Monday 21 February 2022

Does road user charging harm electric vehicle sales?

RUC makes little difference to EV sales.

Under a year ago I wrote this piece Would RUC for EVs harm sales? largely in reference to concerns in Australia from the Electric Vehicle sales lobby that Australian states investigating road user charging (RUC) based on distance, for light vehicles; would significantly reduce consumers' propensity to buy such vehicles. 

Dr Jake Whitehead from the University of Queensland apparently conducted a study which was cited in an article in Driven that claimed it would hurt sales by 25%, claiming that it would be perceived by drivers as adding $4,000 on the cost of owning and operating an electric vehicle. I demonstrated in my previous article that this was highly questionable, because no purchasers of petrol powered cars estimate the fuel tax they would pay on top of the cost of owning and operating the car. If they did, it might be around A$367 a year for a new petrol Hyundai Kona driving the average annual 13,838km of a car in Victoria. This compared to around A$346 for an equivalent EV paying the Victoria state RUC. Sure I understand Whitehead's estimate, but it simply isn't a valid comparison to think anyone buys a car thinking about the lifecycle costs of charges associated with using the road (although buyers of many commercial vehicles do make such estimates). 

However, there is now some actual evidence, with the latest sales data on electric vehicles in Australia.  You see Victoria introduced RUC (known as the ZLEV road-user charge) at A$0.025 per km for battery-electric EVs (BEVs) and A$0.02 per km for plug-in hybrid vehicles (PHEVs). Both also obtain a registration fee discount of A$100 a year. 

The data for 2021 indicates that 6,396 EVs were sold in Victoria, with 7,430 sold in NSW and 5,342 in Queensland. Given RUC was introduced in July 2021 if Dr. Whitehead's assessment were valid you would expect a significant decline in EV sales relative to other states, but that isn't what happened.

You see the 7,430 sold in NSW compared to the total passenger vehicle fleet in NSW is around 0.17% of all vehicles. Whereas the 6,396 sold in Victoria, compared to the total passenger vehicle fleet in Victoria is around 0.16% of all vehicles.  If it made a difference it was barely discernible, and undoubtedly less relevant than other factors.

In the ACT, the number of EVs sold comprised 0.37% of all passenger vehicles registered in the territory, the highest of any state and territory, but that's not so surprising. The ACT has relatively high affluence, motor vehicle registration is free for two years for EVs, with a 20% ongoing reduction.  

Queensland has a slightly higher number of EVs as a proportion of all passenger vehicles at 0.18%, but that's close enough to NSW and Victoria to suggest that there isn't a significant difference in policy impacts.

Western Australia, Tasmania, South Australia and Northern Territory all have much lower sales, as a proportion of the total passenger vehicle fleet, than the other states, ranging from 0.1% to 0.05%.  In shortsthere is a great deal of reluctance to buy EVs in Australia, but the wealthiest eastern states/territory have better sales.  Note that both NSW and South Australia announced they would introduce RUC from 2027.

So from that I conclude that RUC in Victoria has made virtually zero difference to EV sales in the state. 

That's with a RUC rate of A$0.025 which is equivalent to US$0.029 per mile or £0.02 per mile or €0.016 per kilometre. That is higher than some US states have implemented or are proposing, but much lower than would be a replacement rate for fuel taxes in any European countries.

A higher rate might make a difference to sales. Certainly New Zealand has maintained an exemption from its RUC system for EVs, which would have resulted in them being charged NZ$0.076 per km (A$0.071 per km, €0.045 per km, US$0.08 per mile or £0.06 per mile). Exemption EVs from RUC has undoubtedly contributed to greater sales per capita than in Australia, but I suspect other factors, including price of electricity and the network of charging points matter as well (plus less range anxiety in a smaller jurisdiction).

What it means is that RUC, as a policy tool, needs to be considered within the context of a wide range of pull and push factors for jurisdictions that want to encourage sales of EVs. 

Thursday 17 February 2022

UK House of Commons Transport Committee Road Pricing report released - so here's my review of it

Nothing new in the UK investigating road pricing

The UK's journey towards national road pricing has been long and arduous.  It once started with the Lorry Road User Charging programme in 2002, essentially a more ambitious version of the highway heavy vehicle RUC systems in continental Europe, but that was abandoned because the business case didn't stand up at the time, given the costs.  It was replaced in 2004 with National Road Pricing, which was to pioneer TDP (Time Distance Place) based road pricing, to better manage congestion.  However, that collapsed due to public opposition, primarily out of concern about how much motorists would have to pay, and a lack of trust that fuel duty and vehicle excise duty would be cut at the same time.

That was 2007.

One of my big questions is whether UK politicians and policy makers have learned from that failure, and maybe learned from the successes of jurisdictions elsewhere that have partially implemented forms of RUC, such as in continental Europe, the United States and New Zealand. 

House of Commons Transport Committee report

Last year, the Transport Committee of the House of Commons (UK) held an inquiry into road pricing (and also electrification of the road vehicle fleet in the UK).  It released the findings of its work on 2 February 2022 with a full copy of the report available here (PDF).

The Committee notes in its report that it "launched an inquiry in December 2020 called Zero emission vehicles and road pricing. We chose to split the inquiry into two parts. We reported our findings on Zero emission vehicles in July 2021. That Report addressed the opportunities and challenges presented by the advancement of the ban on the sale of new petrol and diesel vehicles to 2030. This Report on Road pricing covers the second part of our inquiry. It examines the consequences of the shift to electric vehicles, including tackling the decline in fuel duty and vehicle excise duty."

Before I review the report's key findings, it is important for those unfamiliar with the UK to recognise that the report is influential but not binding on the UK Government. It is a matter for Cabinet and ultimately Parliament to make decisions on implementing any form of road pricing or road user charging (RUC) for the UK.  

What is behind the inquiry?

  • Revenue
  • Demand impacts of no charges on electric vehicles

It's fairly simple, because it follows on from the primary reason RUC is being implemented in US and now Australian states. Revenue from fuel taxes is threatened by the shift away from fossil fuel use by road vehicles. £28 billion is the annual revenue from both fuel duty and with the UK looking to prohibit sales of petrol and diesel vehicles from 2030, it will drastically erode revenues from motor vehicles. Add the £7 billion raised from vehicle excise duty (an annual registration fee for motor vehicles) that is hypothecated to the National Roads Fund (which fully funds National Highways and contributes towards spending on local roads) and the total revenue is equivalent to 4% of all tax revenue, and this is more than just a problem for spending on roads, let alone transport.  That revenue is five times what is spent on roads, so it is a general source of revenue.

In other words, unlike the US or Australia where motoring taxation barely or doesn't at all cover the cost of paying for roads, in the UK it does so and so much more.  Setting aside externalities (and given no other sector pays for externalities either), the roads make a significant fiscal surplus.

However, it's not all just about money, it is also about the demand impacts. Electric vehicles pay nothing to use the roads, which means that driving is substantially cheaper for those vehicle owners than others. This is a matter of both equity, but also overall impacts on the network. If a third of vehicles are paying nothing in say ten years' time, the effects on congestion are likely to be considerable. I recall a study around twenty years ago that estimated the demand impacts of fuel duty on driving in the UK was around 10%, in other words, 10% fewer miles were driven in the UK because of the effect of fuel duty. As fuel duty has not increased in 11 years, that impact is likely to have weakened somewhat, but it still exists. So for the UK, RUC is also about sending a price signal to drivers to think about whether to drive or choose another option.

Key conclusions

  • It is difficult to predict the timescale of the impact on revenues. In 2021, 11.6% of new cars were battery electric and it is estimated by 2026, only 4% of all registered cars will be electric. (comment: Indeed, but hybrids and more fuel efficient vehicles are eroding revenues easily as much as electric. It would be wrong to focus solely on pure electric vehicles) 
  • However, it gets much harder to introduce RUC when there are many electric vehicles compared to when there are few, so the message needs to be sent that electric vehicles will have to pay to use the roads. (comment: agree)
  • A shift to electric vehicles could increase "traffic levels" by 51% by 2050 according to the Department for Transport (DfT), with the average driver spending an additional nine hours a year in traffic by 2040. (comment: "on average" hides some disparities, but the aggregate impact is likely to be true. Cheaper driving will mean more driving).
  • "Any alternative road pricing mechanism must be revenue neutral to the Government rather than causing drivers, as a whole, to pay more than they do currently. Such a mechanism should be phased in before fuel duty and vehicle excise duty decline to zero. The situation is urgent; work must begin without delay" (comment: YES absolutely)
  • Technology already built into vehicles (in-vehicle telematics) has potential to collect and deliver the data necessary for a road pricing system (comment: Yes, and this happens already today on a small scale in Utah).
  • "The Government must assess the potential effect of a road pricing mechanism based on telematic technology on high-mileage drivers, such as road hauliers and those in rural communities, and on those least able to adapt to increased motoring costs." (comment: Yes there should be modelling and even piloting of road pricing to assess those impacts. Noting that road hauliers should see the HGV Levy replaced by RUC, but rate setting needs some serious policy analysis. A cost allocation study should be carried out).
  • "The successful implementation of a national, technology-based road pricing scheme is contingent on the Government explaining how data capture will work in practice, ensuring that data management is subject to rigorous governance and oversight and reassuring the public that their privacy will be protected" (comment: Data privacy is critical, but there needs to be acknowledgement of some obvious ways to ensure this, such as not having a government owned RUC system collecting trip data, specifying the level of location disaggregation needed, consider having a non-location aware option (at least for a transition period))
  • The DfT and Treasury should set up an arms-length body to investigate and evaluate options to replace fuel duty and vehicle excise duty (comment: This seems reasonable, but there isn't an actual need to replace vehicle excise duty. Vehicle excise duty is only an issue because electric vehicles are exempt, but that exemption need not be maintained. However, there are merits in shifting from ownership based to usage based charging of motor vehicles).
Conclusions that are questionable
  • (para. 25) A claim that the UK can be a "world leader" in road pricing, is bold but also maybe fanciful, especially given how little progress has been made in the UK compared with other jurisdictions. The report doesn't note that there are three and soon to be four jurisdictions charging light vehicles (in all cases but one, electric vehicles) by distance instead of fuel duty. There are two jurisdictions that charge heavy vehicles by distance, weight and configuration instead of fuel duty. There are twelve jurisdictions that charge heavy vehicle by distance and weight as well as fuel duty, using GNSS technologies. However there is very little citing of relevant international experience. Is that because most of those who gave evidence have no such experience> There are multiple companies in Europe, the US and New Zealand that run RUC account management operations, including supplying technology for vehicles (heavy and light). If the UK WAS able to implement a national road pricing scheme that included congestion pricing based on distance it might be world leading (although I suspect Singapore might beat it), but it has so far proven that it cannot even roll out a basic cordon or area based congestion charge beyond London and Durham after 21 years. 
  • (para. 28) The claim is that "The devolution of road pricing could lead to the introduction of clunky, unconnected schemes that charge users the same price for driving one mile into the zone as those who drive across it for hours in a day. The more regional schemes that are created, the harder it will eventually be for the Government to implement a functional national system". This is simply nonsense, not only because it is based on the idea that there is only ONE type of congestion pricing scheme (cordon or area charges) and only one way to implement it (assuming there is a flat all day charge). Why can local schemes not work with a national system? The US looks like it may have multiple state RUC systems and ultimately federal system, which may all be interoperable and function seamlessly. Several European countries have toll schemes plus heavy vehicle RUC systems operating in parallel (see Belgium which has a national heavy vehicle RUC scheme, with Brussels developing a local congestion based scheme for light vehicles). New Zealand has a national RUC system and is investigating a local congestion pricing scheme for Auckland. There MIGHT be a problem if systems require multiple accounts, or multiple pieces of on-board equipment, but this is unnecessary and avoidable. It is likely to be lower risk for congestion pricing to be locally defined and implemented, even if nationally regulated.  It seems to be an excessively engineering rather than policy and business based perspective.
  • (para. 32) The report cites John Siraut from Jacobs who says:
"He stated that a system where charges vary dynamically based on the road being used and the time of travel are “certainly perfectly feasible to do”, even though they do not exist at the moment"

This is simply wrong. There IS a system that varies by location, time of day and charges by distance, in the Czech heavy vehicle RUC system. For heavy vehicles, the charges vary based on the type of road (motorway vs. 1st class highways) and the time of day.  Arguably, Singapore also has a system that varies by road being used and time of travel since.. 1997 (albeit not charging by distance), but certainly the can rates vary across 78 charging points by small increments of time of day. Strange in particular when he is previously cited as saying "Singapore comes closest" to the model discussed, which suggests he may have been misquoted at least once. . (Disclaimer: I used to work for a wholly owned subsidiary of Jacobs, in the UK from 2012-2015).

  • (para. 34) The report says "The Government must assess the potential effect of telematic technology on changing drivers’ behaviour and delivering its wider policies on air quality, congestion, public transport and public health.".  No, the effects of telematics technology on drivers' behaviour should be minimal. It is about the policy that is implemented, not the technology (although technology is the enabler).  Furthermore, if you think introducing road pricing should also address a whole host of externalities, then you'll spend another five to ten years investigating it.  This seems like an unwelcome and unnecessary distraction. The effect on environmental and transport policy outcomes will depend on the policy selected. 

Where to from here?

There is one thing missing from the report, which is a thoughtful and considered view on why road pricing failed last time it was tried in the UK. Perhaps it is because most of the politicians were not in Parliament at the time, perhaps it is because the public servants don't want to look into what were largely mistakes by public servants and politicians.  However, it is important to remember how sensitive the issue is.

The number one reason road pricing failed before was that the policy was not designed to benefit those who would pay, and there was no leading of the narrative to address the two biggest concerns:
  • Road pricing means paying a lot more to drive
  • Road pricing means tracking everywhere that you drive
It cannot be repeated too much that road pricing needs to be introduced in a way that generates the same net revenue as fuel duty and vehicle excise duty. It has to be done in stages and steps that are achievable, but it is almost impossible to introduce any form of congestion pricing until all vehicles are on a road pricing system that measures location and time of day.

So congestion management should be very much a secondary concern, it should be about replacing the current system. There are various pathways for doing this, ranging from starting with electric vehicles, starting with newly registered vehicles or with heavy vehicles.  

There should NOT be pursuit of a central government owned and managed system that is procured centrally with a single provider of technology and account management. This is NOT a model that any jurisdiction has pursued for network wide RUC for twenty years (see Switzerland).  There should ALSO not be pursuit of a PPP with a single integrated supplier of a system to collect revenue, that is ALSO not a model any jurisdiction has pursued for sixteen years (see Germany).  It should be an open market of certified service providers, supported by a government managed enforcement system.  Revenue should be collected by competing service providers, with multiple technical platforms (such as in-vehicle telematics or plug-in devices), but enforcement managed by government.  

Consideration needs to be given as to how to refund/credit fuel duty paid when vehicles that pay fuel duty transition to RUC.

NONE of this matters unless policy is designed correctly.  That means:
  • RUC rates that are based on a rational economic assessment of what different types of vehicles should pay.
  • Charges that don't vary by time of day until a vast majority of vehicle are on a system
  • Charges that vary by road type and location for traffic management purposes, but not to such an extent as to encourage use of vehicles that do not pay RUC
  • Hypothecation of some revenue for roads, to go beyond what the National Roads Fund does now
  • Care being taken as to how to treat the surplus of revenue beyond what is needed for roads.
  • A feasible staged transition path.
It doesn't require investment in alternatives to motoring.

This requires a level of engagement with the public and business that did not happen with National Road Pricing, but is exactly what has been going on in some US States in recent years.  

It would be extremely helpful if the UK looked beyond itself to see that things are moving fast elsewhere.  

Take for example Victoria, Australia.  In a short period of time it went from studying to implementing a basic distance based RUC system for electric vehicles, based on odometer reporting.  Is it the long term scheme? No.  Does it work? Yes. Is it a start beyond which more can be learned and it can be built on? Yes.

I lived in the UK for fourteen years, I worked on Manchester's congestion charging scheme that faltered because of a lack of public support (in part because there was poor communications with those who would pay as to what it would mean for them). I observed the National Road Pricing project going wrong because politicians and civil servants didn't think they needed to seriously and convincingly address concerns of the general public about paying twice, or being tracked. If they still think that, then road pricing isn't going to happen easily.

It's my hope that both DfT and The Treasury can reflect carefully on lessons learned from the past, in the UK and elsewhere, and recognise that for road pricing to be advanced in the UK it might need to think of doing it in incremental steps, that means not all objectives are met at first, and that it doesn't need to re-invent the wheel.

Wednesday 9 February 2022

Can 2022 see congestion pricing moving faster than a glacier?

For all of the fanfare and success of the late 1990s and early 2000s in congestion pricing being launched, it is 2022 and it is still possible for experts in the field to be able to name every single city in the world that has some form of congestion pricing on existing roads.  

However, let me first DEFINE what I mean by congestion pricing.

I do NOT mean the use of time-of-day tolling on toll roads (because toll roads already have pricing applied, it is simply the application of a demand-based rate not a flat rate), nor the introduction of toll or High-Occupancy-Toll (HOT) lanes, which are in most cases the provision of premium priced lanes to bypass congestion on other lanes, not the application of pricing on all capacity in a corridor.  As desirable as time of day tolling can be to manage demand, and as desirable 

It hasn't been because of technology either, it's not technically difficult or expensive to introduce a financial charge on a single or multiple sets of points on a road network.  Nor is it because of doubts about its effectiveness, because pricing is proven to influence demand away from roads at times they are well priced. See Singapore, Stockholm and London (at least in the first few years), followed by Milan and Gothenburg.

Furthermore, when considering the range of issues around urban mobility, congestion pricing is not only proven to be able to relieve congestion, but it also reduces noxious emissions and consequentially CO2 emissions. At a time when so many cities are proclaiming a commitment to addressing climate change, it seems strange indeed that so few bother with a proven policy that enables better use of roads and encouraging modal shift, time of day shifts of travel, and which is more potent that probably any other single intervention to change travel behaviour in cities.

This is not to say there aren't many cities with two other non-pricing interventions:

  • Low emission zones
  • Access control regulations

Many cities in Europe implement these (this excellent website identifies them by country).  However, this isn't pricing. Low emission zones tend to operate 24/7 and are a policy to change the engine type of the vehicles operating within the zone, and access controls are effectively bans or restrictions. 

So which cities have congestion pricing? First let me somewhat discount the ones that exist largely to protect historic centres.  Durham in the UK and Palermo in Italy, and arguably Valletta in Malta (albeit it is more sophisticated than London or Milan, as Valletta charges by the hour).  There aren't many of them, but the characteristics of their local street network are such that it is easier to justify charges that seek to rather bluntly reduce the amount of visitor traffic. An oddity is coastal holiday town Jurmala in Latvia (near Riga), which has a charge imposed during the spring/summer season, largely to control tourist vehicle numbers. It isn't based on time of day, so is similar to the vignette schemes operated in some European countries.

Norway deserves a special mention, because it has seven cities with urban toll rings, not all of which have peak charges. In all cases, they exist primarily to raise money with only Oslo, Bergen, Kristiansand and Trondheim having peak charges to help manage congestion.  Norway has a long history of successfully using tolls to pay not just for new highways and bridges, but also major urban road improvements, by instituting cordons.  So I'd argue that in the cases of those four cities, Norway has congestion pricing as part of the urban toll ring, because it charges ALL trips passing the cordons at peak times.  They are unlike toll roads that are only on a single facility.

Beyond that, the list remains small:

  • Singapore: Remaining the "gold standard", it is on a slower path to transition its technology, but is still the only two cordon plus corridor charging scheme anywhere in the world that prices incrementally purely to achieve an optimum level of service for its road network.  For Singapore, reducing congestion is the goal, because with that comes changes in modal demand that helps sustain more public transport services, reductions in emissions and overall net benefits for the city state in having a more efficient transport network. 
  • London: Despite over 20 years of there being legislation enabling local authorities in the UK to introduce congestion pricing, it is only London (and Durham) that have done so (noting London's powers are under different legislation) in the UK. It remains the only area charge in the world, although Valletta's scheme effectively functions as a more sophisticated one. Its hours are longer, but it is largely the same as it was when it was introduced. Unfortunately, because its gains were not sustained, its example has seemed to inhibit further congestion pricing across the UK.
  • Stockholm and Gothenburg:  These two Swedish cities remain the only two in Sweden to have the "congestion tax", and have managed to evolve their schemes over time. Stockholm's is more about congestion that Gothenburg, which largely introduced the scheme to pay for some large scale transport infrastructure improvements, most of which are not yet complete. Gothenburg's has been far from popular, although resistance has waned in recent years and it has been tweaked to reduce some of the key complaints.
  • Milan: Many Italian cities have permit schemes to access their central cities, but Milan has a congestion charge, which started as a low emission zone. Milan has a complex charging schedule based on vehicle emissions, but is otherwise it is a cordon scheme with a flat charge.
  • Dubai: The Salik scheme in Dubai is one of the few corridor schemes, and in most cases has a flat rate apply on all charging points. It was introduced to manage congestion even though there is no variation of charges by time of day.
  • Abu Dhabi: The newest operational scheme, called Darb, has parallels to Dubai's, but is superior because it is a cordon that does not require tags AND only operates at peak times 0700-0900 and 1700-1900 Saturday to Thursday.  It is still the newest congestion pricing scheme anywhere in the world. 
  • Tehran: Tehran's cordon scheme gives all vehicles 80 days a year of uncharged travel, but beyond that a fee is charged which is dependent on time of day, and a measurement of the duration spent in the charged zone.

Yes New York is under development, and over this year there will further detailed design and consultation undertaken, before a decision likely at the end of the year, but it is still not operational. There are numerous studies underway in cities ranging from San Francisco to Los Angeles and Chicago in the US, and Auckland, New Zealand may yet get approval to proceed this year. Hong Kong is on hold because of the limitations on travel due to Covid 19. Meanwhile, Doha may yet follow the two cities of the UAE in introducing congestion pricing in future years.

So what is it going to take?  Is congestion pricing doomed to be the idea that emerges in one city every few years or so? Has Covid 19 changed travel patterns in ways that make congestion pricing less or more urgent? 

There are a number of factors that affect it. Firstly, the motivations for considering pricing are usually mixed. Many politicians like the idea of revenue that can be used to fund whatever it is that interests them, but the public is understandably reluctant to support any proposals that simply look like a new tax. Few politicians are interested in congestion pricing as a tool to improve conditions for those who drive, to ease congestion and even fewer want it to replace existing taxes. This makes it difficult to get the public on side.

However, there are grounds for some optimism. New York seems likely to proceed, and that will be seen as a demonstrator project for the United States, and may rekindle interest in congestion pricing elsewhere. Yes, the conditions seen in lower Manhattan aren't replicated elsewhere in North America, and in Europe there is always some scepticism about concepts implemented in the US (although it already exists in the UK, Sweden and Italy), but New York will be a step forward in demonstrating how pricing can positively improve traffic and the local environment. 

Hong Kong is unlikely to proceed until Covid19 is behind the region. Auckland has a reasonable chance of proceeding to become the first city with a population over 1 million that is predominantly car dependent to have congestion pricing, but there probably won't be news on that for a few months.

Can we hope that maybe ANOTHER European city might advance congestion pricing? Maybe Copenhagen (again), maybe Dublin, maybe Brussels (already looking to replace registration fees with a time and distance based RUC)? It seems utterly extraordinary that Gothenburg is the last so far in Europe.

In the US, it seems unlikely any more will progress seriously until after New York has implemented its scheme, but with San Francisco advancing a cordon (with charges in some cases based on driver income) maybe it can get further, but the real reward in the US will be for something beyond cordons, towards corridor based charging.

In Asia, with Hong Kong on hold, Jakarta is constantly in the "nearly implementing" phase, but has major structural issues holding it back (particularly enforcement of number plates). 

Is there anywhere else looking promising?