Wednesday 28 March 2018

How should the UK Government reform the HGV levy?

In January 2018, the UK Government consulted on how it could reform it's heavy goods vehicle vignette scheme, also known as the HGV Levy.  The HGV Levy charges all trucks 12 tonnes and above for using any public roads in the UK, based on either one day, one week, one month or a year of use of the network (see this PDF for the rates).

For UK registered trucks it is relatively simple, as the Levy is only available as an annual charge collected in parallel with Vehicle Excise Duty (VED) - the annual vehicle registration tax.

The HGV Levy was introduced as a relatively simple way of raising revenue from foreign trucks using UK roads (as VED on the trucks liable for the HGV Levy was reduced on almost all those vehicles, meaning most paid nothing more).  

The HGV Levy was introduced on 1 April 2014 and in its first year earned £46.5m (US$65m) in revenue from foreign trucks (it also earned £146m from UK registered vehicles, although this largely corresponds to revenue lost from reductions in VED).

As a vignette, the HGV Levy is basically a prepaid fee for using UK roads based on time, as is similar to such schemes operating in several European countries.  The well-known Eurovignette is levied by Luxembourg, the Netherlands, Denmark and Sweden for using their roads.  Latvia, Lithuania, Romania and Bulgaria all also have heavy vehicle vignette charging systems.

What's wrong with the HGV Levy?

It's better than not having it at all, but it has the same fundamental limitation of a charge, for an activity that generates externalities which isn't based on usage, but on charging a right of use.  It effectively means that vehicles using roads the most are cross-subsidised by those using roads the least.

The HGV Levy will generate more revenue if the number of HGVs registered in the UK increases, but not if the current fleet is used more intensively.  Similarly, if more foreign HGVs enter the UK, there will be more revenue, but not if those that already visit the UK travel more often.   Obviously, the vignette may be said to incentivise operators to maximise utilisation (because they pay the same whether a truck is idle or not).

However, to properly understand whether or not the HGV Levy should be reformed, or not, there needs to be some clear policy purposes to it.  For all of the information in the consultation documentation, there isn't a clear stated policy purpose for the Levy itself, just some related policy goals of the UK government around the road freight sector. 

Those policy goals are:
  • Encouraging individual HGV operators to plan more efficient route operation and use the most modern equipment.
  • Helping to drive more efficient use of our roads.
  • Reducing emissions which contribute to poor air quality and climate change.
These are all very well, but before any of that, I'd suggest the core objective of any such charge should be:
  • In association with VED, seek to recover the fixed and marginal infrastructure costs of road use from vehicles that are charged.
As it stands, it doesn't do that particularly efficiently.  For a start, it charges vehicles more according to the number of axles they have.  Although this might be seen as a proxy for increasing wear and tear (based on Equivalent Standard Axle (ESA) Loading), it also penalises operators for having lower ESA loadings for the same tonnage.   However, in a vignette, it doesn't make a lot of difference.

How could the HGV Levy better meet the UK Government's policy goals?

The last of the goals could be better met by having emissions ratings included in the existing charging scheme.  EEVs and the highest Euro rating vehicles could pay less, as with the German system.  That would also, indirectly, encourage use of the most modern equipment (although it's far from clear exactly what public policy objective that achieves in and of itself).   That would be relatively simple, but the other goals (more efficient route operation and driving more efficient use of the roads) are difficult to achieve with a charging scheme that only distinguishes by vehicle classification and dates of permitted access to the network.

Distance, mass, location based charging is the logical next step

For any scheme to incentivise better use of the road network there needs to be a geographic element.  This is near impossible to do with a prepaid vignette.  There could be separate vignette products for use of the strategic road network (Motorways and major highways) compared to the local road network, but it is unclear what that would achieve? Virtually no trucks can usefully operate in the UK without access to the local road network, so that would be universal.  Charging separately for motorways would disincentives use of that network, which is hardly going to optimise road use.

Far more effective would be to charge for actual road use.  That means charging for distance consumed, based on vehicle class and with a location element, so that route choices can be influenced.

For example, it would be logical to incentivise greater use of motorways and major highways over local roads, because such roads tend to be built to standards that mean that the marginal wear and tear generated by a heavy vehicle axle passing over the road surface is lower.  Furthermore, such roads tend to be safer on a vehicle km basis and avoid built up areas.  As a further step, rates could vary between individual strategic and local roads at a later date, with lower rates for A compared to B roads and unclassified roads, to discourage use of residential streets to avoid congestion.

Environmental objectives could also be better met through a distance based charged, because it better links payment with road use, so that those that drive the most pay the most.  As in many European countries, rates could vary according to Euro engine rating, with higher charges for the lowest rating. See the charge rate structure for the German LKW-Maut system below:

German LKW-Maut vehicle classification
A vehicle has a emissions classification as above, which places it into a category.  This category determines whether a pollution charge is levied on top of the infrastructure charge.  The highest quality emissions category pays nothing for pollution, but the lowest quality pay €0.083/km (US$0.167/mile).  Infrastructure categories reflect numbers of axles, which is a simple proxy for vehicle mass.  It can argued that this is not strictly efficient, in that the more axles there are for a vehicle of similar mass, the lower the marginal wear and tear that vehicle imposes on the network.  However, given Germany only charges for use of the motorways and expressways, it is likely that the difference this makes to road maintenance costs in Germany is low.

LKW Maut per kilometre charge rates

So having emissions based charging will both reduce emissions and encourage use of newer vehicles.  Allowing distance and location based charging will enable more efficient route choices to be made.  The bigger question then becomes: How could the UK transition to mass/distance/location based heavy vehicle charging?

A possible path

Other European countries have moved from vignette based charges to distance charging, and the UK could combine this with a reduction in Vehicle Excise Duty to match the rates for light vehicles, so that charges shift from ownership to usage.  However, the UK will probably also need to offer some refunds for fuel duty as well, given how significant fuel tax is as a revenue source (and how unsustainable it is over the long term).  There are revenue benefits in replacing a proportion of fuel duty, because a shift to distance charging will protect it from fuel efficiency and changes in engine technology (albeit limited for many HGVs in the short term).   Any revenue collected from a new distance based charge from heavy vehicles would have to be hypothecated, and it would make sense for this to be a source of funding for both local roads and strategic roads for maintenance funding, alongside Vehicle Excise Duty from all vehicles.

Rate setting for the heavy vehicle charge should also be undertaken independently, by an independent regulator seeking to raise revenue to meet the needs of the road networks across the country, using a cost allocation approach that means heavy vehicles pay only their share of long run maintenance and capital costs of the network.  This also means reforming the whole funding and governance system for roads.  Highways England is a good start, but a similar reform should apply to local roads and the road authorities of the devolved administrations should follow as well.

The way to deliver a heavy vehicle charge would be to adopt an open market approach, certifying service providers to a standard of performance to measure road use and bill for it.  There would have to be procurement of a service around compliance and enforcement, to ensure vehicles paid the charge.  Refunds of fuel duty as part of the system may help increase compliance, but there will need to be considerable efforts made to promote compliance.

Any change should start with HGVs 12 tonnes and above first and transition down to 3.5 tonnes and above over several years.  What this should do is provide a platform for further change in the longer term as fuel tax becomes increasingly unsustainable, and mean that heavy vehicles are charged efficiently and effectively for wear and tear they impose on the road network, with rates able to be adjusted at a geographical and time of day level, to help encourage road use outside congested periods.

I doubt the current UK government will go down this path, but I can certainly see a future one doing so.

Tuesday 27 March 2018

Did London's congestion charge increase pollution?

There has been some news coverage today of a presentation at the Royal Economic Society's annual conference from , Professor John Heywood, Dr Maria Navarro and Professor Colin Green (Lancaster University, University of Wisconsin- Milwaukee and the Norwegian University of Science and Technology respectively, I believe) which allegedly claims that the central London congestion charge caused an unintended increase in some forms of pollution, namely nitrogen dioxide  (by 20%), although it reduced carbon monoxide, nitrous oxides and particulates.

The Sun reported this, but TLE (The London Economic) reports it in more detail. The Royal Economic Society (RES) has a press release about the presentation.  

The obvious simple reaction to this is to say the congestion charge "doesn't work" in terms of pollution.  However, it is much more nuanced than this and worthy of more detailed investigation.  Since I don't have a copy of the presentation, my comments are limited to what has been reported.

Reviewing the key findings

The report says that there were "significant reductions across a range of pollutants" in comparison to "comparison cities" and of course, because the congestion charge did reduce road traffic volumes and there were reports at the time of reduced emissions, it is a fair assumption that the net effects should be positive environmentally.   The key element of this research was to consider the composition of pollution.   What isn't clear is how this was actually measured.

The RES press release said:

While some trips to central London simply may not take place, congestion charging policies seem more likely to change the method of transit. The charge made driving in London more expensive and generated improvements in bus services.

Improvement in bus services came about primarily because subsidies for bus franchises were increased so service frequencies could be increased, and some new routes introduced, and also the introduction of new bus lanes (and of course reduced congestion improved the ability to operate buses on existing roads).  The congestion charge helped raise revenue to support increased subsidies for buses, made reallocation of road space to bus lanes less disruptive for existing traffic and enabled some improvement in bus trip reliability.  

It continues: 

As might be anticipated, more travellers used buses and taxis in central London. This caused a move away from predominantly petroleum-based transport (private vehicles), towards diesel based transport (black cabs and buses).

This is true, although it's important to note that the mode-share for private car travel in central London was 12% in 2001 (before the charge was introduced).   The number of such trips dropped by around 33% between 2002 and 2006, but as a share of trips, it is still small.


The new study demonstrates reductions in three traditional pollutants: carbon monoxide (CO), particulate matter (PM10) and nitrous oxide (NO). These reductions are as large as 25% to 30% for PM10 and NO.  The researchers show that the reduced pollution per mile travelled in the zone exceeds that expected from the reduction in traffic flows alone. Thus, the reduction in these pollutants reflects, in part, removing very high levels of traffic congestion as well as reducing miles driven.

The evidence that reducing pollution is not just about reducing trips, but also reducing congestion (so the trips that remain operate more efficiently, and burn less fuel to travel the same amount of time).  This ought to have significant policy implications, because some environmentalists have treated traffic congestion as a tool to suppress car traffic, but congestion is not a positive.  It wastes time and energy, but also increases pollution.  Unfortunately, in recent policy debates about addressing air pollution in UK cities, the idea that policies that reduce congestion are worth pursuing seems to be ignored. 

Of course the study claims the NO2 increase of 10-20% is due to an increase in bus and taxi traffic.  TfL's own figures indicate around a 13% increase in the number of black cabs entering the charging zone between 2002 and 2006 (bear in mind black cabs** do not pay the congestion charge, neither do minicabs**.  There was also a 23% increase in bus trips, presumably almost all of that was due to TfL increasing services.

Traffic entering central London comparing 2002, 2003 and 2006 (Source: TfL)

From that there come a series of statements, which I think are worthy of questioning.

The authors argue that this increase is likely to reflect the shift towards diesel-based transport.
Thus, the congestion charge may have actually increased the harm from pollution.
They conclude that the reduction in congestion associated with charge simultaneously reduced some forms of pollution but had the unintended consequence of increasing more damaging forms of pollution.

Is this analysis damning of congestion charging as a policy from an environment perspective?

Of course, without having access to the full paper, it is difficult to be certain of my hypothesis here, but this is what appears to have happened.

1. The congestion charge co-incided with policies to increase bus services.
2. Black cabs, being exempt from the congestion charge, were able to undertake more trips under conditions of lower congestion than before.
3. With more buses (a deliberate policy initiative) and more black cabs (because more trips were easier and there was no congestion charge applying),  NO2 emissions increased.
3. Ergo "the congestion charge increased pollution".

This would appear to be an overly simplistic conclusion which could cause some policy makers or lobbyists to think congestion charging, per se, is not a good policy environmentally.  However, such a conclusion would be wrong.

What could London have done differently?

Politically, the congestion charge couldn't have been introduced without an exemption for black cabs, and the expansion of bus services was intended to offer an alternative for some private car trips (although it is unclear whether there was much of a modal shift from car to bus, as it is more likely the underground and rail services took some of the trips).  However, the congestion charge as a policy in itself was not specifically responsible for an increase in diesel vehicle trips and NO2 pollution.  

The congestion charge introduced a flat charge on private cars and goods vehicles (the latter saw a 13% decline in trips between 2002 and 2006), but did not charge buses nor taxis (of any kind).  So it is hardly surprising that while private car and goods traffic declined, the vehicles not subject to a charge increased in trips.  In other words, the congestion charge reduced pollution from the vehicles that it charged, but not the vehicles that it did not charge, in terms of NO2.  Bear in mind that the proportion of diesel cars in the UK vehicle fleet increased during this period (not least because of misguided changes to vehicle excise duty intended to reflect CO2 emissions only).

Technology has progressed since 2003.  30% of London buses are now hybrids, with 0.8% now zero-emission electric and fuel-cell vehicles. So a similar increase in bus services today would mean a lower increase in emissions, particularly if priority was given to putting new hybrid vehicles on central city routes.  However, although increasing the capacity and frequency of bus services is a logical complementary policy to congestion charging, and may be desirable (depending on the design of the charging scheme itself) it is not absolutely essential and certainly today, with the engine technology available (notwithstanding cities with trolley buses who can always claim zero emissions from such vehicles), it does not necessarily follow that more buses means higher NO2 emissions.   I suspect that had London not increased bus capacity (by increased frequencies, but rather taken the gains from reduced congestion to improve trip reliability), although the congestion charge itself may have been marginally less effective, NO2 would not have increased as was reported.  

For black cabs it is simpler, as there isn't a compelling policy reason why people who choose to travel by taxi into or around central London should be preferred over those who bring their own car.   Bear in mind that black cabs circulate "empty" searching for fares, contributing to both congestion and pollution, whereas a private vehicle always undertakes a "useful trip" in transport utility terms and when it isn't, it is not polluting or contributing to congestion.  Stockholm, Gothenburg and Singapore do not make any taxis exempt from their urban congestion charging schemes. Transport for London has been consulting on removing the exemption for minicabs (but not black cabs).   To be fair, had black cabs not been made exempt, they would have opposed the congestion charge and been a problem for the Mayor in introducing the charge.  However, if the exemption were removed, it is not clear how much of a difference it would make with the current charge structure.

The congestion charge is set at £10.50 (US$14.93) for unlimited trips within the zone, for vehicles registered with Autopay (which means vehicles are automatically charged when detected entering or circulating within the zone, rather than having to consciously pay for a trip each day in a separate transaction).  It seems likely that most black cabs (and minicabs) would absorb this in their fares and spread that cost, although there will probably be a small incremental reduction in vehicles.


The congestion charge did NOT increase pollution, in fact the report quoted states that with the exception of NO2, other pollutants reduced.  However, given the information presented, it is not the congestion charge that is to blame. It is the increase in bus services, approved by the Mayor of London (using the engine technology of the time) and the increase in black cab trips (exempt from the congestion charge) that appears to be the evidence for a causative relationship in increasing NO2 pollution being attributed to the congestion charge.

It would be highly misleading to draw the conclusion that introducing congestion charging today or expanding it would have the same result.  Hybrid or electric buses can mean bus services can be increased without the same result, and taxis (both kinds) need not be exempt at all (they aren't in other jurisdictions).

* The press release stated the City of London introduced the congestion charge, but it was Transport for London (as the area charge crosses the boundaries of several other boroughs, most notably Westminster, but also Camden, Southwark, Lambeth, Islington, Tower Hamlets and Hackney).

** For those unfamiliar with the split in the taxi industry in the UK: Minicabs (legally "private hire vehicles") tend to offer fixed fares, are not metered and are only allowed to pick up passengers on pre-booked trips (which may be by phone, online or by asking at a minicab office, in which case a minicab may be available immediately). Black cabs (legally "licensed hackney carriages") are metered and are legally entitled to pick up "hailing" passengers (people who flag down a taxi in the street or request a trip from a taxi rank).  Black cabs can also offer fixed fares to passengers and be pre-booked, just like minicabs, but the licencing requirements for black cabs are stricter.  Both types of taxi are licensed by local authorities (but in London by Transport for London).