Congestion pricing-charging a price to use highways that is high at peak hours and low at off-peak times-holds great potential for easing traffic congestion and reducing auto emissions in Southern California. New technology for nonstop electronic toll collection, coupled with increased public concern over congestion and auto emissions, makes implementation of such pricing more feasible today than in previous decades. But the political risks of charging for freeway use suggest the need for demonstration projects, both to introduce the public to the idea and to collect data so as to quantify its effects.
I. INTRODUCTION: WHY CONGESTION PRICING?
For more than three decades, economists have urged that direct pricing of road-use be employed in an effort to bring demand and supply into balance.,, To date, pricing for congestion-control (as opposed to the use of tolls to pay for road construction and operation) has seen only limited use, and only overseas. Singapore is thus far the only location of pricing instituted to limit vehicular traffic entering the central business district.
The implementation-or even serious consideration of the implementation-of congestion pricing in the United States has been held back by two problems, one technical and one political.
The technical problem has been the difficulty of instituting variable pricing with conventional methods-either toll booths or access-control stickers. Stickers-as used in Singapore-permit only a single price to be charged for access to a certain region or facility. Toll booths, in addition to being unpopular with users and causing additional congestion, do not lend themselves to variable pricing (being set up with fixed-price exact-change lanes, for example). The advent of automatic vehicle identification (AVI) systems eliminates this technical obstacle and makes it feasible to implement sophisticated pricing schemes in user-friendly ways.
But increased concern over vehicle emissions and congestion levels has made both ride-sharing and mass transit popular causes in the past decade. The idea that drivers should pay the full costs of their auto use has gained respectability-especially when seen in the context of achieving overall air quality goals. Increased awareness of the costs of congestion has diminished political opposition to the idea of congestion pricing.
These changes are beginning to affect transportation policy overseas. Norway is considering the conversion of its toll rings around central business districts from flat-rate tolls (for raising highway funds) to peak and off-peak pricing for congestion controls. Trondheim has implemented electronic toll collection, and Bergen and Oslo are converting their manual toll-systems to electronic toll collection, as well. The Dutch government has announced plans for electronic congestion pricing in Amsterdam, Rotterdam, Utrecht, and The Hague as part of its National Environmental Policy Plan to reduce urban air pollution. Singapore plans to convert its sticker-based central-business-district pricing system to ...
This is a preview of the whole essay
These changes are beginning to affect transportation policy overseas. Norway is considering the conversion of its toll rings around central business districts from flat-rate tolls (for raising highway funds) to peak and off-peak pricing for congestion controls. Trondheim has implemented electronic toll collection, and Bergen and Oslo are converting their manual toll-systems to electronic toll collection, as well. The Dutch government has announced plans for electronic congestion pricing in Amsterdam, Rotterdam, Utrecht, and The Hague as part of its National Environmental Policy Plan to reduce urban air pollution. Singapore plans to convert its sticker-based central-business-district pricing system to a full-fledged congestion pricing system using electronic toll collection. Cambridge and Edinburgh plan to be the first cities in the United Kingdom to implement congestion pricing on city streets, and London transport officials are also studying the idea.
II. THE NEED FOR DEMONSTRATION PROJECTS
The idea of charging for freeway use is still unfamiliar to most Californians and their public officials. Less familiar-even within the toll-road community-is the idea of using prices as a means of managing traffic demand (as opposed to simply a means of financing the road). The theoretical work carried out to date suggests that region-wide congestion pricing in Southern California would have large positive impacts on vehicle miles traveled and vehicular emissions. But even if the theoretical benefits were overwhelmingly persuasive, it is unlikely that state or local officials could be convinced to implement such a far-reaching measure on every congested freeway in the region.
Demonstration projects are therefore an attractive next step. There is much that economists and transportation planners still do not know about possible behavioural response to the choices posed by congestion-priced facilities. Also unknown are the political dynamics of congestion-priced projects: which groups will support and oppose such projects and why.
In selecting demonstration-project sites, the least desirable place to start would be on an existing freeway, no matter how congested. Putting a price on something that has traditionally been offered free at the point of use risks major public and political resistance, akin to that encountered when an existing freeway lane is "taken away" to create a high-occupancy vehicle (HOV) lane (e.g., the infamous Diamond Lane episode on the Santa Monica Freeway). The two best types of facilities for introducing the concept are:
(1)existing toll roads and bridges, where off-peak discounts and peak-hour surcharges can be introduced as fine-tuning the existing pricing in order to benefit users via reduced congestion and incentives to ride-share, and
(2)completely new facilities which give users a choice compared with existing, unpriced facilities. New toll roads, in particular, offer an ideal setting.
Designing and carrying out demonstration projects is important because there is a great deal that the transportation community does not know about user response to congestion pricing. A controversy has developed between traditional toll-road planners and economists modeling congestion pricing over the appropriate modeling techniques to use. Thus far, the financial community is very cautious about revenue projections based on anything other than traditional analysis using flat-rate tolls, because they have no empirical data on which to make judgements about congestion-pricing revenue projections.
Specifically it is unknown what the response of drivers in auto-oriented California will be to peak-hour pricing incentives. What fraction of users will shift their travel to off-peak times? To what extent will time-sensitive drivers be attracted to a less-congested highway? What fraction of people will opt for ride-sharing or transit, and how will this vary by income level and other socio-economic factors? To what extent will there be displacement of traffic onto non-priced or conventionally priced facilities? And finally, despite the expected environmental benefits (reduced emissions, reduced auto use, etc.), to what extent will environmental and pro-transit organizations support congestion pricing?
These significant unknowns can best be assessed by means of carefully designed demonstration projects. Because of the very large potential gains from congestion pricing, it is important that such experiments be designed and implemented in the near future. If congestion pricing on the entire freeway network truly is a more cost-effective way of achieving important transportation goals (e.g. increased vehicle occupancy, reduced vehicle-miles-traveled, increased demand for transit, etc.), then it is vital to quantify those effects in order that this information be available for use in transportation and air-quality planning.
This section discusses the theory of congestion pricing and welfare measurement under tolling. Congestion reduction and efficient resource allocation are cited as some of the main benefits of road pricing, particularly peak period pricing. Qualitatively, the idea behind congestion pricing is this: person R has a high value of time, person P a low value of time. Without pricing, persons R and P both travel at a slow speed. But if roads are priced, person R will be able to pay money and travel faster, while person P will not pay the money and not travel at that time (or travel on more congested and slower free roads). To increase the welfare of travelers (or potential travelers), the money collected needs to be redistributed to persons R and P in some fashion, either through lowering other taxes, through direct payments, or by reinvesting it in transportation. If person R's value of time saved plus the amount returned is greater than the amount paid, R is better off. If the amount of money returned to person P is greater than the cost of deferring the trip (or traveling at a slower speed), then P is better off. Road pricing will inevitably create both winners and losers (and usually losers) without redistribution of the toll that was collected. However, under the right redistribution policy, most people can be made better off.
The conventional explanation of road pricing uses a variation of Figure 1. On the y-axis is a measure of generalized cost (e.g. price plus monetized time), on the x-axis is flow in vehicles per hour. In the absence of any toll, equilibrium occurs at (Qo, Po), where demand intersects the short run average cost curve. Any traveler who values a trip more than Po will travel, anyone who doesn't won't travel. The shaded (red) area on the graph is considered the welfare loss, the benefit which is lost when tolls are not imposed. The loss is due to the difference between the cost a driver imposes on society (the short run marginal cost) by making everyone else's trip take a little bit longer, and the cost that driver bears personally which is spent in traffic congestion due to all the other cars on the road (short run average cost). The imposition of a marginal cost toll moves the equilibrium to (Q*,P*) and eliminates the welfare loss due to the congestion externality.
Figure 1: Optimal Congestion Toll and Welfare Loss Without Toll
However, the use of a single demand curve on the graph confuses the issue, since clearly moving from the short run average cost to short run marginal cost has a welfare implication. Raising the price lowers the demand, and thus the area that is conventionally thought of as consumer surplus gets smaller. But whether consumer's surplus gets larger or smaller depends on how individuals value the time savings. The conundrum results because individual drivers would pay more for a better level of service (LOS). Therefore the movement from short run average cost to short run marginal cost implies a movement from a demand curve with poor LOS (DF) to a demand curve with a better LOS (DA).
The demand for a graded commodity at a given price depends on the grade of the commodity. In the case where the commodity in question is road use, the grade is the level of service, the time it takes to traverse the road. At better levels of service (lower travel times) the demand will be higher at the same money price than at lower levels of service. We will describe LOS as ranging from SA to SF, with SA being best.
Suppose that there is some money price (a toll) charged by the agent managing the road, such that, even if the travel time is zero, the quantity demanded will be very small or zero. At a zero price, even if the travel time is small or zero, the quantity of travel will be limited. Similarly, there is a travel time at which demand will be small or zero, even at zero price.