5

Design, Evaluation, and Research

Although relatively few proposals for congestion pricing are moving forward in the United States, there appears to be growing interest. The proposals that are enacted will provide unique opportunities to learn about what will and will not work. A poorly designed initiative that fails, however, may spell the end of this approach to managing congestion or at least set it back by another 10 to 20 years.

Offered in the first section are a few design issues that need to be considered in any congestion pricing proposal. Given the lack of experience with congestion pricing in the United States, there are many questions about design that do not have clear answers. Learning from early efforts is critical to determining the effectiveness of this approach and for improving future proposals.

In the second section, the basic features that should be part of the evaluation of these early experiments are outlined, and in the final section, other important research is suggested that would improve the understanding of congestion pricing.

DESIGN ISSUES

Several issues have been discussed in previous chapters that are critical to the design of any congestion pricing project, such as mitigating distribu-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion 5 Design, Evaluation, and Research Although relatively few proposals for congestion pricing are moving forward in the United States, there appears to be growing interest. The proposals that are enacted will provide unique opportunities to learn about what will and will not work. A poorly designed initiative that fails, however, may spell the end of this approach to managing congestion or at least set it back by another 10 to 20 years. Offered in the first section are a few design issues that need to be considered in any congestion pricing proposal. Given the lack of experience with congestion pricing in the United States, there are many questions about design that do not have clear answers. Learning from early efforts is critical to determining the effectiveness of this approach and for improving future proposals. In the second section, the basic features that should be part of the evaluation of these early experiments are outlined, and in the final section, other important research is suggested that would improve the understanding of congestion pricing. DESIGN ISSUES Several issues have been discussed in previous chapters that are critical to the design of any congestion pricing project, such as mitigating distribu-

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion tional consequences and the use of the revenues earned. Discussed in this section are several other issues that need careful consideration: allowing sufficient time for a proposal to develop political acceptance, setting an appropriate price, and providing alternative routes or modes of transport for those who prefer not to pay a congestion fee. Proposal Development Congestion pricing has gained currency only in a few places, and even in these cases the process of gaining political acceptance has been quite slow and conditioned by almost unique circumstances. Interest in California has surely been fostered by the fiscal restrictions that voters have placed on state and local government spending, the continued rapid growth in population and automobile travel, and the stringent requirements of the California Clean Air Act. In the San Francisco Bay Area, congestion pricing has also been furthered by broad public concern about the environment and by the sophisticated advocacy of a business-sponsored organization dedicated to resolving regional issues in the Bay Area (Dittmar, Vol. 2). Even with these conditions influencing adoption of congestion pricing, the Bay Area congestion pricing proposal was some years in the making (and at the time of this writing requires state enabling legislation to raise the toll). In San Diego, interest in congestion pricing was fueled by accelerated growth in traffic congestion over the last few years, the leadership of local elected officials, and the general acceptance of a philosophy of charging the user directly for the cost he or she imposes on others (Duve, Vol. 2). Even so, acceptance of the concept of congestion pricing emerged over a considerable period of time. In contrast, an effort to build consensus for a congestion pricing proposal for Manhattan during the late 1980s was summarily dismissed by the public and by some downtown commercial interests (Altshuler 1990; Zupan, Vol. 2). These cases suggest that interest in congestion pricing— driven by concerns about congestion and the environment—will emerge in a few places, but sufficient public and political acceptance to allow it to go forward will occur slowly, if at all. The congestion pricing pilot projects allowed under the Intermodal Surface Transportation Efficiency Act of 1991 (ISTEA) may be developed at any time during the 6-year authorization period, but the time frames for development and implementation of proposals to meet the first and second rounds of solicitations

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion assumed an ability to move forward at a speed that few jurisdictions could meet. In the interest of political feasibility, the development of congestion pricing proposals requires time to design features that address the concerns of groups resisting change. In both San Francisco and London, detailed studies of the possible impacts and costs of congestion pricing are being used to further public awareness and acceptance. In addition, the ongoing London study (a 3-year, $5 million effort) includes a considerable effort to gauge and inform the opinions of local leaders as well as the public at large (Gomez-Ibanez and Small forthcoming). Such efforts highlight the difficulty of convincing political leaders and the public of the potential of congestion pricing and the considerable effort required to move proposals forward. Because of the need to build consensus slowly and to fashion a politically feasible congestion fee, the first year of the San Francisco Bay Area congestion pricing pilot project will be devoted to extensive analyses of travel patterns and how they might be affected by toll increases (with particular emphasis on the impact on low-income individuals), consideration of off-peak discounts for freight operations, estimates of the potential mode shifts to transit, focus group and survey research of potential public reaction to different price increases, and review of the proposal with various interests that could benefit from or block the proposal. Setting the Price Setting a proper price is important to the success of congestion pricing. If the fees are too low, they will not divert sufficient traffic to relieve congestion, which would undermine the efficacy of this proposal in the minds of motorists and voters. If the prices are too high, the road system will be underused, and motorists and voters can be expected to be critical of that as well. One of the criticisms of Singapore's area licensing proposal is that it set the fees too high initially (Toh 1977). Whereas this greatly reduced automobile congestion in the downtown area, it apparently diverted too many travelers to facilities unable to accommodate them, such that little or no improvement occurred in total trip times to the core (Gomez-Ibanez and Small forthcoming). Given the lack of experience with congestion pricing in the United States, however, there is no perfect way to predict the appropriate prices to charge (Vickrey 1993). Vickrey (1993) and Wohl and Hendrickson (1984) suggest deriving an estimate for the peak fee using standard approaches to

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion calculating demand elasticities and short run marginal costs, and then subsequently adjusting the fees in response to observed traffic patterns. With the administrative and political constraints that congestion pricing proposals face, however, it may be difficult to obtain public and political approval of a proposal without publishing the fees in advance. One way to predict the right price is first to determine an appropriate traffic flow for a facility and then to set the price according to estimates of motorists' sensitivity to price increases in analogous situations. Although the models applied to develop this estimate are quite complex, this is essentially the method used in estimating the benefits of congestion pricing for the San Francisco Bay Area and Southern California (Harvey, Vol. 2). Another method for setting the right price is to estimate those costs and benefits that can be quantified and then to calculate the price level that maximizes the net benefits. This approach is being used in the London study currently under way (May, Vol. 2). In the case of the Riverside Freeway project in Southern California, the private developers are being allowed to set and vary the prices according to demand, with no prior commitment to the public about what those prices might be. Rather than attempting to regulate the prices that the developers could charge motorists, the state chose to regulate the maximum return that developers could receive on their investment (Fielding, Vol. 2). The developers will obviously have to be sensitive to consumer demand in setting and varying their prices and are also required to maintain free-flow conditions for high-occupancy vehicles. Because this is a private project, however, there is less public information available about how the developers will estimate demand and how frequently the prices might be changed. During the first year of the San Francisco Bay Area congestion pricing pilot project, public and political opinion will be surveyed on the receptivity to different price increases for the Bay Bridge tolls. It appears that the price increase for the Bay Bridge will be on the basis of both extensive modeling of the price increase required to improve traffic flow to a desired level of service and an appreciation of what is politically feasible. Another pricing issue to be confronted is whether the structure of the congestion fee would be a simple, flat increase or smoothly varying increases. For the latter, a congestion fee might increase 10 percent at the beginning of a morning peak hour and then increase substantially in 15-min increments to its highest level when traffic reaches its normal peak. The toll might then decrease gradually during the shoulder of the peak

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion period. The developers of the Riverside Freeway project intend to use a smoothly varying price structure. Flat congestion tolls lasting for the duration of the peak period have the advantage of being simple to understand, but have other disadvantages. Motorists who typically travel during the middle of the peak but who would prefer to shift to an earlier time to avoid the congestion fee may have to adjust their departure by as much as an hour. A flat fee structure could incorporate a substantial off-peak discount to induce such adjustments, but such a large time shift may reduce motorists' willingness to change. With a smoothly varying toll, in contrast, motorists could adjust their departure time in order to pay the level of toll they find acceptable. In theory, if motorists all make small adjustments according to their differing values of time, substantial time savings would occur for all without requiring a few motorists to make very large time adjustments (Vickrey 1993). Another disadvantage with a flat toll is that some motorists who want to avoid the toll will arrive just before or just after the change and will thereby increase congestion. In Singapore, for example, during the half hour before the congestion fee took effect, congestion initially increased by 24 percent (Watson and Holland 1978, 43).1 A major question about variable tolls is whether drivers will be able to predict with sufficient certainty the fee that they will face when they reach the tolled facility, thus ensuring the traffic smoothing predicted by theory. There is no technology on the horizon that would predict, in real time, the congestion that a traveler might encounter during the trip and the fees that the traveler might pay (May, Vol. 2). The actual effects on motorist behavior and traffic flow in response to the variable pricing proposed for the Riverside Freeway (SR 91) will be of great interest in this regard. A number of different variable pricing proposals have been made in an attempt to optimize the efficiency gains of congestion pricing. Additional proposals to the ones outlined here, and the possible technologies for implementing them, are reviewed in the paper by May in Volume 2 of this report. 1   Because the increase represented a relatively small number of vehicles, the authorities did not believe that the increased traffic represented a substantial problem. A similar increase in traffic initially occurred just after the peak-period fee ended. In this case the authorities extended the period during which the charge was imposed until mid-morning.

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion Available Alternatives Travelers will make adjustments to congestion pricing in ways that cannot be fully predicted. Assuredly, however, some will seek alternatives to avoid the fee. Underestimating the size of these shifts could result in substantial dislocations. In Singapore, for example, it appears that the high price for entering the central area diverted a substantial portion of through traffic to a ring road that could not handle the capacity. Planners also underestimated the shift to transit. The large number of new transit patrons apparently increased bus boarding times and therefore slowed the travel times of the buses (Gomez-Ibanez and Small forthcoming). On the other hand, planners overestimated the demand for park-and-ride facilities on the edge of the central business district, which resulted in construction of some expensive and unnecessary parking facilities and purchase of small buses to operate between the lots and the city center. In the aggregate, the congestion on the ring road and slower travel times for bus patrons negated many of the time savings enjoyed by those who paid the fee to travel into downtown. Over time, the net benefits became more positive with the expansion of the ring road and bus fleet (Gomez-Ibanez and Small forthcoming), but the dislocations caused early in the Singapore experience suggest that considerable care should be paid to providing alternative modes or routes. In the example of the A-1 motorway in France, considerable care was taken to minimize any tendency to divert travelers to adjacent, signalized arterial routes. When congestion pricing is applied to a single facility, there can also be the problem of traffic diverting from a tolled route to untolled routes. This diversion may slow traffic on the untolled routes. Motorists may also begin using neighborhood streets to avoid tolled routes, and fear of such effects by neighborhood groups could lead to rejection of proposals for congestion pricing. Any proposal for congestion pricing should make explicit the plans for handling traffic diversion and incorporate design features to minimize its impact. Avoiding the creation of congestion on unpriced routes would be particularly important for avoiding localized increases in automotive emissions. EVALUATION Several key policy questions emerge from the previous chapters of this report, the answers to which will only be gained through experience. The major questions about congestion pricing can be summarized as follows:

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion What is the range of behavioral responses at different prices? How will behavioral changes affect congestion? What will the impacts be on different groups (considered by income, gender, and geographic area)? What will the air quality and energy effects be? What will the effects on urban form and development be? How do all of the above affect public receptivity and political feasibility? In the sections that follow, some basic evaluation elements are outlined that could be employed to learn substantially more than is now known about each of these questions. In an individual corridor, in which a priced route would be parallel to a free road, an evaluation could gauge fairly readily motorists' willingness to pay to avoid congestion. The privately financed and operated toll road in the median of the Riverside Freeway, which will have congestion tolls for single-vehicle occupants, offers just such an opportunity. The extensive evaluation of the Singapore congestion pricing program, conducted by staff at the World Bank with financial support from international organizations and from the Environmental Protection Agency, U.S. Department of Transportation, and U.S. Department of Energy, provides a useful framework (Watson and Holland 1978); the following text draws upon that design.2 Travel Behavior Questions about how personal, business, and commute trips would be affected by congestion pricing are among the most intensely debated within the transportation community, and for which, aside from mode shifts between automobile and transit, little empirical information is available (Kain, Vol. 2). Very little information is available at present to guide estimates of how shifts in the timing of trips would affect the total loading on the transportation system. This is a particularly important area for evaluation. One approach to measuring travel behavior is to collect detailed household travel diaries for some time period (24 hours or 7 days). This was the approach taken in the Singapore study and in the various personal transportation surveys undertaken by the Bureau of the Census. 2   The Federal Highway Administration has contracted for a guidebook on project development and evaluation (Bhatt forthcoming).

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion Travel diary data are expensive to collect, however, which helps explain why such surveys are conducted infrequently. The Singapore study, for example, collected before-and-after travel diaries from a sample of 2,000 travelers (the “after” survey results were supplemented with less extensive surveys of another 10,000 residents). These data allowed evaluators to measure changes in overall trip making, mode shifts, and elapsed travel times during the peak, and route shifting. All these elements were important to gauging whether the program was a success or failure. Less extensive information about travel behavior can be collected through telephone surveys. Either approach would allow for an estimate of demand elasticities, trips forgone, trip chaining, shifts to other modes, and the use of travel substitutes, such as increased telephone use, catalogue ordering, and telecommuting. The World Bank evaluation of Singapore relied on traffic counts of commercial vehicles, but such data are not likely to address the ways in which commercial activity might be affected positively and negatively. It would be much more useful to draw a sample of carriers and business entities dependent on the transportation system and collect travel and cost data before and after congestion pricing was put in place. Effects on Facilities Another of the most important questions is whether and how much congestion pricing reduces congestion on specific facilities. Traffic counts and vehicle speeds could be collected by time of day and day of week on tolled and untolled routes with the automated technologies routinely used for these purposes. Because of seasonal and economic effects on travel, however, it would be important to develop sufficient trend data to control for these external effects on travel behavior and to take repeated measurements of traffic effects over several months or quarters after a new policy was put in place. Simple before-and-after surveys could easily be distorted by seasonal effects unless they were carefully timed to occur at the same time of year, but even this degree of care would not be sufficient to control for other external effects on traffic. In addition, the before-and-after approach does not address what the traffic would have been in the absence of the intervention. Automated traffic counts will also not provide measures of changes in vehicle occupancy and in the composition (personal/commercial) of ve-

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion hicular traffic. Measures of such changes can be inferred from the travel diary data but should be confirmed by visual observation. In cases where transit options are available to travelers, shifts to transit can be estimated from travel diaries but should be confirmed with ridership surveys. Observations of sample routes and trips would also provide important information about increased crowding, if any, and the effects this might have on speed. The effects of congestion pricing on transit service would be of considerable interest. The improvement in traffic flows would provide faster and more timely bus service; this and the direct monetary incentive for people to shift from automobiles to transit should improve ridership and revenues (Kain, Vol. 2). Such improvements would benefit the lower-income riders who rely on this service and could mitigate any adverse effect congestion pricing might have on the poor. The effects on transit service reliability and frequency should be closely monitored. Impacts on Specific Groups A panel survey, in which data would be collected from travel diaries or telephone surveys, would enable researchers to address many of the questions about the equity consequences of congestion pricing. The data collected should include important household and life-style characteristics: income, residential location, commuting pattern, number of vehicles, number of workers, family size, children's ages, annual travel, and employment type and location. Without these data, it would be difficult to measure the consequences on groups that could be most affected. Environmental and Energy Consequences Given the concerns about the environment and energy conservation that help motivate the interest in congestion pricing, it is important to collect data on both travel behavior and actual emissions. For example, some estimates that congestion pricing would produce energy and environmental savings assume a net decline in travel, but if a substantial share of trips are diverted to other times, the environmental and energy benefits would be lower. Data collected from travel diaries or surveys would provide an assessment of how the pattern of travel activity was affected by congestion pricing.

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion Congestion pricing is assumed to improve the flow of traffic and thereby reduce emissions because of less stop-and-go traffic. Such effects can be inferred from traffic engineering surveys of vehicular flow, but the actual measures of emissions on the road during peak periods would also be important. When individual facilities are tolled, for example, traffic diverted to other routes can increase the congestion and emissions on those routes. Whether congestion pricing results in net environmental benefits requires an empirical answer. New technologies, such as the Stedman scanning device, allow emissions to be measured while vehicles are in use (Bishop et al. 1993). The Stedman device directs an infrared beam through the exhaust emission of individual vehicles (or a traffic stream) to a measuring device. The characteristics of the light indicate carbon monoxide and hydrocarbon content. This kind of technology would permit sampling of vehicle emissions before and after the toll was put in place. It is also important to measure changes in concentrations of pollutants, particularly carbon monoxide. Emission monitors could measure changes in emissions on tolled facilities before and after the tolls were put in place. Data from these monitors, combined with traffic counts and traffic flow estimates, would allow inferences to be drawn about the net effects on pollutants. Effects on Urban Form The effects that congestion pricing might have on urban form are among the most difficult to measure. Changes in business and personal location occur over several years, during which time the effect of congestion pricing could be tempered by a number of other countervailing influences. Although congestion pricing might lead to more dense development near activity centers, such changes take time and would often be affected by zoning and development restrictions. It would be useful to conduct surveys of managers of commercial activities to determine whether or how they might react, but it is difficult to know what managers will actually do compared with what they think or say they might do. Perhaps more useful would be research on how to properly evaluate the impact of congestion pricing on urban form. Political and Public Receptivity The perception of how a policy intervention works may be as important to its continued operation as the actual effects it has on congestion. Such

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion opinion surveys could also provide insight into issues such as privacy concerns. Specific attention should be given to sample design to ensure that it captures those most affected (for good and ill), but also to ensure that the sample in the aggregate is representative of users of the facility. Summary Careful and extensive evaluation of early congestion pricing programs is critical to developing a better understanding of the advantages and disadvantages of this policy. An appropriate evaluation of a congestion pricing project must be designed carefully to account for different explanations of the possible outcomes other than the introduction of a congestion fee (Campbell and Stanley 1966; Weiss 1972). Considerable data must also be collected. The evaluation outlined in this chapter could easily extend over several years and cost $5 to $10 million. The costs would be high, but it would be a worthwhile public investment. Congestion wastes billions of dollars in metropolitan areas every year, and it has persisted despite the expenditure of many billions of dollars to expand highway and transit capacity. Many of the transportation control measures being considered and implemented to meet clean air standards are estimated to have quite modest effects on travel (Deakin 1993). Despite its political challenges, congestion pricing has more promise than most other policies for significantly reducing congestion. The expense of a few million dollars to demonstrate the potential of the theory would be a good social investment. OTHER RESEARCH Although many of the most important questions about congestion pricing —the time savings and potential adverse impact on the poor—can only be answered through evaluation of actual projects, there are issues and questions about congestion pricing that deserve further exploration independent of any project actually going forward. Commercial Transportation and Activities In examining the possible impact of congestion pricing on business activity and commercial transportation in the United States, the committee found no prior studies to inform its analysis. Given the importance of transporta-

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion tion to the economic vitality of metropolitan areas and the widespread perception that congestion increases business costs, it is remarkable how little is known about the magnitude of these costs. Surveys of service firms in congested metropolitan areas reveal that the managers of these firms believe that congestion affects them but have little specific information (Cambridge Systematics forthcoming). Transportation planners have little information about the logistics patterns of businesses dependent on transportation or the costs borne by commercial carriers operating in congested metropolitan areas (Cambridge Systematics forthcoming; Giuliano, Vol. 2). Research is needed on the logistics patterns of metropolitan firms; the frequency, origin, destination, and timing of trips; substitutes for transportation; and the ability of firms to adapt to congestion without affecting costs. All of these elements are important to predicting how firms would be affected by congestion pricing. Research is also needed on how congestion pricing would affect commercial carriers operating in metropolitan areas. Would they experience net savings? Could they pass along cost increases? Impacts on Transit and Other Modes Much more needs to be known about possible increases in transit use and ridesharing in specific metropolitan areas. Kain (Vol. 2) notes that “ideally this paper would have presented the results of simulation studies of alternative congestion pricing schemes for three or four representative metropolitan areas, selected to represent different land use patterns, levels of transit use, and congestion levels. ” Such analyses would provide better estimates of the extent to which congestion pricing would shift demand to carpools and transit and would provide more insight into how transit operators might be able to expand service and how that service might benefit lower-income users. Modeling Most regional travel demand and land use forecasting models in use by metropolitan planning organizations (MPOs) do not incorporate a variable reflecting household income, nor do they include price changes as a predictive variable. Thus most MPOs are not in a position to begin estimating the range of consequences that would result from congestion

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion pricing. (Some regional models can be and are being modified to incorporate price signals to some degree.) Simulation of the possible effects of congestion pricing and parking pricing in the Bay Area has been made possible by the previous development of complex models that incorporate price signals to some degree (Harvey, Vol. 2). These models, however, were not designed with the intent of testing pricing strategies. Additional refinements would be desirable. Given the importance of estimating the possible effects of pricing strategies and the variance of these effects with local geographic, demographic, economic, and transportation system characteristics, considerable effort will be needed to develop models applicable in individual metropolitan areas that would incorporate household demand characteristics and be capable of simulating behavioral changes in response to pricing strategies. Such models have been developed in the United Kingdom and in the Netherlands; U.S. models might be patterned on the Bay Area model. Measures of Congestion Although it is widely believed that congestion is worsening in metropolitan areas throughout the country, there is little empirical basis to corroborate this impression. Only fairly crude data are available, and these do not measure congestion directly. A few metropolitan areas have consistently measured congestion on specific corridors, but few MPOs have any measures of how congestion has changed over time throughout their regions (Meyer, Vol. 2). Research is needed to develop better congestion measures. Meyer (Vol. 2) outlines several characteristics that such measures might have; most important, they should “have a strong functional relationship to the actual costs of congestion.” (Measuring traffic congestion is also reviewed in Appendix A.) Productivity Transportation makes up a substantial share of the U.S. economy—in total it accounts for about 17 percent of the gross national product. If congestion pricing strategies became widespread, they would diminish the need for building capacity to serve peak demand and would thereby reduce tax burdens and free more revenues for use in the private economy. Research is needed to estimate the magnitude of these effects and how the efficiency

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion gains of congestion pricing would improve the efficiency and productivity of regions that adopt it. International Experience Congestion pricing has sparked interest and policy studies all over the world. Evaluations are under way for London, Cambridge, and Edinburgh, and congestion pricing policies are being considered at the highest levels of government in the Netherlands and Sweden. International colloquia on congestion pricing and other demand management experience abroad would be helpful. Project officials can learn from the design of other projects, U.S. policy makers can learn from the experience of their counterparts, and researchers can learn about the successes and failures of projects in other nations. Land Use Changes The relationship between transportation and land use has been difficult to determine. Land use changes occur over long periods of time, and the effects of transportation are difficult to separate from those of other influences. Congestion pricing's effect on land use and urban form is made more difficult to determine by specific policies that private businesses might use to minimize the impact of congestion pricing on their employees. Research is needed on how to measure the short- and long-term relationships between congestion pricing and land use. Shifting of Burdens Much of the debate about congestion pricing has focused on how the burdens might fall on certain groups of users and retail and commercial interests. Research on tax burdens and benefits generally has shown that they shift through changes in land costs and labor compensation. The immediate incidence of a tax is not a good measure of how it affects individuals or the economy over time. Research has not been conducted on the longer-term shifts in land, labor, and retail markets. Such analyses would greatly improve the understanding of the net benefits of congestion pricing.

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion Regional Politics A successful policy change of a controversial nature at the regional level requires a great deal of public education and constituency building. Congestion pricing faces uncertain political prospects, in part because the benefits are large but diffuse, and the costs, though outweighed by the benefits, are concentrated. Research is needed for specific regions to determine which groups would benefit and which might lose (or even perceive that they might lose) and the influence these groups have at the regional level. A broad class of highway users would save several minutes a day, for example, but they might not be sufficiently motivated by this potential gain to fight for it, and their interests might not be represented at the regional level. Environmental groups might be motivated to fight for the air quality benefits and might be arrayed with groups supporting public transportation. Opposing political interests might include some commercial groups, commercial transportation, and low-income advocacy groups. The opposition, however, might be divided since some, if not many, members within these groups stand to gain. Large employers in a region might be more attracted to congestion pricing than to employee trip reduction programs. Low-income advocacy groups might be attracted to the possible benefits that improved bus services would provide their constituency. Research is needed on whether a successful constituency could be put together that would encompass the broad beneficiaries, how such a constituency might be built, and whether it could be expected to counter opposition. Alternatively, research is needed on whether opposition could be minimized by educating interest groups on how congestion pricing would benefit society generally and even groups within their own constituencies. Tradable Permits A new concept for applying congestion pricing would be for the government to distribute “tradable peak driving permits,” which individuals or groups could use or sell (Rom, Vol. 2). This proposal was put forward as a way of uniting interest groups and avoiding the political problem of converting a “free” road to a road with a congestion toll. With such a scheme, everyone would retain the privilege of driving during the peak that they now enjoy, and those that did not value this privilege as highly as its market price could sell it to someone who did. Instead of being “tolled

OCR for page 81
CURBING GRIDLOCK: Peak-Period Fees To Relieve Traffic Congestion off,” a user with a low value of time could be “bought off.” The concept has obvious appeal for overcoming political problems but should be studied further to determine its feasibility. REFERENCES Altshuler, A. 1990. Discussion of C. Winston, How Efficient Is Current Infrastructure Spending and Pricing? In Is There a Shortfall in Public Capital Investment? Proceedings of a Conference Sponsored by the Federal Reserve Bank of Boston, Conference Series No. 34, pp. 206–213. Bhatt, K. Forthcoming. Congestion Pricing: A Guide for Project Development. FHWA, U.S. Department of Transportation. Bishop, G., et al. 1993. A Cost-Effectiveness Study of Carbon Monoxide Emission Reduction Utilizing Remote Sensing. Journal of the Air and Waste Management Association, Vol. 43, pp. 978–988, July. Cambridge Systematics. Forthcoming. Impact of Urban Congestion on Business [NCHRP Project 2-17(5)]. TRB, National Research Council, Washington, D.C. Deakin, E. 1993. Policy Responses in the USA. In Transport, the Environment, and Sustainable Development (D. Banister and K. Button, eds.), E. and F.N. Spon, London. Gomez-Ibanez, J., and K. Small. Forthcoming. NCHRP Synthesis of Highway Practice: Road Pricing for Congestion Management: A Survey of International Practice. TRB, National Research Council, Washington, D.C. Toh, R. 1977. Road Congestion Pricing: the Singapore Experience. Malayan Economic Review, Vol. 22, pp. 52–61. Vickrey, W. 1993. Principles and Applications of Congestion Pricing. TR News, Vol. 167, July-August. Watson, P., and E. Holland. 1978. Relieving Traffic Congestion: The Singapore Area License Scheme. Working Paper No. 281. World Bank, Washington, D.C., June,286 pp. Weiss, C. 1972. Evaluation Research: Methods for Assessing Program Effectiveness. Prentice-Hall. Wohl, M., and C. Hendrickson. 1984. Some Practical Pricing Problems. In Transportation Investment and Pricing Principles, John Wiley and Sons, New York, Chap. 13.