What works and. FFrom Houston to Milan to Bangkok, rush-hour road traffic is appalling, wasting enormous amounts of time and fuel,

Transcription

1 raffic ong esti on C FFrom Houston to Milan to Bangkok, rush-hour road traffic is appalling, wasting enormous amounts of time and fuel, undermining economic productivity and increasing air pollution. To most people, such congestion is viewed as an immutable fact of modern life the price that must be paid for personal mobility. To most economists, it is a source of irony a failure to heed the basic lessons of market economics in an era in which policymakers routinely pay lip service to the virtues of markets. Research on the efficient pricing of road use along with the information technology to make good use of that research has progressed to the point that practical solutions to urban congestion appear promising. Applications are still rare, however, largely because politicians and their constituents see congestion charges as unfair taxes. What works and The idea of pricing road use can be traced back to Arthur Pigou, an early 20th-century British economist best remembered for the distinction 44 The Milken Institute Review

2 Pwhat doesn t ricing By Georgi n a S a n t o s hagg+kropp/age fotostock Fourth Quarter

3 congestion pricing he drew between private costs borne by individual economic actors and social costs borne by parties not directly part of the transaction. The concept is fairly straightforward. Drivers making trips take into account all their own costs fuel, vehicle depreciation and the like, as well as the value of the time spent behind the wheel that could have been used doing something else. They do not, however, take into account the costs they impose on others in making the decision notably, though not limited to, the added delays linked to greater road congestion. Theory then suggests that in order to maximize economic efficiency, motorists should be charged a fee equal to this marginal congestion externality. This would increase the costs faced by drivers, and as a result, only those trips worth making (trips for which the marginal social benefit exceeds the marginal social cost) would be made. But putting theory into practice can be should be used solely to raise traffic speeds. For example, one might widen sidewalks or close some streets and transform shopping areas into pedestrian malls. Alternatively, one might create rights-of-way for light-rail systems or license more taxis or expand the service of bus systems. Indeed, the fierce opposition to Mayor Michael R. Bloomberg s plan for congestion charges in New York City suggests just how difficult the process can be, both in terms of economics and politics. In New York, for example, some of the support for congestion pricing was lost when it was revealed that the subway system was running at capacity at rush hour, and thus couldn t handle additional commuters diverted from private cars. Singapore first imposed congestion charges in London introduced a scheme in 2003, which was extended to include some areas of the city to the west. A more detailed look at how these systems work and what they do and do not accomplish is revealing. With ERP, charges are tied to the number of boundary difficult, to say the least. For one thing, it isn t obvious how far one should go in accounting for costs. Should the unpleasantness for pedestrians of navigating traffic count? For another, it s unclear who should reap the benefits direct and indirect from congestion pricing. Most obvious here is the issue of what the revenue should be used for. Better roads? Better or cheaper mass transit? Or should it go into general government funds? There is also the question of whether the resulting reduction in private vehicular traffic GEORGINA SANTOS is a lecturer in the Transport Studies Unit of the School of Geography and the Environment at Oxford University. road pricing in singapore In the 1970s, the city-state of Singapore was experiencing very rapid economic growth, which inevitably brought with it more surface traffic. If nothing else, this congestion increased the cost of doing business and acted as a deterrent to foreign investment or at least that was the perception of Singapore s leaders. To cope, the government introduced what was known as the area licensing scheme (ALS). Vehicles entering the seven-squarekilometer (three-square-mile) central business district in the morning (and later on, vehicles exiting in the evenings) were required to purchase and display paper licenses on their windshields. 46 The Milken Institute Review

4 crossings rather tha n t o d a i l y f e e s. afp/getty images Over the years a number of changes were made to the scheme. For example, the restricted hours were extended to cover the inter-peak period (10:15 to 4:30) on weekdays and the post-peak period (10:15 to 3) on Saturdays, which was later cut back to 2 p.m. The charges also varied through the years. The inter-peak fees were reduced to twothirds those for the peak. The colors of permits varied according to vehicle type, and the color was changed every month. Enforcement was done manually by officers standing at the boundaries of the restricted zone. Drivers of vehicles lacking licenses were fined. Although the vehicles included in the plan varied over the years, by 1989 the only exemptions were for public transport buses and emergency vehicles. There were no discounts for residents, although they could avoid the charge by driving within the zone without crossing the boundary. The ALS increased average traffic speeds from 18 to 35 kilometers per hour (11 to 22 miles per hour). Traffic entering the RZ initially fell by 44 percent, but it then crept back up. Yet in 1988 the number of licensed vehicles was still 31 percent below the maximum registered before the ALS, despite a 33 percent increase in employment and 77 percent increase in the number of registered vehicles. In 1998, the ALS was replaced with a more Fourth Quarter

5 advanced and flexible electronic road pricing (ERP) system that automatically kept track of vehicle movement. Gantries monitor traffic in the three central business districts (where charges apply from 7:30 a.m. to 7 p.m.), and expressways/outer ring roads (where charges apply only during the morning rush 7:30 to 9:30 a.m.). The rates are published on the Land Transport Authority s Web site, and are adjusted periodically in order to keep traffic to levels that allow target speeds 45 to 65 kilometers (28 to 40 miles) per hour on expressways and 20 to 30 kilometers (12 to 19 miles) per hour on major roads. The most interesting feature of ERP charges is that they vary according to vehicle type, time of the day and location of the monitoring gantry. Charges for passenger cars, taxis and light commercial vehicles range from 50 cents to 4.50 in Singapore dollars (each dollar is currently worth about 65 United States cents). Charges for motorcycles are about half that much, while the heaviest vehicles and trucks pay between one Singapore dollar and nine Singapore dollars. Enforcement is electronic. Motorists passing through a gantry without a smart card to register the transaction (or with one lacking enough credit to pay the charge), receive a letter soon afterward, demanding payment of the outstanding charge plus an administrative fee of 10 Singapore dollars. The finely tuned ERP decreased traffic in the restricted zones by about 10 to 15 percent during its hours of operation. The most important difference between the two systems is that ERP applies at each crossing, whereas the ALS charge allowed multiple entries on the same day. The ERP has therefore deterred multiple trips into the charged zones. The only exempt vehicles are police cars, fire engines and ambulances. This leaves few 48 The Milken Institute Review

6 options for those who want to enter the central business district without charge. They can travel during hours in which the ERP does not apply, or use the light-rail network that covers the whole of Singapore. Or, of course, they can switch to unpriced, uncongested areas for shopping, leisure and the like. Interestingly, although ERP applies to public buses, the charges are not reflected in fares, mainly because the charge per passengercrossing is negligible. As a result, there has not been switching from buses to rail because of the ERP. In 2005, the average daily rail system patronage was 1.3 million, while the bus patronage was 2.8 million. The government of Singapore is committed to increasing rail use and reducing bus use, however, largely with the goal of reducing road congestion even further. This will be achieved not through an increase in the ERP rate for buses, though, but through the expansion of the competing rail network. the london congestion charging scheme In 2003, the London Congestion Charging Scheme (LCCS), designed and managed by the government agency Transport for London (TfL), was introduced in central London. All vehicles that drive into, out of, or inside the charging zone (CZ), or even just park on a public road inside the zone on Monday to Friday from 7 a.m. to 6 p.m., must now pay 8 (about $16) a day. In 2007, the charging zone was extended to the west. But it is worth noting that the whole zone is still relatively small, representing only 2.4 percent of Greater London. Enforcement is built around automatic number-plate recognition. There are cameras at every entry and exit to the congestion charging zone, as well as many within the zone. Cameras read and record the license plates numbers of virtually all the vehicles making use of the zone, sending the data to a processing center. These number plates are then matched against plates in the database that have paid or are exempt. After a second (manual) check, violators are tracked through the Driver and Vehicle Licensing Agency and fined 100 by letter. The penalty is reduced to 50 if paid within 14 days, and increased to 150 if not paid within 28 days. ANNUAL COSTS AND BENEFITS OF THE LONDON SCHEME MILLION AT 2005 VALUES AND PRICES COSTS TfL administrative and operation costs TOTAL BENEFITS Time savings and reliability benefits Loss of initial gain in time savings and reliability due to induced traffic Charge-payer compliance costs Vehicle fuel and operating savings Accident savings Environmental benefits 3.00 Dis-benefit to deterred trips TOTAL source: Transport for London, Fifth Annual Report 2007 A vehicle with three or more outstanding congestion notifications may be clamped or towed if found anywhere in Greater London. If the (very substantial) release fee is not paid, the vehicle is eventually sold at auction or scrapped. Even in that case, though, the owner remains liable for all outstanding charges, including an disposal fee. The standard way of measuring the value of public programs is with cost-benefit analysis. Very roughly speaking, if benefits exceed costs, the program is worthwhile. In July 2007, Transport for London provided its own estimates of these costs and benefits, as shown in the table. Note that, in this context, the fees Fourth Quarter

7 congestion pricing Both London and Singapore began congestion pricing with systems that were easy for motorists to understand. In London, the charge is the same for all vehicles and charging hours, and the charge zone is clearly marke d. paid by motorists are not societal costs since the money is simply transferred from one set of pockets to another. The benefit-to-cost ratio from the table is a healthy But road users are not as interested in this number as they are in direct measures of their own benefits in terms of reduced congestion. The average speed inside the congestion zone increased from 14 kilometers per hour in 2002 to 17 kilometers per hour in 2003 and However, in 2005 it decreased to 16 kilometers per hour and in 2006 it decreased even further, to 15 kilometers per hour. In other words, about twothirds of the initial increase in traffic speed had been lost by The number of vehicles entering the CZ is still substantially lower than before the scheme was introduced. Transport for London offers some reasons why fewer vehicles do not translate into faster road speeds: The high number of road repairs in progress, particularly in the second half of 2006 New traffic management programs to reduce the number of accidents Increased numbers of buses in service New rules (and space) to improve the environment for pedestrians and cyclists Put simply, much of the road capacity freed by smaller numbers of cars has been reallocated to buses, bicycles and pedestrians. But the road repair explanation shouldn t be entirely discounted. Nor can one say that average traffic speeds wouldn t have been even lower in the absence of congestion charging. what works and what doesn t I would argue that, while quite different, the London and Singapore systems are both success stories. And their experience offers some insight in how to get the job done in the teeth of a skeptical public. Safety valves There need to be either alternative uncharged ways of getting around town or alternate uncharged routes in place when the system is first deployed. Transport for London improved the bus network (more routes, bigger and more frequent buses) before the LCCS started. Singapore, for its part, offers an exempt light-rail system as well as a bus system that, while not exempt from the fees, doesn t pass them through to patrons. In Edinburgh, by contrast, where congestion pricing was also seriously considered, it lost in a public referendum. One of the main reasons was apparently public anger at the lack of mass transit to serve residents who couldn t or wouldn t pay to drive. Simplicity Both London and Singapore began congestion pricing with systems that were easy for motorists to understand. In London, the charge is the same for all vehicles and charging hours, and the charge zone is clearly marked. Indeed, the lack of sophistication of top: national pictures/topham/the image works; bottom: martyn goddard/corbis 50 The Milken Institute Review

8 tk Fourth Quarter

9 congestion pricing the London approach was offset by the support it earned from simplicity. Singapore s original ALS was straightforward, but grew in complexity with the years. By the time different charges for different vehicles and hours of the day became too complicated to enforce manually, road users had 23 years of training in congestion charging and were ready to cope with the far more elaborate electronically monitored system. Transport for London is now considering the addition of differential tailpipe emissions charges as an incentive to reduce climatechanging carbon emissions. If the agency goes forward, it will presumably face less public opposition than it might otherwise, because Londoners will have had years of experience with the general concept of charging for vehicle externalities. Political will In the case of London, Mayor Ken Livingston was determined to introduce congestion charging and made it part of his platform in The act restoring the mayoral form of government to London after more than two centuries also gave him powers to introduce congestion charging without consent from the national government. Both the original congestion pricing plan and the extension of the charging zone met fairly stiff public opposition, but the mayor went ahead, nonetheless. Later on, the public indirectly endorsed the plan, re-electing Livingston a year after congestion pricing was introduced. There were no U-turns at the local government level at any point. Singapore has been run by the People s Action Party since it first won an election in As the ruling party, it has dominated most of the political and economic development of Singapore though not without criticism from opposition parties and international organizations for passing laws that deter free speech. Thus, it is not surprising that once the decision was made to charge for road use, there was little dissent. Public consultation Consulting with businesses and local political leaders gave London s planners a chance to reduce confusion and to disarm their critics. What s more, it provided genuinely useful feedback on the plan design. For example, charging times were originally going to be 7 a.m. to 7 p.m. The decision to change the evening end-time was primarily a result of lobbying by the entertainment industry, which argued that having the charge apply as late as 7 p.m. would discourage potential patrons from entering the theater district. By the same token, the proposed heavy truck charge of 15 was reduced to the same as that for cars to placate a critical constituency. The borders of the charging zone were also changed a few times before the plan was first put into effect. This was, again, a response to pressure groups that the mayor decided he could not ignore. Having said all that, Mayor Livingston never gave any group (or, for that matter, the voters) the option to reject the plan entirely. Edinburgh, as noted earlier, left the final decision in the hands of voters, who resoundingly rejected it. Livingston, for his part, managed the appearance of gaining consensus without ever testing public opinion. And in some cases, it is not clear that features of the plan would have ever made it past a referendum. That certainly applies to the increase in the congestion charge to 8 from 5, and the extension of the charging zone to the west. Indeed, it may apply to the initial decision to adopt any congestion pricing scheme. 52 The Milken Institute Review

10 martyn goddard/corbis Use of revenues In Western democracies, surveys suggest that the public won t willingly accept congestion charging unless the revenues are guaranteed to stay within the transport sector in road building and repair, or mass transit. By law, all net revenues from the London charges must be used for transportation in London. More than 80 percent has been spent on bus network improvements. In Singapore, car ownership and driving are discouraged altogether, and congestion charging is only one of many measures for traffic demand management. The measures are not designed to raise revenues but to keep traffic at acceptable levels. Accordingly, the revenues from road transport include a number of items apart from the congestion charge: fuel taxes, import duties, parking charges, taxes on vehicle ownership and vehicle registration fees. Prospective vehicle owners are required to purchase a 10-year quota license or certificate of entitlement. There is a quota for each vehicle category, announced in May each year. Thus, in Singapore, the question of how the revenues from ERP are spent is not as important as it is in Britain, or as it would be in any Western democracy. The experiences of Singapore and London suggest that road congestion pricing is very useful, even though there has been no attempt to measure the actual marginal congestion externality and set the congestion fee to match. Singapore targets minimum traffic speed, adjusting prices (if necessary) every three months to get closer to the goal. London set the 5 and the 8 charges not on the basis of estimates of congestion costs but on the basis of targeted traffic reduction. Stockholm conducted a trial of electronic congestion charging between January and July City residents subsequently voted for permanent implementation. And although the vote was not binding, the national government chose to put the system in place in August The revenue will be spent on road construction in and around Stockholm. Political acceptability does not seem an insurmountable obstacle any longer and the issue of equity seems to be resolved with the provision of some alternative uncharged mode of transport. Singapore led the way, starting 28 years before any other city. London s system is less ambitious, but probably more compatible with Western democracy. The question now is whether other urban leaders will muster the arguments and the political will to break through reflexive public resistance to paying for what they regard as a free good. As the London case suggests, the success of congestion pricing does not entirely turn on its impact on traffic speeds. M Fourth Quarter