Dealing With Traffic Congestion

Download as pdf or txt
Download as pdf or txt
You are on page 1of 45

Dealing with Traffic Congestion:

An Interdisciplinary Perspective

Yu-hsin Tsai

The author is a postdoctoral researcher at the Institute of Transportation Studies, University of California, Berkeley, CA 94720-1782. (ytsai@uclink.berkeley.edu). He holds a Ph.D. degree in Urban, Technological, and Environmental Planning from the University of Michigan, Ann Arbor.

Dealing with Traffic Congestion:


An Interdisciplinary Perspective

Roadway traffic congestion has been a common, irritating problem in many areas and countries. This problem has occurred in both rural and urban areas, and developing and developed countries. Time spent in traffic jam is widely considered as a total waste and irreversible. Traffic congestion influences quality of life, economic growth and environment. In the US the delay on urban freeways is expected to increase 360% from 1985 to 2005 (Institute of Transportation Engineers 1989). This includes an increase of more than 300 % in urban areas of the population of over one million, and 1,000 % in urban areas with population of less than one million. A variety of theories and measures have been developed and intended to solve this problem by transportation planners, urban planners and transportation economists. Different combinations of these three groups of measures were applied to deal with congestion, but unfortunately few cases show congestion was completely eliminated, in particular in the long run. Experience suggests there seems no penicillin that works for all congestion cases. The major purposes of this literature review paper are to understand the causes of traffic congestion, the measures to dealing with traffic congestion from the perspectives of transportation engineers, urban planners and transportation economists, and the interactions between these groups. Through the analysis of the causes of traffic congestion, this paper is intended to shed some light on the question Is traffic congestion an unavoidable condition in urbanized areas in particular? The analysis of

congestion-alleviation measures proposed by transportation engineers, urban planners and transportation economists, provides the information on their pros and cons. Finally, this paper presents the issues regarding the conflicts and collaborations between capacity expansion, demand contraction and pricing. For the sake of analysis, capacity expansion, pricing are assumed to be supported mainly by transportation engineers, and economists; on the other hand, land-use measures are demand-contractionoriented1.

Introduction

Congested highway2 can be considered as club goods (Cornes and Sandler 1986, Levine 1998) specifying the characteristics of impure public goods shared by members of voluntary collectives. In a strict sense, highway is a not a public good since, first, it is excludable3 (e.g., relatively high price of cars prevents the poor). Second, rivalry4 occurs when traffic increases to the point when traffic speed is slowed down5. This rational justifies the government intervention (consisting of market-force-based and regulation-based measures). Intuitively, time wasted and discomfort caused by traffic congestion may be

A great many of demand-contraction measures, however, are proposed by transportation engineers. See details in Demand-contraction section. 2 Highway applies to roads, streets, parkways and highways (Braum et al 1994). 3 Nonexcluxible means that it is not possible (at least at reasonable cost) to exclude consumers who do not pay the price from consuming the good or service (Fisher 1996). 4 Nonrival means one additional person can consume the good without reducing any other consumers benefit (Fisher 1996).
5

Transportation economists define this point as road capacity.

considered as a waste and had better be avoided. Indeed quality of driving constitutes part of life quality, and hence mobility can be regarded as one goal of transportation. Accessibility, however, is prevailingly believed by transportation planners as a major goal of transportation (Johnson 1998), particularly where there are conflicts between mobility and accessibility (see footnote 3). Since most demand for travel is derived from the demand to accomplish certain social or economic activities, lower cost per trip, instead of merely lower cost per mile is the objective of transportation service. Hence, even within the scope of transportation, congestion can be an inevitable tradeoff for better accessibility. In the context of sustainable transportation6, new approaches to congestion mitigation are required. In the 70s and 80s when interstate highway were overwhelmingly developed (the era of the freeway could date back to the Interstate Highway Act of 1956, traditional concerns of mobility and capacity dominated and the movement of vehicles was the focus, rather than that of persons. The underpriced road transportation in part due to lack of accounting for the externality of congestion and environmental pollution, and subsidy (e.g., subsidy on single family housing) led to the overuse of road network, urban sprawl, and air pollution. The experience in these two decades suggested that Americans could not build their way out of congestion through building highways (Transportation Research Board 1994). The desire for changes in transportation policy due to recognition of these external costs of the road system and
6

A sustainable transportation system is one that: 1) allows the basic access needs of individual and societies to be met safely and in a manner consistent with human and ecosystem health, and with equality within and between generations; 2) is affordable, operates efficiently, offers choice of transport mode, and supports a vibrant economy; 3) limits emissions and waste within the planets ability to absorb them, minimizes consumption of non-renewable resources, reuses and recycles its components, and minimizes the use of land and the production of noise (The Center for Sustainable Transportation 1997).

was respected in Intermodal Surface Transportation Efficiency Act of 1991 (ISTEA), and is consistent with the spirit of sustainable transportation. The spirit of sustainable transportation and ISTEA is expected to lead to more comprehensive balance7 among society, economy and environment (The Center for Sustainable Transportation 1997, World Bank 1996). In other words, in planning and assessing a transportation plan, all benefits and costs (internal or external, direct or indirect) should be accounted for in the decision making. In addition, the current transportation objectives per se, focus on accessibility, rather than speed (or mobility) and capacity. This means movement of persons rather than vehicles is of major concern. As a result, the throughput of highways focuses more on persons than vehicles per lane per hour. An underlying idea is to seek alternatives of capacity expansion, such as alternative transportation modes, and land-use or economics based measures to improve accessibility, even at the cost of lower mobility. For example, if the introduction of a mall with mixed land use, or a pedestrian or bicyclist friendly environment could improve the residents overall accessibility, but at the cost of reduced speed on road for motor vehicles (the former could generate more traffic, and the latter could reduce road capacity), it could still be an acceptable measure. Consequently, worsen congestion could occur but still be acceptable if it is the result of the trade-off for better balance among society, environment and economy, or for better overall accessibility. Though, highway congestion is still an irritating condition, and the measures to calming congestion should still be taken as long as they do not contradict sustainable transportation. The influence of traffic congestion on transportation can be described through
The notion of balance is an ambiguous notion, and can be different from case to case. For example the cost of air pollution could be valued less for the countries seeking economic growth. Hence their balance system may lean
7

accessibility and mobility. For a highway system, better mobility can cause worse accessibility; on the contrary better accessibility may result in worse mobility. Transportation facilitates household activities by providing accessibility to different activity locations, such as work places, schools, shopping malls, and recreation areas; transportation also facilitates business and delivery for industries. Accessibility of an individual location can be simply measured as the number of opportunity for travel objective fulfillment that can be reached within an acceptable travel (Lomax and Levinson 1997) or it could be measured as simply as cost per trip. Mobility is the ability of people and goods to move quickly, easily and cheaply to where they are destined at a speed that represents free-flow or comparably high-quality conditions (Lomax and Levinson 1997). Mobility can be characterized as speed of movement, and can be measured as cost per mile or mile per hour. Both the travel distance between the locations of two activities (i.e., land-use factor) and the travel speed (i.e., transportation factor) influence accessibility. A location with high accessibility to another place may be constituted by close proximity or (and) high mobility (i.e. high speed). If the trip length is fixed, higher mobility (i.e., higher speed) leads to higher accessibility (shorter travel time). Under this condition, congestion will degrade accessibility. In contrast, higher accessibility can reduce mobility if trip length is shortened because of changes of locations. For example, the introduction of a local shopping mall into a community could increase local traffic and lower the mobility, but the accessibility could be improved due to shortened travel distance8 for shopping purposes. In sum, congestion is an issue when obstructing accessibility (when travel distance is fixed); but congestion may be
more to economic development.

more acceptable as a result of better accessibility. Congestion has been defined differently by different persons or from different perspectives. For travelers congestion is immobility (Levinson et al, 1997). The traveler-based definition of congestion tends to be subjective since it is subject to their internal rulers. This internal ruler is influenced by personal, socioeconomic and temporal characteristics, so it is very likely that traffic congestion for one person may not apply to another. For example, surveys were carried in Japan by different organizations to define congestion, and their respective definitions are different to certain extent (Hashimoto 1990). From the transportation perspective, congestion is defined as when speed is lower than the speed at the design level of service9 in the U.S. (Pline 1992). This kind of definition is more engineer-oriented though it was initially expected to take users perception into consideration. Furthermore, the acceptable and unacceptable congestion were initially developed in Congestion Management System regulation (Federal Register 1993) . While the speed is still higher than at design level of service but lower than free flow, it is called acceptable congestion. The unacceptable congestion indicates there is a problem, and actions should be taken. From the perspective of transportation economists, the definition of congestion is travel time or delay in excess of that normally incurred under light or free flow travel condition (Levinson et al, 1997). In other words, if increasing traffic flow increases travel time or cost, there is congestion (Fisher 1996). This notion of acceptable and unacceptable

This credit of the generation of this idea is attributed to Levine, Jonathan. Expected speed of a road is influenced by two variables: designed speed and designed level of service (Pline 1992). Designed speed is the maximum safe speed that can be maintained over a specified section of highway, which is a design factor of the road. Different roads can have different designed speed. Level of service is defined as those operational conditions within a traffic stream as perceived by users of the traffic facility. Levels of service are based on density of the mainline traffic, and can be presented by Volume/Capacity. The design level of service can vary
9

congestion can also apply to transportation economists definition. The unacceptable congestion occurs when the travel cost or time is higher than the optimal point where marginal cost equals marginal benefit (Appendix 1). For achieving maximum benefit of society, this economics-based definition of congestion is chosen for this paper. There seem some essential factors that contribute to the mystery of highway congestion that few efficient measures can address. Traffic congestion is a reflection of relative shortage of capacity (supply) in association with travel demand. The underlying variables for this phenomenon are enormous and variant, some of which are avoidable or treatable, others are unavoidable or hard to overcome. Experience suggests that in existing congested urbanized areas, there seems to be no long-term solutions. This paradox drives our curiosity toward the question Is congestion unavoidable and can be regarded as a normal condition in urbanized areas? This research is not intended to find the answer but to shed some light on this paradox based on previous researches. From the perspective of travel demand, a group of essential variables is associated with rapid growth in already urbanized areas, such as population and economy growth, opposed to limited available land. These factors generally are regarded as natural trend sin human society and as a given condition to planners. Another group of variables is distribution associated (in terms of spatial and temporal dimensions) variables: overconcentration of trip origins and destinations caused by land use and density, overconcentration of trips on shortest routes, same periods of time and auto use. In practice, traffic is not evenly distributed throughout a day or days, and on different routes. This unevenly distributed traffic condition raises the issue about what is the optimal road capacity (Hay 1973). Within a day, for example, adequate provision for all demands
by types of roads, where the road is located (Pline, 1992) and time of day or year (Lomax, et al, 1997).

can only be ensured at the price of excess capacity during some other periods; while alternative full utilization of capacity can be achieved only at the cost of leaving demands unsatisfied. So if the capacity is set up and built at the point of financial balance, then the peak-hours, days, or seasons are certainly facing congestion naturally. From the perspective of highway supply, the capacity is decreased by unavoidable conditions due to ambient environment, accidents and routine maintenance. A negative spiral results as the space saved by congestion-alleviation measures other than pricing strategies is usurped by the latent demand. Capacity-usurping effects involves triple convergence10, swamping effect11 (Downs 1992), and change of other travel behaviors (e.g., changes of destination, vehicle-occupancy and trip frequency). In addition, congestion could be a result of good quality accessibility. Sometimes congestion is unavoidable since high quality accessibility attracts more residents or industries. Also, for the sustainable environmental reason a higher level of congestion is accepted. Finally, users (or transportation engineers) definitions of congestion are likely to be different from economists. That is, congestion felt by users might still be regarded as acceptable by economists. All these variables point to highway congestion as a natural social system condition. That is, as long as people follow the same clock or calendar, seek the shortest route and most accessible locations, use the same, convenient travel mode, then under limited resources, congestion seems unavoidable. Congestion could be regarded as a factor for people to trade-off against other factors, such as economic

The triple convergence principle states that any large initial reduction of peak-hour travel times on a major limitedaccess roadway will soon be offset by the subsequent convergence on that roadway of drivers who formerly (1) used alternative routes, (2) traveled at other times, or (3) used public transit (Downs 1992). 11 The principle of the swamping effect of rapid growth states that relatively small reductions in initial traffic congestion in a rapidly growing metropolitan area will be fully offset within a few years by the arrival of more people, jobs, and vehicles there (Downs 1992).

10

development and sustainable environments. This, however, does not mean there is no need to take an action to alleviate traffic congestion. Measures for calming congestion that can improve the efficiency of road network, as well as achieving sustainable transportation should still be taken.

Measures for Congestion Relief

Understanding the variables influencing miles traveled can help understand how congestion relief measures affect traffic. Miles traveled is the quantity of road consumed by travelers, which is determined by supply and demand for highway altogether. Miles traveled is basically composed of personal and industrial trips. For the personal trips, the function of miles traveled can be presented as:

Miles traveled = Population * Vehicle ownership12 (vehicle/population) * Travel mileage per vehicle13 (mile/vehicle)

The variables that influence population, vehicle ownership, and miles traveled per vehicle are numerous, varied from time to time, and from place to place (Downs, 1992). The underlying factors that influence these variables invlove population growth, technology development, economic development, peoples preference and land supply. For example, both population immigration and population fertility increases will increase population. Increasing income may increase buying power of customers, and
12 13

The maximum vehicle/population is close to one (Downs 1992). Mile/vehicle is quite constant (Downs 1992).

technology progress will lower car prices; as a result car ownership may increase. The preference for driving alone is likely to increase miles traveled per vehicle. In terms of industrial trips, population growth, economic development, and business style influence miles traveled. Regarding business style, for example, home shopping and mall shopping will create different miles traveled. Most of the congestion alleviation measures are developed from three main perspectives: economics, land-use and transportation engineering. The scope of economic measures involve measures using price mechanism (Institute of Civil Engineers, 1989); the land-use related measures focusing on changing land-use policy, considered as one kind of demand contraction; the transportation measures contain capacity expansion and demand contraction strategies. Evidently more than one measure or measures of more than one perspective are applied. The measures derived from these three different perspectives are based on different theories, require different ways and different amounts of budget for implementation, have different degrees of effects, and could cause different side-effects or problems. The following contain details of the above measures, organized as capacity expansion (transportationengineer perspective), demand contraction (consisting of land-use- and transportationengineering perspectives), and pricing (transportation-economics perspective).

Capacity Expansion Providing more road capacity is probably the most intuitive idea for alleviating congestion. The variables that may influence the supply of highway are composed of

10

temporary and permanent variables. Road capacity14 is the maximum number of vehicles that can reasonably be expected to use the facility in a given time period under prevailing roadway, traffic, and control conditions (Pline 1992). So a road capacity may decrease if the roadway, traffic or control conditions gets worse. Roadway condition contains geometric and design elements such as width of lanes, lateral clearance, and horizontal and vertical alignment. These factors sometimes may affect speed or delay time, while not affecting the capacity (United States, Federal Highway Administration, Office of Traffic Operations 1994). Traffic condition includes vehicle types and lane or directional distribution. Control condition includes traffic control, curb parking, pedestrian ways, restrictions on vehicle types, ramp metering, posted speed limits, bus stops, and parking regulation. All the above elements are design elements, most of which will influence the design capacity of a road. On the other hand, some conditions will reduce road capacity temporarily, and cause congestion irregularly. These variables include road maintenance, car accidents, break-downs of electronic traffic signals, and ambient environment including wet, snow covered pavement, darkness, and fog. Supply-side congestion alleviation measures are composed of three types: building new roads or road expansion; improving roadway conditions, adjusting to traffic conditions, and improving control system; in the case of temporarily reduced road capacity, restoring the designed capacity. First, the most fundamental one is building new roads or road expansion. This may not be suitable for developed areas since it is hard and costly for land take. Besides, it needs longer time to build physical facilities (Downs 1992). This method, however, could be welcome in areas intended to grow
14

The unit of road capacity is passenger car or passengers per hour per land (Transportation Research Board 1994).

11

since it allows more traffic. To increase the capacity of existing roads there are numerous methods of improving roadway, traffic, and control conditions. About roadway condition, certain design elements could be modified. For example, city streets can be upgraded by widening their lanes. About improving traffic condition, for example, the numbers of the lanes of two opposite directions could be adjusted according to traffic condition. About control conditions, for example, Advanced Traffic Control (ATC) of Intelligent Transportation System (ITS)15, automated highway16, and electronic turnpike system17 could smooth traffic flow, and hence the road capacity increases indirectly. Finally, as for temporarily reduced capacity, the measures to bring it back to original capacity level include: using roving response team or surveillance system to clear roadways quickly after traffic accidents (Systems and Detroit 1996); improving process of highway maintenance. However, Downs (1992) argued that the congestion problem could only be relieved temporarily, and in the long run the congestion problem would recur due to the principles of triple convergence and swamping effect. Hansen (1995) provided evidence that at the metropolitan level in U.S., 1.0 percent increase in lane-miles soon induces an immediate 0.2 percent increase in traffic (Vehicle Miles Traveled) and 0.9 percent increase in traffic four years after. In fact, this represents the popular belief of

FAST-TRAC is an Intelligent Transportation System (ITS) that integrates advanced traffic control with a variety of advanced traffic information systems through centralized collection, processing, and dissemination of traffic data. (University of Michigan, ITS Lab 1996). 16 The National Automated Highway System Consortium is constructing an automated highway demonstration project on a stretch of interstate 15 between San Diego and Escondido, Calif. Equipment in and along the roadway will be used to regulate the speed, steering and spacing of vehicles, creating a sort of automated train of cars and eliminating the stop-and-go conditions of congested highways (ITS Online, 1999 ) 17 The E-Z Pass in the New York Metropolitan area is using electronic scanner as the car rolls through the tollbooth (New York Time 1997).

15

12

transportation planners that adding road capacity does little to decrease congestion because of substantial induced traffic. In addition, in the U.K. most city centers are congested but it is now generally accepted that solution is not new road building (Younes 1993). Nevertheless, Singapore, for instance, is still planning to build 225 lane-kilometers (or 141 lane-miles) claimed in its 1996 White Paper (Phang and Asher 1997). Is Singapore not going to experience capacity-usurping effect, or what will be the reasons underlying its policy of capacity expansion? Here, we might need to look at two things: the optimal investment of roads, and capacity-usurping effect. The optimal scale of road capacity would be acquired when long-run marginal cost equals benefit (Mohring and Harwitz 1962; Hau 1992) (Appendix 2). In other words expanding a road until the additional benefit equals the additional cost of building it would yield maximal net benefit to the community. This might justify building new roads theoretically. One issue here is the other external costs are not counted. Another concern is the side effect of capacity expansion, e.g., high lane-mile elasticity of VMT. Younes (1993) research shows weak capacity-usurping effects were found in his eight case studies in Britain, Germany and Sweden with relatively new road building within the past 2-15 years. In these case studies, traffic flow did not increase significantly enough to remain previous congestion nor did it lead to a significant modal shift from public to private transport18. He discovered building of new road schemes in urban area could still be effective and did not necessarily lead to adverse and worsening conditions, when coupled with appropriate complementary policies (such as improved public transit, parking control and bypasses). This result is consistent with a

18

This paper does not provide quantitative evidence, such as change of speed.

13

case study on the short run impact of a ring road in Amsterdam, Netherlands19 (Kroes, et. al, 1996). The capacity-usurping effects could still hold in the cases of Younes study as long as growth continues in these cities in the long run, but the weak capacityusurping effects strengthens the support for capacity expansion. The ideas underlying the differences between the U.S. and the Younes cases could be explained by the degree of lane-mile elasticity of VMT. In areas (1) with few alternatives of driving (such as convenient, cheap public transport, pedestrian friendly environment), (2) with autooriented land use (such as low density, unbalanced, separate land use), (3) trend of growth (e.g., such as cities experiencing the effect of agglomeration, or fast population and economic growth), and (4) lower cost of driving, the momentum of traffic expansion to fill the space available to it is stronger. In the American metropolitan areas, where roads have been overused, and urban sprawl continues due to the failure of transportation pricing, convenient road network, cheap cars and gasoline, subsidies of single family housing, and relatively cheap land, new road is obviously not a suitable plan. If transportation pricing remains implausible, building new roads can not relieve congestion, but lead to more unbalanced, low-density cities (more unsustainable environment). In turn, this condition may also lead to further demand for driving (i.e., negative spiral). But in those European cities, where induced supply did not induce significant demand, building new road may be an effective measure to calming congestion. Younes attributed this success to complementary transportation policies, efficient public transport and traffic management measure. In sum, capacity expansion as a means to relieving congestion is generally doubted, only not if implemented in

19

This result from this research may deserve less credit since its survey for the impact of the road opening was

14

areas with low lane-mile elasticity of VMT. This conclusion leads to the need for further research on capacity-usurping effect from counties other than U.S.

Demand Contraction: Several conditions encourage the use of demand-side measures (e.g., transportation demand management (TDM)) to relieve congestion. In many areas, It is physically impossible to enlarge current road networks, or congestion pricing is politically or socially infeasible. Besides, English drivers reported that 30% of car mileage, and 43% in Los Angeles, was not at all or not very important (Kessler and Schroeer 1995). These situations encourage applying TDM for more efficient use of current highways20. The fundamental ideas of demand contraction are composed of reducing total trips made and redistribute trips more evenly or efficiently in terms of times, modes and routes of travel. This section is divided into non-land-use- (basically transportation-engineering-oriented) and land-use-oriented for the sake of clarity.

Non-land-use measures: About decreasing travel demand, examples include encouraging people to work at home, and substituting trips with telecommuting technology (Kessler and Schroeer 1995), provision of better public transit, and encouraging ride-sharing (Giuliano 1992). The disadvantages of encouraging working at home include it can not work for all types of work since a lot of works need face to face contact. Telecommuting technology as a means to congestion alleviation may be

conducted only two months later.

15

criticized for its indirect effect, i.e. stimulating more trips due to more social or economic interactions. Next, about redistribution of trips, they will improve the efficient usage of road network or increase the usage of environment association (on foot, bicycle, public transit). For example, high-occupancy vehicle (HOV) lane has been in place on many highways, such as Route 55 in southern Calnifornia (Giuliano 1992), which are designed to encourage more people to use ride sharing by providing the privilege to less congested lane (i.e., HOV) (Downs 1992). Many transportation departments also provide ride-match service for carpooling (Washington 1999). Measures to encouraging the usage of pubic transit are numerous such as park-and-ride, user-responsive system, exclusive bus lane, and subsidies. For instance, establishing exclusive bus lane in Taipei Taiwan, Toronto Canada, and Zurich, Switzerland (FitzRoy and Smith 1993) is to encourage travelers to use bus by providing better mobility. In Zurich, Switzerland, bus ridership increased 33 percent from 1985-1990 attributable to exclusive bus lanes and automatic traffic light signaling that provided absolute priority for public transportation at intersections. However, car numbers on main roads had been stable from 1981-1990, which indicated the shift from car to bus seemed marginal. Building the off-road transit systems, such as people mover, light rail system is good for population dense areas or cities. Provision of convenient public transit, however, does not necessary lead to shift from cars to transit. Younes (1995) found the provision of rapid rails of Victoria Line in London, of S-Bahn and the U-Bahn in West Berlin, Germany only had marginal relief from road traffic since there was insignificant shift from cars to rail, but substantial shift

The approaches to increasing efficiency of existing road usage also include supply-side measures (excluding building or expanding roads) and pricing.

20

16

from bus to rail. Improvement of parking policy could also help relief congestion. Shoups case studies of eight firms that have complied with Californias cash-out requirement, found the solo drivers to work fell by 17 percent, and the number of carpoolers, transit riders, and bikers or walkers increased by 64, 50 and 39 percent, respectively. Also a before-and-after study within a city district of Munich, Germany (Topp 1993) shows the effects of residential parking permits on the modal choice of employee. The share of solo-drivers dropped from 44 to 32 percent, while the share of environment association rose from 54 to 64 percent. Without reducing total traveled mileage, diverting traffic into non-congested time periods and routes are other alternatives for congestion alleviation. In practice, traffic is not evenly distributed during different periods of a day, or weekdays and weekends, and on different routes. Several measures could redistribute the demand into alternative time periods or alternative routes. The measures include staggering working hours (Downs, 1992), route guidance system, the real-time traffic information of Advanced Traveler Information Systems (ATIS), by-pass, and road pricing (described in the economic perspective section below). For example, a research based on simulation programming suggested real-time en route information could reduce the variance in travel time considerably in the context of recurrent congestion (Emmerink, et al. 1995). There are some points requiring notation. Among the measures to redistributing trips to different times, routes or modes, some researches found mode change was less likely to occur. The examples include the provision of rapid rail transit in London, U.K. and West Berlin, Germany (Younes 1995), and the improved mobility of bus system in Zurich, Switzerland. Besides, the capacity-usurping effect also threatens the result of

17

initial reduced traffic by demand contraction, in particular in an over-saturated system.

Land-use measures: Land use is said to be a common way in solving traffic congestion by many transportation planners (Levine, 1998), in particular when capacity expansion measures barely succeed. The relationship between land use and transportation is complex. Land use is an essential factor influencing trips generation, trips distribution, and mode split, or generally called travel (Stover and Koepke 1988); on the other hand, transportation can also influence land use pattern and intensity. For instance, in low-density areas, personal vehicles tend to be more convenient; mass transit is more feasible in high-density areas; mixed land use21 may encourage pedestrians and bicyclists. Furthermore mismatches between land use and transportation system or between different types of land uses may cause congestion problems. For example, high-density areas with insufficient mass transit may cause serious congestion problems; insufficient affordable housing in the nearby of jobsites is likely to increase VMT. Due to the close relationship between land use and transportation, land use measures are largely believed as a significant means for congestion alleviation. Land-use measures to preventing traffic congestion are demand-side measures. They contain Neotraditional Town Planning components22 (e.g., higher density, easier access to mass transit systems and mixed land-use) (Crane 1996), balanced land use,
In Taiwan, stores are located on the lower levels of condominium and apartments, and residences are located on the upper floors. This extremely mixed land use style provides high accessibility of certain services for this population dense country. 22 A growing numbers of urban planners argue that new suburban neighborhoods should be designed more like those of yesterday to get drivers out of car (Bressi 1992). The components of this Neotraditional Town Planning are composed of transportation (such as gridiron and radial streets instead of cul-de-sacs, narrower streets) and land use (such as mixed land use, higher density, easier access to transit) factors.
21

18

and growth management. These measures are explained in the following. The major purposes of maintaining density in certain range are threefold, at least. First, it can support the sustainable public transit, carpooling, biking and walking, and preventing from low-density urban expansion, which encourage more usage of cars. These measures include maintaining minimum residential and commercial density, concentrating jobs in large clusters outside of downtown (for reducing driving to work), and cluster high-density housing near transit or commute rail stations (for reducing home-related trips by cars) (Newman and Kenworthy 1991, Downs 1992). In this way, the demand for road may decreases since more trips are served by public transit or ride sharing. With regard to mixed land use, it may range from mixed land uses in a community, mall, block, to in a building. This is another demand-side measure to alleviating congestion, which functions to reduce trip distance and number of trips since the origin-destination distance has been reduced and many activities can be conduced in the same destinations (say, a shopping mall). Whether these Neotraditional Town Planning components are able to calm congestion is uncertain. Through the implementation of integration of residential and commercial locations in neighborhood, higher density land use around bus stations, a more walking, biking and transit-oriented land use is constructed. The increasing use of walking, biking and transit is expected to reduce the car usage. (Cervero and Gorham 1995) found transit in some cases of their research ridership was higher in transitoriented environment than freeway-oriented. Their research applied a quasi-experiment on transit oriented and freeway-oriented neighborhoods (micro-environment) in the same metropolitan area; two metropolitan areas in California were selected: San

19

Francisco Bay Area with transit-oriented regional patterns (macro-environment), and Los Angeles area with freeway-oriented pattern. Six out of seven transit-oriented neighborhood in San Francisco Bay Area showed higher transit ridership rates than their freeway-oriented counterparts, but all six pairs in LA did not support this relationship. This research, however, does not lead to the conclusion that Neotraditional planning could calm congestion yet. The increased accessibility of walking and biking (due to shorter distance), and higher transit ridership may divert some drivers from cars and hence reduce demand of car at the outset. In the long run, whether the overall trip generation is reduced is ambiguous (Crane 1996). On the one hand, the increased accessibility (or increased speed) may trigger latent demand (from triple convergence, higher travel frequency). On the other hand, the reduced cost in the time and convenience required for trip on foot will both increase the attraction of walking and other modesthe substitution effectand also increase the amount of time available for travel by all modes. Even if the overall trip generation is reduced, the congestion will be reduced or not is still uncertain since space previously used for road could have been substituted for pedestrian and bicyclists. Regarding jobs-housing balance, all else being equal, increased supplies of affordable housing in the vicinity of employment center should fill up largely with households whose members work nearby, as they would tend to outbid similar households employed more remotely (Levine, 1998). In this way, the distance of work trips can be reduced, and the inter-area trips are likely to be reduced. As a result, the demand for travel will fall, and consequently the traffic demanded will be reduced. The role of jobs-housing balance to clam congestion is skeptical even the balance

20

is achieved, in particular there being many obstacles to its implementation. In 1989 Cervero proposed jobs-housing balance as a strategy for reducing peak-hour traffic congestion in American cities. Through the reduction of work trip distance and perhaps switch to walking and biking, internal trips would possibly increase and VMT could possible reduce initially. One successful case is in Toronto, Canada (Nowlan and Stewart 1991, Cervero 1996) showed that serious traffic problems were averted in Torontos central core despite an office building boom in the 1970s and 1980s, through accelerated downtown housing construction. Levine argued (1998) jobs-housing balance would contribute little to calming congestion, since the second run of its impact iteration could come the capacity-usurping effects derived from triple convergence, increasing trips, travel frequency, or new demand from the new locators. Gordon and Richardson (1989) further argued that jobs-housing balancing had little effect on nonwork trips (which already accounted three quarters of all trips in U.S. and the majority of trips during peak hours). In fact, there seems little empirical evidence showing congestion is less serious in better jobs-housing balanced areas. Besides, more concerns have been focused on the obstructions to jobs-housing balance. The success of this measure largely depends on the residential location selection of the households (demand-side perspective of housing) and the availability of affordable housing (supply-side perspective). The ideology of residential location choice-the workers choose homes as close to their jobs as possible (Giuliano 1991) (given others equal) -is unsupported due to tradeoffs with other goals of the household (Levine 1998). The logic is that workers still choose homes as close to their jobs as possible, but have to account for other goals, and to have it balanced with other goals or

21

limitations. The variables influencing residential location choice model are numinous: probably including some trendy ones such as increasing two-worker households, frequent jobs turnover, affordable housing, exclusionary zoning; other general ones such as school district consideration, larger house sites, socioeconomic (e.g., same social class), or environmental (i.e., living near a park); other individual considerations such as closeness to relatives, living with parents or in inherited housing; or planning failure such as lacking of phased infrastructure in nearby residential areas, or road expansion leading to other than nearby areas (Giuliano 1991, 1995, Downs 1992, Cervero 1996, Levine 1998). Levine (1998) further argued that land use regulation, even designed for achieving jobs-housing balance, could deter the balance due to insufficient supply of affordable housing nearby (caused by government or planning failure). In this regard, market force may work better than land use regulation. In sum, seeking jobs-housing balance is merely one goal in residential location choice, and consequently achieving minimal travel cost alone is not consistent with practical condition. Different from the above three Neotraditional town-planning elements, growth management does not eliminate the demand for development and its derived traffic. It transfers, however, the development to other nearby areas (i.e., imperviousness principle) (Downs, 1992). Hence, planners should take actions to deal with the transferred development and the derived traffic. Otherwise, it would just transfer the congestion problem from one place to another. Besides, since the demand may still exist, it may bring a negative effect on other people, say middle and low-income households. For example, the increasing demand for household will increase housing

22

prices. Thus, the high-income households could outbid the middle- and low-income households and they will be then expelled through market force. The success of land-use measures23 to alleviating congestion is also challenged by the capacity-usurping effect. Besides, even if the land-use measures could successfully alleviate congestion, the effects could only be observed long after the implementation. In general, in highly urbanized areas, the current land use could not be changed dramatically, in particular in the short tun. This kind of measures take more time to bear fruits, but once it is implemented, it serves long-term function since generally land use changes at a relatively slow pace. Hence, it is more like a means to preventing from future congestion than to solve existing congestion. Inappropriate current land use plans, on the contrast, could cause future congestion as presented above.

Congestion Pricing One primary principle of economists in dealing with problems is to use voluntary market forces, instead of compulsory administrative regulations. For the purpose of a distinctive perspective boundary from the transportation-perspective measures, the economics-perspective congestion alleviation methods, here, are defined as those means that use price-mechanism of roadway traffic. There are various pricing associated measures, but in this section the focus is on the congestion pricing due to

The credit of land-use measures may exist in improving better accessibility if it succeeds. These measures are particularly good for population dense countries or areas with limited land supply. This kind of measure may not work well for people who prefer country-style, low-density, and separate-land-use residence. The worse condition of these tactics could be in a transit-oriented land use city, the cars are used as primary transportation vehicle, such as some big, population dense city, without proportional transit systems.

23

23

three reasons. First, its theory would cover the major part of other similar measures; second, it influences the congested traffic directly; finally it is one of the most promising measures, in particular capacity-usurping effect has threatened the capacity-expansion and demand-contraction measures. The other pricing related measures contain pricing on complements of road usage (such as taxing on gas and plate), and subsides to its substitutes (such as other transportation modes). In order to maintain traffic speed at the Pareto efficiency when marginal social cost equals marginal benefit, price mechanism is applied to internalize social (externality) caused by congestion (Fisher 1996). This economists theory is derived from the failure to make drivers bear full costs they generate. In this system transportation is taxed primarily as a consumption good (Institute of Transportation Engineers, 1989). About the waste of time as the external cost of congestion, it is fair among all road users since all road users share this cost24 by being slowed down by the congestion. However, it causes the inefficiency of the whole transportation system. At this point traffic flow is higher than at the optimal point. This externality justifies the government invention, for which congestion pricing is probably economists favorite. In addition, jurisdiction of metropolitan areas is an appropriate to implement congestion pricing (Downs 1992). As for the calculation of congestion pricing, some transportation economists provide some theories and formula, such as Decorla-Souza and Kane (1992) (Appendix 3), Borger, et al (1996). In terms of the advantages congestion pricing, first, it could be fully applied-at least technically, if not politically-within a relatively short-time period, say no more than five

This external cost of congestion differs from those of air pollution, noise and vibration, which are shared by nonusers too.

24

24

years (Downs 1992). This is because none of the tactics involving changes in residential or job density or location has that trait. Second, it could be a more reasonable pricing system with the application of ITS technologies, because it could charge commuters according to the car size, weight, and the traffic condition. Next, with this flexibility of floating rates, different levels of congestion can be dealt with at different congestion prices. Furthermore with the advancing ITS technologies, it could be applied to all congested road section, no matter highways or local. About its effects, its initial congestion-reducing effects may not be offset by capacity-usurping effect. In contrast, congestion pricing can divert the transportation demand to alternative times, routes, or modes. Also, It may immediately affect all peakhour movements on main arterial (Downs, 1992). Furthermore, it affects not just workrelated trips or local trips. Besides, in this system people can make their own choice about when, which route or which transportation mode to use. Finally, a side benefit is that it could have income distribution effects if the revenue was used for other nonusers (Hay 1973). All these advantages explain why academics strongly support congestion pricing.

Congestion pricing, however, had experienced significant obstructions from politicians and the generally public, as opposed to the support from academics (Verhoef and Piet, 1997). Since the emergency of theory of congestion pricing in the 1960s, it has been in practice in only a few cities around the world. Singapore started its pricing system since 1970s, and 1995 the system has developed into one area (Area Licensing Scheme) and one liner (Road Pricing Scheme) congestion pricing (Phang and Asher

25

1997). A electronic road pricing (ERC) was scheduled for implementation in early 1998, which was planned to handle vehicles traveling up to 120 (Km/Hr) when passing charging points in a multi-lane environment. After a two-year (1983-1985) experiment with electronic road pricing, Hong Kong dropped out due to the issue about invasion of privacy (Transportation research Board 1994). In cities of Bergen (1986), Oslo (1990) and Trondheim (1991), Norway a pricing system was used for raising funds for needed road improvements, but not based on congestion. The San Francisco Bay Area proposed to increase the toll for using the San Francisco-Oakland Bay Bridge during the peak period to shift traffic to off peak and to transit. In Southern California, it was proposed that solo drivers could pay to use HOV lane on certain highways. Proposals or researches were brought up for congestion pricing in London and Cambridge, England. The obstacles to congestion pricing involve the issues of individual loss, redistributed welfare from road users to the rest of the community, and perverse incentives for government. In theory(Gomez-Ibanez 1992) stated that three groups of people would benefit and four groups would lose from congestion pricing. The wining groups were composed of the (tolled-on) motorists whose gains from improved traffic speeds outweighed the toll cost, travelers who would use bus or HOV, and government (Hau 1992), or recipients of toll revenues. The four losing groups consisting of the tolled-on motorists who had relatively low value on travel time, the tolled-off motorists to other roads, time or modes, the users of the competing roads or time, and those who chosed not to travel. (Hau 1992) attributed the failure of the implementation of congestion in part to this condition that more people lose from congestion pricing. This condition also brings

26

in two inequity issues: first, the poor are tolled off, and the second is the payers (tolledon) are not the same as beneficiaries (Evans 1992). To abandon the free access of roads in favor of access by ability to afford does not allocate to need (Downs 1992), so congestion pricing may penalize the poor. Hence compensation from the toll revenue is suggested to pure into improvement of public transit (which can be an alternative to the tolled-off, in particular the poor). Besides, the traffic it is intended to control primarily single occupancy car commuters merely transfer the increased cost onto their companies. From the perspective of the tolled-on, Evans (1992) stated that congestion pricing was that it might redistribute welfare from road users to the rest of the community, and introduce horizontal inequality between road users who were subject to it and those who were not. Consequently, the revenue is suggested to be used to expand or improve the road system (Evan 1992). The resolutions to these issues include packaging congestion pricing with other appropriate associated measures including the use of toll revenue. Some researches showed congestion pricing alone could gain minor support would gain more major support if accompanied with improvement on roads or public transit. As for the issue of perverse incentives for government, Evan (1992) argued that congestion pricing (as a direct charging mechanism) would give governments the power to exploit monopoly over roads. If this is true, government may try to raise substantial amount of revenue through overcharge on congestion pricing and undersupply of roads. The overcharge will distort the function of congestion pricing (short-term best pricing principle), and undersupply of roads is against the long-term best investment principle (long term optimal point). In addition, another difficulty in implementing congestion pricing is the public opposition to paying

27

twice for the facility (Fisher 1996). Besides, consumers regard congestion fee as immediate cost, but the derived benefit such as reduced congestion is intangible or expansion of transportation can only be achieved in the future. Furthermore, in some cases government officials are not willing to implement congestion pricing because it requires interaction with other organizations, which outweighs the costs of delay (Downs 1992). Congestion pricing may can be better in the society with high demand elasticity. In the areas with high demand elasticity, the amount of reduced traffic caused by congestion fee is larger than in the city with low demand elasticity. The areas with low demand elasticity could be auto-dependent, sparse areas, or in rich society. For example, in area with better service of transit system, it might need lower congestion fee to reduce congestion than in auto-dependent areas. The extent to which congestion pricing changes travelers behaviors in terms of trip route, time, mode, destination, vehicle occupancy, frequency (Hills 1993), and trip chaining, as well as location changes is not clear due to limited research. One of the few cases involves a survey conducted for the morning-peak passengers in the Randstad area, Netherlands in 1995 (Verhoef, et al, 1997). At that time, road pricing had just been discussed in the Dutch Parliament again, and a proportion of Dutch users in the Randstad was expected to be familiar, to some extent, with the concept of road pricing. The result showed if pricing were in place, the most frequent mentioned alternative was no alternative at all and trip rescheduling (some 32%). As for changing residence and changing jobs (location change), using public transit and using carpooling (behavior change), and canceling trip (some 10 %) were lower than the

28

above two alternatives. This, again, strengthens the idea of packaging congestion pricing with other measures to cope for the diverted demand. As for the congestion pricing elasticity of demand, few researches have been conducted, but there are some on the price change in petrol, transit fare (Oum, Tretheway, and Waters 1992; Goodwin 1992).

Interdisciplinary Discussion A general argument about the implementation of capacity expansion (e.g., transportation-engineering-based), demand contraction (including land-use-based) and congestion pricing (transportation-economics-based) associated measures to cater for congestion is that one single measure could barely work satisfactorily. Though some researches suggested certain measures combined could work better than separately, few academic researches provide affluent theories and evidence on it. On the contrast, some measures may work better separately than together due to the cancel-out effect. Theoretically, their push (out of road) effects and pull (to other alternatives) effects, and their combination cause the reinforced or cancel-out effects of the interaction of different measures. The following are some discussions on the interactions of capacity expansion-land use, capacity expansion-pricing, and land use-pricing. Land use to capacity expansion: road building could be an effective measure to calming congestion in a balance transportation environment which contains a wellproportioned mixture of road building and public transport. Road building in U.S. generally is not considered as an effective one since the vacated space is usually filled 29

up rapidly by latent demand, in a system of over-saturated demand for autos (e.g., U.S metropolitan areas). Younes case studies (1993) showed this condition did not occur in certain cities in Britain, Germany and Sweden, which was attributed to the balance of road and transit system. Subsidies to bus, transit-oriented land use, such as higher density (in particular around transit stations), mixed land use were all components of these successful transit systems. On the contrary, in the countries with low populationdensity and segregated land use such as U.S. and Australia, building new roads as a solution to congestion is reasonably less expected. Capacity expansion to land use: To achieve jobs-housing balance a convenient transportation plan providing access to jobsites and housing is necessary. A transportation system provides better access to outer areas than the nearby would explicitly encourage industries and households to move outwards and then increases imbalance. In a sense limitation on road building could be considered as one kind of growth management. For one it could press the demand for development or divert the development to other areas. That is, congestion is used as a tool for managing development. For another it could divert the development from urban periphery to existing developed areas. In the former case the existing urban form might remain the same; in the later case the metropolis might become more compact. The impact of road building on urban form is possibly quite different from that of congestion pricing on land use or urban form shown in the following section. Road building in metropolitan areas could possibly induce development within the affected area that it serves due to the improved accessibility. This land use changes due to colocation of firms and households, or influx from neighboring rural (Kilkenny 1998) or

30

other areas. In fact road building is a necessary factor for outlying development (i.e., urban sprawl) (Gordon and Richardson 1989). Classical location theory indicates the watershed of the development may spread along the new road if land is available. Law of constant travel time (Hupkes 1982) argues that work trip time would possibly remain the same by adjusting travel behavior or locations would help define the possible watershed of development. The resulting land use changes, however, perform in different shapes and degree due to natural and artificial factors. The development of land could be contiguous, or discontinuous to existing urban areas (Weitz and Moore 1998). In terms of their spatial pattern, there are strip, leapfrog, dispersed development (Ewing 1997). The development could occur in existing area (such as urban growth boundaries (used in Oregon project) or outside) (Weitz and Moore 1998). Also the densities of the developments are likely to be different. The prediction of outlying development caused by capacity expansion is beyond difficulty since many factors are involved, such as landscape, land value, land use, economic, and social conditions. From the sake of calming congestion, the U.S. had benefited from road building but it repaid in other aspects. The urban sprawl relieved the congestion pressure in urban core, which led to the current, most popular travel pattern-suburbs to suburbs. The travel time, distance, and speed of work trips remained constant or even improved partly due to this transition (Gordon, Richardson, and Jun 1991). Nevertheless, there are various concerns not only from transportation but more significantly from other perspective. Urban sprawl tends to lead to more auto uses, so congestion might be temporarily relieved or will be transferred owing to the escape hatch of spatial redistribution. Eventually, the pressures of growth and deteriorating congestion would

31

become overwhelming. High-density development (a common factor of transitsupportive environment or walking and bicycling) could hardly achieve without strong planning authorization (Bernick and Cervero 1997) where cheap land is always available through road building. A crucial reason is the failure of pricing mechanism of land, which have been constructed of demand and supply of humans, in particular of this generation. This failure leads to the overuse of land by human beings of this generation. Also travel distance is likely to increase (compared with more jobshousing-balanced land use). Urban sprawl may lead to unbalanced job-housing balance if job centers and housing are not developed simultaneously. All these possible impact on land use add to skepticism on net benefit of road building. Particularly notable, Newman and Kenworthy (1991) found the effect of the provision of infrastructure for automobiles was more significant than pricing on influencing urban forms in their research on 32 worlds principle cities. Congestion pricing to capacity expansion: Probably the most important roles that congestion pricing plays are first, it can internalize the externality caused by congestion, and its potential to resist capacity-usurping effects. Capacity-expansionand demand-contraction-based measures, on the contrast, are usually criticized for being unable to serve these two functions. Underpricing tends to lead to overuse of roads. Without pricing, traffic might turn back to original congestion in urban areas after the implementation of either demand-contraction- or capacity-expansion-based measures. These issues justify the significance of congestion pricing to congestion per se, as well as to the other two groups of measures. Hence, capacity-expansion and demand-contraction related measures would work better together with congestion

32

pricing. For instance, Vuren and Smart's (1990) theory-based paper (1990) concluded that the combination of electronic route guidance and road-use pricing would create greater improvement to the transportation system than if each was implemented in isolation of each other25. However, congestion pricing has been implemented in only a few cities, and furthermore, few research provides empirical evidence on its practical effect. Land use to congestion pricing: Levine (1998) discovered an effect of land use regulation on road pricing. In suburban areas, land use regulation of the lowdensity community tends to push those who are not affordable to live farther away from the suburban job centers26. They could have lived closer, if there was no maximum density regulation and the resulting, affordable housing was available. In the meanwhile, those living in the low-density community are effectively subsidized. Hence, these lowincome employees are double discriminated. He quoted the theory of the second best27 and then argued that the longer-distance commuter was worthy of a subsidy, not a tax. What is worse is that congestion pricing makes the unaffordable pay more, otherwise they have to turn to an alternative, including finding a new job at worse. The issue he raised here was about the efficiency of road pricing in areas where a principle desired outcome-relocation to reduce transportation costs- is effectively barred by land use regulation.

25

Electronic route guidance, as supply-oriented measure, could make more use of available road space;

road-use pricing, a demand-oriented measure, could guarantee all the trips made were socially worthy. Hence, they could complement each other (Vuren and Smart, 1990).
26

This is the phenomenon that land use regulation causes jobs-housing imbalance (Levine, 1998). The theory of the second best suggests equalizing subsidies for substitutes where societal marginal cost pricing

27

is not politically achievable (Bailey 1995).

33

Congestion pricing to land use: The impact of congestion pricing on land use or urban form is still a new area lacking empirical evidence because there are very limited cases of congestion pricing implemented in the world. In theory, congestion pricing is likely to influence travelers behavior at the outset of the congestion pricing implementation, and in the long run there are chances that land use or urban form may change. Congestion pricing may change land use in two ways. The first way is through a simplified process in the order as follows: change in travel cost, in accessibility of locations, in value of land, in location decisions, and finally land use patterns. The second way is through the use of revenue collected from congestion pricing (Elizabeth 1994). In the process of congestion pricing affecting land use, first the increased travel cost caused by pricing is likely lead to the triple divergence of traffic or reduced activities (both social and economic). Consequently in the short run traffic flow during peak hours in congested areas will be reduced to the Pareto point. Besides, the accessibility of location will change after congestion pricing due to increased out-of-pocket travel cost. Based on the equilibrium model of urban location that land values are determined primarily by differential accessibility to the Central Business district (CBD) (Isard 1956), two popular arguments about the impact of transportation pricing on urban form are derived. One is priced areas could have higher-density land use patterns and a smaller metropolitan area (Solow 1994). The other is congestion pricing would have a further decentralizing effect (to non-priced areas) by reducing the attractiveness of destinations within the priced areas (Elizabeth 1994). With the relaxation of uniform value of time, the result will be different. For people with high value of time (e.g., high-income, people with emergency) may regard the accessibility of priced areas is improved due to the

34

savings on travel time. On the other hand, people with relative low value of time (e.g., the low income) may regard the priced areas become less accessible and the nonpriced areas is relatively more accessible since where they could trade time (e.g., longer travel time) for savings on congestion tolls. Then the class segregation may be formed if the high-income could move to the priced areas and the low-income to city center or non-priced areas. As for business location choice, without considering factors other than market and labor input, the population-serving business may probably follow the customers (i.e., market) and the labor-intensive industries might follow its labor. Hence the industrial structure could change in priced and non-priced areas. Taking other variables into consideration, the condition will be a lot more complex than the above. Basically, we would like to know the extent to which the change in accessibility of locations by congestion pricing would influence resident and industry location decisions, and then the land use. The understanding of the following conditions, relationships or puzzles may contribute to this prediction job under the circumstances of few past empirical experience. We would like to know: 1) the response or adjustment (e.g., changes in travel behavior or new location decision) of residents or industries to different changes in accessibility caused by congestion toll (which is related to levels of congestion and demand) and different congestion plans (such as area-wide and arterial pricing, and with or without accompanying plans); 2) the extent to which different residents (such as race, income levels) or industries will change locations in the long; 3) the economic influence on industries caused by the out-of-pocket costs; 4) the area variables that may encourage or impede moving (such as transportation system, land value, availability of new places, market conditions, the restrictions of current land use

35

code, crime rate, and school district); 5) the influence of use of the revenue gathered from congestion toll. This theoretical analysis suggests that it is the prediction derived from the equilibrium model of land value is too simplified since there are lots of variables left out. Transportation Research Board (1994) indicated that retail and commercial establishments in Manhattan in the mid-1980s thought congestion pricing would reduce the attractiveness of the CBD (Central Business District). Singapore (Phang `993), however, does not support this argument. Interviews with businessmen and retail shopkeepers revealed there was no particular influence on sales. Furthermore, labor availability appears to have improved because of improved public transit. It also mentioned that congestion pricing would increase jobs-housing balance since workers would move closer to their work site. Elizabeths (1994) research on San Francisco Bay Area found firms would response more to area-wide scenario than smaller scale plans since more of their employee could be influenced. Also low-income employee would not be affected as much as expected since they generally were more likely to live nearby. However, this conclusion does not apply to the urban-to-suburban work trips of the low-income employee. For some firms other factors such as taxes, crime rates, and the general business climate were more important than transportation in business location decision. Besides, there were a variety of ways that firms will take in response to congestion pricing, such as increased parking subsidy, schedule change, and seeking for exemption from pricing. Furthermore Local government officials were skeptical about the possibilities for increased development due to higher density since current land use regulations usually would not allow so. Deakin also stated that if the

36

use of the revenue from congestion pricing was in the supply of transportation (e.g., building new road), then its impact could be larger than pricing itself.

Conclusion

Through the analysis of the causes of traffic congestion, this paper found there do exist certain conditions that can make highway congestion unavoidable in urbanized areas. The increased interaction in human society in terms of economic and social activities, and the population growth constitute the fundamental power of traffic growth, in particular as opposed to limited supply of road capacity in urbanized. The supply of road capacity is getting difficult in the context of sustainable transportation, when the previously underestimated external cost on environment has been marching into history. As a result, movement of persons (instead of vehicles) and accessibility (rather than mobility) become the goals of transportation professionals. Under these circumstances, congestion turns out to be the tradeoff for the sustainability of society, environment and economy. Besides, traffics are distributed unevenly in terms of time, space. In this case for a balanced financial condition, certain time period or day, certain routes need to be overused to compensate the loss from the underused. The short-turn deficit of capacity supply is even harder to deal with, such as the decreased road capacity caused by weather, accidents and routine maintenance. Even if all the above conditions did not exist, in a transportation system of Pareto efficiency, drivers would always feel congested during peak hours or days. In an over-saturated system, the

37

capacity-usurping effects will bring the congestion back sooner or later after our endeavor (excluding pricing-oriented measures). This, however, does not mean there is no need to try to solve traffic congestion. Measures to calming congestion that can improve the efficiency of road network, as well as achieving sustainable transportation should still be taken. Most of measures for alleviating congestion are developed from three main perspectives: transportation engineering, and economics, and land-use. The scope of economic measures contains all measures using price mechanism; the land-use related measures contain measures by changing land-use policy only, considered as one kind of demand contraction; the transportation measures contain capacity expansion, as well as demand contraction. For purpose of systematic analysis under demand-supplypricing structure, the measures are reorganized into capacity expansion (transportationengineer perspective), demand contraction (including land-use- and transportationengineering perspectives), and pricing (transportation-economics perspective). Two common problems that capacity-expansion and demand-contraction measures may face are capacity-usurping effects and inability of achieving Pareto efficiency. This condition points out the significance of congestion pricing in alleviating congestion. However, congestion pricing have not been widely implemented primarily due to political difficulty. These conditions make it more difficult to dealing congestion. The capacityusurping effect was found, however, not very significant in some European cities. This gap may deserve some researches to fill in. Through some researches suggested certain measures combined could work better than separately, few academic researches provide affluent theories and evidence

38

on it. On the contrast, some measures may work better separately than together due to the cancel-out effect. Theoretically, their push (out of road) effects and pull (to other alternatives) effects, and their combination cause the reinforced or cancel-out effects of the interaction of different measures. For example, road building is likely to induce weaker capacity-usurping effects in areas with transit-oriented land use than with autodependent land use. To achieve jobs-housing balance a convenient transportation plan providing access to jobsites and housing is necessary. On the contrary, capacityexpansion measures may lead to transit-oriented land use, such as low density, urban sprawl. The shortage of affordable housing caused by land use regulations may impede the desired outcome of congestion pricing- relocation to reduce transportation costs (Levine 1998). Two popular arguments about the impact of transportation pricing on urban form are: (1) priced areas could have higher-density land use patterns and a smaller metropolitan area; (2) congestion pricing would have a further decentralizing effect (to non-priced areas) by reducing the attractiveness of destinations within the priced areas (Elizabeth 1994). In the scope of interaction among all three kinds of measures, our understanding are still very limited, not even mentioning that there exist numerous, variant measures in each of the three categories.

39

References 1. Bailey, Strphen J. 1995. Public sector Economics Theory, Policy and Practice. UK: Macmillan Press, LTD. 2. Bank, World. 1996. Sustainable Transport: priorities for policy reform. Wrold Bank, Washington, DC. 3. Bernick, Michael and Robert Cervero. 1997. Transit villages in the 21st century. Cbi . 4. Board, Transportation. 1994. Curbing Gridlock, vol. 1. 5. Borger, Bruno De, Inge Mayeres, Stef and Proost, and sandra Wouters. 1996. A simulation exercise for Belgium. Journal of Transport Economics and Policy January 1996. 6. Braum, Phil and et al. 1994. ISTEA Planner's workbook. 7. Bressi, Todd W. 1992. The neo-traditional revolution. Utne Reader, us . 8. Cervero, R. 1989. Jobs-housing balancing and regional mobility. Journal of the American Planning Association, us 55:136-50. 9. Cervero, Robert. 1996. Jobs-housing balance revisited: trends and impacts in the San Francisco Bay area. Journal of the American Planning Association, us 62:492511. 10. Cervero, Robert and Roger Gorham. 1995. Commuting in transit versus automobile neighborhoods. Journal of the American Planning Association, us 61:210-25. 11. Cornes, R. and T. Sandler. 1986. The theory of externalities, public goods, and club goods. 12. Crane, R. 1996. Cars and drivers in the new suburbs: linking access to travel in neotraditional planning. Journal of the American Planning Association, us 62:51-65. 13. Decorla-Souza, P. and A. R. Kane. 1992. Peak period tolls: precepts and prospects. Transportation 19:293-311. 14. Downs, A. 1992. Stuck in traffic. Washington, DC: Brooklings Insituttion. 15. Elizabeth, Deakin. 1994. Urban transportation congesion pricing effects on urban form. Pp. 334-355 in Curbing Gridlock, vol. 2, edited by T. R. Board. 16. Emmerink, R. H. M., K. W. Axhausen, and P. Nijkamp. 1995. Effects of information in road transport networks with recurrent congestion. Transportation, ne 22:21-53. 17. Engineers, Institute of Transportation. 1989. A tool box for alleviating traffic congestion. Washington DC: Institutes of Transportation Engineers. 18. Evans, Andrew W. 1992. Road congestion pricing:when is it a good policy. Journal of Transport Economics and Policy September 1992. 19. Ewing, Reid. 1997. Is Los Angeles-style sprawl desirable? Journal of the American Planning Association, us 63:107-26. 1

20. Federal Register, Part II, Department of Transportation, Federal Highway Administration, 23 CFR Parts 500 and 626, Federal Transit Administration, 49 CFR Part 614, December 1, 1993. . 21. Fisher, Ronald C. 1996. State and local public finance. 22. FitzRoy, Felix and Ian Smith. 1993. Priority over pricing: lessons form Zurich on the redundancy of road pricing. Joutrnal of Transport Economics and Policy May 1993:209-214. 23. Giuliano, Benevieve. 1991. Is jobs-housing balancing a transportation issue? Transportation Research Record 1305:305-312. 24. Giuliano, Genevieve. 1992. Transportation demand management: promise or panacea? Journal of the American Planning Association 58:327-335. 25. Gomez-Ibanez, Jose A. 1992. The political economy of highway tolls and congestion pricing. Transportation Quarterly 46:343-360. 26. Goodwin, P. B. 1992. A review of neew demand elasticities with special reference to short and long run effects of price changes. Journal of Transport Economics and Policy May 1992. 27. Gordon, P. and H. W. Richardson. 1989. Gasoline consumption and cities; a reply. Journal of the American Planning Association, us 55:342-6. 28. Gordon, Peter, Harry Ward Richardson, and Myung-Jin Jun. 1991. The Commuting paradox: evidence from the top twenty. Journal of the American Planning Association 57:416-20. 29. Hansen, M. 1995. Do new highway generate traffic? Access 1995. 30. Hashimoto, K. 1990. Monitoring road traffic congestion in Japan. Transport Reviews 10:171-186. 31. Hau, T D. 1992. Economic fundamentals of road pricing: a Diagrammatic analysis. The World Bank. 32. Hay, A. 1973. Transport for the space economy. Seatle, WA: University of Washington. 33. Hills, P. 1993. road congestion pricing:when is it a good policy. Journal of Transport Economics and Policy January 1993. 34. Hupkes, G. 1982. The law of constant travel time and trip rates. Futures 14:38-46. 35. Isard, W. 1956. Location and space-economy. Cambridge, Mass: The MIT Press. 36. Johnson, W. C. 1998. Urban planning and politics. Chicago: American Planning Association. 37. Kessler, J. and W. Schroeer. 1995. Meeting mobility and air quality goals: strategies that work. Transportation, ne 22:241-72. 38. Kilkenny, Maureen. 1998. Transport costs and rural development. Journal of Regional Science 38:293-312. 2

39. Kroes, Eric, Andrew Daly, and Hugh Gunn. 1996. The opening of the Amsterdam Ring Road: a case study on short-term effects of removing a bottleneck. Transportation, ne 23:72-82. 40. Levine, Jonathan. 1998. Rethinking accessibility and jobs-housing balance. Journal of the American Planning Association 64:133-49. 41. Levinson, Herbert S., Timthy J. Lomax, and Turver S. 1997. Traffic congestion -past-present--future Pp. 1-13 in Congestion and traffic sasfety: American Society of Civil Engineers. 42. Lomax, Timthy J. and Herbert S. Levinson. 1997. Overview of congestion measurement principles Pp. 549-555 in Traffic congestion and traffic safety: American Society of Civil Engineers. 43. Newman, Peter W. G. and Jeffrey R. Kenworthy. 1991. Transport and urban form in thirty-two of the world's principal cities. Transport Reviews 11:249-272. 44. Nowlan, David M. and Greg Stewart. 1991. Downtown population growth and commuting trips: recent experience in Toronto. Journal of the American Planning Association, us 57:165-82. 45. Oum, Tae H., Michael W. Tretheway, and W. G. Waters, 2nd. 1992. Concepts, methods and purposes of productivity measurement in transportation. Transportation Research. Part A us:493-505. 46. Phang, Sock-yong. `993. Singapore's motor vehicle policy: review of recent changes and a suggested alternative. Transportation Research A 27A:329-336. 47. Phang, Sock-Yong and Mukul G. Asher. 1997. Recent developments in singapore's motor vehicle policies. Journal of transport economics and policy 1997:211-220. 48. Pline, James L. 1992. Traffic engineering handbook. Englewood Cliffs, N.Y.: Institute of Transportation Engineers. 49. Shoup, Donald C. Evaluating the effects of cashing out employer-paid parking: eight case studies. Forthcoming in Transport Policy . 50. Solow, Robert M. 1994. Congestion cost and the use of land for streets. Pp. 116-132 in The Economics of transport, vol. 2, edited by H. Mohring. Aldershot, Hants, England ; Brookfield, Vt.: E. Elgar Pub. 51. Stover, Vergil G. and Frank J. Koepke. 1988. Transportation and land development: Prentice-Hall. 52. Systems, Michigan Intelligent Transportation and Center in Detroit. 1996. Michigan ITS: About ITS. : http://campus.merit.net/mdot/its.html. 53. Topp, Hartmut H. 1993. Parking policies to reduce car traffic in German cities. Transport Reviews 13:83-95. 54. Transportation, The Center for Sustainable. 1997. Definition and vision of sustainable transportation. : http://www.web.net/~cstctd/documents/d&v%20english.pdf. 3

55. United States. Federal Highway Administration. Office of Traffic Operations. 1994. The 1995 Highway capacity manual : a summary. Washington, D.C.: U.S. Dept. of Transportation Federal Highway Administration. 56. Verhoef, Erik T., Nijkamp Peter, and Rietveld Piet. 1997. The social feasibility of road pricing: a case study for the Randstad area. Journal of Transport Economics and Policy September 1997. 57. Vuren, Tom V. and Malcolm B. Smart. 1990. Route guidance and road pricing-problems, practicalities and possibilities. Trasnport Reviews 10:269-283. 58. Washington, University of. 1999. Carpooling options for Faculty/Staff. : http://www.washington.edu/admin/parking/carpool.html. 59. Weitz, Jerry and Terry Moore. 1998. Development inside urban growth boundaries: Oregon's empirical evidence of contiguous urban form. Journal of the American Planning Association 64:424-40. 60. Younes, Bassem. 1993. Roads in urban areas: to build or not to build. Transport reviews 13:99-117.

Appendix 1. Definition of Congestion

Cost or time

Forced Flow

Unacceptable

D Stable Flow

Acceptable Congestion Free

Traffic Flow (V) Capacity (Economists)


Free Acceptable Congestion

Capacity (C) (Engineers)

Unacceptable

D: Demand for travel on highway. Source: this research

S: Supply of highway.

You might also like