"Motorists First" - Findings of the Independent Toll Review for the NSW Government - Part Three: The Recommendations

Given the findings of the Independent Review, and particularly the highly controversial Interim Report (which essentially called for the NSW Government to legislatively override existing toll concession agreements, causing heart attacks at Transurban and among its investors), the recommendations to finally come out of this review are critical. However, equally critical is what, if anything, the NSW Government is going to do in response.

It's worth noting the wealth of data and research compiled in this review, which should help inform discussion and debate about tolling in Sydney for some time.

42 recommendations were made, and I wont repeat them all in detail here. 

However, a key part of the review work was to model the impacts of models of reform that were presented. It's critical to understand that the report recommends "moving towards" the Network Toll Restructure and Reduction model, not necessarily the details of that model exactly, but does not recommend the Network Toll Restructure Model.  Therefore, I will focus on the former.

These models are:

• Network Toll Restructure model: Introduction of standardised network tolls and including application of two-way tolling; and

• Network Toll Restructure and Reduction model: This uses revenues generated from two-way tolling, peak pricing and other sources to reduce tolls where appropriate. A declining distance approach with fixed infrastructure charges is proposed.

The effect of the latter was modelled as meaning:

• 78% of motorists are the same or better off, 17% would pay $3 + more per trip.

• Main losers are those using the Sydney Harbour Crossings

• Western Sydney motorists get some relief as longer trips are reduced in cost

The following table lists the current tolls in Sydney (all in Australian Dollars ~ US$0.66-A$1.00:

As you can see the basis for tolling varies between being two-way or one-way, fixed or distance-based, with rates for different classes of vehicles varying considerably between toll roads, and the basis for adjusting tolls varying as well.

The proposed new structure is as follows:

This is for Class A vehicles only for simplicity in illustration, but would have a consistent toll distance rate, with infrastructure rates that reflect fixed costs for those roads. The declining percentage means that every 4km the per/km distance rate declines 15%.  The effect is to make some shorter journeys more expensive, and almost all longer ones cheaper.

The following table indicates what the modelling of the Tolling Review suggests would be the distribution by trip distance of the "winners" and "losers" of reform. All those travelling longer distances would be better or no worse off, whilst about 40% of shorter (<10 km) toll road trips would be more expensive. 

The difference in average toll for a car would be to reduce from $9.02 today to $5.43, a drop of 40%. The effects on the network are seen in the following map, depicting traffic increases and decreases on the tolled and untolled network. It would increase traffic on the M2, M4, M5 east and south west and M7, as well as River Road, Victoria Road and James Ruse Drive (as traffic either avoids the northbound Harbour Crossing tolls or queues to use the M4 more intensely). 

It's striking that the obvious impacts on reducing traffic are on the Harbour Crossings and Eastern Distributor, as introducing northbound tolls on the Harbour Crossings and southbound on the Eastern Distributor sees some redistribution of traffic to the west, primarily on the untolled crossing at Iron Cove Bridge and Gladesville. There is also reduction on some streets in the CBD and some parallel routes to the M5, as lower tolls make some toll roads more attractive that local streets for some drivers. The M4 and M5 in particular see much higher traffic volumes. It's unclear the impact on congestion overall, as this did not include any peak/off-peak pricing on a network basis.

This table produced with the press release accompanying the Independent Toll Review (PDF) illustrates the effects on some toll trips:

The Toll Restructure and Reduction scenario has significant impacts on all of these examples, notably halving the cost to drive from Campbelltown to the CBD, but more than doubling the price from North Sydney to the airport. 

Before summarising the other recommendations, it is worth going over in some detail the proposed tolling principles (the first set of recommendations).

Tolling principles

These principles are recommended to guide policy measures to reform tolling of existing roads and should inform the implementation of tolling on new roads.

The Tolling Review considered the set of tolling principles agreed in 2014 which were as follows:

1. New tolls are applied only where users receive a direct benefit. 

2. Tolls can continue while they provide broader network benefits or fund ongoing costs. 

3. Distance-based tolling for all new motorways. 

4. Tolls charged for both directions of travel on all motorways. 

5. Tolls charged reflect the cost of delivering the motorway network. 

6. Tolls take account of increases in expenses, income and comparable toll roads. 

7. Tolls will be applied consistently across different motorways, to the extent practicable, taking into account existing concessions and tolls. 

8. Truck tolls at least three times higher than car tolls. 

9. Regulations could be used so trucks use new motorway segments. 

10. Untolled alternative arterial roads remain available for customers. 

The review found that these were rather general and didn’t include some key issues, such as the proportion of costs that should be recovered from tolls relative to taxpayer funds. There was little recognition of the need for tolls to vary by time of day, plus although some of the principles (tolling in both directions) are valid, they were not always applied (see the Harbour Crossings and Eastern Distributor).

The review proposed a set of modified principles with one set about the level and structure of tolls and another on consistency with competition policy.

Proposed New Tolling Principles

On the level and structure of tolls:

• Toll setting should be guided by the objectives of efficiency, fairness, simplicity and transparency. 

• Tolls should have regard to the costs associated with the provision of toll road services as well as benefits. Declining distance-based tolls are consistent with the principle and have efficiency and equity advantages over fixed distance-based tolls or variable zonal distance-based tolls. 

• In general, it is appropriate that beneficiaries pay for toll roads, for example, where benefits flow to the broader community then government contributions are appropriate. The extent of cost recovery achieved through tolls should reflect the extent to which a toll road’s benefits are enjoyed directly by motorists. 

• The process for setting tolls should be transparent to the public to promote understanding and allow for informed comment. 

• The methodology for determining tolls should, so far as possible, be applied consistently across the entire network. 

• Tolls should allow toll road operators to recover their costs incurred in financing the construction of the toll road including an appropriate (i.e. risk adjusted) return, and efficient operating and maintenance costs where relevant. It may be appropriate to apply specific charges to individual parts of the network to allow for cost recovery, for example infrastructure charges to cover the additional costs associated with constructing tunnels or bridges. 

• Tolls should not be set at a level which would allow excessive, monopoly profits, or inefficient cost levels to prevail over time. 

• Maintaining flexibility to adjust tolls over time in response to demand and supply changes is important. 

• Toll setting should take into account fairness as well as efficiency considerations, bearing in mind that other more direct policy approaches may be preferable forms of intervention in relation to fairness. 

• The different vehicle categories for tolls should balance impactor pays (the extent to which vehicles impose costs on the network and other users due to their weight and size set against the costs imposed by such vehicles on ancillary roads) and beneficiary pays considerations (a higher willingness to pay for travel time savings). For example, under this principle setting higher tolls for heavier and larger vehicles is consistent with efficient tolling. 

• The structure of tolls should be simple enough to be readily understood by users and avoid creating perverse incentives for the use of the road network. Inconsistent approaches to the tolling of toll roads can cause distortions to traffic flows. 

• Tolling information should be communicated in real time to inform customer journeys and enable improved decision-making.

On consistency with competition policy:

• Competitive pressure should be harnessed when setting tolls and assessing concessionaire bids (competition for the market) and when regularly reviewing tolls (competition in the market). Bidding for concessions should focus on ensuring tolls are set at competitive levels. 

• Unsolicited proposals for toll road extensions should not be considered in isolation of the possibility of first modifying tolls to better manage traffic flows. 

• Restrictions should not be imposed on the use of any road or public transport in order to enhance the financial viability of a toll road. 

• Tolls should only apply where motorists have reasonable and effective untolled road options, including arterial roads, or public transport alternatives, except where community benefit may necessitate restriction on access to alternatives. 

Other recommendations

Moving to network tolling: The core recommendation is to change the current ad-hoc setting of tolls by individual concession (and the State), to a more coherent and consistent approach. The key recommendation is to have declining distance-based tolls, so that the first two kilometres are charged at a higher rate than the next two and so on.  This is for fairness, but also efficiency to recognise the cost imposed on other users of using toll roads for shorter trips, and disrupting traffic flow.  Network tolling should mean some reductions in tolls, through measures like implementing two-way tolls on one-way toll roads, and more use of peak tolling to lower tolls off-peak. Moving to network tolling should help with steps to phase out or reform toll relief, and how to progress this over time. Other options to lower tolls includes extending toll concessions.

Using pricing to influence demand: Going beyond tolling as an infrastructure cost recovery measure, is to use peak and off-peak pricing, with an initial focus on trialling peak pricing for the freight sector. This is both to reduce congestion at peak times, and to encourage better use of spare capacity off-peak. Included in this recommendation is dynamic pricing, which by the conventional definition is not a good idea in this context (although reviewing peak/off-peak pricing more frequently than annually IS a good idea). 

Updating vehicle classifications and charges: Having uniform classifiers and consistent multipliers for heavy vehicles are the key recommendations, along with exempting public bus services from all (not just some) toll roads.

Expanding toll coverage: Applying two-way tolling on the Sydney Harbour Crossings and Eastern Distributor is the obvious step (and one that has generated understandable controversy in isolation). More strategically, the review recommended evaluating the entire motorway network to see if untolled sections should be tolled (reducing tolls on other sections) or if tolls should be removed from some sections. It seems likely that this will be difficult to sell politically.

Initial assessment of toll reforms: Implementation of the reforms should be carefully monitored with frequent modelling to ensure results meet policy objectives.

NSW Motorways: The review recommended establishing a new entity called NSW Motorways, intended to strengthen governance and accountability over NSW toll roads in order to improve outcomes and transparency for motorists. It would work with concessionaires to set network tolls and adjust them working with concessionaires. It would take over the E-Toll retailing business of Transport for New South Wales and have a focus on innovating to improve the tolling experience in the state. It could also manage future toll roads and contract managers for those toll roads, and bring existing public toll roads within its operations.

Concessionaire negotiations: The Review recommends that the Government negotiate with concessionaires to implement network tolling by the end of 2024 and if not achieved, use legislation to advance it. This raises obvious concerns about legislating over the contracts the state has with concessionaires.

Independent oversight of toll setting: The Independent Pricing and Regulatory Tribunal (which is already price regulator for water, public transport and local government services) should also have oversight for toll rate setting. It should work with NSW Motorways and Transport for New South Wales to monitor prices, including the financial and traffic impacts of network tolling, toll relief schemes, the need for and operation of time-of-day pricing and concessionaire performance. 

Legislative package for toll setting: Essentially a recommendation to legislate over concession agreements if necessary to implement network tolls. This should include a Revenue Adjustment Mechanism so revenues can be “appropriately” shared.

Competition measures: These recommendations seek more competition in future concessions and a long-term view on competition with procurement of future toll roads. Concession periods should be set based on public interest considerations, including competition. Competitive tendering should be favoured over unsolicited proposals. Roaming fees (across retail toll providers) should be regulated.

Transparency for motorists: Motorists should be able to see past and projected future toll road spending.  More information should be provided for trip-planning online and via apps, as well as better signage to inform motorists of toll road prices before they make a decision on whether or not to use a toll road.

Tolling customer advocate: NSW Motorways should have a tolling customer advocate function to consider and manage customers complaints, influence improvements to systems, processes and legislation to minimise future complaints and improve compliance. It should manage awareness and education campaigns, address new “pain points” from the transition to network tolling, and publish reports on the implementation of toll reform. If a toll debt is disputed, debt recovery action should be suspended while the dispute is being addressed.

Industry ombudsman: Proposes that NSW, Victoria and Queensland require toll operators to belong to a statutorily approved independent dispute resolution scheme.

Toll notices: These should be simplified and modernised, calling them “invoices” and removing administration notices, but adding late payment fees to incentivise early payment. Information provided should be user-centric, informing them of the most common reasons for non-compliance (flat tag battery and number plate not linked to an account) so motorists can address such issues to avoid a repeat of unpaid tolls.

Debt recovery:  Reform criminal enforcement so there is only one offence per trip and clearly identify if it applies to the driver or the registered vehicle. At present debt is owed by the vehicle’s owner, but it may be appropriate for that to be the driver in some cases. For civil debt recovery, find ways to improve the accuracy of contact information for registered vehicle owners. Noting that debt collection agencies seem to be able to find debtors easier that toll road operators. Toll road operators should develop and publish customer charters

My thoughts

This is a weighty report, and a lot of thought has gone into it.  The reforms proposed might be categorised into three areas:

• Rate setting/tolling policy
• Business rules
• Competition
• Governance

The most fundamental part of the review is the recommendation to take a network approach, and to apply a declining distance based tariff with an infrastructure fee layered on top of it for the higher capital cost toll roads. There is merit in taking such an approach, albeit it is obvious the biggest challenge is doing this whilst ensuring concessionaires are not disadvantaged, and consent to the changes. The "sword of Damocles" of regulation may be there, but it is not one the NSW Government will want to enforce, as it is likely to make any future PPPs more expensive (as it would have an impact on investor confidence in contracting with the NSW Government).

Two-way tolls and having consistent vehicle classifications and multipliers all make economic sense, but it will be difficult to convince motorists that pay one-way on the Sydney Harbour Crossings that they should pay in both directions, without getting anything for it. Other than by halving existing tolls (so they are split by direction), which will likely exacerbate AM peak congestion, it seems unlikely that this will be able to be implemented due to public resistance, although if it were focused on managing demand (and moderating tolls for the Western Harbour Crossing) there might be more tolerance for it.

Certainly the distributional impacts of tolls in Sydney fall greatest on those in the West, so it is understandable why there is some emphasis in improving conditions for motorists there. I note that significant cutting tolls from Campbelltown to the CBD, a route which has a frequent commuter rail service, might have negative impacts on congestion if the modelling doesn't take into account the risk of modal shift from rail to driving, although the cost and availability of parking is a significant deterrent.

The biggest challenge is going to be getting agreement from Transurban to advance these proposals. This is only going to happen if it can be convinced it will be no worse off, not just today, but over the duration of each concession, because each concession has investors (Transurban does not own 100% of all of them) expecting consistent returns. The willingness to do this is likely to be limited, as it requires forecasting changes in demand for several decades out.

The proposal to enable peak/off-peak tolling is likely to have the greatest impact on congestion on the one-hand, and underutilised capacity on the other, noting that for concessionaires, underutilisation is not a problem but rather maximising yields. If there is a public policy reason to reduce tolls off-peak on some roads, to remove traffic from other roads, this may justify a subsidy, or better yet, justify peak pricing to offset it. 

What all of this suggests is that the proposal to change governance, by creating NSW Motorways, to undertake the analysis and modelling needed to advance negotiations with Transurban, will be important. Assuming the NSW Government is not willing to regulate over concessions, it will need to be able to model the impacts of a range of pricing policy options on each individual concession, and to creatively identify options to ensure that public policy objectives are achievable (reducing congestion, better use of toll roads off-peak) alongside making sure concessionaires are willing, to commercially, to accept changes to their concession agreements.  It is appropriate to set up NSW Motorways in any case, as a road regulator which applies to state toll roads (there are two more being built now on top of the two existing Sydney harbour crossings), and which could be extended to cover a future road user charge...

The Review does allude to the wider issue of how motor vehicles are charges for road use across the network in NSW, and the need for some form of road user charging for EVs. Ultimately, there may be scope for more direct user charging across all roads, but given the Vanderstock decision at the High Court of Australia, that looks likely to be led by the Commonwealth Government. At the very least, the NSW Government should be thinking strategically about tolling in that wider context. It is not that road user charging will replace tolling anytime soon, but if there is to be a shift towards distance based tolling across the board, it should not be inconsistent with applying some form of per kilometre charging for vehicles on all roads. 

On the business rules side, the proposals around debt recovery, transparency for motorists and an industry ombudsman are all good from a consumer protection point of view.  None of this should be particularly controversial.

Given the role of Allan Fels it should not surprise anyone that competition has been a focus of this review. The dominance of Transurban should give cause to seek to diversify the profile of future concessions, but the retention of retail competition is also important. Bear in mind the main competition for toll roads are the untolled roads (and for a small subset of users, public transport on some corridors), and although it is flawed, toll roads do have a form of price control over price increases (albeit it effectively means prices increase by inflation).  However, competition can never really be addressed whilst other roads are priced so indirectly, through fuel tax and fixed charges like motor vehicle registration fees. Perhaps the most effective way of enabling competition for future toll roads is either for such roads to be state owned and concessions issued for operations, or for future concessions to have tolls set by independent regulation.

Finally, although the political will is hardly likely to exist for it, there is likely to be sense in at least considering implementing congestion pricing in the form of a CBD cordon in Sydney in parallel with such changes. Such a cordon could be used to moderate tolls as well as better manage congestion on traffic towards the CBD, but that was outside the scope of this review.

The response

The Government response so far is through this press release, which is not really a response as of yet. According to The Guardian, Transurban has said it wants to take a corridor based approach and does not approve of the full network approach.  Roads Minister John Graham also suggested that taxpayers might pay concessionaires to implement some of the recommendations, which is a good idea, if it results in net benefits to consumers and the economy (noting that it could reduce the cost of existing toll relief schemes if tolls can be reduced for some customers).

The full response will not be clear for perhaps a few months, and it seems unlikely that all recommendations will be accepted. However, there is a strong case for more consistency in tolls across Sydney, and despite the unpopularity of two-way tolls for the harbour and peak tolling, the merits of being able to spread demand more efficiency are likely to be high.

What needs to be behind any reforms are consistent principles and objectives. Discouraging short trips on toll roads is likely to result in more efficient use of the network, declining distance based tolls makes sense up to a point, but the merits of high toll costs for long distance travel come from the signals they send for land use and modal choice. 

There have been enough toll reviews in recent years, as my first post on this topic showed (and I was involved in one of them myself). I sincerely hope the NSW Government acts on much of what this one recommends.

"Motorists First" - Findings of the Independent Toll Review for the NSW Government - Part Two: The Findings

Following on from my previous post, this is a listing of the 16 findings of the review. Not the recommendations, but the findings. I have included some of my own comment on these at the end of each finding. Generally the findings are fair, although I think some of them are repetitive and essentially different sides of the same point. The findings have a strong consumer interest element to them, which is unsurprising given it was led by Allan Fels, but there is also some discussion around public policy implications and a bit around markets and delivery of services. Again it reads a bit like an ACCC series of findings, unsurprisingly.

For me, the main points are the lack of coherence around toll rate setting and structures, the inflexibility to apply time-of-use based pricing to better manage congestion and demand, and the poor policy responses to the current structures.  The dominance of Transurban is valid in the toll concession process, but with the presence of E-toll, its retail market share is not monopolistic. Future envisaged toll roads are not intended to be undertaken as PPP concessions, indicating a willingness to take a different approach, although it should be possible to proceed with PPPs without the restrictions and constraints (including the toll rate escalators) implemented in previous years.  Following this article will be one on the recommendations and what I think of those.  However, for those outside NSW, the main benefit of this report is on lessons to apply elsewhere around toll rate setting, PPP contracts and taking a strategic network view, rather than an ad-hoc approach to separate major projects. 

The structure of the findings is a summary of the findings from the report, followed by my brief comment.

The findings

1: The process for setting tolls has been flawed: Largely because governments determined them in advance of PPP concessions, rather than using competition in procurement to incentivise bidders to propose the lowest tolls needed to fund the roads. Long concession periods and higher than inflation cost escalators mean tolls in early years are lower than they should be, as the cost of the infrastructure is pushed towards future users more than early users.  Efficient in road and toll operations almost entirely benefits owners of concessions and is not reflected in lower tolls. Comment: Ideally tolls should be proposed by project bidders or proponents and be subject to competitive pressure, and rigorous public sector scrutiny. It is worth reviewing the merits of allowing tolls to increase above CPI if costs do not do so, but not there is also no scope for tolls to reflect actual demand. Rigid concession conditions around tolls affect the ability for future tolls to be able to address distortions in pricing between tolled and untolled roads, and changes in demand across the network.

2: PPP details relating to toll setting are not publicly disclosed reducing information available to assist in public understanding: Commercial confidentiality claims around PPP agreements limit this information, and consequently increase public disquiet about toll rate setting. The Review noted that Base Case Financial Models are confidential and commercially sensitive, but said returns from PPPs are “generous”. The Review cannot publish the differences between actual revenue and model forecasts because of this confidentiality, making it difficult to assess whether tolls set are too high and whether excessive profits are being generated from toll concessions. Comment: Future concessions should enable regulatory oversight of the differences between actual and forecast revenue. A careful balance is needed between incentivising PPPs sufficiently and not enabling rent-seeking behaviour.

3: Toll road users bear a disproportionately high proportion of the cost of toll roads: The key issue is when toll roads bypass the untolled network and generate significant local amenity benefits. The Review noted the Cross City Tunnel (which provides a bypass of inner Sydney between east and west) which brings significant benefits to surface traffic, including property owners and pedestrians, but was expected to be fully funded by the users of the tunnel. There is a case for those others benefiting from the project to contribute towards its costs. Comment: Toll roads offering significant local amenity improvement, due to removal of traffic and enhancing of property values ought to be partially supported by revenue generated from surface traffic (through network charges such as fuel taxes) and property taxation from property owners. It is clear the Cross City Tunnel in Sydney is underutilised due to its high toll structure.

4: There is no overall system of tolls: Tolls are all set in isolation of each other, and although they could be set to send price signals to optimise the use of road infrastructure they are not designed to do so. The complexity of tolls as they are, including toll relief schemes, untolled motorway sections (which are often used by many motorists paying tolls on other sections).  Comment: From a network perspective, tolls in Sydney send inefficient price signals that distort behaviour and do not encourage efficient network use. For example, overnight toll prices are far too high and ought to be set to remove traffic from surface streets whilst peak period tolls are often too low, and should be priced to encourage time and modal shift. There are no effective means to enable this.

5: The lack of a unified tolling system creates complexity, inefficiency, inequities and unfairness: With different vehicle classification systems and toll regimes, similar trips are priced differently across the network. Roads with similar levels of service are priced differently. Smaller trucks are in some cases charged the same as larger trucks, discouraging them from using some toll roads. Comment: As above, there should be more efficient pricing applied by location, distance and time-of-day and vehicle class. More standard pricing across the network, unless particularly costly parts of infrastructure are being used, would be rational and efficient.

6: Tolls are too rigid and locked-in for decades without options for review:  No other sector of the economy sets prices for such a long period, certainly no other transport mode. This increases perceptions of unfairness over time, as prices rise faster than inflation. The Review reports modelling that around A$123 billion in tolls will be paid between 2024 and 2060. With no processes or means to review tolls during those concession periods, it raises serious questions as to why it is justifiable to have prices set for well over a generation through contract between the private sector and state government. Comment: Concessionaires like guaranteed toll levels and escalations, but no other investments in the private sector guarantee such revenues without regulatory oversight (see energy and water utilities which are subject to such oversight). This suggests that future PPPs have provision for regulatory oversight of pricing at regular intervals.

7: On most toll roads, time-of-day tolling is not used:  At off-peak periods many toll roads are heavily underutilised, and at peak periods several can be highly congested. Pricing should enable better utilisation of the infrastructure. Comment: Generally, there are wider economic benefits in enabling better use of tolled infrastructure, especially since most of it has natural monopoly characteristics and there are some amenity benefits in enabling it.  However, there is limited elasticity of demand off-peak, in that lower prices will result in lower revenues (as additional traffic is unlikely to offset reduced prices), although at peak times higher tolls that reflect demand profiles should improve congestion on a network basis and encourage modal shift. There are considerable merits in enabling time-of-day pricing, subject to regulation, in ways that do not undermine concession net revenues, but significant improve outcomes for the transport network. 

8: The financial impact of tolls is greatest in Western Sydney: Western Sydney suburbs have the highest proportion of motorists paying over A$60 a week on tolls, reflecting the extent of tolled infrastructure in the West and the lack of useful alternative routes. This arguably affects access to employment and other opportunities for residents in those suburbs. 

9: Transurban’s profitability has not been excessive in recent years, but its NSW toll road portfolio profitability is likely to increase over time in line with traffic and toll rate escalation, and declining construction costs: Sydney generates 50% of the toll revenue for Transurban, but its returns are not excessive when considered against the Weighted Cost of Capital. However, it is expected that profitability will row in future years. Comment: This is critically important, as it is important to ensure that Transurban isn’t extracting excessive rents from Sydney road users. However, it also suggests that the toll rate setting system for future concessions should not enable continued increases above inflation.

10: The level of tolls appears to be higher than necessary and desirable: This is in part, counting earlier points as follows. There was no competitive bidding for PPPs on the basis of toll price, concession agreements allow relatively high returns for multiple reasons including a regulated monopoly price safe from competitive challenge, incentives for efficiency are largely captured by concessionaires (and not shared with users). Toll roads are relatively free-flowing and potentially underutilised (indicating tolls are certain times are too high) and motorists perceive tolls as too high. Most of those surveyed who claimed tolls are too high tend to use alternative non-toll routes or reduce frequency of non-essential travel. 15% use other modes, but nearly 40% do not change behaviour (but pay the toll). Comment: There is clearly a distortion in travel between tolled roads and untolled roads essentially because of underpricing of untolled roads. Surveying the public about tolls is likely to result in an answer that many people think tolls are too high, but the real evidence is that the tolled network has much less congestion, on average than the untolled network. Many complain about tolls but still pay them, but that does not mean that tolls are not too high, but it does mean that this is overplayed. Toll roads take up land, and are high capital cost assets and arguably it is fair they generate a return on capital (even if this isn’t what explicitly happens with other roads). However, the negative externalities of pricing only part of the network are not insignificant, and there is a strong case for enabling time-of-use pricing.

11: Transurban has a dominant market share in the current provision of toll roads in Sydney:  Although this is clearly the case, there is competition from untolled roads and other modes. Restrictions on Transurban include the limits on toll rate increases and the conditions on maintaining network quality during concession periods. Comment: Transurban has been commercial adept in expanding its presence in the market, but the “market” itself has entirely been driven by the State Government issuing concessions and the conditions it sets for those. The presence of the state account manager adds significant competition in terms of customer service, for “some” services, but concern over Transurban’s dominance is within the control of the State Government for future toll road concessions and in future regulation of them.

12: Transurban has been dominant in the NSW market for acquisition of toll road concession contracts: This is due to factors, such as its experience in bidding, the economies of scale of its existing operations and its access to in-house data on traffic and in modelling.  It’s noted that of the four motorways under construction in Sydney today, two wont be tolled and the other two will be state-owned toll roads. Comment: This is essentially a repeat of the previous finding, and what matters is what impact it has on public finances, motorists and the economy. That hasn’t been explained clearly.

13: The significant position of Transurban in the toll retailer market could adversely affect competition for tolling concessions: Until 2019 there were four toll road retailes, but Transurban acquire two of them. Now it is Linkt (Transurban), E-Toll (State Government) and Eastlink (a toll road in Victoria) that hold the entire market, with Eastlink’s presence essentially only for a handful of vehicles that hold such accounts in Victoria visiting Sydney. Barriers to entry are not seen as significant, and clearly the presence of E-Toll makes a difference to Transurban’s performance in the market. Comment: There is a “could” here significantly diluted by the presence of E-Toll, but there aren’t enormous barriers to market entry and future concessions and toll roads should be open to more innovative solutions in providing retail services. This could include the growing mobile phone based suppliers, but longer term the inevitable implementation of RUC in Australia should see providers of such services also being able to supply toll retail services to their customers (e.g. telematics service providers for heavy vehicles). 

14: Current tolling information fails to adequately enable, inform, and educate motorists thus reducing user empowerment and efficient decision-making: There is no “one-stop” platform for motorists to obtain all tolling information (including available rebates) and undertake trip planning in a way that is easy to use. Signage about toll rates is inadequate to give motorists sufficient time to adjust route choice. Retail toll platforms do not allow motorists to project future toll usage. There is little understanding as to how tolls are calculated, or understanding about toll administrative charges, and what revenues are used for on non-PPP toll roads. There is also insufficient information about the rights and responsibilities of toll road customers. Comment:  This is true, although there is nothing stopping there being such an app or platform to do this, other than the lack of commercial interest in doing so.  Signage should better enable route choice, and even could compare travel times by tolled and untolled road, although this would have to be the responsibility of the public road controlling authorities. 

15: Toll reform is preferable to toll relief: The current toll relief schemes are inadequately targeted and underutilised, in part due to overly complex administration. Toll relief is not financially sustainable given the existing pattern of toll escalation and limitations on the availability of government resources to fund relief:  This is focused on the M5 toll relief scheme which is confined by geography, does not have processes for review. It appears to be politically entrenched and is likely to have significantly affected transport and land use decisions along the M5 corridor. This and other toll relief schemes are blunt and likely to be financially unsustainable, and likely to primarily benefit higher income earners. It would be preferable to reform tolls more widely. Comment: Clearly the current toll relief schemes are inefficient ways to address public concerns about toll rates, and it would be much preferable to phase it out and reform the toll system more widely. 

16: Concessionaires are an unintended beneficiary of the current approach to toll relief. Increased traffic and patronage of toll roads, through induced demand created by toll relief, directly benefits operators by increasing their revenues: By subsidising tolls, toll relief effectively benefits concessionaires by subsidising demand for their facilities. It is not enough to generate funds beyond agreed levels that would require upside sharing with government, but is enough to benefit Transurban.  Comment: This highlights the inefficiency of toll relief as a subsidy from other road users and taxpayers to concessionaires and the beneficiaries of relief.

"Motorists First" - Findings of the Independent Toll Review for the NSW Government - Part One: Background

 This is the first in a multi-part series about the epic toll review.

“Motorists First” (PDF) is the title given to the latest report on tolling in Sydney. Led by Professor Allan Fels. Fels is best known as having been Chairman of the Australian Competition and Consumer Commission between 1995 and 2003. The focus of his career has been on breaking monopolies, and he took this opportunity of leading the Independent Toll Review for the Minns’ Government in New South Wales to try to do the same to Transurban – which has a stake in most of the toll roads in Sydney.

The review was announced in July 2023, three months after Labor won the state election, but it follows a long line of reviews of tolling in New South Wales (by which I mean Sydney as there are no toll road outside the greater Sydney metro region).  The review posted this handy list of the reviews undertaken by multiple NSW state governments. 

Major toll road openings and New South Wales tolling reviews over 21 years

Why so many reviews? Sydney has one of the most extensive toll road networks of any cities globally, although I have yet to see any detailed research to identify whether it has the biggest tolled road network of any city (Santiago, Chile has quite a network).

Most people are aware of the Sydney Harbour Bridge, opened with tolls in 1932 and still tolled, but much of Sydney’s urban motorway network has been funded through tolling and financed through a patchwork of PPPs.  The latest review notes that of 320km of motorways,156km are tolled. Although there are alternative routes, it is slow and inconvenient to drive from the north or south of Sydney towards the airport or city centre without using toll roads. 

Sydney's tolled and untolled motorway network

Untolled sections were mostly built in the 70s and 80s, whereas the tolled sections have been built since then, with more under construction (the Western Harbour Tunnel and Stage 1 of the M6, which eventually will bypass Sydney’s southern suburbs towards Wollongong). 

The 382 page review has a lot of information in it, and so is worth pouring over for those who are interested. Here are some highlights that I found of interest:

• Only around 4% of journeys (using any mode) were made using toll roads, and 7.6-8.8% of car journeys are undertaken using toll roads at least once.

• There are 10 PPP concessions and 2 state government owned toll roads, Transurban has some shareholding in all of the PPPs (ranging from 50-100%).

• 56% of the toll retail market is held by the state operator E-toll, and 44% by Transurban operator Linkt. This suggests that a majority of toll road users prefer having the state as account manager, not the operator of most of the toll roads (only the Sydney Harbour Crossings are not at least partially owned by Transurban) 

• Westconnex has the biggest proportion of toll road traffic and revenue.

• Toll rates on all routes, except Westlink M7 and Westconnex are point charges (M7 and Westconnex have flagfalls plus a per km rate)

• Only the Sydney Harbour crossings have prices that vary by time of day (A$4.27 peak, A$3.20 interpeak and A$2.67 off peak), but there are only tolls in one direction on the crossings (and the Eastern Distributor).

• Toll escalation factors for the PPPs tend to be based on the greater of CPI or 1% per quarter.  

The review includes this handy chart describing the ownership and key suppliers throughout the supply chain for all of Sydney toll roads:

Sydney toll road ownership/supplier distribution

What's the problem?

What’s the problem?

Fundamentally there are public and political concerns that tolls are incoherent and unfair, largely because each toll concession has seen tolls set that reflect the cost of supplying each individual segment of tolled road at the time it was built. This has resulted in a network that isn’t priced like a network, but priced ad-hoc. The result of this has seen a range of interventions by the state government which are arguably also heavily flawed, including the M5 Cashback scheme, which gives refunds to regular users of that toll road, and was set up entirely for political reasons. That scheme alone costs around A$127m per annum. 

What did the Review find?

16 findings were published by the review, which I will summarise in my next post... 

London makes two steps forward with congestion charging and tolls

In the past couple of weeks two significant changes have been announced regarding road pricing in London:

• Removal of the Cleaner Vehicle Discount (more widely known as the EV exemption) from the London Congestion Charge from 25 December 2025;
• Proposed time-of-use based tolls for the Blackwall Tunnels and the new Silvertown Tunnel.

Congestion Charge Cleaner Vehicle Discount abolished

Since 2003, the London Congestion Charge has provided a concession, in the form of a 100% discount, for lower emitting vehicles.  Initially the discount applied to Alternatively Fuelled Vehicles, which initially meant vehicles powered by natural gas or hybrid or electric vehicles.  Over time that discount has been tightened, with the most recent change in 2021 restricting the discount to pure EVs (not hybrids) and hydrogen-fuel cell vehicles. 

It is a 100% discount for such vehicles which register for the discount and are identified through the UK’s motor vehicle register. Owners have been required to pay a £10 annual fee for this registration. Over 112,000 cars and light commercial vehicles are registered for the discount, with another 16,000 private hire vehicles (i.e., Uber and app/phone booked minicabs) also registered. Transport for London (TfL) has provided no data on the estimated impacts, either in terms of traffic or in financial terms as it has not prepared annual impact assessments for 19 years. Given reports from several years ago, it seems likely that abolition of the Cleaner Vehicle Discount should increase the number of vehicles that the Congestion Charge applies to by between 10-20% per day, and so should have an impact on reducing congestion within and approaching the Congestion Charge zone.

Note the latest TfL annual report (PDF year ended September 2023) indicates the operating costs of the London Congestion Charge consume around 37% of gross revenues.

Note also that the entire area of the London Congestion Charge has been an ultra-low emission zone since 2019, applying a £12.50 per day charge (on top of the Congestion Charge) for vehicles that do not meet specific standards. At present that standard is a minimum of Euro 4 for petrol vehicles (generally any vehicles manufactured from 2005) and Euro 6 for light diesel vehicles (generally any vehicles manufactured from 2015).

Blackwall Tunnel and Silvertown Tunnel toll/time-of-use pricing announced

The Blackwall Tunnel is a pair of two-lane, one-way tunnels (one built in 1897 and another in 1967) under the Thames that comprise part of a major arterial highway from north-east to south-east London. The route (A102 and A12) is the eastern portion of the never completed Ringway 1 – an inner London orbital motorway proposed in 1966, but mostly cancelled in 1973.  More details on the route here.

Location of Blackwall Tunnel

The Blackwall Tunnel is the eastern-most fixed road crossing of the Thames within the boundaries of Greater London (the next crossing is the Dartford Crossing 26km to the east, which comprises part of the M25 orbital motorway). The northbound tunnel handles around 49,000 vehicles per day and the southbound around 53,000. It is one of the most heavily congested corridors in Greater London with severe delays all day during weekdays and for many hours in the weekends. The tunnels have never been tolled.

To help relieve congestion and improve network resilience, the Silvertown Tunnel is being built to the east connecting the approaches from the south of the Blackwall Tunnel to the suburb of Silvertown and main arterials towards the east and the centre of London. It is currently under construction and is planned to open in 2025. It will have two-lanes in each direction, albeit one lane in each direction is dedicated to trucks and buses only. It will be the first new road crossing of the Thames since the QE2 Bridge at Dartford Crossing in 1991. 

Location of Silvertown Tunnel

The road has been controversial in some circles, concerned that any road building contributes to congestion and emissions, but given half of the new capacity is for freight and buses, and tolls are to be applied, in part, to manage demand, it seems highly unlikely that the new tunnel will make congestion worse. It is far more likely that the combination of pricing and new capacity will improve conditions for both traffic, but also the environment. More detail on the tunnel here.

The Mayor of London had always proposed that the Silvertown Tunnel be tolled to pay for most of the capital costs of the tunnel, but also that the Blackwall Tunnel be tolled, to ensure that demand between the tunnels was not distorted by having tolls on one crossing. Furthermore, it was expected that applying time-of-use pricing to the Blackwall Tunnel would help relieve the chronic congestion on that route.

The Mayor of London has now proposed a time-of-use toll to apply to all tunnels as follows:

• Peak charges between 0600-1000 northbound weekdays and 1600-1900 southbound weekdays only

• Off-peak charges between 1000-2200 northbound weekdays and 0600-1600 and 1900-2200 southbound weekdays, and 0600-2200 weekends

• No charges between 2200-0600 all days.

The price schedule is below. Autopay applies if vehicle owners register their vehicle number plate with TfL to be automatically detected and have their bank account or credit/debit card deducted for crossing each day.  

Proposed Silvertown and Blackwall Tunnel tolls

Taxis, blue badge vehicle holders (disabled), buses, coaches, minibuses, emergency vehicles, military vehicles, vehicles in disabled tax class and NHS vehicles exempt from vehicle tax and zero-emission/wheelchair accessible private hire vehicles are all exempt.

Low income drivers in east London (in any of 13 boroughs) may register for a 50% discount. This is determined based on being in receipt of one of a range of means-tested benefits such as Jobseekers Allowance.

A £1 discount at off-peak times only applies for registered small businesses and charities located in the Royal Borough of Greenwich, and the Boroughs of Newham and Tower Hamlets.

Furthermore, three new cross-river bus services and the new cycle bus will be free for local residents for the first 12 months, as well as further enhanced public transport services and Docklands Light Railway trips between two station pairs.

The cumulative effect of the new tunnel and the time-of-use tolls is expected to be up to a 20-minute reduction in peak period travel times. Part of this relief is from a 50% increase in capacity, but also expected is some demand management as drivers shift travel to off-peak periods and some demand shifts to the new bus services.

Estimated gross revenue from tolling the tunnels will be £123m per annum, which should fully recover the construction cost of the Silvertown Tunnel in just over 10 years (indicating that there is more than adequate demand for the additional capacity and new connection).  

Conclusion

Removal of the Cleaner Vehicle Discount is a useful step forward for the London Congestion Charge, as there is already an Ultra Low Emission Zone applying across of London which is the tool for emissions. The Cleaner Vehicle Discount undermines the effectiveness of the congestion charge in managing congestion, so removing it should help wind back some of the congestion growth in recent years.

It is rational to apply tolls to both the Silvertown and Blackwall Tunnels, as the Silvertown Tunnel will directly relieve the Blackwall Tunnels, so it is fair for users of the latter to pay for the former. The benefits of reduced congestion will be significant.  Furthermore, it is a welcome leap forward for London to effectively trial time-of-use pricing at peak periods by direction. London needs more congestion charging on existing roads to manage demand at peak periods, and this should be seen as a pilot for implementation on other routes.  Hopefully shoulder periods will be introduced too, to encourage peak spreading.