Wellington and the "long tunnel"

Early LGWM "Long Tunnel" option

I don't tend to do parochial posts on here, but this has my interest, not least because I grew up in the eastern suburbs of Wellington, went to school for some years via Mt Victoria Tunnel, and now once again, live in the eastern suburbs of Wellington after a 20 year absence. I've followed this with interest because for all of my life Mt. Victoria Tunnel has been seen as "inadequate", and it is. This post attempts to explore what might be done, and why...

The raising of the "long tunnel" initially made public under Let's Get Wellington Moving, and revived under the National-led Government since the 2023 General Election is all about a fundamental land transport problem in Wellington that has been a weeping sore since the early 1980s - what should the bypass of Wellington's downtown be?

The problem harks back to three major strategic infrastructure decisions from many years ago:

• Rongotai as the site for Wellington's primary airport (replacing Paraparaumu and Evan's Bay (the latter for flying boats!)
• The Foothills Motorway (in favour of the waterfront motorway proposal). 
• Location of Wellington's primary airport at Rongotai.

What this all means is that two key regional facilities are separated from the Hutt Valley, Porirua and Kapiti, as well as Wellington's western and northern suburbs, by central Wellington.  The key traffic problem is that the motorway was never finished to a second Mt Victoria Tunnel, as was originally intended.

None of this was a problem for the first few years after the motorway was built, because until 1984 it was not connected to State Highway 1, but after the Ngauranga Interchange was opened (relieving the Hutt Road south of Ngauranga and the waterfront quays), it was clear that having through traffic snake through Te Aro was unsatisfactory. 

There are two main arterial routes from the north through Wellington, both of which funnel traffic from the motorway (SH1 and SH2), the Hutt Road (from the northern suburbs of Ngaio and Khandallah mostly), the western suburbs via secondary arterials such as Glenmore Street and Aro Street, towards the south and the west.  

These images below depict the volumes of traffic in the AM and PM peak respectively, in 2013 along the main arterial routes, highlighting the urban motorway and SH1 (Vivian Street eastbound and Karo Drive westbound), and along the waterfront route along the Quays, Wakefield and Cable Streets, then Kent and Cambridge Terraces.

AM peak traffic Wellington 2013

 

PM peak traffic Wellington 2013

Between 25 and 30% of traffic exiting and entering the motorway at the Terrace Tunnel is coming from or heading to the Mt Victoria Tunnel, so is using the streets in Te Aro to bypass the downtown area. Another 5-25% of traffic at peak times each is heading onto or coming from Taranaki Street and Adelaide Road south of SH1, indicating that the vast majority of Terrace Tunnel traffic is travelling to and from the southern and eastern suburbs, not Te Aro and the CBD. 

For Mt Victoria Tunnel 30-40% of traffic at peak times is heading to or from the Terrace Tunnel, noting its importance for providing access to the wider region to and from the airport. 

Of course that route as a whole has multiple bottlenecks, particularly southbound with the Terrace Tunnel having only one-lane.  Both the Terrace Tunnel and Mt Victoria Tunnel as bottlenecks, result in diversion onto other routes, notably the waterfront as seen below.

Southbound, around 25-30% of traffic on Aotea Quay at peak times is actually heading for either Mt Victoria Tunnel, Adelaide Road or Oriental Parade, it seems likely another few percent is heading towards Taranaki Street as well. Northbound the figure is 15-20% at peak, likely because there are fewer delays westbound through Te Aro as Karo Drive and the Terrace Tunnel both have one more lane (and two fewer traffic signal delays) than the eastbound direction along Vivian Street.  The 5-15% of traffic heading for Oriental Parade will be in part for leisure and accessing Roseneath, but is also likely to include substantial traffic going "around the bays" to the airport, notwithstanding that it is around a third longer in distance than using Mt Victoria Tunnel. 

In other words, the waterfront route is a backup, and carries between 15 and 30% more traffic than it should because there isn't a reliable bypass to the city. This doesn't just add to congestion along that route, but adds to the cost and hindrance of bringing the city more closely to its harbour. If that proportion of traffic were able to be removed, then there would be scope to convert a lane each way for use for bus rapid transit, but to do so with current traffic volumes would exacerbate congestion elsewhere and add to delays, in the absence of time-of-use road pricing (more on that later).

Weekday traffic between Aotea Quay, Mt Victoria Tunnel, Adelaide Road and Oriental Parade

Defining the problem

Although it would be technically feasible to implement time-of-use road pricing on central Wellington at peak times to reduce traffic entering the central city, it is unlikely to be politically feasible to apply this to through traffic not terminating there, at least as a first step.  While motorists are likely to accept road pricing at peak times entering central Wellington, where there are bus and rail options from most parts of the region, such options are not feasible for traffic heading from the Hutt/Porirua towards the airport and hospital, so an inner city cordon, for example, would need to exclude at the very least, SH1 across Te Aro and roads to its south, and possibly also through traffic along the waterfront. 

That leaves the inadequacies of the current SH1 corridor very clear:

• Terrace Tunnel bottleneck. One south-bound lane, with it being not technically feasible or safe to implement tidal flow operations (this has been previously investigated).
• Vivian Street bottleneck. Besides being only two-lanes eastbound, the presence of SH1 along Vivian Street adds to the blight in that part of Te Aro, as a major highway route severs Willis, Victoria, Cuba, Taranaki and Tory Streets. The congestion and community severance are not going to be resolved by doing nothing. Karo Drive adds to this westbound, but east of Taranaki Street the Arras Tunnel provides relief.
• Basin Reserve: With major flows from the south to the east and from the east to the west, as well as smaller north-south flows, this roundabout's asymmetric traffic flows have been a bottleneck for decades.
• Mt Victoria Tunnel and approaches from the east: Peak and inter peak, Mt Victoria Tunnel is a bottleneck that backs up to Vivian Street at certain times from the west and north, and backs up along Cobham Drive from the airport eastbound.

The attractiveness of the long tunnel was that it would provide a means for traffic from the east, particularly the airport and Miramar Peninsula, but also Kilbirnie, Lyall Bay and part of Newtown, to access the motorway easily. However, its main limitation (notwithstanding the likely cost) is that it would not provide a solution for through traffic to the southern suburbs, like most of Newtown and Island Bay and notably the hospital. The existing road layouts in Te Aro would have to continue, by and large, to service those areas, unless an expensive interchange were built somewhere in Newtown, which itself would add enormously to the cost of the project.

The main benefit of the long tunnel was that it could be built with minimal disruption, but that is about it...

How Wellington got here.

It was not always going to be that way. The De Leuw Cather report that saw the Foothills Motorway approved (which became the Wellington Urban Motorway) proposed the motorway be completed to a second Mt Victoria Tunnel, to the north of the current tunnel.  By and large the motorway that was built to the Hawkestone Street/Tinakori Road off/on ramps resembled the original plan (albeit the flyovers to and from Ngaio Gorge were never built, but the stubs for them remain).  However, budget cuts in the early 1970s saw only half of the motorway built south of there, which is why it suddenly goes from three lanes each way to one plus two.  Only one Terrace Tunnel was built, where there was meant to be two, and it terminated at Ghuznee Street southbound and Vivian Street.

The original urban motorway would have seen a second Mt. Victoria Tunnel, but would have decimated the Basin Reserve.  Te Aro would truly have had a brutalist motorway severing the area, and the cricket ground would have been gone, but that concept was gone within a decade.

Original De Leuw Cather Wellington urban motorway Basin Reserve interchange concept.

In 1972, the Wellington City District Scheme had this option as the revised plan which from the bottom of the image would have seen two-lane each way (extended from two Terrace Tunnels, with a lane gained/lost at the Vivian Street/Ghuznee St one way pair) under Willis and Victoria Streets (Victoria Street had not been extended then), under Cuba Street as well with on and off-ramps at Taranaki Street and Tory/Tasman Street. Curiously the Basin Reserve would have remained two way along Dufferin St , with a second Mt Victoria Tunnel, and the motorway skirting the Basin Reserve.

1972 Wellington Urban Motorway Te Aro extension with duplicate Terrace and Mt Victoria Tunnels

By the 1980s this had been much more simplified, but with a bridge over Taranaki Street that was seen as too intrusive

1980 Wellington Urban Motorway Arterial Extension concept with elevation over Taranaki St

Although it was placed under Taranaki Street that didn't exactly help public acceptability with this sort of rendering.

1991 Wellington Urban Motorway trenched (not covered) under Taranaki St

So, the entire section from Willis Street to the Basin Reserve was to be put into a cut and cover tunnel, called Tunnellink. Going east there would be one on-ramp at Victoria Street and another from Kent Terrace to the Mt Victoria Tunnel, with an off-ramp to Cambridge and Kent Terrace for southbound traffic to Newtown. Going west, one onramp from the Basin Reserve and an offramp at Victoria Street were all, as the road was primarily designed for through traffic.

1994 Tunnellink concept - cut and cover tunnel from Vivian Street to Sussex Street with two-way Basin Bridge to single Mt Victoria Tunnel

This did not proceed, as the funding constraints in the 1990s saw the newly named "Wellington Inner City Bypass" built in 2007, essentially widening and extending Arthur Street west to Willis Street, so that the northbound on-ramp could be relocated a block south, and Vivian Street became the southbound off-ramp as a one-way pair. The Arras Tunnel, built solely to enable the War Memorial Park, provided a wider route with the removal of Buckle Street and the Tory Street westbound intersection.

The Inner City Bypass was meant to be a medium-term, second stage of a three stage programme to relieve congestion between the motorway and Mt. Victoria Tunnel (Tunnellink meaning to the third). The first stage was a simple one-way system on existing streets, and the third-stage was meant to be a cut and cover tunnel across Te Aro to the Basin Reserve.  The second stage (the status quo) was expected to be adequate for ten years at the most, noting that from the early 1990s Wellington has had a strategy of constraining growth in traffic towards the central city.

Constraining traffic

In the early 1990s, what was then called the "arterial extension" to the Wellington Urban Motorway required Ministerial approval, and the then Minister, Rob Storey, a National MP from a rural electorate, required that action be taken to constrain growth of traffic because of concerns of induced demand.  The business case for the project estimated an annual traffic growth factor of 2%. The response was the introduction of the Coupon Parking Scheme, on unmetered car parks close to central Wellington, which Wellington City Council estimated would constrain traffic growth to 0.5% (along with introduction of limits to car parks for all new developments and abolition of the minimum car park requirement).  At the time, Transit New Zealand as the land transport funding agency and state highway manager, recognised this, but the impact on the business case for the project was to knock it back.  In effect, demand management delayed the need for the project.

Tunnellink and the three stages

In the 1990s, there was some controversy over the proposed motorway extension, although there was always a clear majority of city councillors in favour of the project.  Yet the key issue was funding at a time when Government only allowed enough funding to be made available for land transport projects with a benefit/cost ratio of at least 5:1 (lowered to 4:1 in 1998 onwards), for a project that, at the time, had a BCR of 2.6 (although an optimistic scenario saw it raised to 3.2). 

With the change in government in 1999, there was quite a fight to be had as the Clark Government relied on the Greens for confidence and supply, and also agreed to consult the Greens on transport policy. The Greens demanded that the Inner City Bypass be stopped. At the time, the legislation did not allow the Minister of Transport to direct the then, independent funding agency, Transfund New Zealand, to approve or block any projects. In truth, the Clark Government was in favour of the project, and it received approval, although opponents used every legal avenue to delay it, so that it took over a decade from the initial approval of the concept for it to finally get built.

Of course by that time Transit New Zealand, as the state highway manager, had abandoned hope that any additional project would ever get approved across Te Aro, so had shelved the Tunnellink idea and refocused on what would be needed next - and that was the Basin Reserve.

That became the next problem,

Basin Reserve

The Inner City Bypass made a useful difference to traffic flow across the city. Particularly westbound, having shifted traffic a couple of blocks south, and with three lanes, queues towards the Basin Reserve were eased, and delays reduced.  Eastbound, Vivian Street flowed more efficiently as a single one way route than the dog-legged route via Ghuznee Street, but the Terrace Tunnel remained a major bottleneck. By the 2010s, although there were now budget surpluses and governments of both hues committing to more funding for roads, a second Terrace Tunnel still seemed some years away.  The focus shifted to the Basin Reserve, as it would (and still does) back traffic up in both directions.  As the Basin is effectively a major intersection with heavy traffic flows east-west and lesser flows north-south, the case for some grade-separation of those flows has always existed.  

From there came the Ngauranga to Airport study and the Basin Bridge project. The Basin Bridge started as a project to take westbound traffic out of circulating around the Basin Reserve from Mt Victoria Tunnel, and towards Buckle Street. Indeed, the 2012 depiction essentially envisaged it feeding the Arras Tunnel.  The effect of this would have been to significantly change the traffic light phasing around the Basin Reserve to ease flows to and from Adelaide Road, but the backlash, even though the bridge was essentially clear of the Basin (and designed with a 50 km/h limit) was considerable from Mt Victoria Residents. This was also fuelled by many who opposed the Inner City Bypass, seeing this as just another "motorway" to oppose.

2012 one-way Basin Bridge concept to existing Mt Victoria Tunnel only

Perhaps the biggest limitation of this proposal was that it offered little scope to support relocating State Highway 1 south away from Vivian Street to enable two-way traffic.  It solved "half" of the problem and offered no real vision for what a decent bypass of Wellington should look like and how it should function. Hence it would also not satisfactorily support a second Mt Victoria Tunnel, as that tunnel would be constrained by the capacity of Kent Terrace and Vivian Street.  The Basin Bridge would have been short-sighted.

2012 one-way Basin Bridge concept to existing Mt Victoria Tunnel only from west

The history behind the Basin Bridge and the Environment Court rejecting it, largely due to objections of local residents is the embryo of Let's Get Wellington Moving.  As that emerged, in part to try to find a way forward, another concept came up, enabling two-way traffic to bypass the Basin.

This concept enabled a future of a trenched bypass through Te Aro feeding onto a bridge towards a second Mt Victoria Tunnel, with an at grade route close to the Basin, with Sussex Street widened and built over the new road towards a widened Cambridge Terrace.  This would mean two way traffic from Adelaide Road circulating over Sussex Street abandoning the Basin roundabout, although it was not perfect. It offered no access from Mt Victoria Tunnel to Cambridge Terrace with only a rudimentary ramp for access to the schools at the Basin, although that could easily be fixed by making the Dufferin Street ramp larger and retaining "around the Basin access" to Sussex St.  You can see the dotted line for an eastbound cut and cover tunnel  from the motorway with a ramp to Cambridge Tce and continuing to Mt Victoria Tunnel, but again there remains a flaw.  Access from THAT tunnel to Adelaide Road doesn't exist. Without direct access from the bypass to Adelaide Road, the route misses a key connection, so this is not a solution.

2018 alternative Basin Bridge concept two way

So a whole host of options emerged, as Let's Get Wellington Moving saw its objectives amended, largely reducing the value placed on reducing traffic congestion, and more focused on modal shift and encouraging active modes and public transport.  The options illustrated below demonstrated this as none would support moving SH1 off of Vivian Street, and enable a second Mt Victoria Tunnel for more general traffic capacity. That of course was not the objective in the latter years of Let's Get Wellington Moving. 

Let's Get Wellington Moving Basin options

For what it's worth, I think the right solution is to enable what is depicted below from an early concept in LGWM to shift SH1 to the Karo Drive corridor in a trench to connect to a future proofed Basin Reserve grade separation.  

Long term option for cut and cover bypass across Te Aro

and no, tunnelling too deep on the north side of the Basin Reserve isn't an option. There is a stream already running under there and the geology is soft (it isn't called "Basin" for nothing, as the land was uplifted in the 1878 earthquake from what was then considered to be a possible location for docks). So the Basin Reserve either needs to be bypassed far away (Long Tunnel), be bridged over somehow (or with a shallow trench and lower bridge), or be left alone.  However, moving SH1 into a trench to its west, is not going to be worth doing until a solution is found for the Basin, and there is likely to need to be some engineering creativity involved in achieving that. Otherwise there would be a congested at grade intersection.

Maybe this sort of elegance is needed to make it acceptable? Bearing in mind it would need to be a two way bridge, be lower at the western end (as Sussex Street on a bridge over the end of it would make sense), and then assuming it isn't going to be light rail, there is space for bus rapid transit to go around the Basin in both directions to and from Adelaide Road (and the duplicated Mt Victoria Tunnel).

2008 indicative Basin Bridge concept from north

Mt Victoria Tunnel

East of the Basin Reserve the options become clearer.  In 2011, a range of options were considered, including tunnels either side of the existing one, and longer ones from Wellington Road at various points, and new bus tunnels.

Mt Victoria Tunnel duplication options considered 2016

The conclusion was clear:

and it was depicted as follows from Wellington Road to the Basin Reserve

Preferred Mt Victoria Tunnel, Ruahine St and Wellington Rd duplication concept 2017 before cancellation under LGWM

It seems likely that this is the most cost efficient option to proceed, as long as the decision on the Basin Reserve can address issues around visual impact and noise.

Funding and pricing

Tolls aren't likely to be much use for this option, not least because the key point of doing this project is to take traffic off of alternative routes.  Tolling the existing and a new Mt Victoria Tunnel (unlike the long tunnel) would likely see Newtown and the Evans Bay/Oriental Parade routes get filled. Unlike tolling more rural segments of highway, the costs of less traffic through the tolled routes are more likely to be born in higher externalities in built up areas.  

However, time-of-use pricing designed to shift demand during peak times entering the central city (to the east and north of SH1) could certainly be used, to reduce congestion entering Wellington city and free up capacity for traffic bypassing the city. It could also be used to help fund a bypass, as this is what happened in Oslo with implementation of its toll rings through Oslo Package 1.

The amenity value in removing perhaps half of the traffic from Vivian Street (and on what remains of a surface route were Karo Drive now is), and removing around a quarter of traffic off of the waterfront route would be considerable. As such, this whole project needs to be seen as much as an urban amenity and redevelopment project, as a road project. It would enable longer crossing times across the Quays and the Quays to be more oriented towards local access. It would help increase connections by bus, foot and cycling across this route to and from the city, and help to revitalise Te Aro with a significant reduction in traffic. Oslo saw this happen when it built a bypass tunnel under its central city.

Bear in mind the long tunnel would have speeded up travel from the airport to the motorway, but would not have had the same scale of impact on local amenity because it would not have served as much traffic as the more localised option across Te Aro.

What matters is cost, although clearly this all could be done in stages, unlike the long tunnel. This enables costs to be better managed, and for benefits to be realised much more quickly.  

 Mt Victoria Tunnel and the widening to the east could be done in one stage. Other stages include grade-separating the Basin Reserve, trenching across Te Aro and finally, a second Terrace Tunnel (which would be relatively easy to build, as all of the land is reserved and part of the approaches were built before being cancelled). 

It could all be done supported by time-of-use road pricing to manage demand and help support the funding of it, and the sequencing is important.  A second Mt Victoria Tunnel would make some difference, but without a Basin Reserve upgrade would see traffic backing up through the westbound tunnel from the Basin.  A second Terrace Tunnel without a trenched route across Te Aro would not achieve much beyond some shortening of queues.  The trenched route would be highly disruptive while being built, but without a Basin upgrade would see big queues eastbound. 

One thing seems likely, the long tunnel looks like it is a non-starter, but a second Mt Victoria Tunnel looks like it could proceed, more quickly than anything else at this stage. Wellington could do worse than proceed with those plans, which include an upgraded cycling and pedestrian route in the new tunnel, as a first step, while more time and money is spent working out how to take that traffic between there and the Terrace Tunnel.

Conclusion

The rational economic answer would be to introduce time-of-use road pricing for entering central Wellington and then seeing what the traffic looks like, but it seems highly unlikely that this could obtain the political and public support necessary to proceed. So a second best is to implement road pricing to help   pay for a project that has, as a fair proportion of its benefits, public amenity and development of the city by removing through traffic. To do that requires extra capacity at the Terrace Tunnel, Mt Victoria Tunnel and across Te Aro between them.

It would be a significant step forward not just for general traffic for both people and freight between the airport and the region, but also the hospital and eastern and southern suburbs, and the region, but for public transport, cycling and walking to and from the city, and for the development of Wellington as a destination.  The latter comes from it being not just easier to get to and from the airport, but by removing the blight of through traffic in Te Aro and making the waterfront a destination that is not severed by a primary arterial highway.

Japan planning introduce time and location based pricing on expressways nationwide

Japan's nationwide expressway network is run by a series of private businesses. In 1956, the Japan Highway Public Corporation was formed to build and operate a national highway network, using tolls and accessing private financing. At the time, only 23% of Japan's national highway network was sealed including only two-thirds of the Tokyo-Osaka highway.  Tolling was extensively used, and for sections of highway that did not gain private finance, the government guaranteed the loans. Tolling revenue was pooled to cross-subsidise parts of the network that did not generate enough toll revenue to pay for construction (details on the history of highway in Japan is available here (PDF). 

In 2005, the Japan Highway Public Corporation was split and privatised into multiple companies, including the Japan Expressway Debt Repayment Agency (to use toll revenue to repay the considerable debt that remained for the development of the network) and six regional expressway companies. They are:

• East Nippon Expressway Company Limited;
• Central Nippon Expressway Company Limited;
• West Nippon Expressway Company Limited;
• Metropolitan Expressway Public Corporation (Tokyo);
• Hashin Expressway Public Corporation (Osaka-Kobe-Kyoto); and
• Honshu-Shikoku Bridge Authority.

Tolls were authorised to be collected until 2050, recently extended to 2065.  The privatisation was driven by several concerns, in particular:

• As Japan's network had essentially been completed, there was concern about public ownership enabling politicians to authorise new construction that favoured the construction industry, even if projects were not viable. 
• The pooling of toll revenue nationwide was seen to enable this cross-subsidisation where there was no need for new infrastructure. Residents objected to paying higher tolls in their area for projects that were far away from them and of dubious economic value.
• Interest in improving the efficiency of administration and encourage innovation in operations of the network.
• Interest in enabling comparisons between the performance of companies so encourage more productivity and lift standards across the sector.
• Concern about the levels of debt government was taking on for the expressway company, and privatisation was seen as a way to put discipline on costs, debt and the scale of capital spending.

Map of Japan's expressways and major highways

The national expressway network is 9050km long. Tolls in Japan are generally set to reflect distance travelled between interchanges, and vary by vehicle type. Most toll roads still have a mix of electronic and manual toll lanes.

So the announcement in the Japan Times in the past week that the Ministry of Land, Infrastructure, Transport and Tourism will be introducing the ability for expressway companies to introduce time-of-day varying tolls, based on location, to manage congestion, is a significant step for the history of expressways in Japan.  It was trialled during the 2021 Tokyo Olympics with a higher daytime charge, and discounts after midnight, but the idea is that time periods and variations in toll fees will depend upon the specific route and the conditions on it. This is NOT dynamic tolls, but rather targeted congestion pricing to enable more free flowing traffic and reduce pollution.

Also announced was the enabling of commuter passes for high frequency users of toll roads in particular areas, to encourage greater use of expressways to remove traffic from untolled parallel local roads.

Crucial next steps for Auckland congestion pricing

I usually write about road pricing in my Road Pricing Blog, but I have placed this article here primarily because it is a wider transport policy issue around governance, and isn't just about road pricing in Auckland.

The announcement of the removal of the Auckland Regional Fuel Tax (ARFT) by the (NZ) Minister of Transport, Hon. Simeon Brown is implementation of an election promise, and is not a surprise.  The big question is what comes next, and talk of congestion pricing, given it is also part of the National/ACT party coalition agreement, is clearly something that needs to be progressed this year.  However, there are some big questions that need to be addressed.

Don’t miss the regional fuel tax

The ARFT was introduced, with the intention that it operate for a ten-year period in 2018 to enable Auckland Council to fund transport projects in the region. It had been criticised at the time for its distributional impacts, with a study by Sapere Research Group in 2018 (PDF) regarding both the ARFT and national increases in Fuel Excise Duty (FED) concluding that they will:

• give rise to significant variation across households in terms of impacts;

• be regressive in nature- (i.e. disproportionately impact on lower-income households);

• likely impact inequitably on Māori households, especially in the south of Auckland;

• produce even more inequitable impacts when fuel efficiency of vehicles is considered.

In short, the ARFT unfairly burdens those least able to afford it, not least because they are least likely to be able to afford newer vehicles (such as EVs or hybrid vehicles) that are more fuel efficient, less likely to be able to afford to live near the public transport that connects them to their employment and more likely to work shifts or at times when alternatives to driving are less feasible or attractive. 

Removal of the ARFT is notable as it is rare, anywhere in the world, to see a Government willing a scrap a tax on road use. It raises the issue as to how to pay for significant improvements to Auckland’s transport network beyond the capacity of the National Land Transport Fund (NLTF) which has its own issues of funding capacity to meet government objectives (highlighted in the NZTA Briefing to the Incoming Minister) and the political willingness of Auckland Council to increase rates.

When the previous Government held a Select Committee Inquiry into Congestion Pricing, many suggested that congestion pricing in Auckland could be used to replace the ARFT.  

Congestion pricing can’t replace the ARFT revenue in its entirety for some time

Congestion pricing clearly could generate some net revenues for transport in Auckland, but it is important to emphasise that while it can do that, for it to be effective in managing congestion, the net revenues need to be seen as secondary and not the primary objective. Designing congestion pricing to reduce congestion will mean it is designed to deliver net benefits to transport sector users, regardless of revenues generated and how they are spent.  This is how it has worked in Singapore and Stockholm, but it is not what has happened in London (certainly in the past decade or so) nor Gothenburg.

If it is to be assumed that the Government wants congestion pricing to be about reducing congestion and improving trip reliability in Auckland, it needs to be designed and implemented with that as the primary focus. If Auckland Council and Auckland Transport regard congestion pricing primarily as a tool to make up the revenue lost from not having the ARFT, it is at best going to be sub-optimal in addressing congestion at the times and locations where it is introduced, and at worst will deliver insufficient travel time savings for motorists resulting in a backlash against the concept, killing off congestion pricing in Auckland (and NZ) for many years to come.

The Congestion Question report, albeit now slightly dated (as the analysis was undertaken before the pandemic), indicated the likely net revenues from the short-listed congestion pricing options.  The most promising option was the “Combination” which would see introduction of a city-centre cordon scheme, along with pricing the strategic road network (motorways and parallel main arterials).  

The city-centre cordon was estimated to generate around $21m per annum in net revenues, which is far short of the around $150m per annum generated from the ARFT, so simply operating that scheme would do little for revenue, but of course it wasn’t designed to generate revenue, but to support reducing congestion entering, exiting and circulating within the city-centre.

Table 1 - The Congestion Question: Main Findings July 2020

The strategic corridor option was estimated to generate around $205m per annum, and assuming it would be introduced in combination with the city-centre cordon, there would be a combined net revenues of $223m per annum (the effect of one scheme would be to suppress some demand that would otherwise occur if only the other scheme was in operation).

The strategic corridor option cannot be implemented in full in one go.  The suggestion of pricing SH16 between Lincoln Rd and Te Atatu Rd, and SH1 between Greenlane and Ellerslie/Panmure is a feasible first step towards doing so, so it might be expected that the strategic corridor option is phased in over several years. Although these figures come from 2019, it is reasonable to take a conservative approach, as travel patterns in Auckland have changed somewhat since the pandemic. 

In short it means that the ARFT revenue cannot be quickly replaced through congestion pricing, assuming that congestion pricing is implemented with the objective to improve network performance as a primary goal, rather than revenue raising.

Key decisions

What matters now is how congestion pricing in Auckland is advanced.  There are some big decisions to be made on governance, rate setting and use of net revenues. These will have a significant impact on the likely success of pricing and importantly, public acceptability and durability of pricing as a policy. A key element is how much decision-making will be up to road controlling authorities and how much will be up to central government either politically or administratively, from which there are some big governance issues:

1. How will legislation enable congestion pricing? 

Will it be general empowerment, case-by-case scheme approval or something else?

Existing legislation around tolling essentially requires Cabinet approval on a case-by-case basis following recommendation by the Minister. This is suitable for toll roads, which are specific sections of road with rates set to contribute towards infrastructure cost, and so are unlikely to need to be varied much, except to take into account inflation. For congestion pricing, there is a need to have some variability and flexibility in implemention.  Detailed Orders-in-Council specifying precisely where pricing points are, their hours of operation and price levels are unlikely to be suitable for effective congestion pricing.

An inner-city cordon may need some minor variations in the location of pricing points, and both the hours of operation and rates set will need to vary to manage demand effectively (and avoid pricing traffic during times of low demand).  Pricing, after all, works best in all sectors when it is flexible. History tells us that when prices are heavily regulated for goods and services, especially when pricing is intended to send signals around demand and supply, that it distorts usage, whether by underpricing demand (which is the status quo) or in overpricing (and reducing economic activity in an area).

However, caution should be exercised in simply providing general empowerment for road controlling authorities to implement pricing under certain conditions. International experience indicates that congestion pricing is very difficult to implement, primarily because it is easy to quickly undermine public confidence in the concept and the details of any scheme, resulting in overwhelming resistance and pressure to cancel. This has been seen in several occasions, notably Edinburgh, Manchester, the Netherlands (multiple occasions), Helsinki, Copenhagen and London (with removal of the Western Extension of the congestion charge). 

This is not to say that Auckland Transport, Wellington City Council, Christchurch City Council and Tauranga City Council couldn’t implement congestion pricing, or indeed NZTA, but it’s also worth remembering that when congestion pricing fails, it tends to kill the idea politically for many years. Edinburgh rejected congestion charging in 2005 after five years of development, and is only now again looking at the concept. Manchester rejected it in 2008, after considerable design work over the previous two years, and has no interest in reconsidering it.  Although many trump London as a success, the London Congestion Charge was expanded once and that expansion was rolled back three years later in 2010, and even today there is only talk of more congestion charging for the city (don’t mistake the Ultra Low Emission Zone, for which 97% of vehicles are exempt from paying, for congestion charging).  Local authorities are subject to the political whims of elections every three years, and given the timeframes needed to implement pricing, there is considerable risk that pricing could become a serious political issue if not managed effectively.  Bear in mind no NZ local authority has implemented road pricing before, and only Tauranga implemented tolls (albeit some years ago).  There are a long list of implementation risks that if not managed well could undermine congestion pricing not just in the first city that tries it, but in all others. 

Note also that despite the UK having had legislation generally empowering local authorities to implement congestion charging, only one outside London has actually done so: Durham. The Durham scheme was introduced before London’s and is in effect a small charge on driving in the tiny historic centre of the city.  Empowerment of local government to introduce congestion charging in the UK has largely failed, as none have had the political courage to do it, and of those that tried, they have proven incapable of developing and communicating a pricing scheme that would obtain adequate public acceptability. Indeed, the repeated failures of Edinburgh and Manchester to implement pricing has almost inoculated most councils to implementing it. Most recently it is Cambridge that has attempted to introduce pricing and, in no small part to the design it chose, it has repeated the examples of the other cities in generating enormous opposition to congestion charging for that city. 

Although it is possible to conceive of congestion pricing concepts for Auckland and other cities that do not charge the state highways, excluding them from any scheme is likely to be sub-optimal in managing traffic in those cities. Both Auckland and Wellington could have a “first-stage” pricing scheme implemented in the form of inner city cordons without pricing of the state highways, but the “second-stage” would inevitably need to include them. It is also critical that any road pricing scheme not negatively impact the state highways. In all cities congestion pricing is likely to be more effectively managed if undertaken jointly between NZTA as the State Highway manager and the relevant territorial authority managing the local network. 

For example, in Auckland, this means ensuring SH16 to the Ports of Auckland functions well. Of course if the first scheme in Auckland is actually to price segments of SH16 and SH1 as the Mayor of Auckland proposed late last year, then it is inevitable that it be undertaken jointly, as Auckland Transport does not (and should not) have the powers to price roads it does not manage.

The first congestion pricing scheme in NZ is likely to be in Auckland, so it is critical that it is designed to be effective without being punitive. It needs to have sufficient flexibility as it is introduced to make small adaptations quickly if motorists respond in unexpected ways which cause issues (such as congestion on alternative routes. There should be measurement of performance, in impacting travel times, trip reliability, net revenue collection and compliance, as well as wider economic, social and environmental impacts. London undertook such analysis for its first five years of operation. This is critical in building data on how to expand and adapt pricing and to mitigate any negative impacts.

It is far too risky to simply leave all of this up to a single Council without the necessary experience, particularly if its incentives are more around revenue collection than traffic management.

Auckland congestion pricing should be a joint NZTA/AT project. The Auckland Traffic Operations Centre is, for logical reasons, and this should follow on from that. It doesn’t matter if the first scheme in Auckland is a city-centre cordon or a strategic corridor, as the second scheme will be the other. NZTA is already expanding and renewing its tolling back-office system, which could provide the platform to deliver congestion pricing, although it will need to have a significant uplift in capacity for customer service.  

2. How is rate setting to be undertaken, and reviewed and updated over time? Is there going to be regulatory oversight to it?

To be effective in managing congestion, congestion pricing in NZ should follow the Singapore approach of setting prices to ensure a minimum throughput of average speed on priced roads, with the ability to review and revise (up or downwards) such prices on a regular basis. Singapore does it quarterly, with no regard at all as to whether it impacts net revenues positively or negatively. A similar approach would ensure traffic impacts are optimised, and public acceptability is as well. Having shoulder as well as peak rates, so that the price is not simply $0 at 0630 then $5 at 0700 will be important to avoid bunching of traffic before pricing kicks in, and to spread demand. This is not practicable if rate setting is done by Cabinet, so this needs to be devolved to a governance entity which is empowered to vary rates (with perhaps a cap applied), with some regulatory oversight over time.  In due course, multiple pricing systems in different cities ought to have an economic regulator ensuring that road controlling authorities are looking after the interests of consumers both in the prices they set and how they spend the money. 

3. Who decides what net revenues are used for?

In the UK, legislation determines that net revenues from congestion charging are used on local transport spending. In Sweden, central government decides in partnership with local government how net revenues will be used, and in Stockholm it was initially focused on funding a major highway project, and has since included funding for public transport projects. Oslo similarly has used revenue from its toll ring (which resembles a congestion pricing scheme although it is primarily designed for revenue) for various blends of road and public transport projects, but again, this was determined by central government in collaboration with local government.

If revenue is being raised from state highways, it may seem appropriate for NZTA to decide, following guidance from the Government Policy Statement. If raised from local roads, it may seem appropriate for the territorial authority to decide, albeit it would seem difficult to get public acceptability for pricing if such revenue were used for non-transport purposes or on projects or activities unrelated to those subject to pricing. There are clearly a wide range of options on how net revenues might be used such as supporting spending on capital projects or offsetting rates revenue for spending on local capital projects or local road maintenance or public transport subsidies. A more innovative solution would be to pay a dividend to Auckland householders from congestion pricing, but this seems unlikely given the pressure from the Auckland Council and Mayor to find sources of revenue to spend on transport projects (and the Government is also interested in funding some of those projects as well).

What comes first though is who makes that decision. Compared to many other issues related to road pricing, relatively little thought has been given to this.

One chance to get it right

There are other issues of a secondary nature. Clearly the question as to what the first road pricing scheme in Auckland should look like is critical, but before that there are decisions on governance that are needed that haven’t been adequately canvassed, because they are difficult and controversial.  However they are needed soon, because until they are determined, progress cannot be made on detailed design, procurement and implementation of any congestion pricing in Auckland or elsewhere. It is possible for Auckland to be ground-breaking for NZ and indeed for many other cities with high-levels of private car use relative to other modes. However it is also possible for this to go badly wrong. For the last twenty of so years politicians have erred on fearing the latter, now NZ is on the cusp of making the last crucial step, they should address the remaining key issues and advance, bringing Aucklanders, and especially Auckland motorists with them. This Government has shown, by removing the ARFT, that it doesn’t see motorists as the enemy, it has the chance to show this further by implementing pricing in a way that, overall, delivers net benefits to them, and as a result net benefits to those who ride buses, bicycles and walk, and the businesses and communities they work and live in.

Many write about the successes in congestion pricing, few write or even know about the failures. Plenty of cities have advanced studies and developed proposals, and had the legal mandate to implement them, but pulled the ideas because of public backlash, largely due to key elements having either not been decided or having been designed in a way that doesn't deliver net benefits to those who pay. Auckland should not be on the list of failures.