates: i.e. if it were more aligned with the way projects were previously measured, claiming fewer “benefits” over time. Basically, under the old approach, the project always struggled to get its head above water, even with “Wider Economic Benefits” folded in. Thumbs on the scale?

This is really not good. Remember that a BCR of less than 3 is considered low:

BCR Rating NZTA’s 2024 Investment Prioritisation Method

And note that even if you include tolling (which, as Matt noted, wouldn’t even touch the sides in terms of covering the cost of the road), the BCR for this project still appears to be under 1.

As in: a losing bet. A waste of money. The game is not worth the candle. We do not get back as much benefit as we spend.

By the way: note the last line on the image, redacted under the same reasoning as the PPP data.

Is it what the BCR would be if you took into account the enormously more expensive approach of a PPP, perhaps?

This recalls a stark point the Ministry of Transport made about the RoNS in August 2025. We can’t see what’s been redacted, but the document states plainly that the cost to deliver projects is usually higher – and the benefits lower – than forecast by business cases, meaning “a BCR of 1 may not represent value for money, if costs increase and benefits decrease.”

From a Ministry of Transport advice document to Chris Bishop (27 August 2025) regarding the Road of National Significance Programme

Presumably this advice fed into the decision to “reprioritise” the RoNS. But was it applied to Warkworth to Te Hana?

Because, if the cost of delivering this expressway rose above $3.65 billion and/or the benefits were lower than expected, the BCR would be less than 1, and the absurdity of using so much of our collective resources on it would be immediately clear.

Is this why there has been no transparency when it comes to the total cost of the project under the PPP? Is the government worried if they reveal how much funding is being blown on this, that the case for the project becomes untenable to the public?

The road-building juggernaut vs the government of the day: who wins?

It’s important to note that a low BCR doesn’t necessarily tank a project. It just means you’d need other factors in order to “prioritise” it – like, for example, political cover by listing it in the Government Policy Statement.

And even with a dog of a project (no offense to actual dogs), prioritising it might be understandable, if there was overriding and long-running bipartisan political will to deliver it. Maybe?

Except that in 2019, when NZTA was progressing this project – including seeking consent – it seems highly likely that they outright ignored the direction of the 2018 Government Policy Statement on Land Transport, which had an explicit strategic focus on safety, emissions reduction, value for money, and mode shift.

How can we be sure if NZTA overrode direction? It’s a strong inference, based on the Investment Quality Assurance advice re Warkworth to Wellsford from that period, which I’ve acquired under the OIA:

Check out the full advice here

The section above says (emphasis added):

A countervailing reputation risk remains, given the project’s medium results alignment and low prioritisation, that the Agency is viewed as giving insufficient weight to the GPS’ priorities.

[REDACTED]

While it is acknowledged that the Minister cannot direct the Agency’s course of action on specific projects[,] consideration will need to be given to whether continuing with route protection is consistent with the GPS.

Want to guess what’s redacted?

The core point here is that despite political direction asking NZTA to reconsider these enormous wasteful projects, NZTA continued developing this project – for which the BCR at the time was just 0.7.

Serious questions should be asked about the actions of decision-makers over this time. Why was the political direction to reconsider low-value motorways seemingly ignored in order to continue prioritising this project, until suddenly it could be accelerated when the political winds changed?

Why this road, why now?

The question remains: how on earth did this project rise to the top of the to-do list?

Is expediency – and an upcoming election – the reason? It’s a reasonable guess: Cambridge to Piarere aside, Warkworth to Te Hana was likely the only one of the RoNS from National’s 2023 election campaign that was not just a fantasy.

But the point of a BCR as a tool is that it lets you compare options and priorities, to determine the best use of finite resources – not just money, but time, workforce, and foregone opportunities.

So, logically, a “reprioritisation” exercise of the proposed Roads of National Significance, carried out while publicly conceding that the programme as a whole is simply unaffordable and is “keeping you up at night” – would look extremely hard at that angle.

And if you compared the BCR of Warkworth to Te Hana to the other RoNS (flawed as all these projects as proposed may be), you can see it would be nowhere near the ‘best’ one to progress:

Note, this isn’t to justify the other RoNS. These projects as designed are neither affordable nor a priority as currently designed – which is why Chris Bishop has effectively cancelled most of them.

For Thomas Coughlan's article on Roads of National Significance:www.nzherald.co.nz/nz/politics/…

— Guy Body (@bodycartoon.bsky.social) 2026-07-10T23:26:39.717Z

And for a proper cost-benefit comparison and “reprioritisation”, you’d be sure to compare options and stages within each proposed project. So that, if you found a number of lower-cost options offered more bang for buck, you could choose to do a bunch of those – maybe across multiple locations – instead of committing to One Giant Project.

That’s certainly how you’d approach a household renovation, to borrow a favourite metaphor of this government – especially when working with a limited budget in a gloomy economy and uncertain climate.

You’d be looking very hard at leveraging your opportunities to future-proof any of your expenditure, the equivalent of putting in solar panels when fixing your leaky roof. And you’d absolutely want to double-check your fundamental assumptions behind proceeding with One Giant Project, versus all the other things you could do with that time and money. Right?

So, looking at the claimed upsides of Warkworth to Te Hana, the vast majority of the projected economic benefits of this investment (85%) come in the form of ‘travel time benefits’. (Meanwhile, emissions are 0%, funny that, except that this is no laughing matter.)

This is the part that feels like a joke. We haven’t even met the volume of traffic to trigger this project – and won’t for another 25 years, even taking at face value the heroic assumptions of NZTA. But evidently the need for thousands of ghost drivers to save a few minutes somehow seals the case that Aotearoa needs to urgently build this particular road, right now?

The opportunity costs are incalculable – but let’s try

None of this makes sense at all.

Especially when you know how much more you could achieve for road safety, resilience, and yes even travel time savings, all over the country, for $3.649 billion (and more over time).

Well, going by the sticker price, ignoring PPP cost escalations, at $140 million per km of Warkworth to Te Hana, here are a few options we could have for a fraction of a RoNS:

• 540 metres would fund the scrapped Marsden Fund for a year ($76m)
• 4.1km would fund Te Huia train service for 100 years ($5.82m per year)
• 920m would fund the entire country’s walking and cycling budget under the 2024 GPS ($130m)
• 190m would restore the recently cut arts funding for the next four years ($27m – noting that investment in live performance has a BCR of 3.2)
• 270m would fund the Auckland City Mission for an entire year ($38.8m)
• 64m would fill NZTA’s share for the community led fixes to the Hill Street intersection in Warkworth ($9m)

It’s also striking to see many folks doing similar calculations – measuring things of great value in metres of RoNS. What’s your favourite example?

My take on all this: Warkworth to Te Hana will go down as one of the worst deals a government has ever signed.

It is a badly chosen project, vastly over-scoped, and procured via one of the most expensive ways of doing it – in this economy, and in this climate. And it does precisely nothing to solve the actual urgent problems with keeping Northland connected, safe and productive.

That is, even assuming everything goes well with construction – and given the Dome Valley’s geotechnical issues, we can expect it will not be smooth sailing.

Even by this government’s own stated values, it misses the mark; and we will all be counting the cost for years to come.

Share this