Background
On 8 February, the HK government launched a 3-month public consultation on three proposals to improve traffic distribution among Hong Kong's three harbour tunnels. This follows the publication of a consultancy report of the study into the problem, and its recommendations.
The consultants recommended considering three options, of which:
Option 1: the fee charged on private cars will fall by $5 for the Eastern Harbour Crossing, and rise by $5 for the Cross-Harbour Tunnel. A Cross-Harbour Tunnel traffic reduction of 40% is expected.
Option 2: reducing the Eastern Harbour Crossing private car toll by $5, and increasing the Cross-Harbour Tunnel private car toll by $5, and tolls for other vehicles at both tunnels will be proportionally adjusted. A Cross-Harbour Tunnel traffic reduction of 30% of the queue at the Cross-Harbour Tunnel during rush hours is expected.
Option 3: the toll for private cars at the Cross-Harbour Tunnel will rise by $10, while that at the Eastern Harbour Crossing will fall by $5. A Cross-Harbour Tunnel traffic reduction of 38% during rush hours is expected.
Current toll structure
Problem:
The consultants framed the problem thus:
Cross Harbour Tunnel (CHT) is the most heavily utilised crossing amongst the three RHCs, and congestion at CHT is a long standing problem. CHT has a clear natural advantage given its central location and connectivity. CHT enjoys the best connecting road systems compared to the other two RHCs and hence higher practical capacity levels. This advantage is reinforced by the significantly lower tolls that apply to CHT over the years
The government stated its aim thus:
We need to adopt measures to divert traffic from CHT to the other RHCs, thereby improving the traffic distribution among the three RHCs and alleviating the traffic congestion at CHT.
The WHT and EHT are in private hands, being run on a BOT model, whilst the CHT is now controlled by the government after the expiry of a 30-year contract in 2002. It is already a stated fact that all three tunnels are at full capacity. At certain times of the day, the traffic through the feeder roads leading to the tunnels are log-jammed.
The government appears to have ruled out buying back the WHT and EHT franchises as a solution to the problem.
Comment on the consultancy report:
The congestion has been used as a strong argument for building the Central-Wanchai Bypass. However, empirical evidence from all around the world has shown that new roads, instead of alleviating traffic congestion, usually create more traffic, and more bottlenecks elsewhere. Thus, as the bypass opens, we can expect greater arbitrage of tunnel usage depending on the price differentials. The government's argument is akin to playing the game of "whac-a-mole". This will not change so long as the CHT charges below the market rate.
The consultants correctly posited that lowering tolls would not work. When demand is in excess of capacity, any price adjustment needs to be upwards to bring demand back to equilibrium. Demand for all three tunnels is already well above equilibrium, the tolls of the WHT and EHT are governed more or less by market forces, and there is no reason to meddle with them. Each time operators apply for a toll increase in accordance with the terms of their agreement, there is huge public uproar. The problem, is in fact not because the WHT or EHT tolls are high or their increases exorbitant, but that CHT tolls are artificially low and have not kept up with inflation. Note that the WHT tolls are below what they are legally permitted to charge.
Also correct is recognition that the traffic arbitrage between the CHT and the EHT can and should be managed by changing the cost differential between them. However, all three options of the proposal involve reduction in the EHT toll (through user subsidy). The proposed subsidy is neither necessary nor desirable as it meddles with the market mechanism, and transfers benefits to a private operator with uncertain benefits. This is the economic equivalent of trying to make a 'right' with two wrongs. Lowering the toll of the EHT will encourage more marginal traffic (traffic that doesn't already use the tunnel for economic reasons) to use the tunnel, further pressuring the capacity of the EHT. Another reason for caution is that the government has only limited power to influence the rates charged by private tunnel operators; even if there were a user's subsidy, and any 'loss' of the EHT operator would be difficult to quantify. As the government cannot bar the operator from increasing its tolls, thus potentially nullifying or attenuating the effects of any reduction, the subsidy is rather nonsensical.
As private cars represent a relatively large and stable percentage (40 to 60 percent) of cross tunnel traffic, it would certainly make sense to target the tariff for private vehicles. However, that does not mean the private car tariff should be looked at in isolation. All tunnel users are affected by congestion as it is not necessarily in society's interest to favour use of the under-priced CHT by heavier and more polluting vehicles.
There is only one option (option 2) suggesting "proportionate rises" for other vehicles. This is in my view superior to options 1 and 3 because the management is more holistic. But it does not go far enough. As some vehicles' tolls are in my opinion considerably out of line according to the resource management principle (see argumentation below), this would create distortions that would not exist otherwise if the entire fare structure is overhauled. In my view, the price changes would need to allow the CHT to further adjust tolls for operational flexibility, and in case the projection of elasticity of demand for the tunnel proves wildly incorrect.
The consultancy report states that a peak hour surcharge was "considered ineffective", yet no evidence is offered to the effect, or why. As there was no mention of what quantum increases were hypothesised, I have to assume it was the principle that was ruled out.
The report continues: "Restrict the use of CHT or increase CHT tolls as a standalone measure may not be considered appropriate or acceptable by the community." Price rises are never popular. The CHT toll adjustments must not be seen as primarily for revenue-generation but for road resource management. I believe the general public recognises that there are cost implications with peak- vs off- season travel, and that the same principle should apply to tunnel usage.
I'm glad the buy-back scenario seems to have been ruled out, although it was for the wrong reasons: the basic premise was that the situation will be more acceptable to the public if all the tunnels were in public ownership. I am surprised that the consultants' report pinned so many hopes of improvement on the opening of the Central-Wanchai bypass, and on the expiration of the two BOT contracts in 2016 and 2017 respectively. Traffic will increase, and logjams will result without further toll increases when the bypass opens. As the government is weak and lacking in political will to take hard decisions, it is tempting to present soft options in the first instance to avoid having to backtrack later. Up to now, evidence of the government management of the CHT tolls would suggest public ownership of all three tunnels is likely to lead to more resource misallocation rather than less. There will be more congestion, and everywhere, instead of merely at the entrance to the CHT.
I am disappointed the consultants did not conduct pilot schemes to evaluate the demand elasticity of tunnel usage. Not only did they fail to consider a solution based on free-market pricing, they too briefly mentioned the impact of dynamic pricing but failed to quantify. The consultants proposed to tweak an already failed interventionist pricing model instead of proposing any real workable long-term solution.
Principles that need to be observed:
The CHT, with its convenient central location, ought to command a premium; yet because it is the sole tunnel under direct government control, it's fares are the lowest. In economic terms, the CHT and WHT are close substitutes. The EHT is arguably the least convenient. In economics, it could be termed an 'inferior good', and will command a corresponding discount (compared with the two 'superior' goods).
The Hung Hom tunnel is priced significantly below the private tunnels. Except for HGVs, CHT tolls have remained unchanged since 1984. However, based on government statistics, the general RPI increase between 1981 and 2013 averaged 4.6 percent. Taking 1984 as the baseline, the crossing tariffs at the CHT in current HK dollars (i.e. adjusted for inflation) ought to be in excess of $36 for private vehicles and taxis, $55 for light lorries and vans, and $90 for heavy goods vehicles. This would be above the level currently charged by the EHT. The underpricing of CHT tolls amounts to a public subsidy to vehicle owners for each trip taken in the CHT, and a loss of revenue for the public coffers. Overall, arbitrary low pricing seriously contributes to the misallocation of resources on Hong Kong's roadways, and is a direct cause of the congestion now being faced.
The number of vehicles registered in the territory is now at a record level of half a million despite the rise in fuel and car ownership costs. The solution for reduction in urban traffic needs to be both urgent and radical. Since the government assumed ownership of the CHT, its pricing structure is now without merit. It is completely illogical and needs to be re-formulated on a sound economic basis.
Although option 2 is the option I consider most likely to achieve the object of redistributing traffic, I would strenuously object to the subsidy component in the proposals. Until such a time when the CHT tolls are resource-neutral and above those of one of the other tunnels, the desired reduction in traffic will be unpredictable.
Instead of marginally adjusting the pricing, the CHT operator – a government department – should immediately be freed to increase its prices according to market forces with some provisos and some exemptions for public transport. If freed to set its own fares free from political influence, the CHT operator will almost certainly opt to maximise profits. Tolls will be priced according to the RMP, and the CHT will charge tolls comfortably above those currently practised by the EHT.
For the effective alleviation of congestion without adversely impacting the traffic at the WHT, CHT prices must to rise to at least the level of EHT; tolls should however be held below the level of the WHT. Such readjustment will automatically force some CHT traffic to the EHT in preference to the WHT. The latter's tolls are still some 80 percent higher than the EHT; because the EHT will see more traffic from a rise at the CHT, the quantum subsidy will probably be small, if any.
Possible provisos or safeguards could include capping of CHT tolls at the level mid-way between the EHT and the WHT. In the public interest, the government could consider a toll-waiver for empty returning buses, for example. Combined with a strong rise in the toll for the private vehicle, toll exemptions or reductions for school buses could be additional incentive for parents not to drive their children to school.
Bearing in mind the price differential between the WHT and EHT, and assuming the WHT will not further increase its fares, any upward revision of the CHT tariffs to a level between that of the WHT and EHT is likely to have a negligible effect on the traffic through the WHT and its feeder roads. At present, the WHT only charges one-third of the gazetted toll and it therefore has the option of raising tolls to maximise its revenues should CHT tolls increase.
As a final tweak to peak-hour traffic, the Transport Department needs to carefully examine the possibility of dynamic road pricing for the CHT (within a band of 50 to 100 percent of 'basic tarif') according to peak-traffic demand, from 7.30 to 9.30am, and from 6 to 9pm. This will time-shift non-mission critical trips to earlier or later periods.Time-shifting of HGV traffic, achieved by discounts for off-peak travel, say, from 11pm to 6am, would create health benefits through lower day-time pollution.
Further reading:
2011 annual traffic census, Transport Department, Hong Kong
"Tunnel vision", Time Out, 15 Jan 2013
On 8 February, the HK government launched a 3-month public consultation on three proposals to improve traffic distribution among Hong Kong's three harbour tunnels. This follows the publication of a consultancy report of the study into the problem, and its recommendations.
The consultants recommended considering three options, of which:
Option 1: the fee charged on private cars will fall by $5 for the Eastern Harbour Crossing, and rise by $5 for the Cross-Harbour Tunnel. A Cross-Harbour Tunnel traffic reduction of 40% is expected.
Option 2: reducing the Eastern Harbour Crossing private car toll by $5, and increasing the Cross-Harbour Tunnel private car toll by $5, and tolls for other vehicles at both tunnels will be proportionally adjusted. A Cross-Harbour Tunnel traffic reduction of 30% of the queue at the Cross-Harbour Tunnel during rush hours is expected.
Option 3: the toll for private cars at the Cross-Harbour Tunnel will rise by $10, while that at the Eastern Harbour Crossing will fall by $5. A Cross-Harbour Tunnel traffic reduction of 38% during rush hours is expected.
Current toll structure
| Vehicle type | Cross Harbour | Eastern Harbour | Western Harbour |
| Tunnel | Crossing | Crossing | |
| Motor cycles, motor tricycles | $8 | $13 | $25+ |
| Private cars, electrically powered passenger vehicle* | $20 | $25 | $55+ |
| Taxis | $10 | $25 | $50+ |
| Public Light Buses | $10 | $38 | $65+ |
| Private light buses | $10 | $38 | $65+ |
| Light goods vehicles, special purpose vehicle* of a permitted gross vehicle weight not exceeding 5.5 tonnes | $15 | $38 | $65+ |
| Medium goods vehicles, special purpose vehicle (other than an articulated vehicle)* of a permitted gross vehicle weight exceeding 5.5 tonnes but not exceeding 24 tonnes | $20 | $50 | $90+ |
| Heavy goods vehicles, special purpose vehicle (other than an articulated vehicle)* of a permitted gross vehicle weight exceeding 24 tonnes | $30 | $75 | $120+ |
| Public and private single-decked buses | $10 | $50 | $100+ |
| Public and private double-decked buses | $15 | $75 | $140+ |
| Each additional axle in excess of two | $10 | $25 | $30+ |
Problem:
The consultants framed the problem thus:
Cross Harbour Tunnel (CHT) is the most heavily utilised crossing amongst the three RHCs, and congestion at CHT is a long standing problem. CHT has a clear natural advantage given its central location and connectivity. CHT enjoys the best connecting road systems compared to the other two RHCs and hence higher practical capacity levels. This advantage is reinforced by the significantly lower tolls that apply to CHT over the years
The government stated its aim thus:
We need to adopt measures to divert traffic from CHT to the other RHCs, thereby improving the traffic distribution among the three RHCs and alleviating the traffic congestion at CHT.
The WHT and EHT are in private hands, being run on a BOT model, whilst the CHT is now controlled by the government after the expiry of a 30-year contract in 2002. It is already a stated fact that all three tunnels are at full capacity. At certain times of the day, the traffic through the feeder roads leading to the tunnels are log-jammed.
The government appears to have ruled out buying back the WHT and EHT franchises as a solution to the problem.
Comment on the consultancy report:
The congestion has been used as a strong argument for building the Central-Wanchai Bypass. However, empirical evidence from all around the world has shown that new roads, instead of alleviating traffic congestion, usually create more traffic, and more bottlenecks elsewhere. Thus, as the bypass opens, we can expect greater arbitrage of tunnel usage depending on the price differentials. The government's argument is akin to playing the game of "whac-a-mole". This will not change so long as the CHT charges below the market rate.
The consultants correctly posited that lowering tolls would not work. When demand is in excess of capacity, any price adjustment needs to be upwards to bring demand back to equilibrium. Demand for all three tunnels is already well above equilibrium, the tolls of the WHT and EHT are governed more or less by market forces, and there is no reason to meddle with them. Each time operators apply for a toll increase in accordance with the terms of their agreement, there is huge public uproar. The problem, is in fact not because the WHT or EHT tolls are high or their increases exorbitant, but that CHT tolls are artificially low and have not kept up with inflation. Note that the WHT tolls are below what they are legally permitted to charge.
Also correct is recognition that the traffic arbitrage between the CHT and the EHT can and should be managed by changing the cost differential between them. However, all three options of the proposal involve reduction in the EHT toll (through user subsidy). The proposed subsidy is neither necessary nor desirable as it meddles with the market mechanism, and transfers benefits to a private operator with uncertain benefits. This is the economic equivalent of trying to make a 'right' with two wrongs. Lowering the toll of the EHT will encourage more marginal traffic (traffic that doesn't already use the tunnel for economic reasons) to use the tunnel, further pressuring the capacity of the EHT. Another reason for caution is that the government has only limited power to influence the rates charged by private tunnel operators; even if there were a user's subsidy, and any 'loss' of the EHT operator would be difficult to quantify. As the government cannot bar the operator from increasing its tolls, thus potentially nullifying or attenuating the effects of any reduction, the subsidy is rather nonsensical.
As private cars represent a relatively large and stable percentage (40 to 60 percent) of cross tunnel traffic, it would certainly make sense to target the tariff for private vehicles. However, that does not mean the private car tariff should be looked at in isolation. All tunnel users are affected by congestion as it is not necessarily in society's interest to favour use of the under-priced CHT by heavier and more polluting vehicles.
There is only one option (option 2) suggesting "proportionate rises" for other vehicles. This is in my view superior to options 1 and 3 because the management is more holistic. But it does not go far enough. As some vehicles' tolls are in my opinion considerably out of line according to the resource management principle (see argumentation below), this would create distortions that would not exist otherwise if the entire fare structure is overhauled. In my view, the price changes would need to allow the CHT to further adjust tolls for operational flexibility, and in case the projection of elasticity of demand for the tunnel proves wildly incorrect.
The consultancy report states that a peak hour surcharge was "considered ineffective", yet no evidence is offered to the effect, or why. As there was no mention of what quantum increases were hypothesised, I have to assume it was the principle that was ruled out.
The report continues: "Restrict the use of CHT or increase CHT tolls as a standalone measure may not be considered appropriate or acceptable by the community." Price rises are never popular. The CHT toll adjustments must not be seen as primarily for revenue-generation but for road resource management. I believe the general public recognises that there are cost implications with peak- vs off- season travel, and that the same principle should apply to tunnel usage.
I'm glad the buy-back scenario seems to have been ruled out, although it was for the wrong reasons: the basic premise was that the situation will be more acceptable to the public if all the tunnels were in public ownership. I am surprised that the consultants' report pinned so many hopes of improvement on the opening of the Central-Wanchai bypass, and on the expiration of the two BOT contracts in 2016 and 2017 respectively. Traffic will increase, and logjams will result without further toll increases when the bypass opens. As the government is weak and lacking in political will to take hard decisions, it is tempting to present soft options in the first instance to avoid having to backtrack later. Up to now, evidence of the government management of the CHT tolls would suggest public ownership of all three tunnels is likely to lead to more resource misallocation rather than less. There will be more congestion, and everywhere, instead of merely at the entrance to the CHT.
I am disappointed the consultants did not conduct pilot schemes to evaluate the demand elasticity of tunnel usage. Not only did they fail to consider a solution based on free-market pricing, they too briefly mentioned the impact of dynamic pricing but failed to quantify. The consultants proposed to tweak an already failed interventionist pricing model instead of proposing any real workable long-term solution.
Principles that need to be observed:
The CHT, with its convenient central location, ought to command a premium; yet because it is the sole tunnel under direct government control, it's fares are the lowest. In economic terms, the CHT and WHT are close substitutes. The EHT is arguably the least convenient. In economics, it could be termed an 'inferior good', and will command a corresponding discount (compared with the two 'superior' goods).
The Hung Hom tunnel is priced significantly below the private tunnels. Except for HGVs, CHT tolls have remained unchanged since 1984. However, based on government statistics, the general RPI increase between 1981 and 2013 averaged 4.6 percent. Taking 1984 as the baseline, the crossing tariffs at the CHT in current HK dollars (i.e. adjusted for inflation) ought to be in excess of $36 for private vehicles and taxis, $55 for light lorries and vans, and $90 for heavy goods vehicles. This would be above the level currently charged by the EHT. The underpricing of CHT tolls amounts to a public subsidy to vehicle owners for each trip taken in the CHT, and a loss of revenue for the public coffers. Overall, arbitrary low pricing seriously contributes to the misallocation of resources on Hong Kong's roadways, and is a direct cause of the congestion now being faced.
- Price levels: The CHT has not had a price increase since 1999. Under government control, pricing is determined by political considerations; economic considerations seem secondary. Each time the EHT or WHT apply for a toll increase according to the pricing regime, the government seems to further entrench the lower price regime at CHT as a 'social subsidy'; such considerations have resulted in severe road congestion
- Price structure: 'commercial' tunnels seem to reflect economic cost of use – they adopt a "resource management principle" (RMP), where each category of vehicle is charged according to estimated fair wear and tear and maintenance costs. The CHT tariffication is out of line with this principle, leading to an inefficient resource allocation.
- Neutrality/fairness : the resources consumed within each category in use is the same. Individual users drive private cars, and are also consumers of taxi services. I see no economic argument nor objective basis for defending a toll rate for taxis that is different to private cars; similarly, PLBs currently pay the same as private cars, yet consume more resources and ought to pay more.
- Correct identification of user categories: Government figures show that occupancy of cross-harbour taxi journeys is typically higher than private cars, marginally. Taxis by definition have a driver, so it would suggest that most private cars are usually self-driven and carry one or no passenger. Yet private cars represent approximately 40 percent of CHT traffic (and between 50 and 60 percent at peak hours); they carry few occupants, thus control of this parameter is the key to traffic reduction at the CHT. Goods vehicles represent up to 30 percent of day-time CHT traffic volumes, and at least some types of logistics may be susceptible to being time-shifted through peak- / low- hour tariffs.
- Revenue gains: All things being equal, the act of independently increasing the CHT tariff to 'market' levels will increase revenues to the government. A back-of-the-envelope calculation, based on government figures (average daily traffic of 120,000 crossings) and assuming 'market' tolls (quantum equal to the EHT) and unchanged demand, government is under-recovering toll revenues of HK1.5 million per day, or HK$500 million per annum.
- Non-subsidy: Subsidies should only be considered as a last resort, but this does not seem to be a last resort as 'clean' price rises can be envisaged. The subsidy seems to be aimed at containing overall price rises whilst addressing the cost differential. However, the EHT franchisee will benefit indirectly from a CHT toll increase (much more so than the WHT) because some drivers and businesses will arbitrage an increased use of the CHT. A sharp increase in CHT tolls (without corresponding subsidy) should not only reduce the overall traffic volume through it, it should also cause users to migrate to the EHT; the EHT revenues will thereby increase.
- Taxi passengers are private individuals who have a choice of (much cheaper) alternative methods across the tunnel, such as buses and MTR. Government annual traffic census suggest that taxi journeys rarely carry more than 2 passengers – occupancy never reaches 3 people, including driver. This indicates that passengers are relatively well-heeled, yet they benefit from tariff that is half that of private cars (excluding return fare). The return pricing of $10 implies that the return trip does not consume any significant resources or cause wear and tear for the tunnel. However, the taxi driver choosing to return to his place of origin immediately would only pay $10 if travelling by the CHT, compared with $25 by the EHT and $45 by the EHT. Although not all taxi drivers choose to return to their point of origin immediately, those who do will obviously choose the more convenient and cheaper CHT. Given the large numbers of taxis always waiting to cross harbour, I posit that the structure of identical pricing as for private cars will have a significant improvement of congestion.
- As with taxis, public light buses enjoy a $10 toll, yet statistics indicate occupancy is never below 10 persons. The price differential with the EHT and WHT respective is $28 and $55. The PLBs are efficiently used, but the low cost is a hidden subsidy from the taxpayer. PLBs represent 1-2 percent of the CHT traffic and there is a small revenue loss to government coffers.
The number of vehicles registered in the territory is now at a record level of half a million despite the rise in fuel and car ownership costs. The solution for reduction in urban traffic needs to be both urgent and radical. Since the government assumed ownership of the CHT, its pricing structure is now without merit. It is completely illogical and needs to be re-formulated on a sound economic basis.
Although option 2 is the option I consider most likely to achieve the object of redistributing traffic, I would strenuously object to the subsidy component in the proposals. Until such a time when the CHT tolls are resource-neutral and above those of one of the other tunnels, the desired reduction in traffic will be unpredictable.
Instead of marginally adjusting the pricing, the CHT operator – a government department – should immediately be freed to increase its prices according to market forces with some provisos and some exemptions for public transport. If freed to set its own fares free from political influence, the CHT operator will almost certainly opt to maximise profits. Tolls will be priced according to the RMP, and the CHT will charge tolls comfortably above those currently practised by the EHT.
For the effective alleviation of congestion without adversely impacting the traffic at the WHT, CHT prices must to rise to at least the level of EHT; tolls should however be held below the level of the WHT. Such readjustment will automatically force some CHT traffic to the EHT in preference to the WHT. The latter's tolls are still some 80 percent higher than the EHT; because the EHT will see more traffic from a rise at the CHT, the quantum subsidy will probably be small, if any.
Possible provisos or safeguards could include capping of CHT tolls at the level mid-way between the EHT and the WHT. In the public interest, the government could consider a toll-waiver for empty returning buses, for example. Combined with a strong rise in the toll for the private vehicle, toll exemptions or reductions for school buses could be additional incentive for parents not to drive their children to school.
Bearing in mind the price differential between the WHT and EHT, and assuming the WHT will not further increase its fares, any upward revision of the CHT tariffs to a level between that of the WHT and EHT is likely to have a negligible effect on the traffic through the WHT and its feeder roads. At present, the WHT only charges one-third of the gazetted toll and it therefore has the option of raising tolls to maximise its revenues should CHT tolls increase.
As a final tweak to peak-hour traffic, the Transport Department needs to carefully examine the possibility of dynamic road pricing for the CHT (within a band of 50 to 100 percent of 'basic tarif') according to peak-traffic demand, from 7.30 to 9.30am, and from 6 to 9pm. This will time-shift non-mission critical trips to earlier or later periods.Time-shifting of HGV traffic, achieved by discounts for off-peak travel, say, from 11pm to 6am, would create health benefits through lower day-time pollution.
Further reading:
2011 annual traffic census, Transport Department, Hong Kong
"Tunnel vision", Time Out, 15 Jan 2013