Roads and PPPs

Had a chat with RNZ's Wallace Chapman and The Panel this afternoon on the ACT Party's proposal for reform to how roads get built and maintained.

I usually put a few notes together for myself ahead of these things, mainly to sort out my own thinking rather than to be able to convey every nuance for a light afternoon radio talk. 

Those notes are here, in part so I can find them again next time I need to think about this stuff. 

Biggest-picture: there is need for fundamental restructuring to get to a system that’s responsive to user-demand and is consequently able to deliver projects where road users are willing to pay the cost of the service and to bat back projects that aren’t cost-effective. 

First, summary of ACT’s proposal:

1. 30-year plans for major infrastructure set by local, central, and infrastructure commission. Sets out expected timelines for NLTF projects and which could be fast-tracked if PPP;
2. Public consultation on draft plan;
3. Private sector bids to deliver kit, with tolling on the road, are entertained – they’d have to beat public sector timelines/spec;
4. Then Waka Kotahi has to get consent and acquire land;
5. Public sector could reconsider a proposed route if private sector gives it a pass at toll rates that users find acceptable;
6. Could add tolls to existing roads to help cover maintenance, with a focus on congested roads to help spread traffic to other routes;
7. Not in the ACT proposal here but mentioned by Simon Court on LinkedIn: shifting ownership of the state highway system to a new SOE, Highways NZ, which would be expected to be operationally self-funding out of user fees and deliver a return on capital to the government. 

Big picture things this gets right:

1. Funding for land transport and its management are currently a mess that contribute to poor quality roads. NLTF increasingly a bucket for sundry transport funding and spending rather than dedicated mechanism for user-pays
2. Getting price signals into the mix would be really useful. Far too much transport debate ignores it. Is a second harbour crossing a good idea and who should be able to use it (car, bus, bike, pedestrian) ought to depend on whether the kit can pay for itself through user fees collected over the decades of its life, like the Auckland Harbour Bridge, rather than who can make the most convincing case to a Minister. Benefit-cost ratios can be a useful proxy for this, helping to figure out user priorities. If traffic volumes are high enough to justify the cost of collecting tolls, that’s even better. 
3. Toll roads get us closer to user-pays, and it is better to be closer to user-pays. Remember though that RUC for cars is 7.6c a km, and petrol excise averages that. Neither vary by time or location – though RUC could get there for heavy commercial on telemetrics. 
4. Reducing admin costs of running a tolling system would also be very helpful. But it’s a big hurdle. RUC has collection costs of about 3%; tolls have been closer to 30% [at least according to the experts I've talked with]. You need high traffic volumes on a road to justify the kit for monitoring and billing. Those costs could come down as tech improves. But it’s still a high hurdle unless there is a lot of traffic on the road. 
5. Setting interest in running a PPP as one market test of a roading proposal is one way of knocking back bad projects and making sure very valuable ones get built. But there will be others where the cost of tolling, relative to RUC, could prevent good but not superb projects from going ahead.

Potential fishhooks:

1. Acquiring the land for routes only after extensive consultation can make it a lot more expensive to run projects at all. The Infrastructure Commission has pointed to some of these problems in its own work on corridor designation. Normal drill has been that land is only designated and acquired when the project is ready to go, which means that the value of the project gets bid into the price of that land, which makes everything more expensive. Early corridor designation can help, and 30-year horizons could help with that as well. Option contracting on potential routes, ahead of designations, could help. Corridor designation can be done much earlier, well in advance of any project being viable, so the option is maintained. 
2. Rather than 30-year plans, long-term corridor designation and flexibility to press ahead whenever circumstances warrant could do more good. Right now, one of Tauranga’s larger housing growth areas is being held up because nobody’s allowed to build to the density that makes sense because the roads aren’t currently up to it, but the SH29 overhaul isn’t planned until 2050. Housing demand can shift more quickly than 30-year horizons. Lots of lead time in designating corridors, and flexibility to build/upgrade as demand comes into the system, may be a better mix. 
3. The proposal conflates user charging with congestion charging, making the toll charge do both jobs. It’s better to keep the separate objectives separate, even if charges wind up being collected through the same system. A user charge or toll is set to cover the cost of the road – its building and maintenance. A congestion charge should be set to maximise traffic throughput: it should be zero when there is no traffic, and potentially high when there is a lot of traffic. The congestion charge should be designed to encourage changes in times of travel. A dynamic toll that can vary by time of day might be able to do both. But keeping the two separate keeps the incentives clearer and provides other alternatives for dealing with potential equity considerations.  
4. Consultative processes risk leading to gold-plating requirements that make routes unviable. The proposal has the consultation process in place as a way of gauging real user demand. But that could also be done by setting congestion charges on existing roads and seeing what actual willingness to pay looks like. The example I love to use is a second Mt Vic tunnel in Wellington. I have no clue whether a second tunnel makes sense. But imagine if we had a congestion charge on the existing tunnel set to make sure that the thing doesn’t get plugged. If it only took a $0.50 charge to clear congestion and nobody saw any way of building a second tunnel that could cover its costs on a $0.50 user charge, then it would be dumb to build a second tunnel. But if it took a $5 charge to clear congestion, and if that charge would be enough to cover the tunnel’s cost over time, then that could be a reasonable option – noting you’d also want congestion charging over alternative main routes – or a downtown cordon.

Potential alternatives: 

1. ACT’s proposals are a step in the right direction. But it’s high time the overall system be reconsidered. 
2. NLTF is meant to cover road building and maintenance out of payments by road users. But excise is increasingly disconnected from road use / burden imposed. Would be simple to shift away from petrol excise and put everything onto RUC. Note that excise is also increasingly inequitable: a new hybrid imposes no more burden on the road than an old Toyota Estima, but the latter pays a hell of a lot more for the same amount of road use and road burden. There’s weird status-quo bias where people freak out about equity implications of any change from status quo, but never question the biases built into the existing system. But it would also require tighter enforcement of RUC – potentially integrated with WoF. 
3. ACT is looking at toll charges as one way of alleviating congestion, with public feedback mechanisms aimed at improving social license. If people were worried about equity, you could instead run congestion charges cleanly on their own basis – and use the collected revenues to fund a congestion dividend that rebated collected fees to road users irrespective of their time of use. There is already cross-party agreement for congestion pricing in Auckland/Wellington. 
4. In 1998, the government proposed substantial transport reform under Maurice Williamson. “Better Transport, Better Roads.” A lot of what was proposed was before its time – the tech wasn’t there for low-admin-cost road user charge collection, and we’re still not 100% of the way there yet. But the system was elegant. Rates would no longer fund roads. Road use levies would be paid into Transfund, a Crown-Owned entity, responsible for recommending road use rates to Minister of Transport. Regional road companies would take over running local roads; a Crown-Owned company would run the state highways and motorways – and this part sounds like what Simon Court had suggested on LinkedIn. It was a straight user-pays system. Transfund would commission roads based on user needs; the public road companies would operate them. The link between user payments and what they get from those payments would be a lot clearer. 
5. Under the 'Better Transport, Better Roads' option, the road companies would introduce pricing as cost and feasibility were demonstrated – rather than it being a political decision for ministers.
6. Currently, a PPP can help solve two different problems: infrastructure delivery & management, and financing. If the government instead issued debt tied to specific roading projects, so that road users could pay those costs off over time through either tolls or road user charges, that would be a more direct way of solving the financing problem. And then PPPs would only be chosen if the road companies [Under something like 'Better Transport, Better Roads', or Highways NZ under ACT's extended proposal], or Waka Kotahi under the status quo, thought it would be better value. 

I'd also recommend Liberty Scott's excellent thread on all of this.