Printer Friendly

Road pricing in practice.

The urban transport problem is usually perceived as a disequilibrium, at peak times, between the supply of transport infrastructure and the demand for the use of this infrastructure by an increasing population for more journeys. Given the consequent problems of congestion and the environmental constraints acting against any substantial increase in road capacity, the ultimate necessity to restrain the use of the private car is obvious. Policy makers have responded with a range of vehicle restraint measures including petrol taxes, parking controls, traffic management, etc., all of which have contributed to the containment of the problem but are also recognized as short-term palliatives pending the implementation of more radical solutions. Transport economists have generally argued that the only comprehensive long-term solution lies with the introduction of road pricing.(1) This article evaluates Singapore's unique experience in actually employing the price mechanism to ration more effectively the use of roads during periods of peak demand and considers some ramifications of such a policy.(2)


Singapore is actually the smallest and yet most developed country in Southeast Asia. It consists of the island of Singapore itself and some 57 islets within its territorial waters. The main island is about 41.8 km in length and 22.9 km in breadth, has a coastline of 131.5 km, and is 580 |km.sup.2~ in area. The total land area, including the islets, is 620 |km.sup.2~. Urban Singapore comprises approximately 130 |km.sup.2~, of which less than 9 |km.sup.2~ forms the earliest and most densely populated central area. The remaining 450 |km.sup.2~ is essentially rural, although there is little differentiation between urban and rural areas and the whole island is, in many respects, an urban complex. With a population of approximately 3 million, modern Singapore is, therefore, both a city and a state and, simultaneously, a very small and densely populated island.

Since gaining its independence, the Republic has maintained very high economic growth rates and promoted an innovative development program that has resulted in rapid urbanization and industrialization. However, one of the major problems often brought about by such development and affluence is a rapid growth in the number of motor vehicles, and in Singapore the vehicle population has indeed increased from 350,000 in 1980 to about 600,000 today. During the same period, and despite the imposition of onerous fiscal disincentives to discourage car ownership, the number of cars has increased from 150,000 to almost 300,000, representing a ratio of one car to ten persons. Although this is considerably lower than in many developed countries (for example, in Japan the ratio is 1:4, in the United Kingdom 1:3, and in the United States 1:1.7), it has nevertheless created considerable pressure to build more roads and stretched Singapore's scarce land resources to the limit. With 2800 km of paved roads, of which slightly more than 100 km are expressways, the proportion of land set aside for car-related uses is at 10 percent already considerable.

Given increasing car ownership and the limited scope for continuing to provide additional road space, it has always been clear that in Singapore urban congestion would reach intolerable levels in the absence of appropriate remedial policies. At the same time, it has become increasingly necessary to ensure an efficient transport system to service the country's role as a strategic trading and commercial center. In light of these pressures, it has long been evident that an effective solution to the urban transport problem would not be found until a comprehensive set of land-use and transport policies was formulated and a proper assessment made of the respective roles of private and public transport. The result has been the implementation of an integrated strategy which involves, on the one hand, land-use policies aimed at restructuring the urban fabric to ensure a more locationally-efficient and manageable pattern of transport demand, supported by the supply of the necessary transport infracture; and, on the other, traffic policies directed at private vehicle restraint and the promotion of public transport.(3) However, before considering the role and significance of vehicle restraint and, in particular, road pricing, in the overall policy package, it is first worth defining more precisely what congestion is and why it constitutes a problem.


Given the current pricing system for transport, the urban transport problem is simply the result of a supply/demand disequilibrium, at peak times, resulting from the concentration of travel in the morning and evening rush-hours. At given prices, rising demand in the face of relatively inelastic capacity always produces congestion for goods and services and, in this respect, urban transport is no exception. But what exactly do we mean by congestion?

In formal economic analysis, traffic congestion is defined as a level of usage of an existing transportation facility at which the social costs of additional users exceed the private costs incurred by these users. The point is illustrated diagramatically in Figure 1, which shows the social and private costs and benefits associated with the use of a particular stretch of road. The two demand curves represent the horizontal summation of the marginal private benefits of all individuals who want to use the road at two different times of the day, DD representing peak demand and D D off-peak demand. DD is greater than D D not only because more people want to use the road during peak times, but also because peak-hour trips are invariably journey-to-work trips and these are considerably more valuable. MPC and MSC are the respective marginal private cost and marginal social cost curves.

In engineering terms, no congestion exists when the level of usage is less than or equal to E, which can be considered as the design capacity of the road, since cars can maintain a speed consistent with this design capacity. At these flows there is no divergence between private and social costs and E is the "technically efficient" optimum flow. Achieving this traffic level would appear to be a desirable policy objective and is often incorrectly pursued because, in economic terms, it can be shown to represent an under-utilization of resources.

Beyond E, additional entrants to the traffic stream are unable to maintain design speeds and, as average speeds fall, they incur additional costs because of increased operating costs and delays which cause MPC to increase. However, each new vehicle is now also causing some delay and risk to others for which it does not bear the cost, hence the even sharper rise in MSC. In the absence of intervention, demand will settle at OG where the marginal private cost and demand curves intersect, whereas the optimum flow is clearly OF. Beyond OF every additional road user costs other road users (measured by the marginal social cost curve) more than the benefit he himself gains from using the road (measured by the demand curve). The important point to note is that at OF there will still be some delays. However, economic theory tells us that this is the level where resources are allocated most efficiently since the marginal social cost the user imposes equals the benefit he obtains.

The above analysis highlights the necessity to ration more effectively the use of roads during periods of peak demand, and several methods are available to address this issue. In a competitive situation the price mechanism will react to growing demand by increasing prices and helping to ration the scarce resource or commodity. Additionally, such price increases will, via the profit motive, stimulate an increase in capacity. When prices cannot be increased for any reason, then the congestion itself will perform the rationing function. However, in this context congestion obviously functions inefficiently, raising the real cost of the service even given a deterioration in quality and, furthermore, may do so without stimulating the desired increase in capacity and output on the part of suppliers. It is this type of congestion which characterizes the road network in many urban settings. An important point now arises: Even if the demand and supply for urban roads could be equated under the prevailing pricing system, there are clearly environmental constraints which mean that for established cities, beyond a certain level, it is impractical to provide for the unfettered use of the motor car. It can be seen, therefore, that the problem in cities is to find ways of satisfying the demand for transport within cost, capacity, and environmental constraints.

Given society's apparently insatiable demand for personalized transport, the resulting problems of congestion and, as pointed out, the environmental constraints acting against any dramatic expansion in road capacity, the ultimate necessity to restrain the use of the private car is clear. This can be achieved either physically or through the price mechanism, and, in Singapore's case, the authorities have placed the emphasis on pricing. Referring once again to Figure 1, imposing a congestion charge equivalent to CB would increase the perceived cost of a journey to the marginal road user so as to equate this with its real cost to society as a whole and, by acting as a disincentive, reduce vehicle flows from OG to the theoretically desired level OF. This is the logic behind Singapore's Area Licensing Scheme.(4)


The ALS was first introduced in Singapore in 1975, in an attempt to reduce traffic congestion in the central area during peak hours. Although scheme details have been modified from time to time, its operational characteristics are still basically the same as when first conceived. It simply defines a restricted zone and limits access to this zone by requiring all motorists who wish to enter during the designated period of restriction to purchase and display a supplementary license. Only emergency vehicles and public buses are exempted. To encourage higher car occupancy and promote more efficient use of road space through car-pooling, passenger vehicles carrying at least four persons were also exempt until quite recently, when this provision was revoked.

The restricted zone covers an area of some 720 hectares and incorporates the entire CBD. It is demarcated by gantries located at each of the 25 points of entry, with enforcement officers stationed at these gantries during the hours of restriction. Any vehicle entering the CBD during the restricted hours and failing to display an area license on its windscreen is guilty of an offense. The offender is not stopped, however. The vehicle number is simply noted by the enforcement officer, and the car owner is fined at a later date.

When the scheme was first introduced, it was felt that applying restrictions during the morning peak alone would significantly reduce traffic both then and in the evening. Furthermore, by keeping the hours of restriction short, there was the possible added advantage of spreading peak-hour travel through an induced staggering of journey times and work hours. Initially, the scheme was, therefore, designed to operate daily from 7:30 to 9:30 a.m. Following implementation, however, congestion was seen to develop after 9:30 and the time period was extended to 10:15 a.m. Another less predictable observation was that traffic flows in the evening were very little affected by the scheme. It had been assumed that any reduction in traffic flow in the morning would be the result of commuters using alternative modes of transport and leaving their cars at home. Since these cars would not be available for the homeward trip after work, evening flows would be reduced accordingly. This proved not to be the case because of the rescheduling of trans-restricted zone trips, i.e. trips with an origin on one side of the restricted zone and a destination on the other. These tended to bypass the restricted zone in the morning but pass through it in the evening.(5) In any event, the scheme has since been extended to the evenings and now also operates from 4:30 to 6:30 p.m. Because the standard working week in Singapore is five and a half days, the ALS is also in force on Saturday mornings.

Because Singapore was the first city in the world to experiment with supplementary licensing of this kind, there was no previous experience to act as a guide in setting the license fee. Policy makers were forced to exercise their judgement. In doing so, care needed to be taken to charge a high enough fee to act as a deterrent to the marginal road user, but not so high as to discourage so many motorists that roads within the restricted zone were under-utilized during operational hours. Fees have, as a result, been adjusted periodically to reflect changing circumstances and, depending on vehicle type, are currently as follows:
 Daily Monthly

Motor Cycles S$1 (US$ 0.60) S$20 (US$ 11.60)
Company Cars S$6 (US$ 3.50) S$120 (US$ 70)
All Other S$3 (US$ 1.75) S$60 (US$ 35)

Since implementation the scheme has, generally speaking, more than served its purpose, leading to considerable improvements in the traffic situation in the city. For example, in a study commissioned by the World Bank to monitor its impact soon after it was first introduced, Holland and Watson found that the number of cars entering the restricted zone during the morning peak fell by about 73 percent. Flows of other vehicle types decreased by much smaller proportions, but the net result was a 44 percent reduction in traffic. Meanwhile, average vehicle speed within the restricted zone, which had previously been less than 20 kilometers per hour, increased to 33 kilometers per hour.(6) Presently vehicle speeds in the CBD during the morning peak average about 32 kph and 26 kph during the evenings,(7) figures which are considerably better than those for other major cities as listed below:
 Average vehicle speed
City at peak times (kph)

Bangkok less than 10
Lagos 10
New York 10
Manila 11
Calcutta 11
London 12
Hongkong 16
Stockholm 18
Singapore 32 (morning)
 26 (evening)

The scheme also resulted in other important changes in travel behavior, with significant numbers of car-owning households switching to public transport, at least for the journey to work, or forming carpools. The Watson and Holland study found that the proportion of trips that members of vehicle-owning households made by car fell from 56 percent to 46 percent, while the bus share rose from 33 percent to 46 percent. Moreover, within the declining car mode, the carpool share rose by a factor of three from 14 percent to 41 percent of all car trips.(8) The actual breakdown of modal shifts is illustrated in Figure 2.

There were problems, however. Changes in travel behavior for those directly affected by the scheme can and did affect other groups of travellers and the ramifications could be quite unfavorable. For example, diverted traffic attempting to avoid the central area has caused considerable congestion on peripheral streets, a problem which has been exacerbated over time as commercial development has spread beyond the CBD. In Singapore, traffic congestion is no longer simply a center-city phenomenon, and the ALS in its present zonal form is too blunt an instrument which cannot be easily extended to tackle these new problems. What is now needed is a more flexible system that can be easily modified as circumstances dictate and can accommodate the complexities of more differential pricing at various points in the network. Electronic Road Pricing affords this opportunity.


ERP is theoretically the most effective transport planning and traffic management tool available, and the Singapore government has committed itself to implementing such a system within the next three years.

The technology is now well established,(9) and the design of the systems being considered is relatively straightforward, as illustrated schematically in Figure 3. All vehicles will be fitted with a uniquely coded electronic identity unit which consists of a microchip and a transceiver. The unit is interrogated using a multi-lane magnetic loop array buried beneath the road surface at the various electronic toll points. One loop in the array transmits a low-powered signal which is received by any vehicle crossing the toll point and used as a power source and reference frequency to transmit back to an outstation the particular vehicle's unique identity. The intelligent outstation at the toll point carries out a range of housekeeping and validity checks. If a vehicle does not have a valid identity unit, an on-site enforcement camera will capture the vehicle's registration number and transmit it to a central control for further action. Meanwhile, legitimate vehicles will be charged in either of two ways, depending on which of the technologies being considered--passive or active--is adopted.

In the passive system, the in-vehicle identity unit has no processing capability. It is a passive device which simply identifies the vehicle and conveys the information via an on-site microprocessor to a central computer, which determines the amount to be charged. The vehicle owner receives a bill at a later date, probably monthly. In contrast, in the active system, the vehicle's identity unit is basically a smart card with stored value which doubles as a microprocessor, and calculates the amount payable by the vehicle each time it passes a point where a charge is to levied. The amount is then deducted from the card's stored value. The two systems are illustrated in Figure 4.

Whichever system is eventually chosen, the intention is that it will initially replace the ALS and, after a two-year experimental period to iron out any operational problems that arise, be extended island-wide. This seems to be a sensible and highly practical approach, but, be that as it may, two issues that will need to be resolved almost immediately are the preferred pricing structure and the actual level of prices needed to achieve the desired level of restraint.

The pricing structure can be on a zonal, cordon, or route pricing basis. The benefits of route pricing are not simply derived from discouraging motorists per se, but also from diverting traffic from congested to uncongested roads. Since the intention is that ERP will at first be confined to the central area, and given the relatively even spread of traffic therein, route pricing is unlikely to be chosen in the first instance. Similarly, the tidal nature of traffic movements and the resulting directional pattern of congestion tends to undermine the effectiveness of a zonal system. It therefore seems likely that a cordon pricing approach will be adopted at whatever level of complexity policy makers think appropriate to reconcile the conflicting requirements of economic efficiency, implementation cost, and user comprehensibility. Thereafter determining correct prices, i.e., the level of congestion charge, is itself fraught with difficulty. The rather simplistic analysis in Figure 1 ignores the reality of actual driving conditions, with vehicles continually entering and leaving the traffic stream along roads whose complexity means that conditions often change from minute minute. This, of course, doesn't negate the analysis, it simply means that in practice decisions on prices and how much restriction to impose will need to be based more on ad hoc assessment of costs and benefits arising from any specific measures, rather than on any theoretically derived estimates of so-called optimum positions.


Singapore was the first city in the world to adopt a form of congestion pricing when the ALS and its related measures were implemented in 1975 and, in the intervening period, central area congestion has been substantially reduced. Furthermore, area licensing has been responsible for an appreciable shift in peak-hour travel from car to bus and from car to carpool, for an induced staggering of the times of travel and work hours, and for the more economic use of roads in the restricted zone. Indeed, traffic conditions within the restricted zone have improved to such an extent that there has also been a sharp drop in the number of accidents and an appreciable improvement in the ambient air quality because of a decline in pollution levels.(10) However, several of these and other benefits are essentially qualitative in nature and those that are quantifiable are not dimensionally homogeneous and do not lend themselves to more refined economic evaluation. Moreover, there is a growing body of circumstantial evidence to support the proposition that diverted traffic attempting to avoid the restricted zone is leading to congestion in other parts of the network, the net disbenefits of which may soon outweigh the benefits of the ALS itself. This is likely to be particularly true when one compares the net losses in induced delays suffered by diverted traffic with the time savings afforded to travellers entering the restricted zone. As congestion spreads, therefore, the benefits of the ALS are becoming increasingly ambiguous. The necessity to refine the system is obvious--hence the commitment to electronic road pricing.

There are still two big question marks against ERP, however. The first, which applies to any system of congestion charging but is more pronounced with full road pricing, relates to its equity. Is it fair or does it discriminate against less well-off road users? This is clearly a difficult issue, but similar problems arise whenever prices are charged for goods and services. One could argue, given the relationship between car use and income, that the distributional consequences of such a system are more likely to be progressive rather than regressive.(11) Furthermore, congestion charges are not a global panacea for tackling the urban transport problem but, as Singapore has shown, need to be considered within the context of a more broadly based integrated land-use and transport planning strategy. In this respect, the provision of cheap, efficient, and easily accessible public transport is just as important and, to a large extent, should help offset any of ERP's potential distributional anomalies. The second problem relates to the question of privacy. To what extent is the detailed monitoring of private vehicle movements, with all of its surveillance implications, an acceptable intrusion by the state into the personal domain? While this might offend the sensitivities of Europeans, such occidental standards of propriety are unlikely to apply in Singapore's already highly regulated environment. Every country must strike a balance between freedom and responsibility, between individual liberty and public order, between over- and under-regulation. In Singapore, that balance tends to favor more regulation than might be acceptable elsewhere, but the benefits in terms of public order and a clean and safe environment are readily apparent.

Road pricing has long been favored by economists and traffic engineers as the best means of controlling congestion and allocating road space more efficiently. Its theoretical attributes have been well documented in the literature, but more recent developments to make it both technically and administratively feasible have placed it, once again, at the forefront of the transport planning agenda. Unfortunately, and notwithstanding its obvious benefits, social and political constraints have prevented its widespread adoption. What Singapore's experience demonstrates is that given an appropriate institutional setting, such schemes can be made operational and with considerable success.


1 Ministry of Transport; Road Pricing: The Economic and Technical Possibilities. H.M. Stationery Office (London), 1964.

2 This is an edited version of a paper originally entitled "From Area Licensing to Electronic Road Pricing: A Case Study in Vehicle Restraint," which was first presented at an SAE conference on "Future Transportation Technology," Portland, Oregon, August 5-7, 1991.

3 B.G. Field, "Integrated Land-Use and Transport Planning: The Example of Singapore's Mass Rapid Transit System." In Margaret Heraty (ed): Developing World Transport, Grosvenor Press International (London), 1989, pp. 149-152.

4 B.G. Field, "Equity and Efficiency in Transport Planning." In Proceedings of the 2nd International Convention on Urban Planning, Housing and Design, Singapore, July 1989, pp. 275-278.

5 J. Yee, "The Area Licensing Scheme in Singapore." In Proceedings of Transpo-Asia 80 Asean Seminar, Singapore, October 1980.

6 E.P. Holland and P.C. Watson, "Measuring the Impact of Singapore' Area License Scheme." In Proceedings of the World Conference on Transport Research, Rotterdam, April 1977, pp. 5-7.

7 T.H. Phua, "Managing Traffic Congestion: The Singapore Experience." In Proceedings of the International Symposium on Policies for Alleviating Traffic Congestion in Large Cities, Tokyo, December 1990, p. 23.

8 Holland & Watson, pp. 8-9.

9 I. Catling and B.J. Harboard, "Electronic Road Pricing in Hong Kong: The Technology." Traffic Engineering and Control, Vol. 27, 1985, pp. 79-83.

10 R. Behbehani, V. Setty Pendakur, and A.T. Armstrong-Wright, "Singapore Area Licensing Scheme: A Review of Impact," Water Supply and Development Department Mimeo, World Bank, Washington D.C., July 1984, pp. 49-50.

11 C.D. Foster, "A Note on the Distributional Effects of Road Pricing: A Comment." Journal of Transport Economics and Policy, Vol. 9, 1975, pp. 186-188.

Mr. Field is director, Hughs Economic Planning, Nicosia, Cyprus.
COPYRIGHT 1992 American Society of Transportation and Logistics, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:area licensing scheme and electronic road pricing in Singapore
Author:Field, Brian G.
Publication:Transportation Journal
Date:Sep 22, 1992
Previous Article:Super-Trains: Solutions to America's Transportation Gridlock.
Next Article:Birth, marriage, life, and death: a life-cycle approach for examining the deregulated U.S. airline industry.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters