Responding to Simon Wilson of the NZ Herald on Auckland congestion pricing

First a warning. I don't use this blog as a place for debate or polemics, but on this occasion, I have decided to respond to an article which is quite damning, as an example of the sort of debate one endures around the topic of road pricing. This article is by no means the worst I've ever seen, but it is full of assertions with little to no evidence, mistakes and bold claims that appear to be more motivated by a desire to undermine the policy itself (and perhaps this is because it is from a government he himself does not support), than to critically review the merits of it.

The NZ Herald is Auckland’s (New Zealand) newspaper of record. Simon Wilson (no relation) is the NZ Herald’s senior writer on transport issues. Given the NZ Government’s recent announcement of its intention to advance congestion pricing (called time of use pricing), he has written a column on the topic, which is arguably a polemic of weak argument against it. Note it is behind a paywall. I am a subscriber, but it is not worth you paying to read his article though. 

It is possible to critique time of use pricing in Auckland on some grounds, such as whether it could be expanded at reasonable cost sufficiently to significantly address congestion, or whether net revenues should be redistributed through tax cuts rather than spending on infrastructure, or if it is better to introduce road pricing more generally, so driving outside congested time and locations is cheaper.  It is possible to argue that there should be more and better public transport to accompany road pricing, or that it could cause government to delay or cancel some new road projects, but none of that is apparent.  

I’ve spent over 20 years working on road pricing policy around the world and I have seen arguments against road pricing from a range of perspectives. Some on the right see it as an additional tax that intends to limit motorists’ freedom and increases the power of the state against them or is used to invade their privacy through "tracking". Some on the left see it as unfair that a scarce resource (road space) is allocated on the basis of price rather than queuing.  Wilson is on the left.

Simon Wilson has written many columns in the past about the importance of tackling climate change, of getting more people out of their cars and into public transport or active modes. He is a fervent believer in behaviour change in urban transport policy. Now he is making himself perhaps the highest profile campaigner against the one policy that could achieve more of what he claims to want than any other – more efficient road pricing.

He titles his article “Sorry Simeon Brown, congestion charges are not the key to freeing up the roads”. 

He’s wrong. I doubt Wilson can identify a city in the world that has freed up its roads without road pricing, he certainly doesn’t name any. Short of Pyongyang (or Covid lockdowns), I don’t know of any major city in the world without road pricing that has significantly reduced congestion, although certainly some that have it do still have severe congestion (notably London). Singapore and Gothenburg certainly have much less congestion than before either city had congestion pricing, Stockholm has much lower congestion approaching its central city and along its main bypass route (although there remains congestion elsewhere in its network). Milan still has bad congestion, although it is better with pricing.  Wilson does not indicate which cities on a scale of Auckland have freed up their roads without either road pricing or depopulation

So what else did he have to say?

He repeatedly asserts that time-of-use pricing, congestion charging etc is a “tax”, which presumably he is saying for pejorative impact to appeal to readers on the right of politics (new taxes are "bad" from a traditional rightwing perspective).  Whether or not it is a tax would be a legislative matter. I’m not sure if he thinks water meters are a “tax” or electricity meters, but applying a price, that varies by time-of-day, and is periodically reviewed as to user demand (as in Singapore), is not very much like a tax. In Sweden, congestion pricing is called the “congestion tax” for legal reasons, in Singapore the term “Electronic Road Pricing” is used to describe simply a fee for using the roads. Does he see bus fares as a tax? Does he see the existing toll roads in NZ as a tax? This is hardly a major point, but it sets the tone for this article, which is a not particularly coherent piece, sometimes opposing road pricing and in one place saying "it has a role". 

Although he admits that congestion pricing works, he then makes several claims that do not stand up to scrutiny, namely:

• They can “do a lot of harm”;

• They are “not the key to reducing congestion”.

He also says they are not popular, which is hardly a surprise, as reports from The Congestion Question (the last major study into the topic from 2016-2020) indicated that public acceptability is the greatest challenge to implementing congestion pricing. Wilson’s column is of course helping to add to this challenge. I don't think you should complain about something being unpopular by contributing to its unpopularity sans the merits.

What about "do a lot of harm”? He doesn’t elaborate, and in fact provides zero evidence of a "lot of harm" anywhere, so why say it? Why scaremonger?

He then acknowledges that for commercial traffic, congestion is a cost, whether for freight delivery or simply providing services that require getting between sites.  However, he then describes Stockholm, London and Singapore as all being cordon schemes, which isn’t quite accurate. Although he cites The Congestion Question report, he clearly did not read the report on international experience (PDF) (disclaimer: I wrote that with colleagues of mine).

Let's be clear, Singapore is predominantly a corridor scheme (with two small cordons), with most charging points on major roads and arterial routes approaching the downtown. Stockholm is a cordon (PDF), but also now has a corridor charge for the Essingeleden motorway that passes through the city.  London is strictly an area charge (it charges circulation within the cordon as well as crossing it).  

Wilson mentions this because he prefers cordons it seems but sees corridor charging (which the Mayor of Auckland did propose last year) as being flawed because they are easy to avoid by rat-running on local roads. That’s true, if you don’t put in place measures to price that rat-running (e.g. by pricing exiting and then re-entering a road to avoid a priced point). 

This appears to be a very weak attempt to condemn road pricing schemes that aren't cordons, but the Government's policy is not that there should not be cordons, or that there should be any specific type of congestion pricing scheme at all.  

Of course, there has been no decision at all about what proposal to introduce in Auckland, but the Congestion Question did recommend a downtown cordon supplemented by corridor charges, on the most congested routes on the Isthmus and towards the North Shore first. More work has to be done on what the first scheme would look like. So it seems rather premature to be antagonistic to the very concept as a whole at this stage.

The Congestion Question indicative downtown cordon

His next point is to appear to be critical of the timeframe, as he claims the first scheme would not be in place until 2028. This seems pessimistic. Sure the legislation and approvals will not be finished until 2025, but it is entirely plausible for a scheme to be operational within two years of that. He indicates that the timing of elections (local in 2025, national in 2026) is driving this, but it’s unclear that this could be accelerated to be significantly faster. He says it should be operational when the City Rail Link opens, which I agree, but it seems unlikely that even if approved today, that a scheme could be operational in 18 months. Yet surely if that is the best time, the second best time is as soon as possible afterwards?

Then Wilson goes back to how unpopular it would be. He claims it is a “regressive tax”, yet I don’t recall him calling the introduction of the Auckland Regional Fuel Tax that, even though there is a study that explicitly concluded that (PDF). The fuel tax applied 12.5c/l on petrol and diesel sold in Auckland, and of course meant everyone driving paid it, except those able to afford an EV or hybrid vehicle. Wilson did support the regional fuel tax when it was introduced, and he said “It does hurt the poor disproportionately …. But it also targets almost everyone who's clogging up the roads”.

That’s nonsense. It targeted nobody. He also said “one day we’ll have better ways to manage demand” saying essentially road pricing would be that but “we don't have the technology in place to do that yet and it's controversial”. The regional fuel tax was not introduced as a demand management tool, but moreover the technology to do road pricing exists now.

Now the government is now advancing it, and he opposes it, not because of the unavailability of technology. Of course the regional fuel tax is now history, but that measure improving the cost of living for most Aucklanders is unnoticed, because this is a polemic.

He gives no evidence for road pricing being particularly regressive, although as a concept it is no more “regressive” than pricing water, electricity or indeed public transport or food. As part of developing proposals for Ministerial approval, an impact analysis of the proposed “time of use” pricing scheme will need to be undertaken.  Perhaps Wilson could focus on what that analysis should look like, rather than dismissing the whole idea of pricing as regressive. Now the Gothenburg congestion tax IS regressive, there is evidence of this (PDF), but it was a scheme set up to raise revenue and is far larger in scope than the scheme that would have been needed to relieve congestion.  The NZ Government is proposing time of use pricing specifically to improve network productivity, not to raise revenue, but Wilson ignores that, as it doesn't fit his polemic. Of course congestion is regressive, as the richest don't commute at peak times or can buy homes close to work.

He claims “It’s not the key to solving congestion. And one of the most common arguments for it is economic gibberish”. This is rather embarrassing. If you don’t understand an economic argument then the best way of understanding it is not to call it gibberish, in fact it displays astonishing ignorance. His understanding of the value of time and the economic impacts of congestion is poor indeed.

He claims “These charges will be a cost of doing business that companies will pass on to their customers. For the general public, they will raise the cost of living”. Will they? Having claimed correctly that less congestion will make all sorts of businesses more productive, whether it be for freight deliveries or services such as the building trades, they will be able to undertake more jobs for the same cost (wasting less time and fuel). Pretty much all benefit/cost analysis indicates businesses save much more than congestion pricing costs, so it would not be passed onto customers. Of course congestion costs are passed on.

It's his next claim that is the most embarrassing.

Alan McDonald from the Employers and Manufacturers Association (EMA) has said much the same. “Recent traffic monitoring data has found that Aucklanders are losing 22 million hours per year out of their lives while they sit in traffic,” he declared. “That equates to a $1.3 billion annual hit to GPD.”

Gibberish. You can’t link private travel to productivity because very few people drive to work on company time. However long your commute takes, it’s your own time you’re wasting.

Everyone resents it, and fair enough. But the economic value – the “annual hit to the GDP” – is zero.

Wilson claims that the economic cost of congestion to private individuals is zero. He claims this doesn’t impact on GDP.  Let’s set aside the obvious social cost. Congestion means commuters leave home earlier and get home later than they would otherwise. That’s less time with family, less personal time, less time to cook, to exercise, to sleep even. Wilson understands what externalities are, I think, so he could at least acknowledge that.  However, what he misses out is what congestion does to opportunities for individuals.

You see the available job pool for most people is directly related to the duration of commute from wherever they live to wherever jobs are located. Most people are happy to commute for up to half an hour, and many in a larger city for up to an hour, although those with children to look after are more challenged. Beyond an hour those able to spare that amount to time to travel to and from work are much more limited in number. In short, congestion reduces the opportunities people have to increase their incomes with better employment, and it also reduces the labour pool available to employers to improve their productivity. I’m always a little sceptical of the methodologies used to “cost” congestion, but to dismiss traffic congestion as not imposing costs on GDP as it applies to private individuals is ignorant. There is literature to back this up.

Wilson then determines that the answer isn’t road pricing, but more rapid transit. Yet he doesn’t seem to be able to explain why cities like Paris, Amsterdam, Tokyo, New York, Sydney or San Francisco all have chronic congestion WITH lots of rapid transit? The Northern Busway is a great piece of infrastructure, but it hasn’t fixed congestion on the Northern Motorway, although it has absorbed a lot of demand growth. Buses do carry a lot of people over the Auckland Harbour Bridge, but the idea that this is a substitute for road pricing is simply absurd. He may as well say that you don’t need parking fees if there are free buses.  It’s completely false to equate the impacts of the Congestion Question, which was a network wide reduction in congestion, from the effects of the Busway on one corridor.  He claims rapid transit reduces emissions. This only happens if it enables modal shift from driving cars, which of course congestion pricing promotes as well (bearing in mind transport emission in NZ are capped with the Emissions Trading Scheme anyway). 

The article explains all of the benefits from a lot of public transport, without even really noting that the costs of all of this infrastructure he wants need to be paid for, and one way of doing it would be through congestion pricing. That doesn’t mean I think that’s how the money should be spent, but surely that connection could be made? Furthermore, all of the rapid transit he touts does absolutely nothing for freight or tradies or other commercial traffic, as they can’t use it.

He then makes this remarkable failure to connect thoughts:

Our roads are appallingly congested, we are failing to reduce carbon emissions and our road safety record is among the worst in the developed world. The opportunity is for a rethink about how and why we use the roads, so we can build ourselves a more functional, friendlier city. Instead, the Government proposes a new tax.

He claims “a great many people will not be able to avoid a congestion tax”. How does he know? If it is a downtown cordon, where only 13% of employment is based and half of commuters already travel by public transport or active modes then hardly anyone will be affected. Even if it is just the Mayor’s two corridors, that wont affect most commuters either. Again, this is just nonsense. 

If Simon Wilson can’t see the link between road pricing, reducing congestion and emissions, and making a city more functional and friendly, then he is either ignorant or deliberately disingenuous. I fear he is simply a polemicist seeking a headline and he can’t give any credit to a politician he doesn’t like or support for implementing a policy that does more for what he wants than any other single measure at the lowest cost.

Opposing the very concept of time-of-use road pricing at this critical stage indicates he is not really interested in enabling all of the potential tools to reduce congestion, lower demand for emissions and encourage modal shift at all, but rather is just writing polemics for headlines. 

Wilson would be better placed to focus not on opposing the first government in NZ’s history to advance road pricing to implementation, but rather to focus on the design of the first scheme proposal for Auckland, to ensure it has a positive impact on reducing congestion, minimal impact on those with low incomes and limited choice, and to encourage creative solutions, such as those used elsewhere, to address any issues. If he wants a cordon, then talk about it. If he wants a different option, then fine. However, if he doesn’t know anything much about the topic at all, he might prefer to read a bit more, talk to people who do and not try to undermine a policy that actually has general support across the political spectrum from the Greens on the hard-left to ACT on the classically-liberal right. 

Time-of-use pricing could help Auckland look much more like what Simon Wilson wants it to, it’s just a shame he wants to get in its way, on grounds that are spurious and almost entirely baseless.

The Congestion Question evaluation of impacts

Are toll roads bad for New Jersey?

This is the question raised and answered by Bob Ahler. Bob joined an ad-hoc commit-tee called Citizens Against Tolls, whose primary goal was the elimination of tolls on the Garden State Parkway.

Those of us working in the road pricing world have to face opponents to tolling regularly. Of course many of the concerns about tolling are quite legitimate issues about whether the money will be used for what is intended, or misunderstandings about how a system may work. People don’t generally like paying for something that is, to them, free otherwise. Given most roads most of the time don’t have any form of direct charging (although fuel taxes form a kind of indirect charge), putting any form of pricing on a road is going to be controversial.

Once you get beyond highly politicized debates about tolling it is worthwhile to understand what arguments are made against it, and so I though it would be worthwhile reviewing Bob Ahler’s series on tolling in New Jersey in the Cape May County Herald (not that it was easy to find all of the articles as they weren’t linked in a series and there are two "fourth articles" and a sixth as well as a final article).  The ones I found are numbers 1, 2, 3, 4, 5 and 6, oh and another 3rd!

All in all there is a lot of interesting history in his articles, but his anti-toll position doesn’t seem unassailable. I thought the best approach would be to go through his conclusion point by point.  He starts:

With the high cost to collect tolls, additional driving times, wasted gasoline, additional pollution, unsafe toll plazas driving conditions, loss in matching federal highway funds and the subsidy paid by drivers on toll roads, the only conclusion that can be reached is that tolls do not serve the best interest of New Jersey citizens. 

It is therefore recommended that tolls be eliminated from all roads, thus saving $522 million a year. Lost revenues can be made up from a 7.2 cent per gallon increase in the state gasoline tax, thus insuring that road upkeep funding is the same for all roads in New Jersey.

So he proposes eliminating tolls and increasing fuel tax. So everyone pays more, rather than letting people using some roads pay for those roads. Except of course, not everyone pays the same. Fuel efficient vehicles pay a lot less, except that they don’t actually consume less “road” than others. However, it is his financial claims of a saving that don’t bear close scrutiny, especially since around a quarter of these “savings” are “travel time” savings which can be gained by fully free flow tolls. Let’s go one by one: 

The following is a list of toll elimination benefits:

• Savings of over three hundred million dollars a year in toll collection costs.

Now it is true that collecting tolls costs more than collecting fuel tax. Indeed, collecting taxes on basic commodities is always cheaper than billing people for what they use. It would probably be cheaper to charge telecommunications, cable TV and broadband through a tax on electricity, but would it result in better services if such systems were provided by a government monopoly reliant on politically driven funding? I doubt it. Toll collection costs can be brought down. EZPass is one way to do it, but a more concerted effort to move to fully electronic free flow and learning lessons of competitive utilities can do a lot more. This is Bob’s strongest argument, and it is up to the toll sector to show it can collect tolls more efficiently. What he completely fails to even acknowledge is that fuel taxes are a poor proxy for road use, as they can’t reflect wildly differing geographical costs, variations in demand and supply, nor do they reflect well the marginal costs of road use by the heaviest vehicles. 

• The in-flow of $75 million a year in matching federal highway funds.

What this demonstrates is that, if true, the Federal government is not neutral on states that toll. He appears to consider that if tolled roads were untolled, they would gain federal funding support, but that also requires matching funding from the state. I advocate that tolled roads that fuel tax payers use should get funding from the fuel tax paid in using those roads, that would eliminate this distortion and reduce tolls.

• Without toll plazas, there would be less traffic congestion, resulting in less wear and tear on vehicles, a reduction in gasoline usage and a reduction in air pollution.

Don Shula Expressway, North Texas Tollway Authority, Maryland County Connector, Toronto 407, Melbourne Citylink etc etc. Toll plazas are unnecessary for tolling and eliminate all of these concerns. Next!

• Driving on the Turnpike, the Parkway and the Expressway would be safer and more of a pleasure since drivers would no longer have to contend with competing traffic in the selection of a toll lane.

See above. Next!!

• Lengthy and cumbersome entrance and exit lanes could be replaced with shorter and more efficient clover-leave ones.

See above. Next!!!

• The location of entrances and exists would be based on the needs of traffic instead of toll plaza revenue generation consideration (sic)

Electronic free flow tolling can enable this too.

• Drivers would no longer be encouraged to change their driving habits by clogging up adjacent roads in order to avoid toll roads. All roads would therefore be funded in the same fair and efficient manner, thus spreading the cost equally among all drivers.

A road that is clogged is badly priced, so should be subject to pricing that matches demand with supply. Fuel tax may be financially efficient, but it is not economically efficient or fair, nor does it spread costs equally. Indeed, why is spreading costs equally fair? Why should users of an expensive major facility be subsidised by other motorists? The diversion issue is reasonable, but it is an argument also for network wide road pricing, which I believe would give Rob apoplexy.

• The elimination of subsidies paid by tolls.

The thesis here is that toll roads are priced far higher than untolled roads and some of the money is used to subsidise the other roads. A fair point, as tolls should be set at efficient rates related to the infrastructure costs concerned, and demand and supply. On some roads this means a healthy surplus, or a profit. Yet fuel tax offers exactly the same thing. Some heavily used roads generate lots of fuel tax revenue, and others generate little. So why are tolls particularly singled out for this argument?

• The lowering of the operating cost for trucks, thus eliminating the need to pass toll costs on to customers.

Well, only for the trucks using toll roads. He wants to increase fuel tax on everyone everywhere in the state, which puts up their costs. Wouldn’t they pass that onto customers?

• The elimination of the temptation by elected officials to sell or lease toll roads to meet current budget requirements.

He means, to retire some debt or fund some other infrastructure. The argument presented as to why this is “bad” is that the new owners would increase tolls. That can be limited by specifying conditions in a concession of course, which is more effective a limit than it could be under state ownership (as politicians can’t constrain themselves as easily as they can constrain the private sector). Why is state ownership of all roads a good thing in and of itself, unless this is an ideological argument?

• Although the need to eliminate tolls appears to be obvious, there could be pressure against doing so by those having a vested interest in the continuation of tolls. It will therefore take a politically courageous person to eliminate road tolls in New Jersey, one who seeks to run an efficient and fair highway system, so that we can say that we do not have tolls on roads in New Jersey.

Politically courageous? Well it would be popular, except for the need to increase fuel taxes and find the money from somewhere else. The vested interest in the elimination of tolls comes from those who pay tolls, and nobody else. Others would have to subsidise them. The vested interest in keeping them comes from everyone else would avoids having to pay for the roads tolled, and the staff. The former is a big enough group to listen to!

All in all, the argument against tolls rests on an antiquated view of tolling technology, combined with high operating costs and poor pricing. There are cases where tolls don’t make sense, but these often are new roads where the case for the road itself was very poor indeed. Unviable toll roads are often unviable roads. Other cases are those where the benefits to non-users are substantial and come from the diversion of traffic onto the new route. Bypasses being a clear example, where having all through traffic bypassing a built up area, reduces congestion, pollution and accidents, but tolling such a route may dramatically cut those benefits, which when monetized, more than make up the value of toll revenue.

I don’t believe toll roads are bad for New Jersey. They may be run more efficiently with new technology, business processes and commercial imperatives. However, they generate net revenue for the roads concerned that would otherwise have to come from fuel tax. Consider the elimination of tolls on the crossings into Manhatten to think about the effects of that on congestion, and the pitiful fuel tax revenue one would get from the increase proposed. Quite why rural motorists should pay more to make up revenue from those driving into Manhatten is unclear. As fuel tax revenues become ever more difficult to obtain, the likelihood is that tolls will be considered more frequently, but the real benefit comes from better pricing to manage congestion, and send signals to road users and governments spending money on roads, as to how to respond.

Almost all of the economy functions according to price signals, and is the better for it. Roads can be too, as long as the transaction costs involved in tolls are more than offset by the better investment and usage decisions that such signals can help inform. An overly simplistic view of tolls is that they are just another way for the government to make money, but I consider them to be a way to efficiently price infrastructure and everything that comes from that – and this is where debate on tolls should reside. In economics, not politics.

Empty arguments against tolling in San Diego

A columnist called John Van Doorn in the North County Times in San Diego is opposing proposals for a new toll road to bypass a congested section of the I-5.  The arguments against tolls are nothing new, and easily refuted.  Indeed three comments on the column already make good points in response.  

This is what he argues, with my response

1.  Toll booths add to delays:  It is wilful blindness to ignore electronic free flow tolling, even though deployment in the US has been glacial.  This argument should be left in the 20th century where this thinking belongs.  Any new toll road in the USA should be built with electronic free flow technology in mind.

2. Studies claim collection costs can be up to one-third of revenues:  I can say "studies show journalists rarely are experts in things they write about", without quoting a study.  Any toll road with such high costs is either grossly inefficient, heavily underpriced or is not viable (for having too little traffic).  Efficient operations today of any toll road should manage costs well below 10% of revenues. 

3. If there isn't congestion, more will use the free original road congesting that route:  So it is better to not have a new route then?  It is entirely circular.  If the current route is congested, the toll route will relieve it.  It wont be empty, so it benefits users of the original route.  You might believe a new route could induce demand, but if tolled that will put a price on that and besides demand could always increase on the untolled route regardless.  The solution is to price ALL roads efficiently, but Van Doorn shows scant knowledge of economics so he wouldn't consider that.

4. Motorists would be tracked:  After implying a toll road means manual collection, he then jumps on an assumption made for electronic tolls.  It is also empty.  Electronic free flow toll roads are charged usually through a tag and beacon system that only identifies a vehicle crossing a particular point so the account associated with it can be charged.   It reads number plates in other systems either to charge or for enforcement purposes.  It isn't tracking, it is no different from paying for your shopping with a card, leaving a record of your presence in a shop at a particular time.  Yes, in theory if done by government entities they would know that you had, driven past a point at a particular time, but that isn't tracking.  Automatic number plate recognition cameras could be placed all over road networks to do this, and mobile phones do it more effectively than conventional electronic tolls.   Now it is more complex if GPS and distance charging are involved, but that isn't what is being discussed and even if it was, it doesn't involve tracking either.  It involves measuring distance.

Tolling has many opponents, typically they are people confused with technology, with little understanding of transport economics, but also understandably sceptical of government agencies running toll roads.  There are good reasons to be resistant of politicians incompetently using tolls as a quick fix, but not to resist tolling per se.  Tolls can be a low-cost and efficient way of raising funds for new roads and maintaining existing ones, and debate about tolls should be about their viability, not on the basis of ill-informed presumptions.