Paying for the RoNS like Rail

Note: Consultation on the GPS closes at noon on Tuesday 2 April See our previous post on feedback guides and check out the Transport4All guide here.

The Government’s Draft Policy Statement (GPS) on Land Transport contains many ideological burbs and inconsistencies. One of those is around the rail network.

For example, the draft talks about how rail freight investment has seen a significant increase in recent years but rail freight usage has dropped. While this might be technically accurate, it fails to acknowledge that:

• much of the investment is once-in-a-generation spending, such as to replace equipment and infrastructure that is often many generations past its use-by date.
• there is typically a lag between investment and a change in usage.
• rail has many other wider economic benefits, estimated at $1.7-2.1 billion annually from things such as keeping trucks off the roads resulting in less congestion, emissions and better safety outcomes (some of this comes from the metro networks in Auckland and Wellington).
• despite the investment, there isn’t exactly a level playing field, given the larger, heavier trucks introduced from 2010 onwards that don’t pay enough road user charges to cover the damage they do to our roads, which is a key cause of our pot hole problem.

Going forward, the GPS suggests:

This Government’s focus will be to invest in maintaining the network between the busiest and most productive parts of the existing rail network – between Auckland, Hamilton, and Tauranga. While rail freight network investments will remain within the GPS, rail infrastructure will no longer be cross-subsidised from revenue generated from road users. It is unfair to ask people using the roads to fund rail infrastructure. Rail investments will continue to be supported and funded through the Rail Network Investment Programme, as agreed by Cabinet. Track User Charges paid by rail users will be used to support these investments.

The GPS proposes to cap rail investment to the amount of revenue generated from Track User Charges, which are fees paid to use the rail network – currently paid by the two metro networks and from freight users. The TUC from freight is currently around $20 million, but as this paper from 2021 highlights, just to cover the cost of wear and tear on the freight network was estimated at $53 million annually, while the cost to bring it up to, and maintain it at, a modern standard would require $420 million annually.

All of this means that much of the state of the network is likely to decline in coming years, and it won’t be surprising if in a few years KiwiRail are forced to close lines down – meaning more heavy trucks, on the roads, more potholes, and all of the other negative impacts that come with that.

A level playing field?

The GPS also pushes the government’s new Roads of National Significance, which they say will all be “four-laned, grade-separated highways“.

The Roads of National Significance include:

Whangarei to Auckland, with the following stages prioritised:

• Alternative to Brynderwyns
• Whangarei to Port Marsden
• Warkworth to Wellsford.

Tauranga to Auckland, with the following stages prioritised:

• Cambridge to Piarere
• Tauriko West State Highway 29.

Auckland roads:

• Mill Road
• the East West Link.

Roads to unlock housing growth:

• Hamilton Southern Links
• Petone to Grenada Link Road and the Cross Valley Link
• North West Alternative State Highway (SH 16).

Other major routes:

• Takitimu Northern Link Stage 2
• Hawkes Bay Expressway
• Second Mt Victoria Tunnel and Basin Reserve upgrade
• the Hope Bypass
• The Belfast to Pegasus Motorway and Woodend Bypass.

We already know that for many of these projects, the costs used by National in its policy are wildly wrong, in some cases out by billions. Furthermore, at least some of these projects, such as the East-West Link and Mill Rd, haven’t been designed as fully grade-separated roads – meaning their costs will likely increase even more.

Furthermore, most of these projects simply won’t have the traffic volumes to justify such investment.

So, the question is: what would it take to have the RoNS pay their way, like rail is required to do? How many vehicles paying fuel taxes or road user charges would be needed to justify the investment? Alternatively, given the government has also said they’re open to tolls as an option to help pay for these projects, what kind of toll might be required?

Estimating assumptions

To come up with these figures, I’ve made a few assumptions.

• I’ve assumed that around 15% of vehicles using a RONS road are heavy vehicles. This is higher that most of these roads see currently.
• For light vehicles, I’ve used the light vehicle RUC rate of $76 per 1000km
• For heavy vehicles, I’ve used the highest RUC rate for a powered vehicle of $435 per 1000km. Given many heavy vehicles will be lower than this, that should help make these numbers more conservative.
• Combined, this works out at vehicles paying an average of 13c per km.
• I’ve also assumed that these projects would be given a long timeframe to pay for themselves – over 50 years – and that they would be subject to an interest rate of 3%, which is lower than government 10-year bonds normally.
• On top of building the road, operations and maintenance will also be needed. To estimate this, I’ve looked at Waka Kotahi data on road maintenance spending by region and compared that to the lane kilometres of state highway that exist (length of the road multiplied by number of lanes). This gives quite a range, from $132k per lane km per annum in Auckland, to just $20k on the West Coast. I’m likely being quite conservative here, but I’ve assumed $30k per lane km per year or $120k per km of four-laned road.

The Results

I’ve only done the calculatoins for a handful of the RoNS projects where it is relatively easy to get a comparison.

Whangarei to Port Marsden

This project is 22km in length and the Waka Kotahi estimates released by Labour last year suggested the project will now cost $1.41 – $1.67 billion. For this exercise, I’ll use the lower cost. Based on that, just to pay the capital costs of the project, the annual cost works out to just under $55 million per annum, and maintenance costs bring the total up to around $57.4 million per year.

Average daily usage of this section of SH1 is around 18,000 vehicles north of Oakleigh and 15,000 vehicles from there to the Port Marsden intersection. So at current usage levels, road taxes over this section would cover only around 30% of the of the cost. A toll to make up the rest would need to be around $6.40 per trip.

Alternatively, to achieve full cost recovery, the average daily traffic would need to be in the range of 55,000 per day – equivalent to SH18 in Auckland.

Alternative to Brynderwyns

Waka Kotahi are currently busy cutting hillsides out of the existing Brynderwyns route to make it more resilient. Previous work has suggested a couple of possible routes, with the shorter orange route being around 13km. However, it’s unclear whether the government would also require them to tie it into Port Marsden in the north and Wellsford in the south, which combined would add around 28km and therefore a lot more cost.

The last time this route was given even a high-level costing was 2017, a long time ago. Given the terrain, using the lower-end estimate of $130 million per km of the Warkworth to Wellsford section seems appropriate as a guide. This would put the Brynderwyns bypass at $1.69 billion, which would give it a combined annual cost of $67 million.

A lot of traffic drops off at south of Port Marsden, so average daily traffic over the Brynderywns is only around 10,000 per day – meaning road taxes would only cover 9% of the cost. A toll to make up the rest would need to be around $16.70 per trip.

Alternatively, to achieve full cost recovery, the average daily traffic would need to be in the range of 109,000 vehicles per day: equivalent to SH1 south at Upper Harbour/Constellation.

Warkworth to Wellsford

This project is probably one of the most advanced of the RoNS, with a lot of business case work already completed and the designation work complete. The project is 26 km long.

The latest cost estimate for this project is $3.5 – $4 billion, with the higher end cost having almost doubled from just a few years ago. Again, I’ve used the lower of the cost range for my calculation, which gives this route an annual cost of $139 million.

Currently around 11,000 vehicles a day travel this route, meaning road taxes would cover about 10% of those costs. A toll to make up the rest would need to be around $31.30 per trip.

Alternatively, to achieve full cost recovery, the average daily traffic would need to be in the range of 113,000 vehicles per day: equivalent to SH20 south just of the Manukau Harbour Crossing.

Even with wider economic benefits included, the 2019 business case only suggested this road had a BCR of 0.7.

Cambridge to Piarere

This project is about 16km long and the cost estimates revealed last year were $1.5 – $2 billion, more than double the estimates from a few years ago.

Traffic on this section has certainly risen since the completion of the Waikato Expressway around Cambridge and is now around 20,000 per day, so road taxes would cover around 25% of the costs. A toll to cover the full cost of the road would therefore need to be around $6.20 per trip.

Alternatively, to achieve full cost recovery, average daily traffic would need to be in the range of 79,000 vehicles per day.

Even with wider economic benefits, the 2021 business case suggested this route had a benefit cost ratio of just 0.58 to 0.63.

The Belfast to Pegasus Motorway and Woodend Bypass

This project is around 8.5km and National has estimated the cost at $270 million. However, given other projects in the area, $350 million is more realistic and it could be even more.

With all the growth in this part of Canterbury over the last decade, this section now sees around 20,000 vehicles per day – meaning this section is probably one of the best performing on a cost recovery basis, coming in at around 57% of costs covered by road taxes. This means a toll would likely only need to be around a dollar to fully pay its way.

Alternatively, traffic volumes would need to increase to 35,000 per day – but that is almost at the level of the section between Kaiapoi and Christchurch.

As you can see from the examples above, none of these RoNS projects pays for themselves based on just the fuel and road user taxes generated from the vehicles that would use them, and they still wouldn’t do so even with a significant increase in usage. This means they are heavily cross-subsidised from people using other roads; exactly the same as the rail network is being treated.

This is also why historically we don’t build transport routes based on their financial impacts, but rather on their economic impacts – the improved connections and other benefits they enable. And on those measures, these newly proposed RoNS that have been assessed, don’t stack up either.

Share this