Litman_Todd MES 1995.pdf
Media
Part of Transportation Cost Analysis: Techniques, Estimates and Implications
- extracted text (extracttext:extracted_text)
-
Transportation Cost Analysis:
Techniques, Estimates and Implications
Thesis
Masters of Environmental Studies
The Evergreen State College
Olympia, Washington
Todd Litman
3 May 1995
~
Transportation Cost Analysis:
Techniques, Estimates and Implications
by
Todd Litman
Victoria Transport Policy Institute
1250 Rudlin Street
Victoria, BC V8V 3R7
Phone and Fax: (604) 360-1560
ur698@freenet. victoria.bc.ca
3 May 1995
·~---
This thesis for the Master of Environmental Studies degree
by
TODD LITMAN
has been approved for
THE EVERGREEN STATE COLLEGE
by
~~-~- -~~·
Robert Knapp, Ph. .
~~ ~-~~~
tv-.
February 25, 1995
~
Acknowledgments
This thesis is the result of several years of research. It has been an exciting and
rewarding project, and sometimes extremely frustrating. I cannot possibly acknowledge
every person who contributed technical or moral support, though I appreciate them all.
The following is a partial list. Please forgive any omissions.
The Bullitt Foundation (especially Kathy Becker & Emory Bundy), Michael Cameron,
Andy Clarke, Tim Coats, Professor Elizabeth Deakin, Patrick DeCorla-Souza, Mark
Delucchi, Mac Elliot, David Engwicht, Ralph Erickson, Larry Frank PhD, Andrew
Ginsburg, Allen Greenberg, Christie Growdon, Andy Hamilton, Mark Hanson, Stanly
Hart, Brendon Hernily, Ralph Hirsch, John Holtzclaw PhD, Walter Hook, Mirjam Imhof,
Chris Johnson, Professor Robert Johnston, Darius Kanga, Chip Kaufinan, Jon Kessler,
Brian Ketcham, Professor Rob Knapp, Charles Komanoff, Peter Lagerway, Felix Laube,
Jim Lazar, Professor Douglass Lee, James MacKenzie, Ann McAlister, Gil McCoy, Peter
Miller, Ted Miller, Deborah Miness, David Mogavero, Stefan Natzke, Dick Nelson, Jim
Page, John Paolella, Randall Pazdena, Professor Setty Pendakur, Professor John Perkins,
Anthony Perl, Steve Platkin, Franzi Poldy, John Pucher, Michael Replogle, Don Rintoul,
Bill Rogers, Professor Mark Roseland, David Van Seters, Preston Schiller PhD, Professor
Elliot Sclar, Michael Shaffer, Don Shakow, Gui Shearin PhD, Professor Donald Shoup,
Deming Smith, Ryan Snyder, Ruth Steiner PhD, Professor Tony Turrittin, Professor
Richard Untermann, Professor Casey van Kooten, Chris Voigt, Professor Bucan Vuchic,
Mathis Wackemagel PhD, Professor Bill Waters, Joel Woodhall, Chris Zegras.
And special appreciation to my wife, Suzanne Kort, and thanks to our children Graham
and Raviv for their tolerance of my long hours of research.
~~
Note to anyone using this thesis: This version of this report has been specially formatted
with double line spacing to satisfy thesis requirements. I consider this a waste of paper and
more difficult to read than the standard single-space version that has been widely
distributed through the Victoria Transport Policy Institute, and which is continually
updated as new information becomes available. I therefore recommend that anyone using
this thesis contact the Institute for a current, standard format version.
© 1995
Todd Alexander Litman
All Rights Reserved
Vll
~
Contents
Executive Summary
Preface
1.0
Introduction
1.1
1.2
1.3
1.4
1.5
1.6
1. 7
1.8
1.9
Study Outline and Scope
Purpose and Context of This Study
Defining Transport
Defining "Cost"
Treatment of Taxes
Discount Rate in Cost Analysis
Pricing Non-Market Goods
Criticism of Transportation Cost Analysis
Treatment of Variability and Uncertainty
2.0
Transportation Cost Literature Review
3.0
Definitions, Costing Methods, and Estimates
3. 0.1 Measuring Costs
3.0.2 Modes Defined
3.0.3 Avoiding Double Counting
3.1
3.2
3.3
3.4
3.5
3.6
3. 7
3.8
3.9
3.10
3.11
3.12
3.13
3.14
3.15
3.16
User Costs
Travel Time
Accidents
Parking
Congestion
Roadway Facility Costs
Roadway Land Value
Municipal Services
Equity and Option Value
Air Pollution
Noise
Resource Consumption
Barrier Effect
Land Use Impacts
Water Pollution
Waste Disposal
4.0
Cost Totals
4.1
4.2
4.3
4.4
4.5
4.6
Summary Graphs
Cost Distribution
Total Transportation Costs
Cost Ranges
Survey Test of Cost Estimates
Summary
Vll
~
5.0
Generated Traffic
5.1
5.2
5.3
5. 4
5.5
5.6
5.7
5.8
5.9
6.0
Introduction
Transportation Elasticities
Defining Generated Traffic
Types of Generated Traffic
Predicting Generated Traffic
Calculating Internal Benefits of Generated Traffic
Calculating External Costs of Generated Traffic
Property Value Impacts
Applying Generated Traffic External Cost Estimates
Transportation Cost Implications
6.1
6.2
6.3
6.4
6.5
6.6
7.0
Economic Efficiency Impacts
Economic Development Implications ofUnderpricing
External Benefits of Transportation?
Land Use Impacts ofUnderpricing
Transportation Decision Making and Underpricing
Summary: Implications of Underpricing on Households and Individuals
Evaluating Transportation Equity
7.1
7.2
7.3
7.4
7.5
7.6
8.0
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
9.0
9.1
9.2
9.3
9.4
9.5
Defining Transportation Equity
Current Transportation Equity Analysis
Automobile Dependency as an Equity Issue
Comprehensive Transportation Equity Analysis
Comprehensive Equity Analysis Applications
Equity Analysis Conclusions
Applications and Case Studies
Evaluating Transportation Demand Management (TDM) Savings
Price Impacts on User Travel Decisions
Marginalizing User Costs
Critiquing Transportation Investment Models
Evaluating Congestion Reduction as a Transport Improvement Priority
Evaluating Traffic Management Benefits
Least-Cost Transportation Planning
Evaluating Electric Vehicle Benefits
Critiquing Taxation Report
Conclusions and Recommendations
Costs and User Pricing
Equity
Land Use Patterns
Transportation Decision Making
Research Recommendations
Bibliography
Vll
Preface
For most of history society's intellectual and industrial resources were devoted to
overcoming logistical and technical barriers. Leaders in expanding these limits, Daniel
Boone, Thomas Edison, and Henry Ford, are cultural heroes. Progress has eliminated
many constraints that faced previous generations. Whipsaws, horse plows and shoe leather
have been replaced by chainsaws, tractors, and automobiles, greatly expanding the
potential for human control of, and impacts on, the natural and human environment. The
question, What can we do? is now often replaced by What should we do? This presents
different but equally challenging problems. This is one way to explain the major changes
occurring in the field of transportation planning.
Since transportation systems involve a combination of public and private decisions,
transport planning is by nature a political process that involves tradeoffs between
stakeholders. It requires effective communication and accurate accounting. Tools to help
in this process are relatively new and still under development. For all of its weaknesses,
economic analysis offers the potential of addressing the diverse and complex issues that
must be considered in transportation decision making.
The basic formal economic evaluation technique, benefit-cost analysis, has been criticized
for excluding significant costs, particularly those related to environmental and social
impacts. This report explores the potential of incorporating these costs into transport
analysis. It attempts to bridge the gap between people who are concerned about
qualitative "problems," and those who prefer quantitative economic accounting.
Transportation planning and policy making desperately need such tools.
Transportation cost studies frequently begin by acknowledging the tremendous benefits
provided by modern transport systems so as not to appear "anti-transport" or "antiautomobile." Consider this done. We all benefit from transportation and many of us delight
in using various travel modes. But more is not better. Our transport system can provide
even more benefits if costs to users and society are reduced. This study identifies methods
for measuring these costs in order to help determine how to optimize our transportation
system and avoid squandering valuable resources.
Transportation Cost Analysis
~
1.0 Introduction
1.1 Study Outline and Scope
This study explores North American roadway transportation costs.1 It attempts to
consider all potential costs, including social and environmental impacts. It investigates the
hypothesis that significant costs are commonly ignored in transportation decision making,
and explores the implications of such omissions on economic efficiency, equity, and land
use patterns.
This first chapter examines the concept of cost and costing methods. Chapter Two reviews
and summarizes recent transportation cost studies. In chapters 3. 0 through 3 .16, sixteen
specific transportation costs are defined and discussed, specific existing are reviewed, and
"Best Guess" cost values are established for eleven modes under Urban Peak, Urban OffPeak and Rural travel conditions. Chapter Four summarizes these estimates. Chapter Five
considers transportation elasticities and the effects of generated traffic. Chapter Six
explores their implications. Chapter Seven examines transportation equity issues. Chapter
Eight applies the cost estimates to various policy and planning applications. Chapter Nine
summarizes the conclusions of this study and offers recommendations for improving
transportation efficiency and equity.
This study's emphasis on costs is not intended to slight the significant benefits of
transportation. However, there is an important difference between the allocation of
benefits and costs. Most transportation benefits are enjoyed by the user, while many costs
are borne by other individuals or society as a whole. These external costs, if they are
significant, imply a conflict between individual and societal interests, and indicate the
1 Light
rail, walking, and telecommuting costs are also estimated for comparison with roadway modes.
Page 1-2
Transportation Cost Analysis
~
likelihood of economic inefficiency and inequity. To appreciate the importance of these
costs it is useful to consider situations in which travel activities change in a community and
residents must respond to the resulting impacts. For example, imagine that:
•
Automobile ownership and travel in your community was expected to double in a few
years. What economic, social, health and environmental problems might increase?
•
You manage a city that currently has no automobiles. Some citizens want to start
using motor vehicles and offer to pay for all costs incurred. The city council asks you
to develop a user fee schedule that completely compensates the city and its residents
for expenses and damages. What costs would you include? What charges would you
recommend for owning a car and driving?
•
A new technology eliminates a specific external cost of driving, such as traffic noise or
accident risk. How much should the community pay to implement it?
These are slightly exaggerated examples of real issues. This report analyzes transport costs
and their implications to help provide answers to these and similar questions.
1.2 Purpose and Context of This Study
This analysis relates to two current trends. The first is a growing concern over social and
environmental impacts. There are indications that growing resource consumption and
waste production endangers our environment and the quality of our lives. It is important
to develop a vocabulary that describes these costs and methods to measure them,
preferably in monetary units since economics tends to ignore features that are not priced.
"The market sees only efficiency--it has no organs for hearing, feeling or smelling either
justice or sustainability. ''2 Traditional economics does not deny the existence of nonmarket impacts such as air pollution or habitat destruction, but economic models typically
assume that they are small compared with the market costs and benefits.3 If non-market
2
Herman Daly and John Cobb, For the Common Good, Beacon Press (Boston), 1994, p. 145.
Transportation professionals often refer to environmental and social costs as "intangibles," with the
implication that they are subjective and ethereal. With the exception of the remarkable 1975 study The
Full Costs of Urban Transport by Keeler, et al., and components of the 1982 FHWA Cost Allocation
Study, transportation professionals did little to quantify or apply these costs until the late 1980's, although
3
Page 1-3
Transportation Cost Analysis
costs are found to be significant then they must be incorporated into transport decision
making or even the best intended programs may make society overall worse off
The second trend is a growing appreciation that motor vehicle traffic must be managed
and reduced in urban areas to address congestion and air pollution problems with available
financial resources. Transport planning is beginning to consider multimodal, demand
management, and land use management solutions to transport problems. These changes
require a greater understanding of the impacts of possible policies and investments.
As described in Chapter 2, several previous studies review and even quantify transport
costs. This study attempts to incorporate and expand on previous work. It:
•
Includes the latest research and cost data.
•
Provides a description of the economic theory of prices and costing.
•
Covers a broader range of costs than many other studies.
•
Creates a framework for using cost estimates in specific policy and planning decisions.
•
Applies cost estimates to specific analysis to demonstrate their implications and use.
Psychologists tell us that people's behavior influences their belief. 4 This explains why the
debate over transportation costs between modes is often emotional: each user finds
arguments to support their own travel choices and habits. Developing objective cost
estimates will help create a context of fair and rational debate over the proper planning
and investments for each travel mode.
they expended considerable effort quantifying non-market benefits such as travel time savings and
accident reduction. To test whether such costs are truly intangible, consider driving an automobile in
which air pollution costs are internalized by piping the exhaust directly into the passenger compartment.
Even the most skeptical economist will probably agree that environmental costs are quite tangible.
4 Cy Ulberg, Psychological Aspects ofMode Choice, WSDOT (Olympia), 1989, p. 65.
Page 1-4
Transportation Cost Analysis
\ .3
Defining Transport
How we think about and measure transportation depends on how we define it. Transport
is defined as "To convey from one place to another. "5 This implies movement or mobility.
But movement is seldom an end in itself Even recreational travel is primarily intended to
arrive at a destination. The ultimate goal of transportation can be defined as access, which
is the ability to obtain desired goods, services, and destinations. Over the last century,
automobile and truck transport have come to dominate most land transport, so the
quantity and speed of motor vehicle traffic have become de facto measures of transport
system performance (usually measured as vehicle miles traveled, or VMT). But these are
imperfect measures of transport quality because:
•
In urban areas it is impossible to build enough roads and parking to satisfy all potential
automobile trips.
•
Some people cannot own or drive a car due to financial, physical, or legal barriers.
•
Automobile use imposes increasing financial, environmental, and social costs.
Defining transport as mobility (typically measured as person mile traveled or PMT) allows
the benefits of non-automotive travel modes such as walking, bicycling, transit, and ride
sharing to be recognized. While this is an important step toward expanding the definition
of transport, it does not go far enough. If transport is defined by its basic function, access,
then an even greater range of options can be considered, some of which actually reduce
the need for movement. 6 Access is affected by the location of destinations, and availability
of substitutes such as communication technology, as well as the ease of travel. Although
there is no agreement on how to quantified access, it can be measured based on total
transport costs, including travel time. Using this definition, increased travel is not
5 Oxford A merican
Dictionary, 1980.
Elliot Sclar and K. Schaeffer, A ccess For A ll, Columbia University Press (NY), 1980; Susan Handy
"Highway Blues: Nothing a Little Accessibility Can't Cure," in A ccess (University of California
Transportation Center, Berkeley), No. 5, Fall1994.
6
Page 1-5
Transportation Cost Analysis
~
necessarily beneficial, it may indicate an overall reduction in access that requires more
movement for the same level ofbenefits. 7 Professor John Whitelegg states,
"It is the ease of access to other people and facilities that determines the success of a
transportation system, rather than the means or speed of transport. It is relatively easy
to increase the speed at which people move around, much harder to introduce changes
that enable us to spend less time gaining access to the facilities that we need '18
David Engwicht develops a similar concept, emphasizing that transport allows exchange,
and that certain land use, commercial, and social patterns accommodate exchange with
more or less ease.9 The loss of neighborhood stores and delivery services, consolidation of
public services such as schools and post offices, and urban sprawl are examples of trends
which force people to travel more to obtain access to goods, services and jobs.
1.4
Defining "Cost"
Since this report investigates costs and costing, it is important to define these terms. Cost
refers to the tradeoffs that individuals and society must make between use of resources.
For example, time spent traveling is a cost in terms of the opportunity to use that same
time in other activities. This same concept applies to the tradeoffs between transport
investments and other possible expenditures, between roads and other land uses, and
sometimes between transportation activities and environmental protection.
The terms cost and price are often used interchangeably, but in formal economics cost is
defined broadly as any "benefits foregone." This can involve money, time and other
resources, or the loss of an opportunity to enjoy a benefit. Price usually refers specifically
to market costs. Lee states,
"The economist's notion of cost -- which is used here -- is the value of resources (used
for a given input) in their best alternative use. If, for example, less gasoline were used
7 Reid Ewing,
"Transportation Services Standards - As If People Matter," Transportation Research, 1193
Whitehead, "Time Pollution," The Ecologist, Vol. 23, No. 4, July 1993, p. 131.
9 David Engwicht, Reclaiming Our Cities and Towns, New Society Publishing (Philadelphia), 1993.
8 John
Page 1-6
Transportation Cost Analysis
in highway travel, what would consumers be willing to pay for the fuel for some other
purpose, or if it were converted instead to heating oil? If less time were used in travel,
how valuable would the time be for whatever purpose travelers chose to use it? If
clean air were less consumed in dispersing vehicle pollutants, how much would society
benefit from using the air to disperse non-highway pollutants or from breathing
cleaner air? This concept of costs depends, then, on benefits foregone; there is no
separate measure of cost that is distinct from valuation of benefits. "10
Because of their mirror-image relationship, measuring costs often begins by defining a
benefit foregone, while benefits are defined by reduced costs. Costing (also called
monetization) involves quantifying these in monetary units. Important distinctions include:
1. Internal and External Costs
Costs can be divided between internal (also called user) and external (also called
social) costs. Internal costs are borne by the good's consumer. External costs are borne
by others, either individuals or society as a whole. Some costs are external to individual
users but borne by the sector (group) as a whole. For example, accident costs that are
compensated by liability insurance are external to the individual who has the accident,
but internal to all drivers who buy insurance. Which standard should be used to define
externalities in a particular analysis depends on the type of problem being addressed. If
the concern is equity ("People shouldn't have to pay for something they don't use.")
then costs need only be internalized at the sector level. If the concern is economic
efficiency ("People tend to squander resources that they get for free ."), then costs must
be internalized at the individual level in order to give users correct economic incentives.
An external cost can be internalized if the user adequately compensates those on whom
the cost is imposed. If the injured party does not consider the compensation "worth"
the damage suffered, the cost is only partially internalized.
10
Douglass Lee, Full Cost Pricing ofHighways, Na. Transport Systems Center (Cambridge), 1995, p. 7.
Page 1-7
Transportation Cost Analysis
~
2. Variable (Marginal) and Fixed Costs
Variable costs are proportional to consumption. Fuel, travel time and accident risk are
variable automobile costs. Fixed costs do not vary with use, such as depreciation,
insurance, and registration. The distinction between fixed and variable often depends on
the perspective and time horizon. For example, depreciation is often considered a fixed
cost because car owners make the same payments no matter how many miles a year
they drive; but a car's operating life and resale value are affected by how much it is
driven, so depreciation is partly variable. Variable costs are also called marginal costs,
defined as the cost of an additional unit. Past and fixed costs are considered sunk.
3. Perceived and Actual Costs
There is often a difference between perceived and actual automobile costs. Users tend
to be most aware of immediate costs such as travel time, stress, parking fees, fuel, and
transit fares, while costs that are only paid occasionally, such as insurance, registration,
and maintenance are often underestimated. 11 Some costs tend to be ignored by users
altogether, such as parking subsidies and external environmental impacts.
4. Market and Non-Market Costs
Costs can also be divided between market and non-market. Market costs involve goods
that are regularly traded in a competitive market, such as land, cars, and gasoline. Nonmarket costs involve goods that are not regularly traded in markets such as clean air,
accident risk, and quiet. Although many non-market goods have significant value, they
are often ignored or underestimated compared with market costs.
5. Direct and Indirect Costs
A fifth consideration is the degree to which costs are direct or indirect. Quantifying
indirect costs and benefits requires an understanding of the various steps connecting an
11
Cy Ulberg, Psychological Aspects ofMode Choice, WSDOT (Olympia), 1989, p. 20.
Page 1-8
Transportation Cost Analysis
~
activity with its ultimate effects. Whether an activity imposes an indirect cost can be
determined using a "with and without" test. 12 The difference in impacts with and
without a project or policy are considered a result of that project or policy. For
example, the negative effects of land use changes resulting from a transportation
project that would not otherwise occur should be considered a cost of that project.
An important indirect and long term impact of automobile use is a greater dispersion of
land uses which increases the need to travel in order to maintain access to goods and
services, and a non-automotive reduction in travel alternatives. This is called
automobile dependency, 13 and will be discussed further in chapters 3.9, 3.14, and 7.
Costs -- A Primer
Consider the costs of owning a pet dog. A dog can often be obtained for a low price or
even for free . But pet owners quickly discover that a dog imposes many costs. Some, such
as pet food purchased at the store, are market costs. Others, such as the nuisance of
cleaning up after the animal, are non-market costs. These non-market costs can be
estimated using a market cost as a reference, such as the price to hire somebody else to
clean up after the dog. Some pet costs, such as registration fees and vet fees, are fixed, the
price is the same for any size dog, while others such as food, are variable because they
depend on the animal's size or breed. Some costs are not separate expenses; they are price
premiums or extra costs to other expenditures, such as more frequent rug cleaning, or the
added cost of a larger back yard. In addition to the internal costs borne by their owners,
dogs can impose external costs on other people, including noise, smells, messes, and fear.
Table 1-1 shows examples of motor vehicle costs indicating major categories. 14
12
C. van Kooten, Land Resource Economics and Sustainable Dev., UBC Press (Vancouver), 1993, p. 86.
Newman and Kenworthy, Cities and Automobile Dependency, Gower Press (Aldershot), 1989.
14 Of course, these categories are general, not absolute. There are exceptions to these allocations.
13
Page 1-9
Transportation Cost Analysis
~
Table 1-1
Motor Vehicle Transportation Cost Categories (Italics= Non-market)
Variable
Internal
(User)
External
(Social)
Fixed
Fuel
Short term parking
Vehicle maintenance (part)
User time & stress
User accident risk
Road maintenance
Traffic law enforcement
Insurance disbursements
Congestion delays
Environmental impacts
Uncompensated accident risk
Vehicle purchase
Vehicle registration
Insurance payments
Long-term parking facilities
Vehicle maintenance (part)
Road construction
"Free" or subsidized parking
Traffic planning
Street lighting
Land use impacts
Social inequity
How a cost affects transport decisions tends to vary depending on whether it is internal,
external, fixed, variable, market, or non-market.
These various cost distinctions have significant effects on decision making. Consumers
base decisions primarily on perceived internal variable costs. Automobile owners decide
how often and how far to drive based primarily on perceived internal short-run variable
costs. Public agencies tend to be influenced by perceived costs to their constituents,
however defined. Current transport planning and investment decisions focus on short- and
medium term direct market costs.
Ideally, public planning and investment analysis should consider all marginal costs,
including long-term, non-market and indirect costs. Since transport planning is based on
time horizons of many years or even decades, virtually all costs can be considered
marginal, including vehicle ownership, roads, parking facilities.
1.5
Treatment of Taxes
Taxes require special consideration in cost analysis. Economists usually consider taxes to
be transfer payments, not costs, and net them out before calculating costs and benefits.15
However, fuel taxes and other charges dedicated to roadway facilities are often considered
15
Ian Heggie and Simon Thomas, "Economic Considerations," Transportation and Traffic Engineering
Handbook, Institute of Transportation Engineers/Prentice Hall (Englewood Cliffs, NJ), 1982, p. 426.
Page 1-10
Transportation Cost Analysis
user fees, and are frequently treated as such in economic analyses. Also, when one activity
is exempted from a broad based tax, it can be treated as an expenditure. 16 Lee states,
"Referring to these as 'expenditures' derives from the idea that the result would be the
same if all taxpayers paid the tax, and the revenues were then paid out to the favored
subset. "17 Examples of this include exemptions of general sales taxes on motor vehicle
fuel, unique petroleum industry tax loop holes, and the exemption of roadway rights-ofway from property taxes, each of which is discussed later in this report.
1.6
Discount Rate in Cost Analysis
Discount rates affect calculations of future costs and benefits. Discount rates reflect the
time value of money, which assumes that wealth can be invested to generate a profit, so
current resources have greater value than future resources, even after adjusting for
inflation. Discount rates that include inflation are referred to as nominal discount rates,
while those that are net of inflation are called real discount rates. Selecting the correct
discount rate is important when assessing environmental costs and benefits that may occur
decades or generations in the future. The higher the rate, the more weight is given to
present over future benefits. Capital investment discount rates are typically 8-10%. These
rates reflect the return capital could earn in typical alternative investments.
A debate now exists as to the discount rate to use for human health and environmental
costs imposed on future generations.18 Conventional discounting implies that costs many
years in the future are oflittle concern now.19 For example, at an 8% discount rate, costs
16 Examples
include Apogee Research, 1993, p.9; Ridgeway, 1990, p. 14; Hubbard, 1991, p. 21.
Lee, Full Cost Pricing of Highways, Na. Transport Systems Center (Cambridge), 1995, p. 31 .
18 C. van Kooten, Land Resource Economics and Sustainable Development, UBC Press (Vancouver),
1993, p. 96.
19 One justification for discounting costs imposed on future generations is the assumption that they will be
wealthier, on average, than current generations. In recent years this assumption has been challenged.
Herman Daly, Paul Erlich and others argue that future generations may have less wealth than we do.
17 Douglass
Page 1-11
Transportation Cost Analysis
and benefits occurring 20 years in the future (a typical planning horizon) are worth less
than a tenth of their current value. Some analysts argue that these financial assumptions
are inappropriate for evaluating human health risk and irreversible environmental impacts,
and do not reflect society's desire to provide a better world for our descendants. They
recommend using a 0-4% discount rate for human health and environmental costs and
benefits to give fair consideration to future generations' interests.20
1.7
Pricing Non-Market Goods21
Including non-market costs in public decision making is challenging but important.
Excluding them skews decisions toward options with high environmental and social
impacts. The transport planning profession uses established values for travel time savings
and accident reductions for investment analysis. This study expands the list of non-market
goods that are monetized to include environmental and social costs.
Assigning monetary values to nonmarket goods can improve planning and policy making.
It facilitates fairness and economic efficiency. For example, it would be unfair and
inefficient if one firm or sector was required to spend $2,000 per ton ofNOx reduction
while another firm producing comparable emission spends significantly less. Greater total
benefit may be achieved by shifting resources to the more cost effective option. Of course,
there are situations in which different unit costs for environmental protection are justified,
for example, to place a greater burden on firms with more resources, but these should be
conscious decisions. This requires that the cost per unit of benefit be determined as a
reference, which is essentially monetization.
20
Robert Costanza and Herman Daly, "Natural Capital," Conservation Biology, Vol. 6, No. 1, Mar. 1992.
21 David Pearce, Economic Values and the Natural World, MIT Press (Cambridge), 1993; Ismail
Seregeldin, Ed. , Valuing the Environment, World Bank, Washington DC, 1994; Nick Hanley and Clive
Spash, Cost-Benefit Analysis and the Environment, Edward Elgar (Brookfield), 1993; David James, The
Application of Economic Techniques in Environmental Impact Assessment, Kluwer (Boston), 1994.
Page 1-12
Transportation Cost Analysis
There is nothing unusual or mysterious about valuing non-market goods. Individuals and
public officials often make decisions which trade non-market goods, such as clean air,
quiet, and wilderness preservation, against money or market goods. For example:
•
Home buyers must decide how much extra they will pay (in dollars or by giving up
other amenities) for a residence that is subject to less noise or air pollution.
•
Public agencies must decide how much society should spend (either in direct
expenditures or by giving up other benefits) to achieve goals such as improved air
quality, reduced accident risk, or increased speed and comfort for drivers.
•
Individuals choose how much to spend to avoid a hazard (such as using a longer but
safer travel route), obtaining safety (such as buying the latest automotive safety
equipment), or how much compensation they require to work at a dangerous job.
When numerous transactions involving trades between market and non-market goods are
performed it is possible to identify patterns that effectively determine the price paid for the
non-market good. In recent years a number of methods have been developed to measure in
monetary units the value that society is willing to pay for non-market goods. Monetization
of non-market goods is becoming increasingly common in a number of fields including
energy planning, injury compensation, and environmental policy analysis. There are five
general techniques for monetizing non-market costs: 22
1. Hedonic Methods (also called Revealed Preference)
Hedonic pricing infers values for non-market goods from their effect on market prices.
A common strategy is to analyze the effects of impacts on property values and wages.
For example, if houses on streets with heavy traffic are valued lower than otherwise
comparable houses on low traffic streets, the cost of traffic (or, conversely, the value of
neighborhood quiet, clean air, safety, and privacy) can be calculated.
22 Kenneth Button, "Overview of Intemalising the Social Costs of Transport," in Internalising the Social
Costs ofTransport, OECD (Paris), 1994, p. 17.
Page 1-13
Transportation Cost Analysis
2. Control or Prevention Costs
A cost can be estimated based on prevention, control or mitigation expenses. For
example, if industry is required to spend $1,000 per ton to reduce an air pollutant, we
can infer that society considers that emission to impose costs at least that high.
3. Contingent f!:aluation (also called Stated Preference)
Contingent valuation infers costs by surveying a representative sample of society how
much they value a particular non-market good. For example, residents may be asked
how much they would be willing to pay for a certain improvement in air quality, or an
acceptable minimal compensation for the loss of a recreational site. Such surveys must
be carefully structured and interpreted to obtain accurate results.
4. Precedents.
This uses policy and legal judgments as a reference for assessing non-market costs.
5. Travel Cost
This method uses visitors' travel costs (monetary expenses and time) to measure
consumer surplus provided by a recreation site such as a park or other public lands.
A high standard of protection of possible irreversible losses to future generations is often
supported based on the precautionary principle .23 Protection against such losses may add
option value to an environmental good's direct benefits.24 Examples of irreversible impacts
include species extinction and climate change. Even the cutting of old growth forest or
draining a wetlands may be irreversible within the time frame of human lifetimes.
23 Andrew Jordan and Timothy O'Riordan, The Precautionary Principle In UK Environmental Law and
Policy, Center for Social and Economic Research on the Global Environment (London), 1994.
24 Hanley and Spash, Cost-Benefit Analy sis and the Environment, Edward Elgar (Brookfield) p. 153.
Page 1-14
Transportation Cost Analysis
1.8
Criticism of Transportation Cost Analysis
The analysis in this report tends to be criticized from two perspectives. One is from
transportation professionals and automobile industry advocates who argue that too much
emphasis is placed on costs without acknowledging the benefits provided by our
transportation system.25 When confronted with evidence of external transportation costs
their reaction to costs tends to follow the following progression:
1. "The cost does not exist."
2. "It may exist, but is not significant."
3. "It may be significant, but is not related to driving."
4. "It may be related to driving, but cannot be quantified."
5. "It may be quantified, but incorporated it into decision making is impractical."
6. "It may be incorporated in decisions, but to do so would not be politically acceptable."
7. "Benefits to society surely outweigh this cost."
The response to points 1-5 rests on the strength of each analysis to demonstrate that a
particular cost exists, that it may be significant compared with costs currently considered
in transport planning, that transportation contributes to it, and that methods exist to
measure it. Since these are relatively new fields of research there is still uncertainty about
some cost estimates, resulting in the recommendation, universal to all such studies, that
further research is needed. However, a major reason that transportation decision makers
are not aware of the full range of external costs, and fail to use monetized estimates of
non-market costs in planning and investment analysis is simply that they have made little
effort to learn about current research in these field or to expand existing data to fill gaps.
Points 6 and 7 require a different response. A discussion of costs does not deny political
realities or the existence of benefits. It does, however, allow society to be more conscious
and precise in decision making so that the maximum benefits can be achieved. It would be
25
See for example Eric Beshers, External Costs ofAutomobile Travel and Appropriate Policy Responses,
Highway Users Federation (Washington DC), March 1994. Contact the author of this report for his
responses to Beshers' analysis of external transportation costs.
Page 1-15
Transportation Cost Analysis
~
irresponsible to provide a blank check for the purchase of any good, no matter how
beneficial, which is implied in arguments that, ''Benefits surely outweigh the costs. "
The second type of criticism typically comes from environmentalists who consider
economics in general, and the monetization of non-market goods in particular, to be
reductionist and inappropriate. Conventional economic models assume that people are
simply selfish consumers, that all resources are commodities, and that market activity
defines society's well being. These assumptions are clearly wrong. It is incumbent on
economics to incorporate a broader view of humanity, society and the environment. In
recent years the field of ecological economics has developed a vocabulary that reflects
society's non-commercial aspirations such as generosity, cooperation, and spirituality. We
can now discuss, for example, the existence value that people place on sacred objects, and
the importance of option and bequest values.26 Of course, there is still uncertainty
concerning these concepts and methods, and many economists do not acknowledge or
understand them, but the potential exists for economics to address the criticisms raised.
Readers who are uncomfortable with formal economics may prefer to mentally replace the
word "cost" with the word "problem" when it appears in this report. Thus the question, "Is
automobile air pollution a cost?" becomes, ''Is automobile air pollution a problem?" But
after reviewing the various problems (costs) readers may become curious about their
relative magnitude. ''Is this problem significant?" "How does it compare with other
problems?" These lines of inquiry return us to the concept of costs (benefits foregone),
and the usefulness of quantitative estimates. We might begin by assigning degrees of
"badness" (''Air pollution is a 10 on the badness scale, and traffic noise is a 6. ''),but if
any such cost can be valued in dollar units, it becomes possible to monetize them all, based
26
Existence value is the value people place on resources or goods that they but do not use directly but
wish to preserve. Option value is the value of retaining a choice even if it is not immediately used.
Bequest value reflects the desire of individuals to provide goods and benefits to future generations.
Page 1-16
Transportation Cost Analysis
~
on relative "badness." Monetized costs, it turns out, are useful references that are already
widely used in various types of decision making.
1.9 Treatment of Variability and Uncertainty
The cost estimates provided in this and other reports are generic values. Of course, there
is considerable variability depending on many factors, including location, time, vehicle
type, and driver behavior. Ideally, the cost values presented here would be modified as
appropriate before they are used in specific planning applications. For example, if you are
calculating parking cost savings that would result from increased transit commuting in
your community, you should consider whether your community has a shortage of parking
spaces, whether this problem is increasing or decreasing, whether the commuters who
would shift to transit are more or less likely than average to receive free parking, and
whether parking facility costs are higher or lower than average in your area. Depending on
your time, resources and needs, you can use the generic numbers from this report as they
are, do a spot check to identify any significant variations from national averages, or
perform a study of this cost specific to your circumstances, in which case this report can
serve as a reference for your research.
Because transport cost analysis involves new areas of research, limited data sources, and
complex modeling, estimates incorporate various levels of uncertainty. This is not a unique
problem; individuals, businesses, and society often face uncertainty when assessing costs
and benefits. As stated by Professor Richard Ottinger, an expert in environmental costing,
''A crude approximation, made as exact as possible and changed over time to reflect new
information, would be preferable to the manifestly unjust approximation caused by
ignoring these costs, and thus valuing environmental damage as zero. '127
27 Richard
Ottinger, "Incorporating Externalities- The Wave of the Future," in Expert Workshop on
Lifecycle Analysis ofEnergy Systems, OECD (Paris), 1993, p. 54.
Page 1-17
Transportation Cost Analysis
~
A common way to deal with uncertainty in economic analysis is to include only costs that
are commonly accepted and easily quantified. If a cost is difficult to measure, it is often
ignored, even if it is probably comparable in magnitude to other costs. 28 Excluding or
using low estimates of costs that incorporate uncertainty is often defended as being
"conservative," implying that this approach is cautious.
The use of the word conservative in this context is confusing because it results in the
opposite of what is implied. Low cost estimates result in undervaluing damages and risks,
thereby overvaluing relative benefits and assets, which is less cautious and less
conservative in accounting terms. Accountants prefer to use high estimates of risks and
losses and low estimates of benefits and assets when uncertainty exists in order to avoid
careless optimism. For example, if different assessments are made of an asset's value, an
accountant should generally use the lower estimate for calculating net worth because most
individuals and businesses can handle an unexpected abundance of wealth better than its
unexpected absence.
When economists call low estimates of costs conservative they are actually using the word
in its political meaning of maintaining the status quo, not the conservation of resources. In
practice, low estimates of non-market and indirect costs leads to increased social and
environmental damages since these have only recently been included in economic analysis.
For example, a low estimate of air pollution costs will reduce the justification for investing
in emission control efforts, resulting in more pollution and less conservation of natural
resources. In other words, excluding or undervaluing these costs result in less
conservative and cautious analysis results.
28
These "intangible" costs are often addressed through a separate part of the transportation planning
process, such as an environmental assessment or through public participation. But such a duel track tends
to underweigh non-market values and provides little help to planners and designers in making tradeoffs.
Page 1-18
Transportation Cost Analysis
~
Another way to deal with uncertainty is to use a range of costs rather than a point
estimate. For this reason minimum and maximum estimates of average automobile costs
are provided in chapters 3.1 to 3.16 to facilitate sensitivity analysis. However, establishing
ranges requires the same estimation methods as a point estimate, so this approach doesn't
completely solve the uncertainty problem. It should be understood that all point estimates
represent a range of values that depend on time, place, and other variables and
uncertainties.
Some cost estimates with a relatively high degree of uncertainty are included in this report,
provided that the existence of the cost can be demonstrated, and the resulting estimate is
within the expected range with respect to other costs. Assuming that the variation among
the uncertainty is random, the over- and under-estimates among these estimates will tend
to cancel each other out. Including such estimates is more accurate and more conservative
than setting their value at zero, which consistently underestimates total costs.
Page 1-19
Transportation Cost Analysis
~
2.0 Transportation Cost Literature Review
Several previous studies describe, assess, and calculate transportation costs, encompassing a
wide range of perspectives and techniques. Many address only one or two costs. A few attempt
to be more comprehensive but include no original material or perspective. Seventeen cost studies
summarized in this chapter were selected because they include at least some original research, are
comprehensive, or because they represent a unique perspective. Taken together, the 17 studies
indicate current knowledge and trends in this field . Table 2-10 at the end of this chapter
summarizes the costs in these various studies.
I. The Full Costs of Urban Transport; Intermodal Comparisons ( 1975) by Keeler, et al. This
report summarizes research by transport economists at the Institute of Urban and Regional
Development, comparing commuting costs of automobile, bus and rail in the San Francisco Bay
area. It includes calculations of marginal congestion costs, public services, noise, air pollution,
facilities, accidents, parking, and user costs. This is the oldest study of its type. The analysis is
still highly regarded.
2. Transportation Efficiency: Tackling Southern California's Air Pollution and Congestion,
(March, 1991) by Michael Cameron, published by the Environmental Defense Fund. This study
uses estimates of external transportation costs to argue for pricing as a strategy for demand
management. External costs include air pollution, congestion, and parking. The recent follow-up
study, Efficiency and Fairness on the Road, (March 1994) by the same writer and publisher,
extends this research to cover transportation policy equity impacts.
3. The Costs of the Car: A Preliminary Study of the Environmental and Social Costs Associated
withPrivate Car Use in Ontario, (October 1991), by Pollution Probe, an environmental
Page 2-1
Transportation Cost Analysis
~
organization in Toronto. This critique of automobile use and the automobile industry covers a
large number of costs, as described in Table 2-1.
Table 2-1
"The Costs of the Car" Cost Catel!ories
Land Use
I Environmental
I Human Health
Highway expenditures.
Destruction of
agricultural land and
urban greenspace.
Excessive energy
consumption.
Government
environmental
expenditures.
Mining.
Metal smelting.
Energy use.
Petroleum industry.
Air pollutants.
Maintenance of a carcentered infrastructure.
Disoosal.
Road safety
expenditures.
Health care costs.
NOx, VOCs and Ozone.
Carbon Monoxide.
Lead.
Water pollution.
Ozone depletion.
Global warming.
Social
Policing.
Court costs.
Congestion and lost
time.
Stress and decline in
quality of life.
The transportation
disadvantaged.
Death and injury.
4.Making Transportation Choices Based on Real Costs, (1991) by Brian Ketcham. This paper
includes monetized estimates for air pollution, noise and vibration damage, pavement wear and
indirect damage to other vehicles, congestion costs, and traffic accidents. Also mentioned but not
quantified are water pollution, oil spills, global air pollution, roadway and parking land value,
petroleum import costs, vehicle production, and waste disposal external costs.
5. Results of Literature Survey and Summary of Findings: The Nature and Magnitude of Social
Costs of Urban Roadway Use (July 1992) by Mark Hanson, published by the U.S. Federal
Highway Administration. This study identifies external costs of urban roadway transportation and
describes costing methods. It also includes recommendations for better calculating external costs,
incorporating costs into user prices, and applying least-cost planning to transportation.
6. Kjartan Saelensrninde's Environmental Costs Caused by Road Traffic in Urban Areas -
Results from Previous Studies (1992), published by the Institute for Transport Economics in
Oslo, Norway. This study focuses on three costs of urban road transport: air pollution, noise, and
the barrier effect. Table 2-2 summarizes these costs for Norway.
Page 2-2
Transportation Cost Analysis
Table 2-2
Costs of Noise, Air Pollution and Barrier Effects Due to Road Traffic in Urban Areas.
Cost Per Person
Total Costs in Norwa
Million NOK/Year I Million US$/Year
NOK/Year
I
US$/Year
Noise
600-3,700
I
88- 541
180-1,110
I
26- 162
Air Pollution
5,100-26,ooo
1
746- 3,802
770-3,900
I
112 - 569
Barrier Effect
767
I
112
3,300
I
483
NOK = Norwegian Kronor
7. The Going Rate, 1992, by James MacKenzie, Roger Dower, and Donald Chen for the World
Resources Institute. A comprehensive study of U.S. motor vehicle costs. Cost categories include
roadway facilities and services, parking, air pollution and global warming, security costs of
importing oil, congestion, motor vehicle accidents, noise, and land loss. The report's conclusion
that driving incurs as much as $3 00 billion in external costs each year is widely quoted.
8. Getting the Prices Right; A European Scheme for Making Transport Pay its True Costs,
(1993) by Per Kageson, for the European Federation for Transport and Environment. This study
estimates pollution, accident and infrastructure costs in European countries. Cost summaries for
the UK are shown in Table 2-3 . Similar estimates are made for other European countries.
Table 2-3
External Transport Costs ECU/1000 passenger km)
Mode
Car
Electric train
Aircraft
Air pollution
14.6
0.9
7.3
co?
4.5
2.2
9.2
Noise
0.9
0.2
1.2
Accidents
8.9
3.8
0.2
Total
28.9
7.1
17.9
Total ($/mile)
$0.060
$0.015
$0.037
9. The Cost of Transporting People in the British Columbia Lower Mainland (March 1993 ), by
Peat Marwick Stevenson & Kellogg for Transport 2021, a planning effort for the Vancouver
region. This study develops cost estimates for 12 modes using local research and generic
estimates. Costs included in the study are listed in Table 2-4.
1The barrier effect is
discussed in chapter 3.13.
Page 2-3
Transp ortation Cost Analysis
__.
Table 2-4
Costs ofT
Direct User
rtin2 Peoole in B.C.
- -- Cost
- -- -~
Indirect
Transport
Time
Parkin2
Infrastructure
Residential. Road construction. Personal.
Commercial. Road maintenance. Commercial
Government. Road land value.
delays.
Transit land value.
Protection services.
Fixed vehicle
costs.
Variable vehicle
costs.
Parking fees.
Urban Sprawl
Environmental
and Social
Infrastructure. Unaccounted
accident costs.
Loss of open
space.
Air pollution.
Future
Noise pollution.
transport.
water pollution.
10. Land Transport Externalities ( 1993 ), by Works Consultancy Services for Transit New
Zealand. This comprehensive and well-researched study is part ofNew Zealand's efforts to
rationalize transport planning, and possibly implement road pricing. It attempts to describe all
external costs of roadway transportation and identify costing methodologies. Cost categories are
shown in Table 2-5 . Cost estimates will be developed in future reports.
Works Consultancv Cost Cate2:ories
Interference
Pollution Effects I Intrusion Effects
I
Effects
Table 2-5
Air Pollution & Dust
Impacts on the Global
Atmosphere
Effects on Water
Systems
Noise & Vibration
Disoosal of Waste
Visual Effects
Habitat impacts.
Effects on Landscape
Archaeological Sites
Cultural & Spiritual Effects
Recreational Effects
Strategic Effects
Urban Form
and Land Use
Community Disruption
Urban and Rural Blight
and Stress of Change
Lighting Effects
Community Severance
and Accessibility
Hazard Effects
11. The Price ofMobility (October, 1993), by Peter Miller and John Moffet, published by the
Natural Resources Defense Council. It attempts to quantify total annual U.S. costs for
automobile, bus, and rail transport. It is one of the most comprehensive efforts in terms of costs
described and quantified for these three travel modes. Costs included are listed in Table 2-6.
Table 2-6
-
-
-
-
The Full C
Personal
Automobile ownership.
Transit fares.
fT
Government Subsidies
Capital and operating.
Local government.
he U.S.A
Societal
Energy.
Congestion.
Parking. Accidents.
Noise.
Vibration.
Air pollution.
Water pollution.
Page 2-4
Unquantified
Wetland lost.
Farmland lost.
Historic property.
Property value impacts.
Inequity.
Sprawl.
Transportation Cost Analysis
-12. The Costs of Transportation: Final Report (March 1994), by Apogee Research for the
Conservation Law Foundation. This estimates user, accident, congestion, parking, road facilities
and services, air pollution, water pollution, energy, and noise costs. Some additional costs, such
as urban sprawl and aesthetic degradation are mentioned but not estimated. A costing model is
developed which calculates the total cost of trips by nine modes, in three levels ofurban density,
during both peak and off-peak periods. This model is applied to case studies of Boston and
Portland, Maine urban travel costs.
Table 2-7
Federal Railroad Administration Costs
Social Costs
I
Community
Disruption
Direct land use for Divides
facilities.
community.
Alters land use
Impacts local
patterns (sprawl). government.
Visual pollution.
Relocation
impacts.
Land Use
Air
Carbon Monoxide.
VOCs.
S02.
NOx.
C02.
Air Toxics.
Particulates.
CFCs.
Odor.
I
Noise
Construction/
repair.
Night operations.
Engines
Wheels/tires.
Congestion.
Braking/
acceleration
Idling.
Whistles.
Energy
I
Safety
Oil spills.
Accidents cause
Air pollution.
death, injuries,
Political instability
insurance and
from foreign oil.
legal costs, lost
Oil price
productivity,
fluctuations
medical costs,
affecting world
emotional losses,
con_gestion.
economy.
Environmental Costs
I (Electromagnetic
Water
I
Fields M
(Cost of electric
vehicles.)
Air pollution
fallout.
Fuel releases and
Possible biological
spills.
Construction/
hazard.
maintenance.
Possible hazard to
De-icing.
migrating birds.
Runoff from roads Problems to
and parking lots.
electronic
eauioment.
I
Congestion
Wasted time.
Wasted fuel.
Added pollution.
Lost productivity.
Vehicle repair and
insurance costs.
Stress.
Land use imoacts.
Hazardous
Materials
Accidental
releases.
Intentional
releases.
13. Environmental Externalities and Social Costs of Transportation Systems- Measurement,
Mitigation and Costing: An Annotated Bibliography (August 1993), USDOT, Federal Railroad
Administration, Office of Policy. This bibliography describes recent publications on
Page 2-5
Transportation Cost Analysis
---
transportation costing. It includes two charts that describe a taxonomy of costs and mitigation
strategies, summarized in Table 2-7.
14. Full Cost Pricing of Highways (January 1995), Douglass Lee, USDOT Volpe National
Transportation Systems Center, Cambridge. This study analyzes optimal pricing for economic
efficiency. Table 2-8 summarizes Lee's estimates of external costs.
Table 2-8
Estimates of Hi2hwav Costs Not Recovered From Users
Cost Grou
!Highway Capital
Cost Items
Land (interest)
Construction:
Capital Expenditures
Interest
Land acquisition and clearance
Relocation of prior uses and residents
Neighborhood Disruption
Removal of wetlands, acquirer recharge
Uncontrolled construction noise, dust and runoff
Heat island effect
Pavement, ROW, and structure maintenance
Administration and research
Traffic
Parking
Vehicle Ownershi
Vehicle Operation
Fuel and Oil
Accidental Loss
Pollution
Social Overhead
or abandoned vehicles
Pollution from tires
Pollution from used oil and lubricants
Pollution from toxic materials
Strategic Petroleum Reserve
Tax subsidies to oroduction
Government compensation for natural disaster
Public medical costs
Uncomoensated losses
Air
Water
Noise and vibration
Noise barriers
Local fuel sales tax exemptions
Federal gasohol exemption
Federal corporate income tax
State government sales taxes
Local government orooertv taxes
Total
Current User Revenues
Loss
cents/VMT
Page 2-6
Estimate
$74,705
42,461
26,255
20,420
6,876
7,756
52,877
14,890
706
3,000
408
1
4,365
9,000
8,535
5,850
43,444
10,861
6,443
5,117
4,302
1,129
3,389
13,218
15,962
$382,134
52,096
330,037
$0.152
Transportation Cost Analysis
15. "The Costing and Costs of Transport Externalities: A Review," Victorian Transport
Externalities Study, Environment Protection Authority (EPA), Melbourne, Australia, 1994. This
report discusses external cost implications, reviews costing methods, and estimates noise, air
emissions, accidents, and congestion costs.
16. Comparing Multi-Modal Alternatives in Major Travel Corridors, (January 1994), by
Patrick DeCorla-Souza and Ronald Jensen-Fisher. This paper compares total costs of highway,
bus, and subway investments. Costs included are listed in Table 2-9.
Table 2-9
DeCorla-Souza and Jensen-Fisher Cost Estimates
Vehicle
Operating.
Hil!bway
Operation&
Ownership. Maintenance
Parking.
Capacity
-
---
Public
Transport
Bus system.
Subway
system.
---
Safety and
Security_
Public services
Accident
(market).
EnvironTravel
mental
Accidents
Time
Travel time. Air& Water. Accidents
(nonNoise.
market).
Waste.
- - ·- - -·- - -
17. Saving Energy In US. Transportation, Office of Technology Assessment, July 1994. This
study includes estimates of total U.S. motor vehicle costs based on preliminary results of an
extensive research project by Mark DeLuchi ofUC at Davis. Updated cost estimates are
scheduled for release by DeLuchi in 1995.
18. External Costs of Truck and Train, Transport Concepts, for the Brotherhood of Maintenance
of Way Employees, October 1994. This study compares external costs of train versus truck
freight transport to justify increased truck taxes or increased subsidies for rail transport. Table 210 summarizes their results.
Page 2-7
Transportation Cost A naly sis
Table 2-10
External Costs of Train Vs. Bus
Intercity Truck
I Rail
Truck
2.15
Net
I
'O:<>rnl
2.10
I Truck
1.73
Canadian Cents
Rail
I Rail
Con-
0.51
0.33
0.28
Rail
Rail Box
0.27
0.17
External Costs
19. Transportation Sector Subsidies; U.S. Case Studies, DRI/McGraw-Hill for the U.S.
Environmental Protection Agency, Energy Policy Branch (Washington DC), November 1994.
This report summarizes existing estimates of external costs including air pollution (local and
~obal),
congestion, accidents, noise and vibration for rural and urban automobile and truck use.
It estimates the macroeconomic effects of implementing various pricing strategies to internalize
costs and reducing C0 2 emissions.
Cost Estimates Summarized
Table 2-10 summarizes the transportation cost studies, identifying costs that are either described
or estimated in each report. These studies show the range of perspectives and efforts exploring
transportation costs.
Page 2-8
Transportation Cost A naly sis
Table 2-10
Transport Costs in Current Literature (C = Costed; D =Described)
Study No. 1
Cost
Catexories
~ehicle Costs
~ravel Time
fWdents
Parking
Congestion
F.cilities
Roadway Land
~un. Services
~rPoUution
lAical
Air Pollution
Global
Noise&Vibration
Resources/
2
Keeler
eron.,
EDF
c
c
c
c c
c c
c
c
c
c c
c
~nergy
.arrier Effect
Land Use/Sprawl
Inequity
Water
Waste Disposal
~storic Artifacts
3
4
5
6
7
8
9
Cam-
D
Pol.
Probe
D
D
D
Ketcham
Hanson
c
D
D
c
D
D
D
D
D
c
c
D
D
c
c
D
D
D
D
D
D
D
D
D
D
D
c
c
c c c
c
c
c D c
c c c
D
c
c D c
c c c c
c c
c c c c
c
c
c
D
D c
-·
----
D
- --- - -
D
--
- - -·-
-
Page 2-9
11
12
ApoPMSK Works, Miller, gee,
N.Z.
Moffet CLF
D
D
D
D
D
D
D
D
D
.
Saelen Mac
K~sminde Kenzie eson
10
c
D
D
D
D
D
D
D
D
D
c c
c
c c
c c
c c
c c
D c
c c
c c
c c
c c
c c
D
D
13
14
15
16
EPA,
Aust.
De
Coria OTA
Souza
us
DOT, Doug
FHWA Lee,
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
17
c
c
c c
c
c c
c
c
c
c
c
c
c
c
c
c c c
c
c c c
c c
18
19
Tran.
Cone.
DRII
MG ,
EPA
c c
D
c c
c c
D
D
c c
c c
c c
D
D
D
c c
D
D
D
D
D
c c
c c
D
Transportation Cost Analysis
3.0 Definitions, Costing Methods, and Estimates
Each of the next 16 chapters (3. 1 through 3. 16) defines, describes, and estimates a specific
road transport cost. Costs are valued in 1994 U.S . dollars and units (mile, foot, U.S.
gallon) except where noted otherwise. Best Guess cost estimates are provided for eleven
modes under Urban Peak, Urban Off-Peak, and Rural travel conditions.1 The distribution
of driving between these travel conditions is shown in Table 3-1 . Minimum, maximum and
weighted average costs are also provided.
Urban Peak
Urban Off-Peak
Rural
Total
- --- - ---
VMT (billions)
Percent of Total
460
920
920
22300
20%
40%
40%
100%
- - --
This table shows how US. trips are divided into three travel conditions in this report.
These Best Guess estimates are either dollars per vehicle mile or per passenger mile,
depending on what is most appropriate for each cost. In a separate spreadsheet these
values are converted into dollars per passenger mile based on average passenger rates, to
create a fair comparison between modes. These values are the basis for cost comparisons
summarized in Chapter 4, and a variety of analysis in chapters 5, 6 and 7.
Cost Chapter Sections
The following 16 chapters include these sections:
Definition: Defines the cost for this analysis.
Description: Describes the cost as it typically applies. (This is not needed for all costs)
Discussion: The existence of each cost, its relationship to transport activities and specific
modes are explored, and useful background information is provided.
1 Urban includes
suburban areas, since they experience congestion, parking, environmental, and
municipal service costs similar to denser urban areas.
2 Facts and Figures 93, Motor Vehicle Manufacturers Association (Detroit), p. 62, assuming 3% annual
growth in VMT since 1992. Percent Urban Peak is estimated.
Page 3.0-1
Transportation Cost Analysis
Estimates: Existing estimates of this cost are summarized, and in some cases an original
estimate is provided.
Variability: Factors that may affect this cost are described.
Conclusions: The cost is summarized and a Best Guess estimate per vehicle mile is made
for the 11 modes under Urban Peak, Urban Off-Peak, and Rural conditions. The weighted
average of these three conditions (based on mileage estimates in Table 3-1) is also listed.
Automobile Cost Range: Minimum and Maximum costs are defined for Average
Automobile travel. 3 This is based on the highest and lowest reasonable estimates.
In this analysis Urban Peak travel is generally used interchangeably with commuting. In
recent years travel surveys have started counting each link of a trip separately. As a result,
statistics now indicate that a majority of peak period travel is not considered commuting.
However, many links not officially considered commute trips are part of the overall trip
between home and work, such as stops at a daycare center or store. Survey data indicates
that work trips account for 21.6% of all personal trips, while chained trips related to
commutes represent 30% of personal trips.4 These are all considered commute trip in this
analysis since their scheduling and direction are generally determined by employment.
The transportation cost framework and specific cost estimates described here can be used
in many specific analyses. Some examples are explored in Chapter 6. Since these cost
estimates are generic, representing overall North American averages, ideally they should
be adjusted and updated to specific applications. For example, parking facility costs may
be higher or lower in specific situations than these estimates due to variations in real estate
values, rates of free parking, and ratio of parking spaces per vehicle. This framework can
also be expanded to include additional modes or travel situations, such as intercity
passenger rail, or alternative fuels. Users can adjust or expand these cost estimates as
necessary, taking care to be consistent with the framework and costing methods.
3
Since cost estimates for most modes are based on average automobile costs, the range of costs for other
modes can usually be calculated from this estimate.
4 James Strathman and Kenneth Dueker, "Understanding Trip Chaining," 1990 NPTS Special Reports on
Trip and Vehicle Attributes DRAFT, USDOT, (Washington DC), 1994.
Page 3.0-2
Transportation Cost Analysis
"Do I need to read all of the technical stuff?"
Much of this analysis is specialized and tedious, and some sections assume a basic
understanding of economic theory. It is simply a matter of document style that these
chapters are not a separate technical appendix, and many readers may prefer to treat them
as such. Most readers will want to review two or three chapters dealing with costs that
they are familiar with in order to understand in general how the estimates are derived and
how the costing chapters are organized. Few will need to study each cost chapter in detail.
3.0.1 Measuring Costs
It is important to determine whether a marginal or average analysis is appropriate for
ridesharing and transit costs. Marginal analysis assumes that the vehicle will be making
the same trip anyway, so each passenger only incurs additional costs in terms of increased
vehicle weight, increased internal accident risk and additional stops. An average cost
analysis assumes that the vehicle trip would not occur without the need created by the
passengers, so each passenger bears an equal share of total cost.
Marginal costs tend to emphasize the short-term, while average costs tend to emphasize a
longer term perspective. For example, the short term cost of accommodating more transit
passengers with existing bus capacity is simply small increases in fuel consumption, air
pollution, and boarding delay. If increased ridership requires more buses, more or longer
routes, earlier replacement or upgrading of equipment, or other expenses, these are costs
of the additional passengers. Car and van pooling costs seem most appropriately based on
marginal costs, on the assumption that the vehicle driver will take the trip anyway,
although trip length and travel time may increase with additional passengers. Van pool
passengers must also bear a portion of the additional costs of a van rather than a smaller
vehicle that the driver would normally choose in the absence of a van pool program.
Page 3.0-3
Transportation Cost Analysis
Transit rider costs seem most appropriately based on average costs, since the system
exists specifically for them; drivers would not drive and buses would not run without
riders using the system. According to the American Public Transit Association, urban
transit buses carry approximately their full seating capacity in the peak direction during
peak hours, or about 53 passengers on a 40 foot bus, but carry fewer passengers during
off-peak periods, on return trips, and toward the ends of the line. 5
Average bus occupancy rates are used in this analysis even during peak periods when
buses are full since low occupancy off-peak trips result from the need to provide capacity
during peak period trips. Since the system's total number ofbuses, bus size and labor costs
are generally determined by peak trip needs, it seems reasonable to charge passengers on
the system's average costs rather than a simple marginal cost. Thus, although the cost per
rider of operating a full, peak hour bus is lower than for a lightly loaded off-peak bus,
from an overall system perspective the off-peak passenger incurs a lower marginal cost
because of excess capacity. This conclusion is indicated by the increasing tendency of
transit systems to offer off-peak rider discounts.
3.0.2 Modes Defined6
l.
Average Car. A medium sized car or light truck that averages 21 mpg overall (16
mpg city driving, 24 mph highway driving). Overall average occupancy is 1.5, but for
commuting (peak-period travel) is 1.1.
2.
Fuel Efficient Car. A small four passenger car that averages 40 mpg overall (34
mpg city driving, 46 mpg highway driving). Overall average occupancy is the same as
an average automobile.
5 Terry Bronson,
American Public Transit Asso. Research Analyst. Conversation, Aprill4, 1994.
efficiency based on Highway Statistics 1990, USDOT (Washington DC), 1991, Table VM-1.
Occupancy values from Homburger, Kell and Perkins, Fundamentals of Traffic Engineering, 13th Edition,
Institute of Transportation Studies, UCB (Berkeley), 1992, p. 11-8.
6 Fuel
Page 3.0-4
Transportation Cost Analysis
3.
Electric Car. A small four passenger battery powered electric car based on current
technology, which consumes an average of0.5 kWh per mile oftravel.
4.
Van. (including driver) A survey of dealers indicates that 14 passenger vans get 13
to 15 mpg city driving, 18 to 21 mpg highway driving. Assuming that van pool
driving is 2/3 highway and 1/3 city driving, and that actual peak period driving fuel
efficiency is 15% lower than these ratings due to increased congestion and imperfect
driving, overall average fuel efficiency is 15 mpg.
5.
Rideshare Passenger. This is the incremental cost of a car pool, van pool or transit
rider. A survey of dealers indicates that fuel efficiency decreases 2-3 mpg for a van
loaded with 10 passengers (1,500 pounds) compared with no load. This indicates an
average per passenger reduction in fuel efficiency of0.25 mpg. This same value is
used for automobile passengers. In addition, some passengers require additional
driving to be picked up. Assuming that this averages 0.01 extra distance per
rideshare passenger (2% extra miles to assemble a 2 person car pool, 10% extra
miles to assemble a 10 passenger van), this means a 0. 01 x 15 mpg = 0.15 mpg
average fuel consumption premium in addition to the 0.25 reduction in fuel
efficiency, giving a total average fuel cost of0.4 mpg per rideshare passenger.
6.
Diesel Bus. A 40 foot bus (total capacity 53 seated and 20 standing passenger) with
an Urban Peak occupancy rate of 25 passengers, and an overall average occupancy
of9.3 passengers, averaging 6.5 mph.
7.
Electric bus!frolley. A 65 maximum passenger bus or trolley with a peak period
occupancy of30 passengers, an overall average occupancy of 14 passengers7 that
averages 6.5 mpg energy equivalent.
7 1993
Transit Fact Book, American Public Transit Association (Washington DC), p. 78 and 79 indicates
14.0 overall average passengers per trolley mile, and 24.5 overall average passengers per light rail mile.
Page 3.0-5
Transportation Cost Analysis
8.
Motorcycle. A medium size motorcycle that averages 45 mpg for urban driving and
55 mph for rural driving.
9.
Bicycle. A moderate priced bicycle.
10. Walk. A relatively healthy person traveling an average of 10 blocks per trip.
11. Telecommute. This represents two commute trips displaced by allowing an
employee to work from home.
3.0.3 Avoiding Double Counting
It is important to prevent double counting when calculating costs. This is occasionally
difficult because some cost categories overlap. Every effort has been made to prevent this
problem. Two areas require special clarification.
There is potential for overlap between energy (chapter 3.12), air pollution (chapter 3.10),
water pollution (chapter 3.15), and waste disposal (chapter 3.16) costs. To avoid this
problem, emissions occurring during production and distribution are considered air and
water pollution costs, and are not included in the energy cost estimate. Pollution impacts
that occur after use (such as crankcase oil emissions) are considered waste disposal costs.
There is also potential for overlap between equity & option value (chapter 3. 9), the barrier
effect (chapter 3.13), and land use impacts (chapter 3.14). To avoid double counting, the
barrier effect includes only direct costs to pedestrians and cyclists of vehicle traffic; land
use costs focus on problems created by land use changes; transport inequity includes the
problems created for non-drivers by an automobile oriented transportation system.
Page 3.0-6
Transportation Cost Analysis
-----
3.1 Vehicle Costs
Definition:
Vehicle ownership and operating costs.
Description: Automobile user costs include:
Fixed Costs
•
•
•
Variable Costs
•
•
•
Vehicle purchase or lease
Insurance
Registration and vehicle taxes
Maintenance and repair
Fuel, fuel taxes and oil
Paid parking and tolls
Discussion: The key factor in determining whether a cost is internal or external, fixed or
variable, is how it is perceived by users and therefore how it influences purchase and
consumption decisions. Some costs, such as liability insurance, may be partially variable
since the price is affected at least somewhat by annual mileage, but are considered variable
costs only if users typically consider them when make travel decisions. Research has found
that automobile owners often underestimate vehicle operating costs and ignore variable
costs such as maintenance.1
Motorcycle user expenses range from much lower to somewhat higher than an average
automobile, and are significantly affected by insurance costs. Electric cars are currently
relatively expensive to purchase (approximately 150% to 200% the price of a comparable
gasoline automobile), and require replacement of battery packs every 20,000-30,000 miles
at a price of $2,000-$3,000, for an average cost range of $0.07 to $0.15 per mile.2 Other
user expenses, including, tires, insurance, licensing, and parking are comparable to
gasoline vehicles, and maintenance costs are slightly reduced. Typical small electric
vehicles consume 0.25 to 0.5 kWh per mile, so energy costs average $0.1 to $0.06 per
mile based on typical residential energy rates.
1 Cy Ulberg,
Psychological Aspects ofMode Choice, WSDOT (Olympia), 1989, p. 20.
from U.S. Department of Energy, Electric and hybrid Vehicles Program, USDOE,
(Washington DC), 1993 ; Fuel Cell Lab staff at University of California at Davis, conversation 25 August
1994; Gil McCoy, Washington State Energy Office (Olympia.) conversation 16 August 1994.
2 Information
Page 3.1-1
Transportation Cost Analysis
Estimates:
• The American Automobile Manufacturer's Association estimates average user vehicle
annual operating costs in 1993 were $4,514 ($0.45 per mile based on average mileage)
for an intermediate size U.S . car. 3 Ofthis cost, $3,584 ($0.36 per mile) is fixed and
$930 ($0.09 per mile) is variable (including gas and oil, maintenance and tires).
• Apogee Research estimates user costs for various modes in Boston, MA and Portland,
ME, including averages of$0.076 per mile for bicycling and 7.1 for walking. 4
• The Federal Highway Administration uses a different approach from the AAMA to
calculate financing and depreciation, which results in a lower average value per mile. 5
Table 3.1-1 summarizes their estimates for vehicle types used in this report.
Table 3.1-1
Vehicle Size
Sub-Compact
(E-E car)
Intermediate
'(Average Car)
Full-size Van
(Van)
Cost of Ownin ~ and 0
Depree- lnsuriation
ance
Fixed
Fixed
~eratin~
Selected Motor Vehicles (1991 ¢/mile)
Fuel
Maint- Parking
Finance Lie. & Fuel
& Oil Taxes Total
enance & Tolls Tires Charges Reg.
Variable Variable Variable
Fixed
Fixed Variable Variable
8.6
7.1
4.0
1.3
0.7
1.6
0.8
3.5
1.3
28.9
10.7
7.0
4.2
1.3
1.0
2.0
0.9
4.6
1.7
33.4
14.2
8.5
4.2
1.3
1.4
2.9
1.2
8.1
3.0
44.8
• The Canadian Automobile Association estimates fixed automobile costs at $4,975
Canadian per year, and variable costs at $0.083 per kilometer, (US$0.10 per mile).6
• Kenneth Small estimates that vehicle operating costs on urban arterials average 40%
higher per mile than driving on highways, and this cost increases proportional to travel
time when congestion reduces traffic speed to 30 mph on an highway or 20 mph on an
arterial.7 He states that vehicle depreciation is only slightly affected by mileage.
• Electric car user costs are estimated based on a sub-compact's costs in Table 3.1-1, by
increasing depreciation and finance by 75%, increasing maintenance by $0.09 per mile
(for battery replacement), reducing fuel costs to $0.03, and eliminating fuel taxes. 8
3
Facts and Figures '93, American Automobile Manufacturers Association (Washington DC), 1993.
The Costs ofTransport, Conservation Law Foundation (Boston), 1994, p.83 and 92.
5 Jack Faucett Associates, The Costs of Owning and Operating Automobiles, Vans and Light Trucks, 1991 ,
FHWA (Washington DC), 1992.
6 1992-1993 Car Costs, Canadian Automobile Association (Ottawa), brochure.
7 Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992, p. 76.
8 The exclusion of electric vehicle users from paying dedicated road taxes may eventually present a
revenue shortfall problem when such vehicles begin to represent a significant portion of road use, which
may eventually require some sort of tax on electric vehicle use.
4 Apogee Research,
Page 3.1-2
Transportation Cost Analysis
1a01e s.1-~
Electric Car
Lost ot
Depreeiation
Fixed
15.1
uwnm g and UJ~eratmg :small Electric Lar {l~n ~/mlleJ
Insur- Maint- Parking
Finance Lie. & Fuel Fuel
ance enance & Tolls Tires Charges Reg. & Oil Taxes Total
Fixed Variable Variable Variable Fixed Fixed Variable Variable
1.3
0.7
2.8
0.8
3.0
0.0 43.8
_1~!__ 13.0
-----
----
----
---
--
-
------
Variability: There is considerable variation in user costs depending on the vehicle and its
use. An old but reliable, uninsured automobile may cost the user only a few hundred
dollars a year, although in practice some of the depreciation savings from driving an older
car are lost through higher repair costs. An expensive vehicle may incur an order of
magnitude higher user costs, totaling many thousands of dollar each year, however, some
of this cost may be considered to provide luxury or prestige, not transport.
Conclusions: Ownership and operating costs for average car, and vans are calculated
using FHWA data. The FHWA data for an sub-compact car is modified to reflect greater
fuel efficiency ofthe energy efficient car assumed in this study.9 Electric vehicle costs are
calculated as described above. Rideshare passengers incur no additional fixed cost and a
0.4 mpg reduction in fuel efficiency, as described in chapter 3.0. Transit fares average
$1.00 per 8 mile trip, or $0.125 per mile, times 25 average passengers during peak periods
and 9.3 average passengers at other times for buses, and 30 average passengers during
peak periods and 9.3 aver passengers at other times for trolleys.10
Motorcycles costs are half of an energy efficient automobile, except for insurance which is
doubled. Bicycling costs $0.07 per mile, most of which is fixed, and walking costs $0.04
per mile. Telecommuting is estimated to cost $0.20 per mile displaced, assuming an
employee or employer spends $400 annually extra in equipment and utilities to
9 The FHWA estimates
that subcompacts consume 76% the fuel of an intermediate size car, equivalent to
28 and 21 mph respectively. In this study an energy-efficient car is assumed to average 40 mpg.
10 Transit Fact Book 1993, American Public Transit Association (Washington DC), p. 13.
Page 3.1-3
Transportation Cost Analysis
telecommute 100 days a year, at 20 commute miles per day displaced. Fixed costs are
applied equally to all driving conditions; variable costs are assumed to represent Urban
Off-Peak driving, and are increased 15% for Urban Peak travel and decreased by 15% for
Rural travel. 11
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Motorcycle
Bicycle
Walk
Telecommute
Best Guess
Vehicle Class
Bicycle
Walk
Telecommute
Urban Off-Peak
Rural
Averae:e
0.206
0.181
0.258
0.268
0.00
0.00
0.00
0.252
0.05
0.00
0.20
0.206
0.181
0.258
0.268
0.00
0.00
0.00
0.252
0.05
0.00
0.20
0.206
0.181
0.258
0.268
0.00
0.00
0.00
0.252
0.05
0.00
0.20
0.206
0.181
0.258
0.268
0.00
0.00
0.00
0.252
0.05
0.00
0.20
Variable User Vehicle Operating Costs Dollars per Vehicle Mile)
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Moto~cycle
Urban Peak
Urban Peak
Urban Off-Peak
Rural
Average
0.147
0.107
0.207
0.207
0.003
3.125
3.75
0.062
0.020
0.040
0.00
0.128
0.093
0.180
0.180
0.003
1.160
1.160
0.054
0.020
0.040
0.00
0.109
0.079
0.153
0.153
0.002
1.160
1.160
0.05
0.020
0.040
0.00
0.124
0.090
0.175
0.175
0.002
1.553
1.678
0.054
0.020
0.040
0.00
In addition to user costs, public transit vehicles receive operating subsidies. On average,
63.5% oftransit operating expenses are subsidized from taxes.12 Significant electric
vehicle development costs are funded through government programs and through cross
11 Based on fuel efficiency ratings which indicate that urban driving incurs about 30% higher fuel costs
per mile than highway driving. These same ratios are assumed to apply to other variable costs.
12 Transit Fact Book 1993, American Public Transit Association (Washington DC), p. 44.
Page 3.1-4
Transportation Cost Analysis
subsidies within automobile companies that are required to sell a certain number of zeroemission vehicles to meet upcoming emission standards, but these subsidies will not be
included in this estimate due to uncertainties.
-
-
-
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
Urban Peak
Urban Off-Peak
Rural
Avera2e
0
0
0
0
0
3.18
3.18
0
0
0
0
0
0
0
0
0
3.18
3.18
0
0
0
0
0
0
0
0
0
3.18
3.18
0
0
0
0
0
0
0
0
0
3.18
3.18
0
0
0
0
Automobile Cost Range: The Minimum value is a rounded lower estimate and the
Maximum is based on the AAMA estimate. Of course, the cost ranges for other types of
automobiles, such as older, used cars and luxury vehicles are much greater.
Fixed
Variable
Total
Minimum
$0.18
$0.10
$0.28
Page 3.1-5
Maximum
$0.36
$0.15
$0.51
Transportation Cost Analysis
3.2 Travel Time Costs
Definition: The value of travel time.
Description: The value of travel time includes the cost to travelers of unpaid time, and the
cost to employers for work time spent in travel. Although a small amount of recreational
travel time has zero or negative costs (people would rather be traveling than engaged in
other activities), the vast majority of travel time imposes a cost, either to the traveler
during unpaid time, or to an employer for travel occurring during work time. 1 Travel time
can also impose costs on perishable or urgently needed goods, and as capital costs of
equipment, such as trucks. Travel time should be measured door-to-door, meaning that
time spent parking and walking to and from a vehicle is considered part of the trip.
Discussion: Travel time is often determined to be the largest single cost oftransport, and
travel time savings the greatest potential benefit of transport facility improvements. Table
3.2-1 shows typical estimated benefits for UK. highway improvements: 2
Table 3.2-1
Tvoical Hi2h
I
Benefi
Benefit
Vehicle Operating Cost Savings
Travel Time Savings-Work
Travel Time Savings-Nonwork
Accident Savings
Total
Percent of All Benefits
0%
51%
29%
20%
100%
Travel time savings are the largest single benefit in most road improvement projects.
Because of this importance, the values of travel time and travel time savings have been
widely studied and various estimates are recommended for use in economic analysis. The
1 Some people
have suggested incorrectly that automobile travel time costs are reduced by amenities such
as portable phones. Although these may reduce some discomfort and inefficiency, there is little indication
that individuals would prefer to sit in a car (even with a telephone) than be at their destination. The high
charges associated with portable telephone use (about $30/hr.) indicate a very high value of travel time.
2 Ian Heggie and Simon Thomas, "Economic Considerations," Transportation and Traffic Engineering
Handbook, 2nd Edition , Institute of Transportation Engineers/Prentice-Hall (Englewood), 1982, p.419.
Page 3.2-1
--
Transportation Cost Analysis
earliest formal application of travel time savings in transport investment analysis was in
1960 of a proposed freeway through London, in which travel time savings represented
60% to 80% of total expected benefits. 3 A 1963 study of expanding London's subway
system also incorporated travel time saving values.
Table 3.2-2
Estimated Val
fT
Author and Year
Country
Beesley (1965)
UK
IQuarmby (1967)
UK
Stopher (1968)
UK
Oort (1969)
USA
Thomas & Thompson (1970)
USA
Lee & Dalvi (1971)
UK
II
UK
Wabe (1971)
UK
Talvittie (1992)
USA
Hensher & Hotchkiss (197 4) Australia
Kraft & Kraft ( 1974)
USA
McDonald (197 5)
USA
Ghosh et al. (1975)
UK
Guttman ( 1975
USA
II
USA
Hensher (1977)
Australia
II
Australia
Nelson (1977)
USA
Hauer & Greenough (1982)
Canada
Edmonds (1983)
Japan
Deacon & Sonstelie (1985)
USA
Hensher & Truong (1985)
Australia
Guttman & Menashe (1986)
Israel
Fowkes 91986)
UK
Hau (1986)
USA
Chui & McFarland (1987)
USA
Singapore
Mohring et al. (1987)
Cole Sherman (1990)
Canada
II
Canada
IT'
s
% of Avg. Wage
33-50
20-25
21-32
33
86
30
40
43
12-14
2.7
38%
45-78
73
63
145
39
35
33
67-101
42-49
52-254
105
59
27-59
46
82
60-120
93-170
116-165
F
v·
Trip Purpose
Commuting
Commuting
Commuting
Commuting
Interurban
Commuting
Commuting
Commuting
Commuting
Commuting
Interurban
Commuting
Interurban
Leisure
Commuting
Commuting
Leisure
Commuting
Commuting
Commuting
Leisure
Commuting
Commuting
Commuting
Commuting
Interurban
Commuting
Commuting
Leisure
Stud'"'"4
Mode
Auto
Auto, Transit
Auto, Transit
Auto
Auto
Bus
Auto
Subway, Rail
Auto, Transit i
Hydrofoil, Ferry!
Bus
'
Auto, Transit
Auto
'
Auto
Auto
Auto
Auto
Auto
Subway
Auto, Bus, Rail
Auto
Auto, Transit
Auto, Bus
Rail, Coach
Auto, Bus
Auto
Bus
Auto
Auto
Numerous studies have been performed to estimate the value of travel time savings.
3 Dr. W. Waters, Value of Time Savings for Economic Evaluation of Highway Investments in British
Columbia, B.C. Ministry of Transportation and Highways (Victoria), 1992, p.4.
4 Dr. W. Waters, Value of Time Savings for Economic Evaluation of Highway Investments in British
Columbia, B.C. Ministry of Transportation and Highways, (Victoria), 1992, p.42.
Page 3.2-2
1
Transportation Cost Analysis
---Table 3.2-2 summarizes travel time values relative to wage rates from different studies.
Various time value schedules have been developed based on such studies. The American
Association of State Highway and Transportation Officials (AASHTO) has published such
a schedule, as have several European national transport agencies. Many road agencies use
the AASHTO values or similar schedules.5 Most of these schedules include higher rates
for commercial travel, some by class of vehicle. Some have slightly higher time values for
commuting compared with other personal travel. Some schedules have different rates
depending on road conditions. Drivers' time is often valued at a higher rate than
passengers', due to their higher stress, and higher time values are sometimes used for bus
passengers who must stand on the bus or while waiting at a bus stop.6
There is some indication that time values are non-linear; time spent on commutes that are
less than about 20 minutes seem to incur lower costs in terms of driver stress than time
spent on longer commutes, but this has not been quantified.7 A recent study indicates that
unexpected delays impose much higher costs than predictable delays.8 Cy Ulberg
emphasizes that the value of time should be calculated based on perceived time, which is
not always the same as chronological time.9 Walking and bicycling seem to incur relatively
low time costs under favorable conditions, indicated by their popularity as recreation
activities and the willingness of some commuters to walk or ride despite longer travel
5 Dr. W. Waters,
Value of Time Savings for Economic Evaluation ofHighway Investments in British
Columbia, B.C. Ministry of Transportation and Highways (Victoria), 1992, p.4, 92.
6 For example, The Netherlands uses separate time values for "Commuting," "Business" and "Other." The
United Kingdom's Cost Benefit Analysis guidebook states that in-vehicle time should be doubled for
walking to and waiting for a bus. New Zealand uses 25% of wage rate for bus rider, but 40% while
walking to the bus stop and 50% for waiting at a bus stop. As cited in Waters, 1992.
7 Raymond Novaco, Commuting Stress, Ridesharing, and Gender; A nalysis From the 1993 State of the
Commute Study in Southern California~ Transportation Research Board General Meeting, January 1994;
Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992, p. 46.
8 Robert Noland and Kenneth Small, Travel Time Uncertainty, Departure Time Choice, and the Cost of
the Morning Commute, TRB Annual Meeting (Washington DC), Paper 950206, January 1995.
9 Cy Ulberg, Psychological A spects ofMode Choice, WSDOT (Olympia), 1989, p. 17.
Page 3.2-3
Transportation Cost Analysis
__..--'
°
times, and relatively high travel time costs under poor conditions.1 Kenneth Small cites
research by Bruzelius indicating that walking and waiting time costs are double typical
travel time costs, but acknowledged considerable variation among studies.11
These factors have implications for valuing modal shifts. Although riding a bus, car
pooling, bicycling, or walking often take more travel time, under favorable conditions this
additional time is charged at a lower rate than driving alone because bus and car pool
passengers can relax or perform productive work, and bicyclists and walkers benefit from
exercise. However, bicycling, walking, or riding a bus incur higher time costs when
conditions are unpleasant, for example, ifbus riders must wait in uncomfortable conditions
or stand in a crowded bus. Even at the lower price ranges, the value of vehicle occupants'
time significantly exceeds variable vehicle operating costs for most travel. For example, an
automobile averaging 3 0 mph incurs marginal vehicle operating costs of about $0.10 per
mile, but time costs of$0.25 per mile if travel time is valued at $7.50 per hour, and this
cost increases substantially as congestion increases or if the car carries passengers.
Average travel times, distances and speeds vary between modes, as shown in Table 3.2-3 .
Table 3.2-3
C
Trio T'
, L,engt h and Soeed bv Mod"' 12
Automobile
Commute Travel Time (min.)
Commute Trip Length (miles)
Commute Average Speed (mph)
Transit
19.0
11.0
34.7
49.9
12.6
15.2
Walking
All
9.6
0.5
3.1
19.7
10.7
32.3
Average travel time, distance and speed vary by mode.
10 Todd Litman, Bicycling and Transportation Demand Managemen( Paper presented at the
Transportation Research Board General Meeting, January 1994.
11 Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992, p. 45.
12 Alan Pisarski, Travel Behavior Issues in the 90's, USDOT (Washington DC), July 1992, p. 70.
Page 3.2-4
Transportation Cost Analysis
Estimates:
• The AASHTO manual values average travel time savings at $10.44 per vehicle hour in
1985 dollars, which represents a mix of private and commercial vehicles.
• Apogee Research developed estimates of travel time costs per passenger mile for several
modes under urban peak and urban off-peak travel at high, medium and low densities in
Boston, MA and Portland, ME, based on average travel times. 13 Time values were based
on 50% of average local wages for commuting and 25% for other personal travel.
- ·---- --- -
--
. -- ------
Expressway
Boston
High
Medium
Low
Portland
High
Medium
Low
----
NonExpwy
--- - .. -
-----
Comm.
Rail
~r-
.-- -
Rail
Transit
.---- ---o-- ------I
Bus
Bicvcle
Walk
Peak Off-P Peak Off-P Peak Off-P Peak Off-P Peak Off-P Peak Off-P Peak Off-P
24.3
15.2
11.0
9.6
8.0
8.0
40.4 23 .9 28.9 22.7 40.1 28.6 50.5 39.8 60.6 47.8 243
24.3 15.9 19.8 14.0 28.1 25.3 50.5 39.8 60.6 47.8 202
20.2 13.6 19.0 13.3 n/a n/a 50.5 39.8 60.6 47.8 202
11.1
10.0
7.7
7.8
7.1
6.0
19.9 13.1 n/a
16.6 11.2 n/a
12.4 9.8 n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
159
159
159
42.6 33.5 49.8 39.2 166 131
42.6 33.5 49.8 39.2 166 131
30.2 23 .8 49.8 39.2 166 _131_
• The California Energy Commission calculated the value of congestion delay reductions
at $10.60 in their Personal Vehicle ModeP 4
• Travel time savings values developed by Professor W . Waters are shown below. Average
wage rates are currently about $15/hr Canadian (U.S. $11.)
Table 3.2-5
Travel Time Saving Values for British Columbia Ministry of Transportation
and Highways Road Improvement Analysis15
Commercial Vehicle Driver
Personal Vehicle Driver
Adult Car or Bus Passenger
Child passenger under 16 years
Travel Time Values
Wage rate plus fringe benefits
50% of current average wage
35% of current average wage
25% of current average wage
Congestion increases travel time costs for drivers by the following amounts:
Level of Service (LOS)
D: multiply by 1.33
E : multiply by 1.67
F : multiply by 2.0
This travel time schedule includes higher rates for drivers under congested conditions.
13
The Costs ofTransportation: Final Report, Conservation Law Foundation (Boston), 1994, p. 119-120.
California Energy Commission, 1993-1994 California Transportation Energy Analysis Report
Technical Appendices, California Energy Commission (Sacramento), Feb. 1994, 3C.
15 The Value of Time Savings for The Economic Evaluation of Highway Investments in British Columbia,
Dr. W. Waters, British Columbia Ministry of Transportation and Highways (Victoria, B.C.), March 1992.
14
Page 3.2-5
Transportation Cost Analysis
--
Variability: User travel time values vary considerably depending on who is traveling, for
what purpose and under what conditions.
Conclusions: The British Columbia value of travel time schedule is used as a basis for
costing because it is current and comprehensive, for an automobile driver time value of
US$6.00 (50% of$12.00 average wage) and passenger travel time value ofUS$4.20
(35% of$12.00). These values are used for average automobile, fuel efficient cars, electric
cars, vans and motorcycles. Urban Peak driving speeds are estimated to average 30 mph, 16
and incur a congestion premium of 16.5%, assuming that half the trip experiences LOS D
congestion. Urban Off-Peak and Rural travel costs are based on average speeds of35 and
40 mph respectively, and no congestion premium.
Rideshare, bus, and trolley trips typically take longer than driving alone, 17 although on
congested routes with HOV facilities these modes can actually be faster. Rideshare
passengers such as car and van poolers are assumed to incur 20% additional travel time in
order to collect riders. Buses and trolleys are estimated to incur 40% additional travel
time, including waiting and slower average travel speeds due to stops. A travel time rate
of$4.20 per hour is used for vehicle passengers, and no congestion premium is charged.
Unpleasant conditions, such as overcrowded buses would significantly increase this cost.
Walking and bicycling travel time is charged at $3 .00 per hour, which is halfofthe
standard rate for SOV drivers, due to enjoyment, health and moral satisfaction benefits,
although this costs is sensitive to sidewalk and road conditions, and personal preference.
Walking is assumed to average 3 mph. Bicycling is assumed to average 10 mph, and incurs
the 16.5% premium for Urban Peak travel. Telecommuting incurs no time cost.
16 This
speed is roughly calculated based on the 1990 National Public Transportation Survey results
showing that average commute time is 19.7 minutes and distance is 10.6 miles, averaging 32.4 mph.
17 Especially since these trips are defined to include walking and waiting at the stop.
Page 3.2-6
Transportation Cost Analysis
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Motorcycle
Bicycle
Walk
Telecommute
Urban Peak
Urban Off-Peak
Rural
Average
0.23
0.23
0.23
0.23
0.18
0.225
0.225
0.23
0.35
1.00
0.00
0.17
0.17
0.17
0.17
0.154
0.193
0.193
0.17
0.30
1.00
0.00
0.15
0.15
0.15
0.15
0.135
0.169
0.169
0.15
0.30
1.00
0.00
0.174
0.174
0.174
0.174
0.152
0.190
0.190
0.174
0.31
1.00
0.00
Automobile Cost Range: Studies in Table 3.2-2 are used for minimum and maximum.
Minimum
Maximum
$0.11
$0.34
Page 3.2-7
Transportation Cost Analysis
...---
3.3 Accident Costs
Definition: Automobile accident costs net insurance disbursements.1
Description: Automobile accident costs include deaths, injuries, pain, disabilities, lost
productivity, grief, material damage, and accident prevention.
Discussion: Every year about 50,000 Americans die, about 3.4 million are injured, and
millions more experience financial losses from automobile accidents.2 Although the
accident rate per VMT and the fatality rate per accident have decreased over the years,
increased mileage has kept pace so the number of deaths has stayed relatively constant.
Although nobody considers human life a commodity, many individual and social decisions
are made that trade risk of injury and death against market goods. Recent research
estimates appropriate monetary values for risk and risk reduction by tracking such
tradeoffs.3 There are two general approaches to monetizing these costs.4 The Human
Capital method measures only market costs, including property damage, emergency
services, medical treatment, lost productivity, and accident prevention expenditures. This
typically places the value of saving a human life at approximately $500,000, with lesser
values for various injuries. The Comprehensive approach adds non-market costs, including
pain, grief, and reduced quality of life, as reflected in people's willingness-to-pay to avoid
such injuries. This approach typically places the value of preventing a human death at
$2,000,000 to $5,000,000, with related values for injuries.
1 Insurance
disbursements are deducted from the cost defined here prevent double counting insurance
payments included in chapter 3 .1. Other accounting models could handle this overlap differently.
2 Traffic Safety Facts, 1992, National Highway Traffic Safety Admin. (Washington DC), 1993, p. 86.
3 Frank Haight, "Problems in Estimating Comparative Costs of Safety and Mobility," Journal of
Transport Economics and Policy, January 1994, p. 14-17, gives an excellent summary of this field.
4 Ted Miller, The Costs of Highway Crashes, FHWA (Washington DC), publ. No. FHWA-RD-055, 1991
Page 3.3-1
Transportation Cost Analysis
Accident Cost Distribution
An important and challenging problem is to determine what portion of accident costs are
external and which are internal. The existence of external accident costs can be determined
by asking, 11 Should society care
if automobile accidents occur, given current
compensation?" Empirical evidence indicates that society makes a significant effort to
reduce automobile accidents, so external costs appear to exist. Even an accident that only
injures the driver who caused it imposes external costs in terms of emergency and medical
expenses, lost productivity, and griefto family and friends.
Accident costs imposed directly on the vehicle user are considered internal. Accident costs
imposed on non-users are internalized to the degree that they are compensated by users,
either directly or through insurance. Many accident costs (especially non-market costs
such as pain and suffering) are not fully compensated, leaving residual external costs. Even
people who never use an automobile may be forced to pay these costs. Liability insurance
pools accident costs among all insured drivers, so accident costs that are compensated by
insurance disbursements are external at the level of individual drivers but internal at the
sectorial level (all drivers).5 About half of the market costs of roadway accidents are
covered by insurance disbursements, averaging $0.032 per vehicle mile (50% of$137
billion divided by 2,147 billion miles traveled).6
Accident analysis usually assigns accident cost to the heavier vehicle, no matter who is
legally responsible for an accident. For example, if an automobile injures a pedestrian, the
responsibility is allocated to the automobile since it is the heavier vehicle. Costs of crash
between a train and an automobile are assigned to the train. Table 3.3-1 summarizes
accident cost categories and how they are allocated.
5 As
discussed in section 1.4, which standard to use in a particular analysis depends on whether the
concern is equity or efficiency.
6Ted Miller, The Costs of Highway Crashes, FHWA (Washington), 1991, Figure 38.
Page 3.3-2
Transportation Cost Analysis
Table 3.3-1
Allocation of Accident Costs
Allocation
Internal
Insurance
External
Market
Vehicle damage deductible.
Damages and lost income compensation.
Uncompensated damages and lost income.
Non-Market
Uncompensated injuries.
Pain and grief compensation.
Uncompensated pain and grief.
Some accident costs can be considered internal, some external, and some internalized
among all drivers who pay for insurance (internal at the sector level).
Ted Miller estimates that accident costs are divided as shown in Figure 3.3-1
Figure 3.3-1 Accident Cost Distribution 7
400
350
300
1:'!
.!!
8
250
.
200
~
150
'0
c
iii
100
50
Comprehensive
Human Capital
This graph compares accident cost allocation based on two estimates of total accident
costs. The Comprehensive estimate includes pain, grief and reduced quality of life costs.
As an example of the use of this analysis, consider the accident cost implications of a
program that allows people who currently drive to use a safer form of transportation,
thereby reducing automobile accidents.8 Individual drivers would benefit from a reduction
in their own (internal) accident risk. The reduction in insurance disbursements would
benefit insurance companies in the short run, but in a competitive market savings would
eventually be passed on to drivers through lower insurance charges. The reduction in
7 Ted Miller, personal
communication, July 30, 1994, based on cost estimates in his Cost of Highway
Crashes study published by the Urban Institute. This is considered a conservative estimate because a
relatively low value of human life was used.
8 In addition to the direct reduction in accidents, reducing traffic volumes on roadways can also
significantly reduce the accident rate per VMT, according to Forkenbrock, et al., Safety and Highway
Investment, Midwest Transportation Center (Iowa City), 1994, p. 35.
Page 3.3-3
Transportation Cost Analysis
external accident costs would also benefit society in general through reduced medical and
disability costs that are not compensated by insurance premiums.
Peter Miller and John Moffet develop an accident cost allocation model based on marginal
risk. Their results imply that accident costs are divided about equally between internal and
external components.9 Transport Concepts cities estimates that 3% to 47% of accident
costs are external, and argue that the higher range is most appropriate when all costs are
considered, especially if users have no safer travel alternative.10 They therefore treat all
non-market costs of accidents between freight vehicles and other road users as external.
Kenneth Small considers the additional accident risk resulting from additional vehicles in
the traffic stream (he estimates that accidents rise with the square of traffic flow based on
the number of two-vehicle interactions), and the fact that many accident costs are not
borne by the user to estimate that about 50% of accident costs are external. 11
Jan Jansson developed a model of marginal external accident costs which emphasizes the
risk imposed by motor vehicles on "unprotected road users" (pedestrians and bicyclists),
and direct costs to society, such as emergency services and medical expenses.12 He states
that about two-thirds of automobile accident fatalities and half of all injuries in European
cities are unprotected road uses, which is higher than in North American cities. He also
discusses, but does not resolve, the question of whether accident risk between automobiles
imposes external costs.
Robert Davis also emphasizes the difference between external (transitive) and internal
(intransitive) accident risk. 13 He explores the equity and ethical implications of accident
9 The
Price ofMobility, National Resource Defense Council (Washington DC), Oct. 1993, p. 30.
12.
10 External Costs ofTruck and Train , Transport Concepts (Ottawa), October 1994, p.
11 Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992, p. 80.
12
Jan Jansson, "Accident Externality Charges," Journal ofTransport Economics and Policy, January
1994, p. 31-42.
13 Robert Davis, Death on the Streets, Leading Edge (North Yorkshire), 1992, p. 23 .
Page 3.3-4
Transportation Cost Analysis
risk imposed by automobile drivers on "vulnerable" road users such as pedestrians and
cyclists. Davis argues that the true costs of accidents is understated by official analysis,
both because many pedestrian and bicycle accidents are not captured in police statistics,
and because a major portion of this cost is the loss of mobility and security to non-drivers.
Bicycle and Pedestrian Accident Risk
Studies indicate that walking and bicycling incur higher injury per mile than driving,
although the exact value is difficult to determine with because total pedestrian and bicycle
mileage is not measured, and because many such injuries are unreported. 14 Many ofthese
accidents result from rider careless; the accident risk for a responsible (trained, sober, and
wearing a helmet) adult bicyclist is significantly lower.15 Since bicyclists tend to travel
shorter distances than drivers, 16 the relative accident risk per trip is lower than per mile.
If accident risk is defined in terms of total health risk, the aerobic benefits of walking and
bicycling compensate for accidents.17 One estimate concludes that the aerobic exercise of
bicycling outweigh accident risk by 20 to 1 in average life expectancy.18
Estimates:
• Apogee Research estimated accident costs in Boston, MA and Portland, ME for
several modes. Costs were allocated to users, government and society. Totals are
shown in Table 3.3-2.
Table 3.3-2
Total Accident Costs in Two Cities (¢ per passenger mile) 19
14 Charles Komanoff and
Cora Roelofs, The Environmental Benefits of Bicycling and Walking, FHWA
National Bicycling and Walking Study Case Study #15 (Washingon DC), January 1993.
15 Todd Litman, Bicy cling and TDM, Transportation Research Board General Meeting, Jan. 1994.
16 For example, drivers frequrently travel several miles to regional shopping centers while pedestrians and
bicyclists use local shops and services.
17 Benefits ofBicycling and Walking to Health, National Bicycling and Walking Study #14, USDOT,
FHWA (Washington DC), 1992.
18 Dr. Mayer Hillman, "Reconciling Transport and Environmental Policy," Public A dministration, Vol.
70, Summer 1992, pp. 225-234.
19 Apogee Research, The Costs of Transportation: Final Report, Conservation Law Foundation (Boston),
1994, p. 112-118.
Page 3.3-5
Transportation Cost Analysis
Boston
High
Medium
Low
Portland
High
Medium
Low
Expressway
1.2
1.2
1.2
NonExpwy
6.3
6.3
6.3
2.0
2.0
2.0
5.0
5.0
5.0
Comm.
Rail
2.6
2.6
2.6
Rail
Transit
1.9
1.9
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Bus
1.8
1.8
1.8
Bicycle
3.2
3.2
3.2
Walk
11.6
11.6
3.7
3.2
3.2
3.2
1.4
1.4
1.4
1.4
1.4
1.4
• The Californian Energy Commission estimates automobile accident costs at $0. 118 per
VMT, and bus accident costs at $0.26 per VMT ($0.014 per passenger mile based on
18.5 average passengers). Of the bus accident costs, 22% is estimated to be internal.20
(This may somewhat overstate true bus accident risk, since transit systems are
reported to be vulnerable to false but successful liability claims.)
• Chirinko and Harper find that automobile safety features (seat belts, air bags, etc.)
encourage risky behavior which offsets much of the safety gain and significantly
increases pedestrian and bicycle accidents.21 They predict that over 1/3 of the 33%
potential fatality reductions due to mandated air bags would be offset by increased
driver risk, and pedestrian and bicycle fatalities would increase by 2%.
• Jan Jansson calculates the marginal external cost of driving to unprotected road users
(pedestrians, bicyclists, and motorcyclists) with three risk levels and three estimates of
the marginal increase of accident risk to these modes with respect to increased motor
vehicle travel, as shown in Table 3.3-3 .
-
- -
-
- -
-
- -- -- -- - --.-- - - -- --- - --- --- - - - - - -- - ---- Unprotected Road User Accidents Per lOOM Motor Vehicle Km
-- - ------- -
- - - - - - --
Accident!VMT Ratio
10
20
30
1/3
2/3
1/1
$0.02
$0.04
$0.06
$0.04
$0.08
$0.12
$0.06
$0.12
$0.18
This table shows the costs of unprotected road user accidents from automobiles with
three accident rates, and three ratios of accident rate to motor vehicle travel volumes.
• Per Kageson estimates that European fuel taxes would need to increase an average of
0.23 ECU per litre (about $0.05 per mile) to internalize all accident costs. 23
20
1993-1994 California Transportation Energy Analysis Report, Technical Appendices, California
Energy Commission (Sacramento), Feb. 1994, p. 3D-7.
21 Robert Chirinko and Edward Harper, Jr., "Buckle Up or Slow Down? New Estimates of Offsetting
Behavior and their Implications for Automobile Safety Regulation," Journal of Policy Analysis and
Management, Vol. 12, No. 2, 1993, pp. 270-296.
22 Jan Jansson, "Accident Externality Charges," Journal ofTransport Eco. and Policy, Jan. 1994, p. 40.
The overall U.S. pedestrian fatality rate per lOOM Motor Vehicle Km is approximately 0.15, but is higher
in cities, based on NHTSA, Traffic Safety Facts, 1992.
Page 3.3-6
Transportation Cost Analysis
• Brian Ketcham estimates traffic accident costs at $0.043 per vehicle mile, citing the
Urban Institute's estimate.24
• Mac Elliott estimates the overall average fatality risk for bicyclists to be 4 to 4.5 times
higher per mile than automobile occupants, but many of these accidents involve
children or careless bicyclists.2s
• Dr. Mayer Hillman estimates that the fatality rate for walking is 18 times higher than
for car travel,26
• Ted Miller estimates the total value of 14.8 million motor vehicle accidents in 1988 at
$358 billion (1988 dollars), a major component ofwhich is pain, suffering, and lost
quality of life.27 An approximately $2 million value of human life is used, which is
considered low. Table 3.3-4 shows his estimates of total accident costs.
Table 3.3-4
FHWA Reoort Total Accident Cost-Estimates
bv Mode
--~~
·-
--
~- -.-----
Vehicle Type
-
-
-
--
----
1994 $/VMT
0.14
2.57
0.29
0.19
0.13
0.23
AutomobileNan
Motorcycle
Bus
Light Truck
Med/Hvy Truck
Combination Truck
• Miller recently provided updated vehicle accident cost estimates shown in Table 3.3-5.
Table 3.3-5
Miller's Estimate of Accident Costs28
Mode
Bus
Commercial Air
Car
Car, Drunk Driver
Car, Sober Driver
Motorcycle
23 Per Kageson,
$ Per Vehicle Mile
$0.32
$0.28
$0.12
$5.50
$0.06
$1.50
Getting the Prices Right, European Fed. for Transport & Env., (Bruxelles), 1993, p. 130.
Konheim & Ketcham Inc.
24 Brian Ketcham, Making Transportation Choices Based on Real Costs,
(New York), Oct. 1991, p. 10
Mac Elliot, chair of the Human Powered Transport Subcommittee of the American Society of Civil
Engineers, Committee correspondence, June, 1993.
26 Dr. Mayer Hillman, Cycling: Toward Health and Safety, British Medical Association/Oxford Press
(NewYork), 1992.
27 Ted Miller, The Costs of Highway Crashes, FHWA (Washington DC), pub. No. FHWA-RD-055, 1991.
28 Presented at FHW A Colloquium on Social Costs of Transportation, 12 Dec. 1994, Washington DC.
Also see Miller, et al., "Railroad Injury: Causes, Costs, and Comparisons with Other Transport Modes,"
Journal of Safety Research, Vo. 25, No. 4, 1994, pp. 183-195.
25
Page 3.3-7
Transportation Cost Analysis
• Peter Miller and John Moffet estimate annual external auto accident costs at $0.043
per urban mile and $0.03 per rural mile, and $0.007 per passenger mile for buses.29
• A 1992 USDOT National Highway Traffic Safety Administration study places 1990
accident costs at $137.5 billion (averaging about $0.065 per vehicle mile) because
pain, suffering, and loss of quality of life were not included.30
• ANational Research Council study concludes that small car occupants have a greater
fatality risk than large car occupants in some types of accidents, but this difference has
decreased due to improved designs and safety equipment and is partly compensated by
reduced accident risk to other road users.31
• Rene Neuenschwander and Felix Walter estimate external accident costs in
Switzerland to average 0.024 ECU per passenger kilometer (about $0.03 per rnile).32
• The Office of Technology Assessment study places annual external market accident
costs to individuals at $3 3 to $3 5 billion, plus about $4 billion per year in government
costs.33 Pain, suffering and lost quality of life inflicted on others (non-market external
costs) are estimated to be worth $132 to $139 billion per year, for a total external
accident cost average of$0.076 per vehicle mile.
• Emile Quinet summarizes accident costs by mode as shown in Table 3.3-6.34 Based on
this analysis he concludes that accident costs per passenger mile is about 10 times
higher for cars than for buses, and that cars accident costs in passenger miles is
virtually the same as accident costs for trucks in tonne miles .
1 ame .1•.1-o
Study
Planco, 1990
Tefra, 1985
Tefra, 1985
EcoPian, 1991
Hansson, 1987
HansS()!J._,_J2_87
1\.CCIOent LOStS D .travet!Y.looe \ u.;:,. oouarsJ
Location
FRG
France
Belgium
Switzerland
Sweden, Urban
Sweden, Rural
Passen2ers (passen2er-km)
Car
Bus
Rail
.020
.004
0.003
0 .030
0 .050
0 .088
0.007
0.013
0.001
0.004
0.001
Frei2ht (tonne-km)
Water
Road
Rail
0.000
0.012
0.008
0 .007
0.00
0.003
0.001
0 .070
0.000
0 .013
----
29 Peter Miller and John Moffet, The Price ofMobility, NRDC, Oct. 1993 , p. 32.
30 Economic Cost ofMotor Vehicle Crashes 1990, National Highway Traffic Safety Admin., 1992.
31 Automotive Fuel Economy: How Far Should We Go, National Academy Press (Washington DC), 1992.
32 Felix Walter, Social Costs of Swiss Transport A ccidents, ECOPLAN (Bern, Switzerland), p. 2.
33 Saving Energy in U.S. Transportation , U.S . Office ofTechno1ogy Assessment, 1994, p. 106- 108.
34 Emile Quinet, "The Social Costs of Transport: Evaluation and Links With Internalisatiion Policies," in
Jnternalising the Social Costs ofTransport, OECD (Paris), 1994, p.38.
Page 3.3-8
Transportation Cost Analysis
• Kenneth Small estimates that total accident costs average $0.179 per mile, $0.09 of
which are external costs. He estimates that external accident costs are 35% higher for
trucks and 175% higher for buses.35
• Daniel Shefer concludes that traffic fatality risk declines with increased congestion due
to reduced speeds, although he provides no specific function of this relationship. 36
• D. Teufel estimates the following accident costs for different modes:
Table 3.3-7
Mode
H
fLost H
Lift
Car
Bus
11.5
1
1.000 p
Rail
0.4
37
Kil
Pedestrian
Air
0.01
1.4
• Transport Concepts cites European estimates that road travel is 8 times more
dangerous for fatalities and 100 times more dangerous for injuries than rail. 38 They
estimate truck accident risk to be six times greater than for train per unit of fright
travel. A significant portion of rail fatalities result from accidents with motor vehicles
at crossings, so cost estimates are sensitive to how the responsibility for these
accidents is allocated. Based on various assumptions this study estimates freight
accident costs at approximately $0.50 per ton mile for truck and $0.076 per ton mile
for rail ($0.40 and $0.06 Canadian per tonne kilometer respectively).
Variability: Accident rates vary significantly with driver behavior and vehicle type.
Although accident rates are higher in urban areas due to increased vehicle interactions,
high-speed rural accidents are more dangerous. The FHWA study figures indicate that
urban and rural accident costs per mile are approximately equal. David Greene and K. G.
Duleep conclude that the additional injury and death risk from smaller, fuel efficient cars is
relatively minor, especially when reduced risk to other road users are considered.39
35 Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992, p. 80-81.
36 Daniel Shefer, "Congestion, Air Pollution, and Road Fatalities in Urban Areas," Accident Analysis and
Prevention, Vol. 26, No. 4, 1994, pp. 501-509.
37 D. Teufel, Die Zuykunft des Autoverkehrs (The Future of Motorized Transport), Umwelt- und Prognose
Institut, Heidelberg, 1989, in Transportation, The Environment and Sustainable Development, p. 184.
38 External Costs of Truck and Train , Transport Concepts (Ottawa), October 1994.
39 David Greene and K. G. Duleep, "Costs and Benefits of Automotive Fuel Economy Improvement",
Transportation Research A, Vol. 27 A, No. 3, p. 221 .
Page 3.3-9
Transportation Cost Analysis
Conclusions: Accidents impose significant costs on individual road users and society. The
Urban Institute/FHWA cost estimates are used as a starting point for calculating costs per
vehicle mile because they are considered accurate and comprehensive. To avoid double
counting insurance payments covered in Chapter 3.1, insurance disbursements are first
subtracted from these estimates. Separate estimates are made for internal and external
costs, based on Ted Miller's additional calculations. Internal accident costs are assigned
per vehicle occupant, while external accident costs are assigned per vehicle. This is
necessary because internal accident risk is relative to VMT and vehicle occupancy, while
external risk is proportional simply to VMT. For example, a vehicle carrying only the
driver is considered to impose only 10% of the internal accident risk as a vehicle carrying
ten people, but the external accident risk is considered the same for both. Although rural
driving has fewer accidents per mile, they tend to be more severe due to higher speeds, so
rural and urban driving accident costs are considered equal.
Internal Accident Costs: Internal accident costs for average automobile and van
occupants, including rideshare passengers, are estimated at $0.05 per passenger mile,
calculated as the FHWA's average accident cost estimate of$0.14NMT, times 75%
internal costs, minus insurance disbursements of $0.031 , divided by 1.5 average
°
passengers.4 Fuel efficient and electric cars are estimated here to impose 10% higher
internal accident costs than an average car due to their smaller size. The California Energy
Commission's accident cost estimate of$0.014 per passenger mile is used for buses and
trolleys, 22% of which is internal, for a cost of$0.003 per PMT.
Motorcycle accident costs estimated at $1.50 to $2.57 per mile reflect this mode's high
accident and injury rates. This extremely high value results in part because motorcyclists
tend to be risk taking young men who have an accident rate 3 times higher than average
40 Alan Pisarski,
Travel Behavior Issues in the 90's, FHWA, (Washington DC), 1992, p. 52.
Page 3.3-10
Transportation Cost Analysis
when driving any type of vehicle, so a lower cost estimate can be used to represent the
accident costs normalized for the average rider. 41 Also, the motorcycle fatality rate per
VMT has declined since the FHWA study was produced. For these reasons, a
demographically average driver who currently rides a motorcycle is assumed here to have
an accident cost 1/Sth of the FHWA's study's estimate (about 1/3 of Ted Miller's more
recent estimate), equal to $0.514. Even with this modification accident risk dominates
motorcycle costs. Internal motorcycle accident costs are estimated to represent 85% of
this cost (a higher ratio of internal costs since motorcycles are less likely to injure other
road users) minus $0.07 for insurance disbursements (twice that of cars) resulting in
$0.437 per mile. Bicycles and walkers are estimated to incur internal accident risk equal to
that of automobile occupants. Telecommuting is not considered to incur any accident risk.
-
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Bus/Trolley
Motorcycle
Bicycle
Walk
Telecommute
Urban Peak
0.05
0.055
0.055
0.05
0.05
0.003
0.003
0.437
0.05
0.05
0.00
-
--
-
~
~
---- -
Urban Off-Peak
0.05
0.055
0.055
0.05
0.05
0.003
0.003
0.437
0.05
0.05
0.00
-
Rural
0.05
0.055
0.055
0.05
0.05
0.003
0.003
0.437
0.05
0.05
0.00
Avera2e
0.05
0.055
0.055
0.05
0.05
0.003
0.003
0.437
0.05
0.05
0.00
External Accident Risk: Based on the FHWA report, the average external cost of driving
average cars and vans is estimated by taking the 25% external portion of $0.14 per vehicle
mile, for an average of$0.035 per mile. Note this is per vehicle rather than per passenger.
Small, fuel efficient and electric cars incur a slightly lower external risk, estimated here at
5% less than a standard car. Rideshare passengers incur no additional external cost. The
41 Traffic Safety Facts 1992, National
Highway Traffic Safety Administration (Washington DC), 1993.
Page 3.3-11
Transportation Cost A nalysis
California Energy Commission accident cost estimate that external bus accident costs
represent 78% of$0.26 per VMT is used for buses and trolleys. Motorcycles are
estimated to have external accident costs of$0.077 per mile, representing 15% of$0.514.
Since the costs of accidents between vehicles and pedestrians or bicycles is normally
allocated to the motor vehicle, pedestrians and bicycles are estimated here to impose only
5% the external accident cost of average automobiles.
-
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Motorcycle
Bicycle
Walk
Telecommute
-- -----
-- -
---- ----
- -- --
-
- - -- ----
--
-
--- -
- -
-
- · -
-
Urban Peak
Urban Off-Peak
Rural
Avera2e
0.035
0.033
0.033
0.035
0.00
0.20
0.20
0.077
0.002
0.002
0.00
0.035
0.033
0.033
0.035
0.00
0.20
0.20
0.077
0.002
0.002
0.00
0.035
0.033
0.033
0.035
0.00
0.20
0.20
0.077
0.002
0.002
0.00
0.035
0.033
0.033
0.035
0.00
0.20
0.20
0.077
0.002
0.002
0.00
Automobile Cost Range: Accident cost estimates range from $0.05 to $0.20 per
automobile mile, the maximum based on the Urban Institute estimate with a higher value
ofhuman life. 15% to 50% of these costs are considered external based on studies cited.
Internal
External
Minimum
$0.03
$0.01
Maximum
$0.17
$0.10
Discussion
Despite the fact that accident costs are greater than other variable costs such as fuel or
parking, they seldom seem to discourage driving. It is therefore interesting to consider
how this cost is perceived by users. Since approximately 1/3 of accident costs are caused
by drunk drivers, the overall average overstates the risk for a sober driver (drunks are
Page 3.3 -12
Transportation Cost Analysis
irrational, so the fact that they take significant risks is not surprising). Accident costs for
sober drivers are approximately $0. 10 per vehicle mile. About $0.031 of this is
compensated through insurance disbursements, a fixed cost since insurance companies
provide little savings for reduced driving.42 Fixed costs give users an incentive to
maximize driving to reduce average costs. About $0.023 of accident costs are external
($0.035 per mile times 2/3 for sober drivers), so the actual internal uncompensated
accident cost for a sober driver is $0.046 per mile, only 1/3 of total accident costs.
Since accident occur infrequently, individual drivers are likely to ignore or understate this
cost. Surveys find that most drivers consider their driving skill above average, so typical
drivers underestimate their actual accident costs. It may therefore be common for drivers
to assume that their insurance payments cover the full risk they impose. Drivers may also
tend to deny the possibility that they may experience uncompensated costs or impose such
costs on others. Thus, it is not surprising that through a combination of fixed costs,
external costs, optimism, and denial drivers are not usually influenced by the risk of
accidents they impose on themselves and others, despite the high total cost it imposes.
42
Compensated accident costs are not included in this chapter's Best Guess estimates to avoid double
counting the insurance payments in Chapter 3.1.
Page 3.3-13
Transportation Cost Analysis
3.4 Parking
Definition: Automobile parking costs.
Description: Automobile parking costs include capital, operating and opportunity costs of
off-street employee, commercial, municipal and residential parking.1 Cost estimates should
be based on the full opportunity costs of the parking lot real estate, which is often underassessed for tax purposes. Tax exemptions for employee parking are also considered a
parking cost in terms of public revenue foregone.
Discussion: Subsidized parking is a significant cost, and a major incentive for driving.
According to the 1990 Nationwide Personal Transportation Survey (NPTS), motorists
reported receiving free parking for 99% of all automobile trips.2 Approximately 95% of all
workers drive to work, only 5% of whom pay their full parking costs, and about 9% pay
at a subsidized rate. The majority of these parking subsidies are income tax exempt. As
Donald Shoup points out, the average value of employee parking subsidies exceeds the
average value of the fuel spent on a commute, yet nobody expects employers to provide
free fuel, and if provided it would be taxed. Parking subsidies are estimated to increase
driving by 20%-40% over levels that would occur if no subsidy were provided, 3 and are
inequitable since transit riders, bicyclists, and walkers receive no comparable benefit.
1 To avoid
double counting costs in chapters 3.1 and 3.6, on-street parking and user paid non-residential
are not included as a cost in this chapter.
2 1990
NPTS, Summary ofTrave/ Trends, USDOT (Washington DC) 1992.
Shoup, "Cashing Out Employer-Paid Parking," in Curbing Gridlock, 1994, pp. 152-199.
3 Donald
Page 3.4-1
Transportation Cost Analysis
Figure 3.4-1 Employee Parking Subsidy Patterns4
Subsidized
Pay Full Cost
Don't Drive
(Including car
pooling)
Free On-street
Parking
Free Employer
Parking
Most commuters who drive enjoy free or underpriced parking.
Typical parking stalls are 8-10 feet wide and 18 to 20 feet deep, totaling 144 to 200
square feet. When access lanes are included, the total parking lot area per vehicle is
approximately double this amount (276 to 340 square feet), allowing a total of about 125
spaces per acre.5 The Institute of Transportation Engineers estimates surface parking lots
cost an average of $1,600 per space in 1994 dollars, in addition to land costs, and parking
structures average about $9,000, in addition to land costs. 6 Wegmann places the average
cost for a new employee parking space at about $5,300 in 1994 dollars. 7 In addition to
these market costs, parking facilities also impose non-market costs, including aesthetic
degradation, stormwater concentration, and increased automobile dependency.8
4 Miller and Moffet, The Price ofMobility,
5 James Hunnicutt, "Parking, Loading, and
National Resource Defense Council, Oct. 1993, p.24
Terminal Facilities," in Transportation and Traffic
Engineering Handbook, Institute of Transportation Engineering/Prentice Hall, 1982, p. 651.
6 James Hunnicutt, "Parking, Loading, and Terminal Facilities," in Transportation and Traffic
Engineering Handbook, Institute of Transportation Engineering/Prentice Hall, 1982, p. 651.
7 Frederick Wegmann, Cost Effectiveness ofPrivate Employer Ridesharing Programs: An Employer's
Assessment, Transportation Center, 1985.
8 PeterNewman and JeffKenworthy, Cities and Automobile Dependency, Gower, 1989, pp. 122-126.
Their research indicates that higher parking provisions as a percentage of urban area increase automobile
dependency, promote sprawl, and reduce urban attractiveness.
Page 3.4-2
Transportation Cost Analysis
Parking subsidies paid by employers and commercial businesses raise the cost of
employment and consumer goods. Since the area devoted to parking typically equals the
floor space of a building, current parking requirements are approximately equal to a 100%
tax on building land. Some communities also directly subsidize municipal parking lots and
structures. Of 1,284 cities that responded to a 1972 survey, 51% own off-street municipal
parking.9 Parking is required by most zoning laws and investment agencies, making it a
fixed cost imposed on users and non-users alike. Employee parking is exempt from U.S .
income tax, a benefit to drivers that can be worth up to $1,800 per year. These foregone
taxes can also be considered a parking subsidy.
The number of parking spaces per automobile varies depending on land use patterns, with
more parking and a higher percentage of free parking available in suburban and rural areas
than in urban conditions.10 Most commercial zoning codes require a generous amount of
parking (Table 3.4-1), which is usually provided for free, resulting in a cross subsidy from
non-drivers to drivers. This may be explained in part by a "prisoner's dilemma," whereby
any individual store that tries to charge for parking will lose more than it can make up in
customers who use other modes from slightly lower overall prices.11 Michael Cameron
emphasizes the need to reduce subsidies for non-employee parking:
"Studies show that when shoppers pay for parking they are more likely to shop in areas
where all of their stops are within walking distance of each other, and that they are less
likely to drive from place to place. This has important implications for both congestion
and air quality -- especially given the impact that cold-starts have on emissions. "12
Estimating commercial parking subsidies is difficult because each car tends to use many
such parking spaces for short periods. Lee applied parking generation rates to gross
9 Hamburger, Kell
and Perkins, Fundamentals of Traffic Engineering, Institute of Transportation Studies,
UCB (Berkeley), 1992, p. 27-3 .
10
Herbert Levinson, "Urban Traffic Characteristics," Transportation and Traffic Engineering Handbook,
Institute of Transportation Engineers/Prentice Hall (Englewood Cliffs), 1982, p. 298, 300.
11 Doug Lee, Full Cost Pricing of Highways, Transport Research Center (Cambridge), 1993, p. 28.
12 Michael Cameron, Transportation Efficiency, Environmental Defense Council (Oakland), 1991, p.42.
Page 3.4-3
Transportation Cost Analysis
leasable square footage of shopping centers, and scaling that to all retail parking using
dollar volume of sales to estimate that 25 million parking spaces are provided for retail
customers, about one for every 7.5 registered automobiles.13 Additional non-retail
activities, such as schools, daycare centers, medical and other professional service
buildings, recreational centers, and municipal facilities such as courts and post offices
would increase this estimate of commercial parking.
Table 3.4-1
Tvoical Z
R,
for Off-S
Building Type
Office Buildings
Retail
Single Family Dwellings
Apartments
Hotels/Motels
Public Services (museums, libraries)
Hos_pitals
Theaters
Park.inn14
Unit
per 1000 sq. ft.
per 1000 sq. ft.
House
Unit
Unit
per 1000 sq. ft.
per 1000 sq. ft.
per seat
··~
Spaces
2.5
5
1-2
1
1
3.3
10
0.25
I
Most local zoning laws require property owners to provide significant amounts of
parking, which is usually provided free, effectively subsidizing driving.
In addition to direct costs, automobile parking requirements impose indirect costs in terms
of environmental impacts, urban sprawl, automobile dependency, and increasing land
costs. Since parking lots are typically larger than the buildings they are intended to serve,
parking costs are a major reason that many businesses choose low price land at the urban
edge rather than a centralized location. In commercial and industrial areas large parking
lots separate buildings, reducing the viability of walking, and provisions for parking have
been statistically correlated with increased automobile dependency.15
Increased parking requirements are unfair and regressive, since lower income households
on average own fewer automobiles and pay a much larger portion of total household
13 Full Cost Pricing ofHighways, National
Transportation Research Center (Cambridge), 1995, p.18.
and Perkins, Fundamentals of Traffic Engineering, Institute of Transportation
Studies, UCB (Berkeley), 1992, p. 27-2.
15 Joel Garreau, Edge City, Doubleday, 1991 ; Peter Newman and Jeff Kenworthy, Cities and Automobile
Dependency, Gower Press (Aldershot), 1989.
14 Homburger, Kell
Page 3.4-4
Transportation Cost Analysis
expenditures per parking space than wealthier households, and because parking
requirements reduce the availability oflower priced housing.16 One study found that
requiring just one parking space per housing unit (many communities now require 2 or
more) increased construction costs 18%, significantly reduced the land available for
housing, and gave developers an incentive to develop fewer, larger and more expensive
units.17 As Donald Shoup concludes, "Form no longer follows function, fashion, or even
finance; instead, form follows parking requirements. "18
Estimates: (Note: Although many ofthese estimates are presented in per mile units, this
cost is correctly measured per trip. Parking costs are not affected by trip length.)
• Apogee Research estimates these parking costs in two cities:
~
-
--
-
-
Automobile
User Cost
External Cost
Bicvcle
External Cost
User Cost
Boston, MA
High
Medium
Low
Portland, ME
High
Medium
Low
15.7
12.6
6.9
9.3
5.5
3.7
0.6
0.5
0.3
2.1
0.7
0.6
5.8
1.5
1.2
10.0
2.8
2.5
0.2
0.1
<0.1
0.8
0.5
0.4
• Michael Cameron estimates parking subsidies at $3 .00 per day for typical commuters
in Southern California, and about one cent per minute for commercial parking.20
• Patrick Hare estimates that parking costs account for about 10% of a typical
apartment rent of$500 per month, or about $600 per year. 21
16 Todd Litman, Parking Requirment Impacts on Housing Affordabi/ity, Victoria Transport Policy
Institute (Victoria), 1995.
17 Wallace Smith, The Low-Rise Speculative Apartment, Research Report 25, University of California
Center for Real Estate and Urban Economics (Berkeley), 1964, cited in Shoup, 1994.
18 Donald Shoup, "Cashing Out Employer-Paid Parking," forthcoming, Journal of the American Planning
Association, June 1994, p. 18.
19 The Costs of Transportation: Final Report, Conservation Law Foundation (Boston), 1994, p. 99-111.
20 Michael Cameron, Transportation Efficiency, Environmental Defense Council (Oakland), 1991, p 41.
Page 3.4-5
Transportation Cost Analysis
• Douglass Lee updated and expanded estimates by Don Pickrell22 to calculate 80.9
million total free employee parking spaces worth $53 billion annually (Table 3.4-3 and
Figure 3.4-2), and retail customer parking subsidies to be $18 billion. Together,
employee and retail parking subsidies total $71 billion, or $0.03 per vehicle mile.23
This total is probably low because it is based on 1980 census commute mode values
(solo automobile commuting has increased), and because non-retail commercial client
parking (professional services, medical services, etc.) are not included.
Table 3.4-3
E ··
Location
Urban
Population
Units
Million
Not Reported
Rural
Suburbs
<50,000
Under 1
1 to 3
Over 3
Under 1
1 to 3
Over3
Under 1
1 to 3
Over3
City
CBD
Total/Average
Parkin2 Subsid.
fE
Commuters Car Drivers
Million
8.3
19.6
13.4
ll .5
7.7
15.7
8.8
6.2
2.3
1.5
1.5
96.7
Avg. Daily
Cost
Park Free
ParkFree
Dollars
1.00
1.00
2.00
2.00
3.00
4.00
5.00
6.00
8.00
10.00
12.00
3.07
Percent
100
100
100
100
100
80
80
70
75
75
65
93
Million
4.1
19.6
11.4
9.2
5.2
9.3
5.0
2.6
1.1
0.7
0.5
68.7
Percent
50
100
85
80
67
74
72
60
62
58
49
77
Figure 3.4-2 Average External Parking Cost Per Automobile Commuter
:
~co
$8.00
$7.00
$8.00
c
:i ..
:;; ~ $5.00
~ E
..~ 8E
~~
~<
~
&
!!
~
--Suburbs
·--- - Ex urban (Rural)
$4.00
$3.00
$2.00
$1 .00
$0.00
<1 million
1-3 million
>3 million
Urban Area Population
21 Patrick Hare, et al, Trip Reduction and Affordable Housing, Transportation Research Board, 1991
22 Don Pickrell, "Eliminating Employer-Subsidized Parking" in Climate Change Mitigation:
Transportation Options, National Transportation Research Center (Cambridge), for USEPA, 1993.
23 Full Cost Pricing ofHighways, National Transportation Research Center (Cambridge), Jan. 1995.
Page 3.4-6
I
I
I
I
Transportation Cost Analysis
• James MacKenzie et al. estimate that 86 million autos receive an average $1,000 per
year in free parking, a subsidy worth $86 billion annually, or $0.039 per mile. 24
• Peter Miller and John Moffet estimate external parking costs to range from $.008 to
$.032 per vehicle mile.25
• Terry Moore and Paul Thorsnes estimate the total annual subsidy of non-residential
off-street parking totals $200 billion, averaging 9.5¢ per mile, based on the assumption
that commercial parking costs approximately equal those of commuter parking. 26
• The Office of Technology Assessment study estimates the annual value of free offstreet employee parking ranges from $37 to $66 billion, and parking for non-work
trips is $64 to $132 billion. They recommend using a range of$43 to $185 billion for
total external parking costs, averaging $0.02 to $0.08 per vehicle mile.27
• Donald Shoup and Richard Willson calculate the average employee parking subsidy in
central Los Angeles in 1986 to be $3 .87, worth $5 .04 in 1993 .28 Willson estimates that
building owners in the Los Angeles region would have to charge from $31 to $134
monthly per parking space, averaging $109 per occupied space.29
• Transport 2021 estimates residential parking stall costs average $746 Canadian per
house and $743 per apartment. Total parking costs average $0.037 total Canadian per
km (about $0.046 U.S . per mile).30
Variability: Parking costs and the portion of costs vary considerably depending on
location and the type of driving. Costs per space are highest in large urban areas,
especially in areas which require multi-story parking facilities . For example, the estimated
cost per parking space is about 4 times higher in the central business district of a large city
than the overall average. However, taking into account the lower portion of free parking
in such areas, the average parking subsidy per automobile commuter is only about half
24 MacKenzie, Dower & Chen, The Going Rate, World Resources Inst. (Washington DC), 1992, p. 10.
25 Miller and Moffet, The Price ofMobility, National Resources Defense Council, Oct. 1993, p.24
26
Teny Moore and Paul Thorsnes, The Transportation/Land Use Connection, American Planning
Association, Report #448/449 (Washington DC), 1994, p.49.
27 Saving Energy in U. S. Transportation , U.S. Office of Technology Assessment, 1994, p. 106.
28 Donald Shoup and Richard Willson, Commuting, Congestion and Parking: The Employer-Paid
Parking Connection , Draft Paper, 1992, cited in Apogee Research, 1993.
29 Richard Willson, Suburban Parking Economics and Policy: Case Studies of Office Worksites in
Southern California, FHWA (Washington DC), Sept. 1992.
30 Transport 2021, Costs of Transp ortation People in the British Columbia Lower Mainland, Greater
Vancouver Regional District, (Vancouver), 1993, pp. 13-16.
Page 3.4-7
Transportation Cost Analysis
again higher than the overall average. Parking costs tend to be relatively high per commute
trip since employees typically need a space for 8 or more hours. The total value of parking
for other non-work trips probably equals or exceeds that of employee parking. The cost
per trip or per mile is lower however, since it is averaged over more travel.
Conclusions: Parking is a substantial cost of driving, most of which is external. Although
parking imposes both market and non-market costs, only market costs will be considered
here because of a lack of a lack of data, and because many non-market costs are already
captured in chapters 3. 10 (Equity and Option Value), 3. 14 (Land Use Impacts) and 3 .15
(Water Pollution and Hydrologic Impacts).
Internal Parking Costs: To avoid double counting user parking fees that are included in
Chapter 3.1, only residential parking costs are considered here. Patrick Hare's estimate
that an automobile parking space costs approximately $600 per year, is used to estimate
an average of$0.05 per mile for a vehicle driven 12,500 miles per year. Although this
estimate is for multi-family residences, it is considered a reasonable conservative estimate
for the amortized cost of single family parking costs. Some residents park their cars on the
street, but this seems to be balanced by others who have more off-street parking spaces
than cars, so one off-street space is assumed to exist for each registered car (about 190
million in the U .S.). Rural parking space costs are estimated at half of urban due to lower
land values. As described below, small cars, motorcycles, and bicycles are estimated to be
5%, 25%, and 95% cheaper to park than an average automobile. Rideshare passengers,
buses, trolleys, walking and telecommuting incur no user parking costs.
Page 3.4-8
Transportation Cost Analysis
-
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
-
Urban Peak
Urban Off-Peak
Rural
Avera2e
0.050
0.045
0.045
0.050
0.00
0.00
0.00
0.040
0.003
0.00
0.00
0.050
0.045
0.045
0.050
0.00
0 .00
0 .00
0 .040
0.003
0.00
0.00
0.025
0.023
0 .023
0.025
0.00
0.00
0.00
0.020
0.001
0.00
0.00
0.042
0.038
0.038
0.042
0.00
0.00
0.00
0.033
0.002
0.00
0.00
I
External Parking Costs: Several estimates place average off-street parking costs around
$750 per year or $3 .00 per day per space, 31 and place total U.S. employee parking
subsidies at between $50 and $70 billion per year. A value of $55 billion is used. Dividing
that amount by 460 billion peak period32 miles gives an average employee parking subsidy
of$0.12 per commute mile. An alternative approach is to divide the $3.00 average parking
space cost by 22 average commute miles and subtract 8% for commuter paid parking,
which gives an estimated average external commute parking costs of$0.125 per commute
mile. Based on these estimates, $0.12 per commute mile is used for Urban Peak driving.
Commercial parking subsidies are estimated using the Office of Technology Assessment's
figures which indicate total average external parking costs (work plus non-work) range of
$44 to $185 billion, with a $115 billion mid-point. Subtracting the $55 billion estimated
for work parking from this figure leaves $60 billion. Divided by 1,840 Urban Off-Peak and
Rural trips, this averages about $0.03 per vehicle mile. An estimate of$0.04 is used for
Urban Off-Peak driving and $0.02 for Rural driving, to represent differences in land value.
31 This $750 annual cost equals about $7,500 in capital cost, $1,500 of which is surface preparation
leaving $6,000 for land. Assuming 125 parking spaces per acre, this implies an average land value of
$750,000 per acre for commercial real estate used for parking, not an unreasonable figure.
32 Urban Peak travel is used to represent commuting in this exercise.
Page 3.4-9
Transportation Cost Analysis
Small gasoline and electric cars can use "Compact Car" spaces, offering an estimated 20%
space savings 25% of the time, for 5% total saving. Ride share passengers, buses and
trolleys incur no incremental parking cost. 33 Motorcycles are estimated to use half-size
parking spaces 50% of the time, for a 25% saving over an automobile, while bicycle
parking costs are estimated at 5% of an automobile, due to minimal space requirements,
and the ability to use otherwise unused space. Walking incurs no parking cost.
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
Urban Peak
Urban Off-Peak
Rural
Avera2e
0 .120
0 .114
0 .114
0.120
0.00
0 .00
0 .00
0 .09
0.006
0 .00
0 .00
0 .040
0.038
0.038
0.040
0.00
0 .00
0.00
0.03
0 .002
0.00
0.020
0.019
0.019
0.020
0.00
0.00
0.00
0.015
0.001
0.00
0.00
0.048
0.046
0.046
0.048
0.00
0.00
0.00
0.036
0.002
0.00
0.00
- --- ---------
_____ O.QQ
- - --- --
Testing these estimates: This cost can be checked by multiplying these costs by mileage:
-
Urban Peak
Urban Off-Peak
Rural
Total
-
VMT (billions)
Parking Cost/Mile
Total Cost (billions)
460
920
920
2,300
0.12
0.04
0.02
$55 .2
$36.8
$18.5
$110
Dividing this total by 2,300 billion total miles, external parking costs average $0.048 per
mile overall. These are comparable to previous estimates.
33 Curb side bus stops and pullouts do use space that might otherwise be available for on-street parking, so
there is a tradeoff between buses and parking. However, this is considered a road cost rather than a
parking cost.
34 Facts and Figures 93, American Automobile Manufacturers Association (Detroit) p. 62, assuming 3%
annual growth in VMT since 1992. Percent Urban Peak is estimated.
Page 3.4-10
Transportation Cost Analysis
Automobile Cost Range: Minimum and maximum estimates are based on cited estimates.
Internal
External
Minimum
$0.03
$0.03
Page 3.4-11
Maximum
$0.08
$0.10
Transportation Cost Analysis
3.5 Congestion
Definition: Incremental costs resulting from interference among road users.
Description: Each additional vehicle on a road can interfere with other road users,
especially when traffic volumes approach a road's capacity. This results in lost time,
increased pollution, increased vehicle operating costs, and driver stress.
Discussion: The capacity of a road depends on various design factors such as lane widths
and intersection configurations. Typical performance values are shown in tables 3.5-1 and
3.5-2 in reference to Level Of Service (LOS), a measure ofroadway congestion. These
tables assume ideal conditions and roads with no or few intersections. Many factors
decrease this optimal performance. Traffic speed and flow on urban streets are determined
primarily by intersection capacity, which is affected by traffic volumes on cross streets and
the need for designated left turn signal phases on medium and high volume roads.
1aote .i.J-1
LOS
A
B
typical Koaaway
Speed Range
(mph)
~
r>eea, .111ow ana vensitv Keianonsmps•
Flow Range
Density Range
(veh./hour/lane)
(veh./mile)
Over 60
57-60
54-57
46-54
30-46
Under 30
c
D
E
F
Under 700
700-1,100
1,100-1,550
1,550-1,850
1,850-2,000
Unstable
Under 12
12-20
20-30
30-42
42-67
67-Maximum
This table shows the speed, flow and density of traffic under each Level of Service (LOS)
rating, a standard measure of traffic congestion.
e)2
Table 3.
oa
4-lane Freeway
2-lane Highway
4-lane Highway
LOSA
LOSB
LOSC
LOSD
LOSE
700
210
720
1,100
375
1,200
1,550
600
1,650
1,850
900
1,940
2,000
1,400
2,200
·-
This table shows maximum traffic volume per lane for various types of roadways.
1 Homburger,
Kell and Perkins, Fundamentals of Traffic Engineering, 13th Edition, Institute of
Transportation Studies, UBC (Berkeley), 1992, p. 4-4.
1 Homburger, Kell and Perkins, Fundamentals of Traffic Engineering, 13th Edition, Institute of
Transportation Studies, UBC (Berkeley), 1992, p. 8-3.
Page 3.5-1
Transportation Cost Analysis
Figures 3.5-1 and 3.5-2 illustrates these relationships. Because faster traffic requires more
separation between vehicles, increased traffic volume reduces a road's carrying capacity
and average speed, so each additional vehicle imposes costs to other road users.3
Asignificant portion of congestion delays are associated with "traffic incidents" that are
randomly distributed by time and location. According to Federal Highway Administration
estimates, incidents (80% disabled vehicles and 10% accidents) account for 60 percent of
delay hours.4 Although these are random events, they only cause significant delays on
roads that are already congested so are considered congestion cost. Under uncongested
conditions an incident causes little or no traffic delay, but a stalled car on the shoulder of a
congested road can cause 100-200 vehicle hours of delay on adjacent lanes.
Figure 3.5-1 Speed-Density
Figure 3.5-2 Speed-Flow Relationship
Relationship
80
80
:2
-so~
.c
60
Q.
Q.
E
E
,-40
140
8.
(J)
Ill
20
UJ 20
o1-~~==~--~~
0
0+------+------~----~
0
1
2
3
400
800
1200
1600
2000
Traffic Flow (veh/hr)
Volume-to-capacity Ratio
Increased traffic reduces traffic speed and flow capacity.
3 A Policy on Geometric Design of Highways and Streets (Green Book),
AASHTO (Washington DC),
1990, pp. 53-97. Timothy Hau's Economic Fundamentals ofRoad Pricing, Working Paper, The World
Bank (Washington DC), 1992 provides an excellent description of these functions.
4 L. Grenzeback & C. Woodle, "The True Costs of Highway Congestion," ITE Journal, Mar. 1992, p.16.
Page 3.5-2
Transportation Cost Analysis
Calculating Congestion Costs
There are various ways to calculate congestion externalities.5 The analytically most correct
but difficult approach is to calculate marginal delay costs to other road users resulting
from an additional vehicle in the traffic stream, taking into account the speed-flow
relationship for each road segment. 6 Another approach is to determine the price drivers
must be charged to reduce demand to roadway design capacity. A third approach is based
on the cost of increasing road capacity to an optimal level. In theory these three methods
should provide converging cost values, assuming that roadway capacity is expanded based
on vehicle delay costs as reflected in vehicle users' willingness to pay, but in practice they
often provide different results.7 A common but crude method for calculating congestion
costs is to sum the additional travel time over free-flowing conditions. 8
Aproblem with modeling congestion costs is the effects of generated traffic, as will be
discussed in Chapter 5. If travel demand were fixed, each additional vehicle would impose
a specific cost and each less vehicle would provide a specific saving. But on many roads
traffic congestion maintains a self-limiting equilibrium.9 As Wardrop observed, "The
°
amount of traffic adjusts itself to a barely tolerable speed "1 Congestion delays cause
drivers to use other routes, travel at other times, shift modes, and avoid some trips.
Uncongested roads attract traffic and encourage more and longer motor vehicle trips than
if the same road is congested. This is called generated traffic.n
/__
Is Mark Miller and Kayin Li, An Investigation of the Costs ofRoadway Traffic Congestion, California
PATH, UCB, Berkeley, 1994. Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992,
PPJS-94; Michael Cameron, Transportation Efficiency, Environmental Defense FUfld (Oakland), 1991,
XZ.
1!0r an overview see Anthony Downs, Stuck in Traffic, Brookings Institute (Washington DC), 1992.
7 Terry Moore and Paul Thorsnes, The Transportation/Land Use Connection: A Framework for Practical
Policy, American Planning Association (Chicago), Report# 448/449, 1993 .
8 This is unrealistic since an economically optimized road system has at least some congestion.
9 Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992, p. 112.
10 "Wardrop's Third Principal," David Holden, Journal of Transport Economics and Policy, 9/89, p. 239.
11 Anthony Downs, "Law of Peak-Hour Expressway Congestion," Traffic Quarterly, Vol. 16 , July 1962.
Page 3.5-3
Transportation Cost Analysis
Generated traffic has three implications for assessing marginal congestion costs. First,
generated travel has relatively low value because these are trips that users don't make
unless traffic conditions are favorable.12 Second, generated traffic reduces the congestion
reliefbenefits of increased road capacity. Third, generated traffic increases total motor
vehicle external costs. Many traffic models and transport investment analyses fail to
incorporate these factors, which overvalues congestion reduction benefits and
underestimates total costs.13
Internal or External Cost?
Whether congestion is an internal or external cost depends on the perspective. Since
congestion is borne primarily by the same people who cause it (road users), some analysts
consider it internal, 14 but as discussed in section 1.4, from an economic efficiency
perspective it is external because users do not bear costs proportional to what they
impose, so there is no incentive toward optimum consumption. As Franzi Poldy states,
"While it is true that road users bear congestion costs collectively, they make their
decisions to travel individually. For each individual, a decision to travel requires only
that the benefits exceed the delay (and other) costs that each traveller would expect to
face on the congested road network. .. By deciding to join the congested traffic flow, the
marginal traveller adds to the congestion, and causes a small increase in the delay
experienced by each of the other users. The sum (over all road users) of these
additional delays can be very much greater than the average delay (experienced by
each individual) which formed the basis of the decision to travel. It is because cost
bearing and decision making are separated that these costs are appropriately
considered external. "15
Traffic congestion imposes inequitable costs on non-drivers, another reason to treat it as
an externality. Bicyclists, buses, and automobiles all contribute to traffic congestion, but at
12 Generated
traffic benefits should be assessed using the "Rule-of-Half' as described in the Manual on
User Benefit Analysis of Highway and Bus Transit Improvements (the Red Book), AASHTO, 1977, p.26
13 See further discussion see Chapter 5 of this report.
14 Mark Hanson, "Automobile Subsidies and Land Use," APA Journal, Winter 1992, pp. 60, 68; Per
KAgeson, Getting the Prices Right, European Fed. for Transport and Environment (Broxelles), 1993.
15 BTCE & EPA, "The Costing and Costs of Transport Externalities: A Review," Victorian Transport
Externalities Study, Vol. 1, Environment Protection Authority (Melbourne, Australia), 1994.
Page 3.5-4
Transportation Cost Analysis
different rates per traveler. Thus, SOY drivers impose costs on car poolers and bus riders
who are equally delayed in traffic (except on HOY facilities) despite the much lower cost
they impose. Congestion causes delays to pedestrians, and imposes increased noise and air
pollution on nearby residents. The external nature of congestion costs is also demonstrated
by the considerable resources society spends to increase road capacity and implement
demand management programs, only part of which are paid by automobile user fees.
Estimates:
• The AASHTO 11 Green Book 11 Table 111-33 shows that buses have the congestion
effects of 1.6 Passenger-Car Equivalents (the unit of relative congestion impacts) on
highways with grades up to 4%, and higher equivalents on steeper grades.16
• Michael Cameron cites overall average congestion costs of $0.11 per vehicle mile in
Southern California, and $0.3 7 per vehicle mile under congested conditions. 17 He also
cites road capacity expansion costs of $.10 per average vehicle rnile. 18
• Table 3.5-3 shows marginal arterial congestion costs for various Australian cities.
Table 3. 5-3
-
Peak Period
Traffic
Poor
Fair
Good
Weighted Avg.
M
--- - -
I Ext
IC
-----------------------
Sydney
tion Cost"' 19
- - - - - - - - - ---
Melbourne
Brisbane
Adelaine
AUS/km US/mile AUS/km US/mile AUS/km US/mile AUS/km US/mile
2.46
3.12
2.28
2.90
2.63
3.34
2.23
2.83
0.55
0.70
0.37
0.47
0.57
0.72
0.50
0.64
0.05
0.04
0.05
0.05
0.04
0.05
0.04
0.06
1.04
0.60
0.76
0.19
0.24
1.32
0.48
0.61
• According to a 1986 report by the Institute of Transportation Engineers, 10% of urban
driving and 16% of principal arterial driving occur under congested conditions, and
these percentages are increasing.2o
j176A Policy on Geometric Design ofHighways and Streets, AASHTO (Washington DC), 1990, p. 261.
Michael Cameron, Transportation Efficiency, Environmental Defense Fund (Oakland), 1991, p.19
18 Steve Morrison, "A Survey of Road Pricing," Transportation Research, 20A/2, p.91.
19 J.J. Dodgson, "Benefits of Changes in urban Public Transport Subsidies in the Major
Australian
Cities," The Economic Record, Vol. 62, No. 177, 1986, pp. 224-235.
20 Urban Traffic Congestion: What Does the Future Hold, Inst. of Transportation Engineers, 1986, p. 7.
Page 3.5-5
Transportation Cost Analysis
• Theodore Keeler, et al. estimated marginal congestion costs for San Francisco area
highways. Their results are summarized in Table 3.5-1 with cost estimates updated to
1994. This is still considered one of the most comprehensive analysis of its type.
-
-
~~
-.
- ---
-
~---
Rural-Suburban
Urban-Suburban
Central City
..
Peak
8.1
15.6
9.9
21.0
45.6
80.1
-~--
Interest
6%
12%
6%
12%
6%
12%
---~-------
---- ,,.... ------
Near Peak
3.3
4.5
3.6
4.8
5.4
5.4
Day Avg.
1.8
2.4
2.1
2.4
2.7
2.7
, ............. _.............._
Night Avg.
1.2
1.5
1.5
1.5
1.8
1.8
....,. .... _. . .... ...,.I
Weekend
0.3
0.3
0.3
0.3
0.6
0.6
• Brian Ketcham estimates national average congestion costs at $.072 per automobile
mile, citing reduced productivity, and increased vehicle operating and freight costs.22
• Douglass Lee cites congestion costs of about $.10 per vehicle mile in urban areas, with
higher spot estimates of$.30 per vehicle mile.23
• Peter Miller and John Moffet estimate national congestion costs greater than $0.035
per passenger mile for all driving, with much higher costs on congested roads.24
• Herbert Mohring and David Anderson estimate average congestion costs for Twin
City roads shown in Table 3.5-2.
Table 3.5-2
A
--
-
M-- --
All Road Links
Expressways
IC----
----
tion Cost.,2s
- - -- -
---- - --
Morning Peak
$0.207
$0.236
Afternoon Peak
$0.17
$0.201
• U.S. research cited in an OECD report indicates that motorcycles on urban freeways
impose 0. 5 passenger car units (PCU) of congestion when traffic per lane is less than
600 vehicles per hour (VPH), but this increases to 1 PCU at 1,800 VPH. 26 Buses on
urban freeways are estimated to impose 1.2 PCU at less than 1,000 VPH, and 1.8 at
1,800+ VPH. On urban arterials, buses are estimated to impose 1.2 PCU at LOS B,
1.3 PCU at LOS D, and 1.7 PCU at an intersection. This does not appear to include
stopping to pick up passengers.
21 Theodore Keeler, et al., The Full Costs of Urban Transport: Part Ill Automobile Costs and Final
Intermodal Cost Comparisons, Institute of Urban and Regional Development (Berkeley), 1975, p. 47.
22 Brian Ketcham, Making Transportation Choices based on Real Costs, Oct. 1991, p. 9
23 Douglass Lee, An Efficient Transportation and Land Use System, January 1989, p.5
24 Miller and Moffet, The Price ofMobility, National Resources Defense Council, Oct. 1993, p.23
25 Herbert Mohring and David Anderson, Congestion Pricing for the Twin Cities Metropolitan Area,
Dept. ofEconomics, University of Minnesota (Minneapolis), January 1994.
26 Jmpacts ofHeavy Freight Vehicles, OECD (Paris), December 1982.
Page 3.5-6
Transportation Cost Analysis
• The Office of Technology Assessment study estimates annual congestion costs at $129
to $150 billion, averaging $0.056 to $0.065 per vehicle mile, but also points out that
some estimates may overstate national congestion growth and total costs. 27
• Robert Repetto, et all. modeled congestion costs on five classes of congested
roadways, covering about half of total U.S . vehicle traveJ.28 They concluded that
appropriate congestion fees average $0.04-0.05 per vehicle mile over these roads, and
range as high as $0.21 per vehicle mile. They estimate total direct national congestion
costs at $44 billion annually, and as high as $98 billion annually when additional
accident costs are included.
• Transport Concepts estimates truck interference costs (congestion and delays to other
traffic) at $0.62 per ton mile for intercity semi-trailer trucks and $0.79 per ton mile for
B-Train trucks ($0.52 and $0.64 Canadian per tonne kilometer respectively). 29
• A Transportation Research Board special committee report cities several congestion
pricing studies that provide estimates of optimal congestion prices (which are
considered to represent congestion costs) ranging from about $0.05 to $0.36 per
vehicle mile on congested urban roads, with averages of $0.10 to $0. 15.30
• A U.S. General Accounting Office estimates that productivity losses from highway
congestion cost the nation as much as $100 billion annually.31
• A recent US DOT/FHWA study estimates annual congestion costs at $4 3.2 billion.32
• Ken Small, et al. estimate that large vehicles such as trucks and buses contribute 1.5 to
5 times more to congestion than automobiles, depending on road conditions.33
• One traffic model estimates that bicycles going straight through an intersection cause
0.2 of the congestion of an average car.34 However, this overstates bicycles' overall
congestion impacts since they are prohibited from freeways, where congestion costs
are highest, and tend to use alternatives to congested roads when possible. 35
27 Saving Energy in US. Transportation , U.S. Office of Technology Assessment, 1994, p. 108, 114.
28 Robert Repetto, et al. , Green Fees: How a Tax Shift Can Work of the Environment and the Economy,
World Resources Institute (Washington DC), 1992.
29 External Costs ofTruck and Train , Transport Concepts (Ottawa), October 1994, p.23 .
3 Curbing Gridlock, TRB, National Academy Press (Washington), 1994, Appendix B.
31 Smart Highways: An Assessment of Their Potential to Improve Travel, U.S. General Accounting Office
°
(Washington DC, 1991). The quality of this estimate is uncertain.
D. Schrank, S. Turner and T. Lomax, Estimates of Urban Roadway Congestion, USDOT, 3/1993
33 Kenneth Small, Clifford Winston and Carol Evans, Road Work, Brookings Institute, 1989, p. 12.
34 Homburger, Kelland Perkins, Fundamentals of Traffic Engineering, 13th Edition, Institute of
Transportation Studies, UCB (Berkeley), 1992, p. 8-11 .
35 Todd Litman, "Bicycling and TDM," Transportation Research Record 1441, 1994, p. 134-140.
32
Page 3.5-7
Transportation Cost Analysis
Variability: Congestion varies by location, time, and, to a lesser extent, the type of
vehicle. Congestion is greatest in urban areas, but is increasingly a problem in suburban
and some rural areas.36
Conclusions: The magnitude of traffic congestion and how this problem compares with
other transportation costs depends on how congestion is measured and various
assumptions about its impacts. Congestion is clearly a significant cost and an externality in
terms of economic efficiency. This cost is primarily associated with Urban Peak travel
(including suburban areas) but a moderate amount of congestion is associated with Urban
Off-Peak travel in many areas. The cost is probably highest on urban highways, but
congestion costs on major urban arterials appear to be almost as significant.
The simple existence of congestion costs does not necessarily demonstrate that road
capacity needs to be increased. From an economic efficiency perspective eliminating all
congestion is inappropriate since it would require a vastly over-built and therefore
suboptimal road system. Urban traffic congestion must be expected because providing
capacity to accommodate unlimited peak-period travel demand would not be cost effective
and because congestion is self limiting.
Viable estimates of total U.S. congestion costs range from $43 .2 to $150 billion per year.
$100 billion is used as a starting point for this study.37 Assuming that 20% of all driving
and 80% of congestion costs occur under Urban Peak conditions, 38 and 2,300 billion miles
are driving annually, the average cost is about $0.17 per Urban Peak mile ([$100 x 80%] I
[2,300 x 20%]). Urban Off-Peak driving represents 40% of driving and is estimated here
36 Robert Cervero, Suburban Traffic Congestion: Is There a Way Out?, City and Regional Planning, UCB
(Berkeley), 1991.
37 Since the total value of U.S. vehicle occupant travel time is estimated at $900 billion per year, this
indicates that congestion increases travel costs overall by about 11%. See Saving Energy in U. S.
Transportation , Office of Technology Assessment, 1994 p. 108.
38 About 60% of driving is urban and about 33% occurs during peak periods, Facts and Figure '92.
Page 3.5-8
Transportation Cost Analysis
to incur 20% of congestion costs, for an estimate of$0.02 ([$100 x 20%] I [2,300 x
40%]). Rural driving is not considered to experience significant congestion costs. This
estimate is somewhat lower per mile than marginal highway congestion cost estimates
such as those by Keeler, et al (Table 3.5-1), but not unrealistic considering that it includes
non-highway driving, which is likely to have lower congestion costs. However, this
probably represents a lower bound and significantly higher congestion costs are likely to
exist in heavily congested areas.
Fuel efficient and electric cars, vans and motorcycles have the same congestion costs as an
average automobile. Additional passengers impose no additional congestion. Buses and
trolleys are considered to impose twice, and bicycles 5% of the congestion costs of an
average automobile. Walking and telecommuting impose no significant congestion cost.
----
-
-- --
----o--- - --- - - - -
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
,.
-
.-
-
-
- -
Urban Peak
Urban Off-Peak
Rural
Averae:e
0 .17
0.17
0 .17
0.17
0.00
0.34
0.34
0.17
0.009
0.00
0.00
0.02
0.02
0.02
0.02
0.00
0.04
0.04
0.02
0.001
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0 .042
0.042
0.042
0 .042
0 .00
0.084
0 .084
0.042
0.002
0.00
0.00
Automobile (Urban Peak) Cost Range: Minimum and Maximum estimates are based on
the literature cited above.
Maximum
Minimum
$0.06
$0.02
Page 3.5-9
Transportation Cost Analysis
3.6 Road Facility Costs
Definition: Roadway facility costs required for automobile use not borne by user fees .
Description: Road facility costs include the costs of road construction and maintenance,
land acquisition, financing expenses, and the portion of roadway support facilities and
programs required for automobile traffic. To avoid double counting costs in Chapter 3.1,
only the portion of these costs not paid by users are included here.
Discussion: Many people assume that fuel taxes and vehicle fees pay all roadway facility
costs, but this is not so. Although these charges cover most highway construction and
maintenance costs, a major portion of road and street costs in North America are funded
by local property and sales taxes. If revenues from driving met all road construction costs,
society could be indifferent to increases in traffic volumes on uncongested roads because
increased revenue would offset costs. In practice, marginal roadway costs almost always
exceed marginal revenues. Communities must either endure increased traffic congestion or
subsidize roadway construction.
The roadway facility costs imposed by different road users has been widely studied and
various models have been developed determine the allocation of these costs. 1 Capital and
operating costs are often handled separately.
Capital Costs
Capital costs include land, facilities (roads), and equipment (signals, signs, etc.). Since
most current capital improvements are associated with reducing congestion, cost allocated
1 Ken
Small, Clifford Winston and Carol Evans, Road Work, Brookings Institute (Washington DC}, 1989.
Page 3.6-1
Transportation Cost Analysis
between vehicle classes is based primarily on the road space each requires. The majority of
this charge is allocated to private automobiles since they account for about 90% of traffic.
Capital costs borne by general taxes (typically, local property taxes, sales taxes, and
special local assessments) represent an external cost since they are not based on road use.
People who place heavy costs on the road system, such as those who drive large vehicles
during rush hour may pay less of these taxes than others who never drive. This implies a
subsidy of driving. Current North American road funding involves another more subtle
subsidy. Most capital road funding does not rely on cost recovery required in most other
economic sectors or for other transport modes, such as rail. Instead, current road users
finance improvements for the benefit of future users, which may or may not include
themselves, and capital assets are written off when completed. Douglass Lee points out
that this practice of pay-as-you-go roadway funding, and treating past road capacity
expenditures as sunk cost represents an undercharging of road users:2
"Current highway finance practice finances most improvements out of current
revenues, eliminating the need for borrowing. If highway users -- who are also
highway investors-- don't have to pay interest on capital improvements, why should
they be chargedfor it? The reason is that money deposited in a highway trust fund
earns interest at whatever rate the US. Treasury is paying, and that interest is
foregone when money is spent. There is no way to pretend that capital investments
have no opportunity cost to the funds committed to them. Equally important, the
amount spent one year bears little relationship to the value of the capital consumed in
that year. If the system is wearing down faster than it is being rebuilt, for example,
current users are living off ofprevious users/taxpayers who built up the capital stock.
If the original investment is worthwhile,
it should be earning -- over its lifetime -- a
rate of return at least equal to the market rate for low-risk investments. If the asset
continues to be used as a highway, then implicitly it is worth what it cost, including
interest on the outstanding balance. To fail to charge users enough to cover the
interest, then, is a subsidy to users, in the form of a zero-interest loan. An upper
bound on the opportunity cost, using this method, would be the (depreciation)
2 Herbert Mohring and Mitchell Harwitz also emphasize that road user charges should incorporate
amortized values for all construction, maintenance, and depreciation costs in Highway Benefits: An
Analytical Framework, Northwestern University Press (Evanston), 1965.
Page 3.6-2
Transportation Cost Analysis
replacement cost of the facility, times the current interest rate. A neutral approach,
then would be to measure the replacement costs of the existing system, annualize that
cost, and recover that amount each year. Replacement cost would be stated in dollars
of the current year, hence revenues would keep pace with inflation. '13
Operating Costs
These are the short run marginal costs of the road system, which consist primarily of
maintenance expenditures. Most road maintenance cost, including resurfacing, bridge
replacement and other repairs are attributed to motor vehicle impacts and needs. Road
wear increases by approximately the third power of vehicle axle weight, so a heavy truck
imposes maintenance costs hundreds of times greater than an automobile.4 The cost of
increasing road surface thickness to accommodate heavy vehicles is also allocated to those
vehicles. Studded tires also incur costs that are significant in many areas, estimated at 15%
of road maintenance costs in Norway, which is probably representative of colder areas in
North America.5 Some road deterioration occurs from weathering, which varies from
about 2% per year in mild climates up to 7% or more in areas with extreme winters.6
Road facility costs are greater than current expenditures due to deferred maintenance that
will increase future costs. According to the U.S. Federal Highway Administration, annual
investments of over $60 billion (three times current expenditures) are needed to maintain
the national road system at acceptable standards.7
Equitable user charges should therefore include recovery of capital investment based on
each vehicle's marginal demand for road space, road durability, and maintenance. The
3 Douglass
Lee, Full Cost Pricing ofHighways, National Transportation Systems Center (Cambridge),
Jan. 1995, p.l3 .
4 Kenneth Small, Clifford Winston and Carol Evans, Road Work~ Brookings Institute, 1989, p. 11.
5 Per Kageson, Getting the Prices Right, European Fed. for Transport & Environment, 1993, p. 150.
6 Kenneth Small, Clifford Winston and Carol Evans, Road Work~ Brookings Institute, 1989, p. 11.
7 1991 Status of the Nation's Highways and Bridges, USDOT (Washington DC)~
Page 3.6-3
Transportation Cost Analy sis
distribution of costs is summarized by Small et al. in the book Road Work. 8 With this
infonnation it is possible to develop road user charges based on marginal costs that does
nor require cross subsidies between user classes, or from users in one time period to
another. If user fees are insufficient to cover the cost of a roadway, society may be better
off abandoning the road and using the resources elsewhere. Douglass Lee comments:
"a capital asset that continues to function as a highway should be earning revenues at
least as great as the interest on the invested capital plus depreciation, plus operating
costs. To earn less implies that the long run costs are not justified, and the road ought
to be phased out of use. What is desired is a capital cost that includes actual
depreciation plus interest, and which will recover the replacement cost of the asset
over its lifetime. '19
Lee is simply suggesting that standard investment criteria be applied to roads. Road users
should pay a return on the capital expenditures that created and maintain roadways. Roads
that cannot be justified economically should be decommissioned so their resources are
available for other productive uses. Failing to do this results in over-investments in roads
and under-investments in other forms of transportation, other economic activities, and
other uses of land. Underpricing and over investment in roads may have been justified
decades ago when developing a networked road system was a strategic national goal, but
contemporary sensibilities are attuned to the significant environmental damage caused by
roads, the need to limit access to the few wild areas left, and the economic inefficiency
implied by building and maintaining roads where social costs exceed social benefits.
Other Road Uses
It is sometimes argued that roads serve purposes other than automobile travel, so a
portion of their cost should be borne by all of society. Even people who never use an
automobile need access to their residence for delivery of goods and services, emergency
8 Ken
Small, Clifford Winston and Carol Evans, Road. Work, Brookings Institute (Washington DC), 1989.
Full Cost Pricing of Highway s, USDOT, National Transportation Systems Center
(Cambridge), Sept. 1993, p.24.
9 Douglass Lee,
Page 3.6-4
Transportation Cost Analysis
vehicles, for walking and public transit, and because road rights-of-way contain public
utility lines such as wires and pipe. One way to address this problem is to establish a
standard of "basic access" that residents require no matter how little they use an
automobile directly. In practice this need can be satisfied by a 10 or 12 foot right-of-way
with a single lane of gravel or light pavement, which is the quality of road typically chosen
when users pay for their own driveway and on campus-type developments. Roadway costs
beyond this should be allocated to motor vehicle use.
Pre-automobile cities typically devoted less than 10% of land area to roads, while newer
automobile dependent cities devote up to 30% of land area to roads, implying that 50 to
75% of road area in such cities is needed specifically to satisfy the needs of automobiles. 10
Since nearly all communities have well-developed roadway systems that easily satisfy
minimal access needs, the costs of increasing existing road capacity can be charged to
motor vehicle use. Pedestrian and bicycle facility costs could also be charged to driving if
automobile traffic degrades the bicycling and walking environment, requiring facilities
separated from the roadway. This implies that most current roadway construction and
maintenance costs are the responsibility of motor vehicle users.
Estimates:
• The American Automobile Manufacturers Association states that 1992 highway user
fees total $54.8 billion, including the federal excise tax on trucks, while road spending
totaled $81 .6 billion, leaving $26. 8 billion in external costs.11
• Apogee Research estimated state, federal and local capital and operating costs for
various modes in Boston, MA and Portland, ME at high, medium and low densities.12
10 Harry Dimitriou,
Urban Transport Planning, A Developmental Approach, Routledge (NY), 1992, p.
136; Herbert Levinson, Transportation and Traffic Engineering Handbook, Institute of Transportation
Engineers/Prentice Hall (Englewood Cliffs, NJ), 1982, p. 256.
11 AAMA, Facts and Figures '93, p. 62, 78.
12 Apogee Research, The Costs of Transportation: Final Report, Conservation Law Foundation (Boston),
1994, pp. 121-137, 155-157.
Page 3.6-5
Transportation Cost Analysis
The sum of road costs for automobile use range from 5.4¢ per vehicle mile for
expressway driving in Boston, to 0.6¢ for non-expressway driving in Portland. They
also found that needed facility maintenance is being deferred, adding 1.2¢ per
expressway vehicle mile, and 2.1¢ for non-expressway driving. Their analysis includes
the estimates of relative costs per vehicle mile for three modes listed in Table 3.6-1.
Mode
Automobile
Bus
Bicycle
State Capital Costs
0.683
1.81
0.34
State Maint. Costs
0.719
3.42
0.014
Local Costs
0.701
2.62
0.23
These factors show relative costs of three vehicle classes.
• The California Energy Commission estimates infrastructure maintenance and road
repair costs at $0.006 per mile for automobiles and $0.12 per mile for buses. 13
• The California Department of Transportation's 1987 Highway Cost Allocation Study
estimated the cost per vehicle mile shown in Table 3.6-2.
Jaote .1. o-~
Laitrans Hignway Lost Allocation vaaues (per vemcie)
Vehicle Class
Automobiles
Motorcycles
Pickups and vans
Recreational Veh.
Buses
Trucks
All vehicles
Capital
Costs
Maintenance
Costs
Total
Cost
Annual
Miles
(millions)
(millions)
(millions)
(billions)
$1,035.4
$8.2
$263 .2
$48.4
$11.3
$613 .7
$1980.2
1,150.2
12.3
376
59.1
11.1
490.4
2,099
2,185.6
20.5
639.2
107.5
22.4
1,104.1
4,079.2
158
1.5
36.3
4.4
0.9
13.9
215
Unit Cost
Unit Cost
( 1987$NMT)
(1994$NMT)
0.014
0.008
0.018
0.024
0.025
0.079
0.019
0.018
0.010
0.023
0.031
0.032
0.101
0.024
• Ken Casavant and Jerry Lenzi estimate the roadway damage costs of overloaded
trucks to average from about $0.08 per mile to $2.50 per mile, depending on how
much they are overloaded. As a base cost they cite estimates that 40-ton trucks impose
road damage costs of$0.01 to $0.06 per ton on state highways, and 50% higher costs
on county roads.
• Automobile user payments (fuel taxes vehicle registration fees) cover 56% of roadway
network expenditures in Wisconsin. 14 Fuel taxes would need to increase approximately
$0.35 per gallon to fund all current road expenses.
• Theodore Keeler et al. estimate road maintenance costs at $.0015 per vehicle mile.15
13
California Energy Commission, 1993-1994 California Transportation Energy Analysis Report,
California Energy Commission (Sacramento), Feb. 1994, p. 29.
14 Cambridge Systematics, Highway Cost & Pricing Study, Winconsin DOT (Madison), Sept. 1994.
15 T. Keeler, et al, The Full Costs of Urban Transport, 1111975, p.52, Estimates updated to 1994 dollars.
Page 3.6-6
I
I
Transportation Cost Analysis
• Brian Ketcham estimates U.S . roadway maintenance costs at $.001 per automobile
vehicle mile and $.45 per vehicle mile for heavy trucks.l6
• Douglass Lee identifies the road system externalities described in Table 3.6-3 . He also
recommends charging a state level service tax for road use to be consistent with other
economic activities, totaling $15.9 billion a year, or about $0.007 per vehicle mile. 17
Table 3.6-3
Lee's Estimates of Road Svstem Externalif
Costs
Construction Expenditures
Interest
Pavement, ROW, and structure maintenance
Administration and research
Total roadway expenditures
Minus $55 billion road user payments
Subsidy per mile (assuming 2,300 Billion VMT)
Billions of Dollars
$42.5
26.3
20.4
6.9
$96.1
$96.1-$55.0 = 41.1
$0.018
• Peter Miller and John Moffet give maintenance costs per mile in Table 3.6-4, and total
U.S. maintenance costs of$48.3 billion per year.l 8
Table 3. 6-4 Road
Maint
Road Class
Urban Car
Cost Estimat
Urban Truck
$/VMT
$/ESAL
0.29
Interstate
0.014
Arterial
0.038
0.76
0.74
Collector
0.037
0.046
0.92
Local
0.677
0.033
....... Average
ESAL =Equivalent Standard Axle Load of 18,000 lbs
- - - ---
Rural Car
$/VMT
0.005
0.012
0.016
0.029
0.015
Rural Truck
$/ESAL
0.10
0.24
0.32
0.58
0.31
• Ken Small et al. use U.S. Federal Highway Administration statistics to determine that
user taxes and tolls accounted for 62% of 1985 road disbursements. 19
• The Office of Technology Assessment study estimates U.S . annual road facility costs
at $77 billion, and user taxes and fees to average $44 billion per year, for a net external
cost averaging about $0.014 per vehicle mile.2o
• Transport 2021 estimates road maintenance costs to average $0.013 Canadian per km
($0.016 U.S . per mile).21
16 Making Transportation Choices Based on Real Costs, Konheim & Ketchem (New York), Oct. 1991
17 Douglass Lee, Full Cost Pricing ofHighways, USDOT, National Transportation Systems Center,
(Cambridge), January 1995, p. 12.
18 The Price ofMobility, National Resource Defense Council (Washington DC), Oct. 1993, p. 10.
19 Kenneth Small, Clifford Winston, Carol Evans, Road Work, The Brookings Institute, 1989, p. 2.
20 Saving Energy in U. S. Transportation , U.S. Office of Technology Assessment, 1994, p. 105-107.
Page 3.6-7
Transportation Cost Analysis
Transport Concepts estimates that trucks impose infrastructure costs averaging $0.82
per ton mile ($0.67 Canadian per tonne kilometer), including road capacity, road
maintenance and roadway services.22 They estimate that although big trucks make up
only about 9% ofvehicle traffic they account for about 25% of roadway costs. Rail
infrastructure costs are considered completely internalized by rail companies.
• The USDOT's most recent cost allocation study concludes that an equivalent standard
axle load (ESAL) of 18,000 lbs. incurs a road maintenance cost of$0.09 per mile on
rural interstate highways, $0.66 on urban arterials and $0.80 on local urban streets. 23
Variability: Road costs depend on the type of vehicle, how much it is used, and how
much it contributes to traffic congestion.
Conclusions: Roadway costs include current capital and maintenance expenditures, a
return on past capital investments in roads, and future costs from deferred maintenance.
These costs can be allocated between different vehicle classes based on their use of road
space and road damage. Although even people who never travel by automobile use roads
for walking and other purposes, virtually all current road expenditures result from vehicle
needs so costs are allocated to them. The 1987 Caltrans cost allocation study provides
relative costs per vehicle. However, these understate total costs because they exclude
deferred maintenance and return on investment costs. The Caltrans $0.024 per mile
average cost estimate times 2,300 billion total miles gives a total cost of$55 .2, which is
57% ofDouglass Lee's $96.1 billion estimate of total roadway costs.24 Scaling up the
Caltrans gives total cost estimates shown in Table 3.6-5. The third column shows this cost
net of $54. 8 billion annual user fees, to indicate the external cost component.
21 Transport 2021, Costs ofTransporting People in the British Columbia Lower Mainland, Greater
Vancouver Regional District (Vancouver), 1993, pp. 25-26.
22 External Costs ofTruck and Train, Transport Concepts (Ottawa), October 1994, p.26.
23 FHWA Cost Allocation Study, 1982 p. E-25, table 3, cited inRoad Work, p. 15.
24 Facts and Figures 93, American Automobile Manufacturers Association (Detroit) p. 63 .
Page 3.6-8
Transportation Cost Analysis
Table 3.6-5
Roadway Cost Allocation (1994 dollars per mile
Mode
Caltrans Estimate
Total Cost Estimate
Automobiles
Motorcycles
Pickups and vans
Buses
Trucks
0.018
0.010
0 .023
0.032
0.101
0.031
0 .017
0.04
0.056
0.175
External Costs
0.013
0.007
0.017
0.024
0.075
This table shows Caltrans vehicle cost allocation scaled up to include total costs
estimated by Douglass Lee. The left column shows this cost net offuel taxes.
These values are used to calculate Best Guess cost estimates. Miller and Moffet indicate
that urban road costs are higher than rural road costs per vehicle mile, so urban driving
costs are increased by 25% and rural costs are decreased by 25%. Since electric vehicles
do not pay fuel taxes dedicated to roads, all of their road costs are considered external.
The marginal maintenance cost of a ride share passenger is considered too small to count.
Since public transit buses are exempt from some fuel taxes their total cost is used, but this
would not be appropriate for private buses that do pay fuel taxes or where such
exemptions do not exist. A trolley that travels on tracks does not incur road wear costs,
but comparable public costs are required to maintain rails and right-of-way. Bicycles cause
no pavement wear and impose minimal demands on road space so their cost is estimated
to be 5% of an automobile. Walking and telecommuting incur no road facility costs.
- - - - ------
--
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Motorcycle
Bicycle
Walk
Telecommute
-·-- -
- - ' -
- -
....
'I
Urban Peak
Urban Off-Peak
Rural
Averae:e
0.016
0.016
0.038
0.021
0.00
0.070
0.070
0.009
0.001
0.00
0.00
0.016
0.016
0.038
0.021
0.00
0.070
0.070
0.009
0.001
0.00
0.00
0.010
0.010
0.023
0.013
0.00
0 .042
0.042
0.005
0 .000
0 .00
0.00
0 .014
0.014
0 .032
0.018
0.00
0.059
0.059
0.007
0.001
0.00
0.00
Automobile Cost Range: Minimum and Maximum values are based on estimates cited.
Minimum
$0.01
Page 3.6-9
Maximum
$0.03
Transportation Cost Analysis
3.7 Roadway Land Value
Definition: Opportunity costs efland used for roadways.
Description: Roadway land value costs include the value efland used for road rights-ofway and other public facilities dedicated for automobile use. This cost could also be
defined as the rent that users would pay for roadway land if it were managed as a utility,
or at a minimum, the taxes that would be paid if road rights-of-way were taxed.
Discussion: Approximately 60,000 square miles ofland in the U.S. are devoted to
roadway rights-of-way, about 2% of the nation's total surface area. In Europe this is
estimated to be a somewhat smaller 1. 3 percent. 1 It is much higher in developed areas
where land values are highest. In modern urban areas, 25%-30% ofland is devoted to
streets, and even higher percentages in commercial centers.2 Unlike other public land uses,
such as parks and forests, roadways provide little secondary environmental benefits such
as wildlife habitat or timber production. Roadway land value is often considered a "sunk"
cost. Emile Quinet for example argues that land which has long been used for roads incurs
no social cost, but there is no reason that the opportunity cost of this resource should be
ignored. Douglass Lee states, ''Land in highway right-of-way has alternative uses, and
this value is included in published figures only when the purchase of new land is a part of
current expenditures. Normally, any long-lived business investment is expected to earn a
rate of return at least equal to the interest rate on borrowed funds.
1 National
''3
Academy of Sciences, Policy Implications of Greenhouse Warming, 1991, cited Miller, 1993.
Urban Transport Planning, Routledge (NY), 1992, p. 136; Herbert Levinson,
Transportation and Traffic Engineering Handbook, Prentice Hall (Englewood Cliffs, NJ), 1982, p. 256;
Emile Quinet, "The Social Costs of Transport: Evaluation and Links With Intemalisatiion Policies," in
Intema/ising the Social Costs ofTransport, OECD (Paris), 1994, p. 54.
3 Douglass Lee, A n Efficient Transportation and Land Use System , National Transportation Research
Center (Cambridge), 1992.
2 Harry Dimitriou,
Page 3.7-1
Transportation Cost Analysis
Some authors argue that since roads often increase adjacent real estate values, that
roadway land provides a positive rather than negative social value.4 It is true that access
(defined in Chapter 1) increases land value. But to assume that driving is the only form of
access ignores differences in the amount of land required by different modes. 5 Modes such
as driving, which require more land, should be charged this incremental costs.
As discussed in Chapter 3.6, a portion of the road system can be considered necessary for
basic access to residents (including utility service lines such as water, sewer and power)
whether they drive or not, the cost of which could be allocated to all community members.
This is estimated at 25% to 50% of the total roadway area by various sources. For
example, Harry Dimitriou states that pre-automobile cities typically devote less than 10%
ofland to streets, while modem, automobile-oriented cities devote up to 30%. 6 This
portion can be subtracted before assigning roadway land value costs to automobile use.
Estimates:
• Ketcham and Komanoff assume that streets constitute 33% of urban land area, half of
which is needed for basic access, to calculate the annualized value of land for
automobiles to be $66.1 billion, or $0.03 per vehicle mile.
• Douglass Lee applies the FHWA's prototypical land acquisition cost per mile for 9
roadway classes to the entire U.S . road system to estimate total land value and
calculate annual interest forgone to be $74.7 billion, or $0.034 per VMT. 7 He
considers the FHWA's estimate to be conservative.
4 Emile Quinet,
"The Social Costs of Transport: Evaluation and Links With Internalisatiion Policies," in
lnternalising the Social Costs ofTransport, OECD (Paris), 1994, p. 55.
5 For an
interesting discussion of relative land use requirements of various modes see Eric Bruun and
Vukan Vuchic, Introduction and Historical Review of Time-Area Concept with Example New Application
in Urban Land Use A naly sis, Transportation Research Board Annual Meeting, Paper 950296, 1995.
6 Harry Dirnitriou, Urban Transport Planning, Routledge, (NY), 1993, p. 136.
7 Full Cost Pricing ofHighways, National Transportation Systems Center (Cambridge), 1995, p. 11.
Page 3.7-2
Transportation Cost Analysis
• Emile Quinet provides a European estimate of the relative land use area of different
modes shown in Table 3.7-1 .8 This indicates that automobiles require approximately 4
times the road space as a bicycle or motorcycle, and 10 to 40 times that of buses.9
-
-
Mode
Bicycles and Motorcycles
It
It
Automobiles (1.33 passengers)
It
It
Bus (daily average: 20 pass.)
It
Bus (peak period: 80 pass.)
It
-
-- --
-
- - -- -
---
2
---
Use
Work (9 hours)
Leisure (3 hours)
Shopping (1.5 hours)
Work (9 hours)
Leisure (3 hours)
Shopping (1.5 hours)
Normal Roads
Bus Lane
Normal Roads
Bus Lane
------
Parkin2
13.5
4.5
2.5
68
23
11
0
0
0
0
Traffic
7.5
7.5
7.5
17
17
17
7.5
30
2
7.5
Total
21
12
10
85
40
28
7.5
30
2
7.5
• Transport 2021 calculates the value of road land dedicated to motor vehicle use in the
Vancouver area to be worth $578 million a year when amortized at 10%, based on
30% of adjacent land's assessed values.10 This averages $0.047 Canadian per km, or
$0.059 U.S. per mile, or almost $0.20 if the land is assessed at its full value.
Variability: Road land costs are based on vehicle use (which creates demand for roads)
and varies depending on location, with higher land market values in urban areas, and
higher non-market values in areas with high environmental worth.
Conclusions: Land used for roads has an opportunity cost. There is no reason to exclude
the value of this resource in transport cost analysis. To determine what should be charged
to motor vehicles, it is appropriate to first subtract the portion of the road system that
provides basic access, which is defined as a single lane. In most cases this represents about
25% of paved road area and a smaller portion of road rights-of-way. The remaining 75%+
is charged based on VMT. Although large vehicles may require more road space under
8 Emile Quinet,
"The Social Costs of Transport: Evaluation and Links With Intemalisatiion Policies," in
Interna/ising the Social Costs ofTransport, OECD (Paris), 1994, p. 55.
9 This appears
to underestimate motorcycle roadspace needs, at least in the U.S. where they are large and
normally take a full lane, and overstate bicycle roadspace needs.
10 Cost of Transporting People in the British Columbia Lower Mainland, Greater Vancouver Regional
District (Vancouver), 1993, p. 27.
Page 3.7-3
Transportation Cost Analysis
congested conditions, this is not considered significant in terms of the total amount of land
allocated to road right-of-way.
Douglass Lee's estimate that roadway land is worth $75 billion per year is a reasonable
starting point. Subtracting 25% of this cost for basic access leaves a $56 billion annual
cost. Divided by 2,300 billion VMT, this averages $0.024 per mile, which is applied to all
motor vehicles. Although urban land values are higher, urban roads receive greater use per
lane mile, so average costs per vehicle mile are considered to be comparable for both
urban and rural travel. Bicycles are estimated to require 5% of an average automobile's
road space, while rideshare passengers, walking, and telecommuting require none.
-
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
-
Urban Peak
Urban Off-Peak
Rural
0.024
0.024
0.024
0.024
0.00
0.024
0.024
0.024
0.001
0.00
0.00
0.024
0.024
0.024
0.024
0.00
0.024
0.024
0.024
0.001
0.00
0.00
0.024
0.024
0.024
0.024
0.00
0.024
0.024
0.024
0.001
0.00
0.00
Avera2e ·
0.024
0.024
0.024
0.024
0.00
0.024
0.024
0.024
0.001
0.00
0.00
.
.
Automobile Cost Range: The minimum represents a low estimate of property values.
The upper range is based on a full value assessment of the Vancouver Region road system,
taking into account that roads outside of urban center have lower value.
Minimum
$0.01
Page 3.7-4
Maximum
$0.10
Transportation Cost Analysis
3.8 Municipal Services
Definition: Costs of municipal services for motor vehicles not funded by user fees.
Description: Municipal service costs include policing, emergency response, planning,
courts, street lighting, parking enforcement, and traffic safety education provided for
motor vehicle use but are not funded from driver fees or fines .
Discussion: Automobile use requires a variety of public services. At the municipal level
these are funded largely through local taxes. A number of studies examine these costs.
Some costs in these studies overlap other cost categories in this report, such as road
facility costs, and must be subtracted to avoid double counting.
According to Stanley Hart's analysis of the City of Pasadena's 1982-83 budgets,
approximately 40% of police department, 15% of the fire department, 16.4% of paramedic
services, and a major portion of public works, capital improvement, and debt service
budgets should be charged to automobile use.1 In his analysis Hart subtracts the cost of
providing minimal access for pedestrians, public service, and emergency vehicles when
calculating automobile roadway costs. He concluded that automobile-related expenditures
totaled $15 .7 million, 75% ofwhich came from local general taxes instead ofuser fees.
Daniel Ridgeway's 1990 analysis ofDenver City and County budgets indicates a similar
portion of municipal costs are devoted to automobile services. He calculated that about
40% of police department activities, 15% of fire department, and a major component of
public works, capital facility expenditures, and municipal debt should be allocated to
1
Stanley Hart, "An Assessment of the Municipal Costs of Automobile Use," self published graduate
studies report (Pasadena) 1985, p.14.
Page 3.8-1
Transportation Cost Analysis
automobile use.2 Additional costs were also mentioned but not included in these estimates,
such as locally funded medical care for accident victims, parking facility costs, air
pollution control efforts, and planning activities.
Estimates:
• Stanley Hart estimated that Pasadena automobile subsidies equal about $270 annually
per household or $0.0 13 per vehicle mile. When roadway construction and
maintenance expenditures are subtracted out to avoid double counting costs in
Chapter 3.6, the remaining shares oflaw enforcement, emergency services, and public
administration expenditures total $7.7 million, averaging about $0.008 per vehicle
mile. In 1994 this is worth approximately $0.012 per mile.3
• Apogee Research estimated police, fire, and justice motor vehicle costs in Boston, MA
and Portland, ME summarized in Table 3.8-1 .4
Table 3.8-1
Public :service Losts ot JJnvm2 m Two Lltles (! per vehicle mile)
Boston
High density
Medium density
Low density
Expressway
NonExpwy
2.4
1.1
1.1
1.0
0.4
0.5
1.3
0.9
0.6
0.5
0.4
0.2
Portland
High density
Medium density
Low density
• The California Energy Commission estimates roadway service costs, including a share
of law enforcement, safety, and administration at $0.0 12/mile for all vehicles.5
• Theodore Keeler et al. estimate the average cost of municipal automobile services
(including police, fire, planning, court, public health, and power) at $0.012 per VMT. 6
• Peter Miller and John Moffet estimate average municipal costs at $.0045 per VMT,
with a higher value of$.01 in congested urban areas and $0.002 for rural travel. 7
2 Daniel Ridgeway,
"An Assessment of the Cost of Private Motor-Vehicle Use to the City and County of
Denver," self published graduate studies paper, Denver, March 1990.
3 Stanley Hart, "An Assessment of the Municipal Costs of Automobile Use," 1985, p.14.
4 The Costs ofTransportation: Final Report, Conservation Law Foundation (Boston), 1994, p. 138-144.
5 1993-1994 California Transportation Energy A nalysis Report, CEC (Sacramento), Feb. 1994, p. 29.
6 The Full Costs of Urban Transport, Institute of Urban and Regional Development (Berkeley),
Monograph #21 , 1975, p. 51.
7 The Price ofMobility, National Resource Defense Council (Washington DC), Oct. 1993, p.15
Page 3. 8-2
Transportation Cost Analysis
• Fadi Nassar and Fazil Najafi estimate that law enforcement and risk management costs
average about $5,000 annually per lane mile, and two to three times that in urban
areas.8 If an average lane carries 7,500 vehicles per day, this cost averages $0.002 to
$0.005 per VMT.
• The Office of Technology Assessment study indicates average externalities of$0.007
to $0.04 per vehicle mile based on the following costs (billions):9
Police protection
Fire protection
Court and judicial system
Corrections
Government pollution control
Totals
Low Cost
$7.9
1.4
4.0
2.5
High Cost
$76.5
3.2
10.0
3.5
.LQ
3.0
$16.8
$96.2
• Daniel Ridgeway estimates municipal automobile subsidies exceed $.003 per VMT. 10
• Ken Small describes a study that estimates automobile municipal service costs in San
Francisco average 2.8¢ per motor vehicle mile.11
• Transport 2021 estimates "protective services" of traffic law enforcement and
emergency services (based on 10% of police and 5% of fire department costs) at
$0.004 Canadian per vehicle kilometer, or about $0.005 U.S . per vehicle mile.12
Variability: Miller and Moffet indicate that service costs are significantly greater per mile
in congested urban areas than for rural driving. This is supported by Nassar and Najafi's
estimates of police patrol costs.
Conclusions: Several estimates indicate that municipal services not funded by vehicle user
fees average more than $0.01 per mile, with higher costs in urban areas. Urban Peak travel
is estimated to incur municipal service costs of$0.015 per mile, Urban Off-Peak $0.01 per
8 Facti Emil Nassar and Fazil Najafi, "Quick Approach to Estimate Law Enforcement Cost on Urban
Roads," Transportation Research Record # J262, 1989. p. 39-47.
9 Saving Energy in U. S. Transportation , U.S. Office of Technology Assessment, 1994, p. 104-108.
10 "An Assessment of the Cost of Private Motor-Vehicle Use to the City and County of Denver," 1990.
11 Ken Small, Urban Transportation Economics, Harwood (Chur}, 1992, p. 82.
12 Cost of Transporting People in the British Columbia Lower Mainland, Greater Vancouver Regional
District (Vancouver), 1993, p. 29.
Page 3.8-3
Transportation Cost Analysis
mile, and Rural travel $0.005 . This cost is applied equally to all motor vehicles. Rideshare
passengers are estimated to incur no additional municipal service costs. Bicycling,
walking, and telecommuting are estimated to cost 10% as much per mile as an automobile.
Vehicle Class
Average Car
Fuel Efficient Car
Electric Vehicles
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Motorcycle
Bicycle
Walk
Telecornrnute
Urban Peak
Urban Off-Peak
Rural
Average
0.015
0.015
0.015
0.015
0.00
0.015
0.015
0.015
0.002
0.002
0.002
0.010
0.010
0.010
0.010
0.00
0.010
0.010
0.010
0.001
0.001
0.001
0.005
0.005
0.005
0.005
0.00
0.005
0.005
0.005
0.00
0.00
0.00
0.009
0.009
0.009
0.009
0.00
0.009
0.009
0.009
0.001
0.001
0.001
Automobile Cost Range: Based on estimates cited:
Minimum
$0.003
Page 3.8-4
Maximum
$0.015
Transportation Cost Analysis
3.9 Transportation Equity and Option Value
Definitions: Transportation Equity: Adequate transportation for people who are
economically, socially, or physically disadvantaged.1 Transportation Option Value: The
value of having a variety of transport choices.
Description: Transportation equity and option value are based on the quality and quantity
of access, especially for people who are already socially disadvantaged because they are
poor or disabled. They are affected by the transport system, land use patterns, facility
design, and social habits that affect travel requirements, and the quality of public transit,
trains, ride sharing, bicycling, walking, and special mobility services.
Discussion: Since access to goods, services, jobs and other destinations is important for
economic, personal and social activities, inadequate mobility and access impose a variety
of costs on individuals and society. These costs are social in nature because they are
affected by social decisions and policies, such as transportation investments, zoning codes,
and the location of public services, and because many of the costs are ultimately borne by
society in terms of reduced productivity and general inequity.
Fieure 3. 9-1
Income
Physical Ability
~
1 Transportation
Community Access
There are three dimensions to
transportation quality: income, physical
ability and community access. Community
access is defined as the ease with which
people can access goods, services, and
destinations in their community. Low
income, physical disability and poor
community access contribute to
transportation disadvantage.
Improvement in one factor partially
compensates for low levels in another.
equity is explored further in Chapter 7.
Page 3.9-1
Transportation Cost Analysis
Transportation prices, transit subsidies, special mobility services, and handicapped access
have been the focus of most discussions of transportation equity and option value, but
other factors such as urban form, social habits, and personal security are also important.
Being transportation disadvantaged is affected not only by a person's physical and financial
abilities, but also on how much travel is required to access goods and services, and what
travel choices are available in their community. Most North American communities now
have a high level of automobile dependency.2 A "personal" car is required to participate in
most activities and there are few travel alternatives. This imposes two sets of costs. The
first are impacts on people regardless of their wealth or physical ability. Even those who
own and drive a personal car suffer from a lack of alternatives. For example:
• Households must own extra cars to guarantee that one is available for every trip, or
rely on expensive alternatives such as car rentals and taxies.
• Household members who are too young or too old to drive, and out of town guests
must be chauffeured to all destinations.
• Irresponsible drivers who commit multiple traffic offenses are still allowed a drivers
license because of the tremendous social cost of not driving.
• Drivers are immobilized when their car breaks down or is unavailable for any reason.
Economists use the term option value to describe the benefit of maintaining an option that
is not immediately used.3 The concept of option value has been applied in a variety of
situations including transport policy analysis to explain why people are often willing to
support programs and facilities they seldom or never directly use.
The second set of costs of automobile dependency are the extra burdens imposed on
people who are already socially, economically, or physically disadvantaged. The poor,
disabled and elderly are further disadvantaged by a lack of access to public services,
2 Peter Newman and Jeff Kenworthy,
3
Cities and A utomobile Dependency, Grover (Aldershot), 1989).
Johansson, The Economic Theory and Measurement ofEnvironmental Benefits, 1987, p. 5
Page 3.9-2
Transportation Cost Analysis
employment, and social activities. This reduces social equity, defined as increased
discrepancy between advantaged and disadvantaged people. Since policy makers and
planners are generally drivers and car owners, non-drivers' needs are underrepresented in
public decision making. 4 As use of public transit, bicycling, and walking has reduced and
users are increasingly from disadvantaged populations, these modes become stigmatized,
to the point that a U.S. transit system executive has stated, "Show me a man over thirty
who regularly takes the bus, and I will show you a life failure. "5 Residential, employment,
commercial, and municipal services are increasingly located for optimum parking and
proximity to freeways, with poor access for nondrivers.6 Elmer Johnson states,
"[T]he dominance of the automobile has reduced the availability ofpublic
transportation. For those too young, too old, too poor or too infirm to drive, the
paucity of mobility alternatives severely limits their opportunity for education and
their ability to share in other essential everyday activities. Moreover, as more
employers have moved to the suburbs, more jobs require car mobility. ''7
Poverty and being transportation disadvantaged are closely linked. According to the 1990
National Personal Transportation Survey only 9% of U.S . households were without an
automobile, but ofthose 42% were below the poverty level in income.8 Rural residents are
especially impacted.9 As described by Meyer and Gomez-Ibanez:
"It is widely believed that poor, handicapped, and elderly persons who cannot use an
automobile and do not have access to high-quality, low cost public transportation
cannot participate fully in society--especially given the dispersal of residences,
workplaces, and shopping and recreational centers in US. metropolitan areas. "10
4 James
Kunssler, The Geography of Nowhere, Simon and Schuster, 1993 .
Garreau, Edge City, Doubleday (NY), 1991, p. 130.
6 The exceptions are residents of a few large cities such as New York and Philadelphia where
transportation is relatively diverse and wealthy non-drivers who can afford unlimited chauffeuring.
7 Elmer Johnson, Avoiding the Collision of Cities and Cars, American Academy of Arts and Sciences
(Chicago), 1993, p 8.
8 Charles Lave & Richard Crepeau, "Travel by Households Without Vehicles, 1990 NPTS Travel Mode
Special Report, p. 29
9 Stephen Nutley in Modern Transport Geography, Hoyle and Knowles, ed., Belhaven Press (London),
1992, chapter 8.
10 John Meyer & Jose Gomez-Ibanez, A utos, Transit and Cities, 1981, Twentieth Century Fund, p. 230
5 Joel
Page 3.9-3
Transportation Cost Analysis
Altshuler adds:
''It is unusual to come across a situation in which circumstances of the disadvantaged
have deteriorated absolutely over a sustained period of time. For many Americans
without cars and/or drivers' license, however, the absolute level of mobility has fallen
sharply over the past several decades. Given the dramatic mobility improvements
experienced by most Americans in this same period, it follows that the relative
deprivation of those left behind has worsened acutely. "11
Hillman examines the equity impacts of increased automobile dependency on children:
"In the pursuit of adult mobility, and of the impacts on children of its growth have
been overlooked... the motor vehicle has totally colonized the most convenient local
places in which children could play and socialize. Most importantly, there is the issue
of road accidents-the most frequent cause of accidental mortality among children. "12
Automobile dependency and use impose special hardships on the poorest households.
According to Peter Freund and George Martin these include: 13
• Poor families spent twice the proportion of income on automobiles as wealthy families.
• A lack of transportation by inner city residents to the growing portion of jobs in
suburbs is a contributing factor to unemployment and poverty.
• The poor suffer higher than average automobile air pollution and accident risk impacts.
Some analysis of transportation disadvantaged people focus on the problems of
households that own no motor vehicles.14 However, owning a car does not necessarily
eliminate mobility problems. Many individuals in households with cars still experience
significant problems accessing services, jobs and other destinations. Merle Mitchell
describes concern among social policy analysts about "locational disadvantage" associated
with people who live in rural areas or outer suburbs that have poor access to community
11Aian Altshuler, The Urban Transportation System, 1979, MIT Press (Cambridge), p. 253
12 Mayer Hillman, "Foul Play for Children: A Price of Mobility," Town and Country Planning,
Oct. 1988,
pp. 331-332.
13
Peter Freund and George Martin, The Ecology of the Automobile, Black Rose (NY), 1993, p. 46-49.
see Charles Lave and Richard Crepeau, "Travel by Households Without Vehicles" in the
1990 Nationwide Personal Transportation Survey, Travel; Travel Mode Special Report, USDOT, 1994.
14For example,
Page 3.9-4
Transportation Cost Analysis
services, although many have a household automobile.15 The requirement to own motor
vehicles can be a serious financial burden on the poorest households, as indicated by
Figure 3.9-2. The portion of households that are likely to have special needs (the elderly,
the poor, and single parents) are growing demographically, as are automobile dependent
transport and land use patterns, so these problems may increase.16
Figure 3.9-2 Automobile Expenditures as Percentage of Household Income 17
i
21%
....
20%
.5
:g
5I
:z:"
11%
0
ii
c
"
10%
~
li
1
1%
0%
<$5,000
$5·10 ,000
$10.15,000
$15-20,000
$20·30,000
$3HO,OOO
$40·50,000
$50,000+
Annual Household Income
Automobile expenditures decline with income among middle class families, but are
higher for the lowest income class, indicating that automobile dependency places a
burden on the poorest households.
When 1,600 randomly selected New Mexico residents were asked, "Do you believe that
the ability to get where you want ot go in a reasonable time and for a reasonable cost is
or should be a basic right in the same sense as freedom of speech or the pursuit of
happiness?", 63 .8% of responses said yes, 22.7% said no, and 13.5% were uncertain. 18
15
Merle Mitchell, "Links Between Transport Policy and Social Policy" in Transport Policies for the New
Millennium, Ogden et al. editors, Monash University (Clayton), 1994.
16 Susan Hanson, The Geography of Urban Transportation , Guilford Press (NY), 1986, p.7.
17 U.S. Department of Labor, Consumer Expenditure Survey, 1989 (Washington DC), 1990. Although
driving and fuel consumption decline for the lowest income households, as discussed in Section 5.3, costs
do not appear to decline in proportion. The likely explanation is that there is a minimal level of fixed
automobile ownership costs, including registration, insurance and repair costs that cannot be avoided.
18 John Hamburg, Larry Blair and David Albright, Mobility as a Right, Transportation Research Board
1995 Annual Meeting (Washington DC), paper #951001.
Page 3.9-5
Transportation Cost Analysis
The existence and public concern over transportation inequity are indicated by strong
public support for public transit, special mobility services, and handicap access. Much of
the political and economic support for transit comes from people who seldom or never use
it themselves. About 64% of U.S . transit service fiscal costs are directly subsidized,
totaling about $10 billion in 1991 .19 Transit also receives indirect subsidies, including tax
exemptions, and facilities (bus stops, pull-outs, and road maintenance costs) provided by
other budgets.
Economist Jim Lazar comments, "... transit is the most oversubsidized public service in
the state [Washington]; and it needs to be, in order to compete with the most
oversubsidized private service in the state (i.e. cars). "20 Similarly, Elmer Johnson states,
"The case for government subsidization [of transit] rests primarily on equity
considerations: on the ground that a base level of urban accessibility is a primary good
(i.e., one that is instrumental to the pursuit of diverse ends) that should be provided to
the less advantaged at below cost. ''2 1
Robert Cervero identifies mobility for the carless and poor, and transportation option
value in general, as the primary benefits of transit service in the United States. 22 He
concludes that environmental, energy, economic development and congestion relief
benefits are minor except in large cities. 23 Of 5,085 transit and special mobility services in
the U.S ., only 787 are large urban transit systems, 1,077 are small urban or rural transit
19
1992 Transit Fact Book, American Public Transit Association (Washington DC) p.45.
20 Jim Lazar, private communications, January 1994.
21 Avoiding the Collision of Cities and Cars, American Academy of Arts and Sciences, 1993, p. 34.
22 Robert Cervero, "Perceptions of Who Benefits From Public Transit," TRR , #936, 1982, pp. 15-19.
23
Some transit services, such as suburban oriented commuter rail, may provide little equity benefit, and
are justified primarily for avoiding congestion at large employment centers. See David Hodge, Social
Impacts ofUrban Transportation Decisions: Equity Issues, The Geography of Urban Transportation
(Susan Hanson Ed.), Guilford Press, 1986.
Page 3.9-6
Transportation Cost Analysis
systems, and 3,222 are non-profit elderly and disabled service providers.24 These smaller
systems also tend to have the highest percentage of public subsidies per rider mile. These
subsidies demonstrate the importance of transport equity and option value.
Even with these subsidies, transit and special mobility services only partially compensate
for the discrepancy between drivers and non-drivers. In most communities transit service
is infrequent, limited in time and location, sometimes uncomfortable, and carries a social
stigma. Bus and train service between cities is even worse, often taking several times as
long and costing much more than driving. Thus, the unmet demand for transport equity
and option value is greater than what is reflected in transit subsidies. In other words,
current subsidies for public transportation services at its existing quality imply that society
would be willing to pay even more for significantly better non-automotive transportation.
Is this a cost of driving? Applying a with-and-without test, one can ask: "If automobile
use decreased, would the quality and choice of non-automotive transport improve, with
benefits to disadvantaged people?" Studies indicate that provisions for automobiles (per
capita road and parking space) and per capita annual mileage are inversely correlated to
transit service quality and the use of non-automotive travel. 25 If current drivers switched
to other travel modes such as transit, ride sharing, bicycling, and walking, the increased
demand would probably improve the quantity and quality of transport available to nondrivers due to economies of scale and increased political support.
24
25
1992 Transit Fact Book, American Public Transit Association (Washington DC), pp. 16-23.
Peter Newman and Jeff Kenworthy, Cities and Automobile Dependency, Gower Press, 1989, p.38, 54.
Page 3.9-7
Transportation Cost Analysis
Comedy Becomes Tragedy
Acity slicker driving a fancy car pulls up to an old farmer along a Maine back road.
"How do I get to Muggsville? 11 asks the out-of-town driver.
The farmer ponders for a moment then answers,
11
'Fraid you can't get there from here. 11
Asmall joke, but consider a minor variation: The same question is asked by a bus rider,
bicyclist or pedestrian, and the same answer given. Comedy becomes tragedy because in
practice non-drivers frequently cannot get where they want to go, at least with any degree
of ease, safety and economy. Bicycling and walking on many roadways is unsafe or even
illegal. Many small communities have no public transportation service, or the service is
irregular and slow. Intercity bus and train service typically take twice as long as driving a
private car and costs several times as much in out-of-pocket expenses.
These are not necessary conditions of a modern transportation system. Many countries
have transportation and land use patterns that provide more travel choices, and much
better transportation services for non-drivers. The extreme degree of automobile
dependency in most North American communities results from public policies that were
motivated by political manipulation as much as economic efficiency or other social goals.26
The U.S. Congress recognized the problems caused by automobile dependency in the
Intermodal Surface Transportation Efficiency Act, which is based on the premise that
developing in a wider range of travel modes, including walking, bicycling, buses and rail
transit, will increase the efficiency and equity of the U.S. transportation system.
26
Stephen Goddard, Getting There , Basic Books (NY), 1994.
Page 3.9-8
Transportation Cost Analysis
Estimates: Although there are many indications that this cost exists, no quantified
estimates have been found . One approach for measuring these costs is based on current
U.S. financial transit subsidies, which total approximately $10 billion annually. Assuming:
1. Indirect subsidies including tax exemptions, special facilities such as bus pullouts, and
road wear equal 10% of financial subsidies.
2. Two thirds of transit subsidies are justified on the basis of transportation equity and
option value (to put this another way, society would maintain 213 of current subsidies
if equity and option value where the only benefits transit provided.)
3 Transit only captures 1/2 of all transportation equity and option value demand (in
other words, society would be willing to double existing subsidies if transit provided
the same quality of service as personal automobiles).
4 Driving is 75% responsible for the current lack of transport equity and option value.
Results: Automobiles' share of reduced transport equity and option value = $10 billion x
1.1 x 0.66 x 2 x 0.75 = $10 billion I 2,300 billion annual miles= $0.005 I vehicle mile.
Because so little research is available to help quantify this cost, this estimate is extremely
uncertain. The high cost per trip of special mobility services that are justified specifically
for equity value, and survey results described later in Section 4.5 indicate that this estimate
of equity and option value may significantly understate the true cost.
Variability: This cost is greatest in communities with the greatest degree of automobile
dependency in terms of land use, transportation options, and social patterns. Although
transit and special mobility services receive the greatest attention as ways to improve
transport equity and option value, other modes, including ride sharing, bicycling, and
walking may be equally important in many circumstances. Transit systems that are
oriented toward upper-income commuters may provide little or no equity value.
Telecommuting can provide transport equity and option benefits, although it has
Page 3.9-9
Transportation Cost Analysis
sometimes been criticized by labor organizations as a threat to job equity if it is required
by employers.
Conclusions: Although transportation equity and option value can be demonstrated both
theoretically and empirically to have value, and a lack of these attributes incurs various
costs, there are currently no models that measure them or determine how much
automobile use is responsible.27 The estimate developed above based on transit subsidies is
probably low but will be used until better methods are developed. It is applied to private
vehicles, but not to van pools, bus, trolley, bicycle, walk or telecommute, which are viable
alternatives for non-drivers.
nesr uuess
Equn:v ana upnon vame Losts (.uonars per vemc1e !YUle)
Vehicle Class
Urban Peak
Urban Off-Peak
Rural
Avera2e
0.005
0.005
0.005
0.005
0.00
0.00
0.00
0.005
0.00
0.00
0.00
0.005
0.005
0.005
0.005
0.00
0.00
0.00
0.005
0.00
0.00
0.00
0.005
0.005
0.005
0.005
0.00
0.00
0.00
0.005
0.00
0.00
0.00
0.005
0.005
0.005
0.005
0.00
0.00
0.00
0.005
0.00
0.00
0.00
Average Car
Fuel Efficient Car
Electric Vehicles
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Motorcycle
Bicycle
Walk
Telecommute
I
Automobile Cost Range: Due to the uncertainty of this cost, its minimal value is zero
and the maximum is somewhat arbitrarily set at an order of magnitude larger than the
estimate developed above.
Minimum
$0.00
Maximum
$0.05
27 For additional evidence of the existence of transportation and equity values see survey results discussed
in Section 4.5.
Page 3.9-10
Transportation Cost Analysis
3.10 Air Pollution Costs
Definition: Costs of air pollution caused by motor vehicle use.
Description: Motor vehicles produce a number of air pollutants, including carbon
monoxide (CO), particulates (PM), nitrogen oxides (NOx), volatile organic compound
(VOCs, also called hydrocarbons, or HC and reactive organic compounds or ROG), sulfur
oxides (SOx), carbon dioxide (C0 2), methane (CH4), road dust, and toxic gases such as
benzene. These have a variety of negative effects including human illness, disability and
deaths, crop and material damage, global warming, ozone depletion, acid rain, reduced
visibility, and increased cleaning costs. Motor vehicle's share of some major pollutants is
shown in figures 3.10-1 and 3.10-2.
Transport Contribution to Air Pollution 1
Figure 3.10-1
111%
•u SA
•southern California
•europe
K
CJFrance
•u
: ""
E
>
~
t
""
•&
<
.1:e .. ,
~
1-
E
•~
l
""
'"
co
voc
NOx
sox
Em lsslon
Transport activities are the largest overall source of many harmful emissions, especially
in urban areas where they impose the greatest cost due to population density.
1 Ken
Small and Camilla Kazimi, "On the Costs of Air Pollution from Motor Vehicles," Journal of
Transport Economics and Policy, January 1995, Table 1.
Page 3.10-1
Transportation Cost Analysis
Figure 3.10-2
Motor Vehicles Contribution to U.S. Air Pollution2
Ill
• Other Transport
• Roadway- Diesel
• Roadway- Gasoline
5100%
'iii
.!!
E
80%
'ii
60%
w
..
.......
..
0
40%
0
c
cu
...cu
u
0.
20%
0%
co
NOx
VOCs
C02
PM
SOx
Transportation, especially roadway travel, is a major contributor to air emissions. These
percentages are higher in urban areas where pollution problems are greatest.
Discussion: Air pollution is one of the most often cited external costs of motor vehicle
use. Estimating this cost requires information about the relationships between driving,
emissions, distribution, and impacts. Pricing this cost requires placing dollar values on
human mortality, morbidity, discomfort, loss of recreation, aesthetic degradation, damages
to crops, wildlife, materials and increased cleaning. Many studies focus on human health
impacts, but research indicates that other air pollution costs, including global warming and
aesthetic damage may also be significant. One study, for example, estimated U.S . aesthetic
costs of smog to be $7.9 billion annually in 1982, worth about $11 .5 billion in current
dollars ($0.005 per vehicle mile).3 Some estimates of global warming rank C02 emissions
as the highest automobile air pollution cost.4
Increasingly sophisticated vehicle engine controls and fuel changes mandated by state and
federal laws have reduced running tailpipe emissions for hydrocarbons by 91%, CO by
96%, and NOx by 85% since 1970.5 The 1990 U.S . Clean Air Act Amendment requires
2 Homburger,
Kelland Perkins, Fundamentals of Traffic Engineering, 13th Edition, Institute of
Transportation Studies, UCB (Berkeley), 1992, p. 30-3, 32-1.
3 Robert Crandall, et al, Regulating the A utomobile, Brookings Institute (Washington DC), 1986.
4 William Cline, The Economics of Global Warming, Institute oflntemational Economics, 1992.
5 Jon Kessler and William Schroeer, Meeting Mobility and Air Quality Goals, USEPA, Oct. 1993
Page 3.10-2
Transportation Cost Analysis
additional emission reductions through the year 2004. These include a 63% reduction in
NOx from diesel trucks and buses, a 84% reduction in truck particulates, and a 92%
reduction in urban bus particulates starting with 1994 models.6 However, since catalytic
converters are only effective when hot, significant tailpipe emissions still occur during the
first few miles (±5 miles) of operation while the vehicle warms up, and "hot-soak"
emissions occur after the engine stops.7 Emissions also occur while vehicles sit unused
(diurnal emissions), and during petroleum processing.
Several approaches are used to calculate air emission unit costs. Some estimates are based
on damage costs.8 Others use control costs, based either on the cost of emission control
equipment or the price needed to reduce emissions to an established goal, such as by
charging an emission tax.9 The cost of growing forests to sequester carbon is sometimes
used to estimate C02 costs, but that approach is inappropriate for large scale policy
analysis due to limits oftotal potential reforestation. Table 3.10-1 shows air emission unit
values used by several researchers and organizations.
Laote
Air
~.1u-1
ronunon lYionenzanon vames
\:111~~,
Source
C02
PMlo
NOx
CA Energy Commission, in state
CA Energy Commission, outside state
KPMG, B.C. Lower Mainland
0.01
0.01
NA
0.024
0.06 to 0.13
0.02 to 0.058
0.011 toO.ll
0.046
9.79
1.00
4.27
4.40
3.60
2.08
NA
3.97
14.56
1.50
3.20
7.15
0.64
2.06
2.20
4.89
MA Dept. of Public Utilities
Miller and Moffet
Pace University
Greene and Duleep
Average
per KitogramJ
SOx
voc
14.44
1.50
7.93
1.65
1.00
5.09
NA
5.15
4.14
0.38
2.93
5.83
3.60-7.20
NA
3.68
2.43
This table shows air emission unit costs developed by various agencies and researchers.
6 John Schiavone, Retrofit of Buses
to Meet Clean Air Regulations, Synthesis of Transit Practice 8, TRB,
National Academy Press (Washington DC), 1994.
7 Homburger, Kell and Perkins, Fundamentals of Traffic Engineering, 13th Edition, Institute of
Transportation Studies, UCB (Berkeley), 1992, p. 30-10 ..
8 William Nordhaus, "Economic Policy in the Face of Global Warming" Energy and the Environment in
the 21st Century, MIT Press (Cambridge), 1990; William Cline, Economics of Global Warming, Institute
of International Economics, 1992
9 For example see Per Kageson, Getting the Prices Right, European Federation for Transport &
Environment (Bruxelles), 1993, p. 68-70.
Page 3.10-3
Transportation Cost Analysis
Estimates:
• Apogee Research estimated air pollution costs in Boston, MA and Portland, ME for
peak and off-peak travel at high, medium and low densities, shown in Table 3.10-2.
-
- -
-
- - -
-
- - -- -
- --- --- -- - -- -- --- - ..
Expwy
-
Non-Expwy
-----
'
.--- r· -----e-- ------,
Comm. Rail
Peak
7.9
6.6
7
Off-P
6.6
9.5
9.5
Peak
10.6
7.9
7.3
Off-P
8.9
7.3
6.9
Peak
0.9
1.0
2.0
Off-P
2.2
2.5
4.9
6.5
6.6
___
_
12.1
~-------
6.9
7.0
12.1
7.9
7.3
6.6
7.3
6.9
6.6
n/a
n/a
n/a
n/a
n/a
n/a
Boston
High
Medium
Low
Portland
High
Medium
Rail Transit
Off-P
Peak
<0.2
<0.2
<0.2
<0.2
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Bus
Peak
0.8
2.4
2.4
Off-P
4.4
5.8
5.5
5.2
5.2
11.0
4.7
4.6
11.0
• Robert Ayres and Jorg Walter estimate market damage costs of global warming at $30
to $3 5/ton of C02 equivalent, plus significant non-market environmental damages.11
• The California Energy Commission estimates air pollution costs in that state to
average $0.012 to $0.014 per vehicle mile, based on 1992 CEC values for NOx, SOx,
ROGs, PM 10 and C0.12 In addition, they apply a climate change charge of$0.0042
($0.084/gallon of gasoline), based on $28/ton of carbon released.
Michael Cameron cites the report Exhausting Clean Air: Major Issues in Managing
Air Quality, 13 which estimates total Southern California air pollution costs at $7.4
billion annually, for an average cost of driving in the region at $.06 per VMT.14
• James Cannon concludes that total U.S . automobile emission costs are approximately
$50 billion annually; averaging $.025 per vehicle mile.15
•
William Cline estimates damage costs for a 2.5°C average temperature increase by
2050 at 1% ofU.S. GDP, which he doubled to account for non-market cost impacts.16
He estimated that a 10°C average temperature rise by 2100 would cost of6% ofGDP,
and up to 20% if non-market goods are included.
10 Apogee Research, The Costs ofTransportation, Conservation Law Foundation (Boston), 1994, p. 148.
11 Robert Ayres and Jorg Walter, "The Greenhouse Effect: Damages, Costs and Abatement,"
Environmental and Resource Economics, Vol. 1, 1991, p. 237-270.
12 1993-1994 California Transportation Energy A nalysis Report, California Energy Commission
(Sacramento), Feb. 1994, p. 29.
13 California Assembly Office of Research Exhausting Clean A ir, 1989.
14 Michael Cameron, Transportation Efficiency, 3/91 , Environmental Defense Fund (Oakland), p.21.
15 Los Angeles Times, Jan. 20, 1990, p. A18, based on "The Health Costs of Air Pollution; A Survey of
Studies Published 1984-1989", for the American Lung Association.
16 William Cline, Economics of Global Warming, Institute oflntemational Economics, 1992.
Page 3.10-4
Transportation Cost Analysis
• Convergence Research reviewed air emission unit costs used by 37 regulatory and
research sources. 17 Table 10-3 summarizes this data.
Table 10-3
Air Emission Unit Val
F
37 Resmlat
-
--- -
dR~
-- -
- - --
-
hS
-
-
- --
-
(1990 US Dollars/Ton)
CH,..
co
C02_
H 2S
N2 0
NOx
SOx
TSP
~nimurn
~mum
$100
$740
$326
$375
9
$500
$1 ,000
$842
$907
6
$2
$84
$25
$20
26
$1,800
$1,800
$1,800
$1,800
1
$3,700
$4,158
$3,880
$3,700
5
$42
$40,000
$8,212
$4,209
36
$405
$21,185
$4,0ll
$1,793
34
$167
$8,780
$3,401
$2,496
20
Average Value
Median Value
Count
• DeLuchi, Sperling and Johnson estimate total annual U.S . human health costs from
motor vehicles range from $5 -$150 billion (1993 dollars), averaging $0.002 to $0.068
per vehicle mile.18
• The Greater Vancouver Regional District estimates the emission rate per passenger for
various modes under average and peak urban conditions, shown in Table 3.10-4.
1aote J.l tJ-4
EmiSSion Kates tor :Selected Modes U rams per passenger-muey"'
Passengers
SOx
PM
HC
co
NOx
Mode
Avera2e
Automobile
CarPool
Van Pool
Diesel Bus
Articulated Diesel
Methanol Bus
Trolley Coach*
Articulated Trolley*
Rail Transit*
Rush Hour
Automobile
CarPool
Van Pool
Diesel Bus
Articulated Diesel
Methanol Bus
Trolley Coach*
Articulated Trolley*
Rail Transit*
--
1.0
2.4
5.0
20
23
20
20
32
25
3.15
1.31
0.72
O.ll
0.12
0.01
0.00
0.00
0.00
23 .57
9.82
5.42
1.50
1.67
0.02
0.001
0.001
0.001
1.91
0.80
0.44
0.67
0.74
0.49
0.006
0.007
0.006
0.07
0.03
0.02
0.09
0.10
0.00
0.00
0.10
0.00
0.10
0.04
0.02
0.17
0.19
0.00
0.00
0.00
0.00
1.3
3.6
7.2
37
44
37
37
44
53
2.42
0.88
0.50
0.06
0.06
0.01
0.00
0.00
0.00
18.13
6.55
3.77
0.81
0.87
0.10
0.001
0.001
0.001
1.47
0.53
0.31
0.36
0.39
0.27
0.003
0.004
0.003
0.05
0.02
0.01
0.05
0.05
0.00
0.00
0.00
0.00
0.08
0.03
0.02
0.09
0.10
0.00
0.00
0.00
0.00
*Electric Vehicles
• Per Kagsen estimates that NOx, VOC, and S02 costs in Europe average $0.03 per
automobile passenger mile (14.6 ECU/1 ,000 km).2o
17 Kevin Bell,
Valuing Emissions from Hermiston Generating Project, Convergence Res. (Seattle), 1994.
ofFuture Transportation Fuels, UCB (Berkeley), 1987.
19 GVRD A ir Quality Management Plan: Stage 2 Draft Report: Priority Emission Reduction Measures,
Greater Vancouver Regional District (Vancouver), May 1992, Table 5-8, p.5-43 .
18 Comparative Analysis
Page 3.10-5
VOCIROG
$340
$21,175
$5,986
$3,300
15
Transportation Cost Analysis
• Keeler provides a range of pollution impact values based on human health impacts
ranging from $.0024-.059 depending on vehicle age, with an average of$.014.21
• James MacKenzie et al. estimate motor vehicle air pollution costs to be at least $10
billion, or about $0.005 per motor vehicle mile, and global warming costs at $60 per
ton of C02 equivalent ($0.012 per mile) based on control costs.22
• Air pollution unit cost estimates by Peter Miller and John Moffet for automobile
elnissions under urban and rural conditions are shown in Table 3.10-5. 23
-
-
-
-
Urban
Rural
-
-
- - -----
----- - - -- ---- - - - -
-
---------
-----
'~ ·
--oj
C02
HC
co
NOx
TSP
so2
0.06-0.13
0.06-0.13_
7.20
3.60
12
0
0.60-8.40
0.05-0.06
0.08-0.013
0
0.01-0.36
0.0003
L ...
--
• The Office ofTechnology Assessment study estimates U.S. annual automobile air
pollution costs (including human health effects, global warming, agricultural losses,
material, visibility and aesthetic losses) to range from $47 to $242 billion, for an
external cost average of $0.02 to $0.10 per vehicle mile.24
• After reviewing various European studies Emile Quinet concludes that car transport is
about 10 times more polluting than railways for passenger transport; and truck
transport is about 10 times more polluting than rail per unit of freight transport.25
• Ken Small and Camilla Kazimi provide a comprehensive analysis of Southern
California motor vehicle air pollution impacts.26 For monetization they focus on human
morbidity from particular and ozone, and human mortality from particulates. Based on
a $4.87 million value of per statistical life and 1992 fleet mix they calculate average air
pollution costs for gasoline cars of$0.033 per VMT, with a sensitivity analysis range
of$0.014 to $0.12. These costs are expected to decline 50% by the year 2000 due to
improved emission controls. Heavy duty diesel trucks are estimated to impose $0.53
per VMT, with a sensitivity range of $0.16 to $2. 19. They state that road dust
particulates may cost an additional $0.043 per VMT, and that global warming costs
may add a comparable cost, but are not confident enough with these estimate to
20
Per Kageson, Getting the Prices Right, European Fed. for Transport & Env. (Bruxelles), 1993, p. 82.
21 Keeler, et al, The Full Costs of Urban Transportation , Institute of Urban and Regional Development
(Berkeley), Monograph #21 (Berkeley), 1975, based on 1972 estimates.
James MacKenzie, Roger Dower and Donald Chen, The Going Rate: What it Really Costs to Drive,
World Resources Institute (Washington DC), 1992, p. 13.
23 The Price ofMobility, Natural Resources Defense Council (Washington DC), Oct. 1993, p.45
24 Saving Energy in U.S. Transportation , U.S. Office of Technology Assessment, 1994, p. 108.
25 Emile Quinet, "The Social Costs of Transport: Evaluation and Links With Internalization Policies," in
lnternalising the Social Costs ofTransport, OECD (Paris), 1994.
26 Ken Small and Camilla Kazirni, "On the Costs of Air Pollution from Motor Vehicles," Journal of
Transport Economics and Policy, January 1995, pp. 7-32.
22
Page 3.10-6
Transportation Cost Analysis
include them. Professor Small has mentioned that emission in urban areas with better
air circulation probably average about 1/3 of Southern California costs. 27
• Sweden has established carbon taxes of$153/tonne for transport and residential fuels,
and $38/tonne for industrial energy, to meet emission reduction goals. 2 8 Swedish
authorities value NOx at $5 .60 per kg. The European Federation for Transport and
Environment applies the same value to hydrocarbon emissions for analysis purposes.29
• Transport Concepts estimates air pollution costs for freight as shown in Table 3.10-6.
-
-
--
Semi-Truck
B-Train Truck
Truck Average
Piggyback
Container
Box Car
Hopper Car
Rail Average
- - - . - - - -- - - - - ------
-----
- - - - --o---
---
-
. --------
voc
co1
Net Payload
Tonnes
Load Factor
Percent
NOx
24.5
44.2
65%
65%
0.28
0.23
0.061
0.050
0.38
0.31
24.5
26.3
71.7
70
60%
60%
36%
60%
0.20
0.16
0.14
0.08
0.13
0.010
0.008
0.007
0.004
0.007
0.15
0.12
0.11
0.06
0.10
Total
Canadian Cents Per Tonne Km
0.72
0.58
0.71
0.36
0.29
0.25
0.15
0.23
• A Union of Concerned Scientists study compares lifetime emissions for new standard
and ultra low emission vehicles (ULEV), and an electric vehicle, based on Southern
California electrical generation mix, shown in Table 3 .10-7.31 About half of electrical
generation emissions produced to power urban electric vehicles occur in urban air
sheds, and about half occur in other regions where unit pollution costs are lower.
-
Pollutant
ROG
co
NOx
PM 111
SOx
Carbon
--- - - - - - - -
- - ---------------- - - - - - -
Average Gasoline
89-119
531-1,072
110-121
2.5
11 .8
19,200
ULEV Gasoline
46-54
198-478
60-66
2.5
11.8
19,200
27
-
-
- --
Electric
0.49
2.76
24.28
1.11
13.8
5,509
At FHWA Colloquium on the Social Costs of Transportation, Washington DC, 12 December 1994.
Getting the Prices Right, European Fed. for Transport & Env. (Bruxelles), 1993, p. 69.
29 Per Kftgeson, Environmental Car Guide 1994/95, Eur. Fed. for Transport & Env. (Bruxelles), 1994.
30 External Costs ofTruck and Train , Transport Concepts (Ottawa), October 1994, p.22.
31 Roland Hwang, et al., Driving Out Pollution: The Benefits ofElectric Vehicles, UCS (Berkeley), 1994.
28 Per Kftgeson,
Page 3.10-7
Transportation Cost Analysis
• USEPA tests show new motorcycles produce over double HC and CO, and higher
NOx than automobile fleet averages, since they lack emission control equipment. 32
• Wang and Santini estimate electric vehicles reduce CO and VOC emissions 98%, with
smaller reductions in NOx and SOx, and 50% reductions in C02 emissions.33
Variability: Automobile air pollution costs vary tremendously depending on where and
when it is used, the type and age of vehicle, the fuel that is used, and how it is driven. A
significant portion of driving incurs minimal local air pollution costs, while emissions in
polluted areas incur extremely high costs. Emissions that contribute toward global
warming, ozone depletion, and acid rain have costs no matter where they occur.
Variation between common vehicle classes is shown in Figure 3.10-3 . Older vehicles
without catalytic converters and those that are not properly adjusted have much greater
emissions per mile than average. These differences will be reduced somewhat in the next
few years as older cars are retired and more areas implement inspection and maintenance
programs. Catalytic converters are inefficient when cold, so emissions are much greater
during the first few miles of a trip. Stop and go driving increases emissions per mile.
Figure 3.10-3 Average HC, CO, and NOx Emissions from Selected Vehicle Classes34
3
j!
2
~
Ill
E
I•Hc I
•Nox
~
...
T
•co
1
I!
(!)
0
SubMidsize
compact
Car
Car
Fullslze
Car
Light
Truck
32
112 Ton
Truck
Compact
Van
Fullslze
Van
Compilation ofAir Pollution Emission Factors; Vo/.11, USEPA, 1/91, tables 1.8.1, 1.8.4.
Quanlu Wang and Danilo Santini, "Magnitude and Value of Electric Vehicle Emissions Reductions for
Six Driving Cycles in Four U.S. Cities," Transportation Research Record 1416, 1993, p. 33-42.
34 Gil McCoy, John Kim Lyons and Greg Ware, A Low Emission Vehicle Procurement Approach for
Washington State, Washington State Energy Office, #92-071 (Olympia), June 1992, Table 5-2, p 32.
33
Page 3.10-8
Transportation Cost Analysis
Conclusions: Air pollution costs are substantial. Pollution control equipment has reduced
tailpipe emissions per VMT, but increased driving and residual emissions (especially from
cold starts and evaporation) result in significant total costs. Estimates of average national
local air pollution costs range from $0.01 to over $0.08 per VMT, depending on
assumptions and data used. Some studies underestimate total costs because they include
only human health impacts and ignore other costs. Adding global warming, acid rain, crop
damage, ozone depletion, and aesthetic damage would increase these estimates.
For this analysis, Urban Peak local air pollution is estimated to cost $0.07 per VMT,
which is slightly lower than Miller and Moffet's high estimate and slightly above Cameron's
overall average estimate for Southern California. Urban Off-Peak costs are estimated at
$0.05 per VMT, which represents the middle of the total range of estimates, and rural
driving air pollution costs are estimated to be an order of magnitude lower at $0.005 per
VMT, based on Miller and Moffet's values shown in Table 3.10-5. In addition, MacKenzie
et al's estimate that greenhouse gases incur a $0.012 per mile cost is applied to all driving,
equal to $60/t C02, representing the middle-high range of current cost estimates.
Using these values, average automobiles are estimated to impose a $0.082 per mile cost
under Urban Peak ($0.07local + $0.012 greenhouse), $0.062 under Urban Off-Peak
($0.05 local+ $0.012 greenhouse), and $0.017 under Rural driving conditions ($0.005
local+ $0.012 greenhouse). Energy efficient cars are estimated to have local emissions
10% lower than an average car, and half the global warming costs. Electric vehicles are
estimated to produce 5% of local emission costs, and 50% of global warming costs.35
Vans are estimated to produce 80% more air pollution than an average automobile.
35 While there is less difference in cost per unit of emission between urban and rural driving for electric
vehicles, (since generators are often located outside of urban areas), stop-and-go urban driving produces
greater emissions per unit of travel, resulting in somewhat higher costs for urban driving.
Page 3.10-9
Transportation Cost Analysis
Motorcycles are estimated to produce twice the local air pollution of a standard
automobile, and half the greenhouse gas.
Rideshare passengers incur an air pollution cost 2% of a van based on a 20% emission
increase for 10 passengers. In the past, buses produced local air pollution costs 10 to 15
times higher per vehicle mile than an automobile, due primarily to the high NOx and
particulate output of diesel engines. This will decrease 75% or more in urban areas as
strict 1995 emission control standards are implemented, so an estimate cost 2.5 times
greater than an average automobile mile is used to represent current and near future local
emissions, and greenhouse gas costs are 3 times higher based on fuel consumption.
Electric trolleys and urban buses are estimated to have air pollution five times greater than
an electric car. bicycling, walking, and telecommuting have no air pollution costs.
--
-
- ---
-
------ - -
Vehicle Class
Average Car
Fuel Efficient Car
Electric Vehicles
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Motorcycle
Bicycle
Walk
Telecommute
--
- ---
Urban Peak
0.082
0.069
0.010
0.148
0.002
0.210
0.050
0.146
0.00
0.00
0.00
-
- - ---
-
---
-
Urban Off-Peak
0.062
0.051
0.009
0.112
0.001
0.161
0.045
0.106
0.00
0.00
0.00
Rural
0.017
0.011
0.006
0.03
0.001
0.061
0.030
0.016
0.00
0.00
0.00
Averae:e
0.048
0.039
0.008
0.058
0.001
0.131
0.040
0.078
0.00
0.00
0.00
To test these estimates, average automobile air pollution costs are multiplied by mileage:
Annual Mileage (billion)
Urban Peak
Urban Off Peak
Rural
Total
460
920
920
Page 3.10-10
Estimate
Total (billion)
$0.082
$0.062
$0.017
$37.7
$57.0
$15 .6
$110.3
Transportation Cost Analysis
This total is within the range of many of the estimate described earlier. It represents a reasonable
estimate of automobile air pollution costs, especially when all impacts (particulate,
aesthetic, ozone depletion, emissions during petroleum processing, and global warming)
are considered.
Automobile Cost Range: The minimum value estimate is based on the lower estimates
described. The maximum is a combination of the highest local air pollution estimate plus
the maximum estimate of carbon global warming costs.
Minimum
$0.01
Page 3.10-11
Maximum
0.20
Transportation Cost Analysis
3.11 Noise
Definition: Unwanted sounds and vibrations produced by motor vehicle use.
Description: Motor vehicles cause a variety of noises and vibrations. Traffic noise
includes engine acceleration, tire/road contact, braking, and horns. Vibration and
infrasound (low frequency noise) are produced by heavy vehicles.
Measuring Noise
Noise is measured in decibels (dB), a logarithmic scale. A 10 dB increase represents a
doubling in noise level. Decibels A-weighted, "dB(A)" units emphasize the frequency
sensitivities of human hearing, and correlate well with subjective impressions of loudness.
Common noise levels range from 30 to 90 dB(A) .1 Decibels are an instantaneous
measurement, so various indexes are used to measure noise over a period oftime:
• Leq represents the equivalent continuous sound level in dB(A) or a period over which
the measurement is taken, usually 8 hours. Leq (8 hours) is used in many traffic noise
standards established by OECD and WHO.
• Lwrepresents the noise level in dB(A) that is exceeded for 10 percent ofthe time over
a one hour period. Analogous measurements, Lob L 05 , L 50 , refer to noise levels
exceeded 1, 5 and 50% ofthe time over a one hour period. L 10 (18 hours) is the mean
ofthe hourly values taken over an 18 hour period, which, is typically from 6 a.m. to
midnight. Lw is often used to define traffic noise in the U.S . and other countries.
• MNL (Maximum Noise Level) is the loudest noise during a certain period. This index
is considered by some researches to correlate with noise annoyance better than Leq
and L 10, but does not address the number of noise events, and is not widely used.
Discussion: According to an OECD report, "Transport is by far the major source of
noise, ahead of building or industry, with road traffic the chief offender. "2 Trucks, buses,
and motorcycles are major contributors to traffic noise.3 At low speeds most noise comes
1 BTCE
& EPA, "The Costing and Costs of Transport Externalities: A Review," Victorian Transport
Externalities Study, Vol. 1, Environment Protection Authority (Melbourne, Australia), 1994.
2 Environmental Policies for Cities in the 1990s, OECD (Paris), 1990, cited in Poldy, p.29.
3 MacKenzie, Dower & Chen, The Going Rate, World Resources Institute (Washington DC), 1992, p. 21.
Page 3.11-1
Transportation Cost Analysis
from vehicle engine and drivetrain, at higher speeds aerodynamic and tire/road noise
dominate.4 Overall traffic noise increases with speed, density, stops (which cause
increased accelerations), and portion oflarge trucks and motorcycles.
Several studies show an average reduction in residential property values of about 0. 5% for
each unit change in Leq.5 Various researchers have used these results to develop general
property value depreciation indexes, some ofwhich are shown in Table 3.11-1. The
OECD recommends a noise depreciation index of0.5% of property value per decibel
increase if noise levels are above 50 dB(A) Leq (24 hours).6 Douglass Lee estimates traffic
noise costs at $21 annually per housing unit per decibel increase.7
Table 3.11-1 NoiseD
· tion Estimat"'"s
Percent House Price Reduction Per dB(A)
Country
Above 50 to 65 dB (A) Threshold
France
Netherlands
Norway
Switzerland, Bas1e
Canada, Toronto
United States
OECD
0.4
0.5
0.4
1.26
1.05
0.15-0.88
0.5
The number of residences impacted by traffic noise is significant in most developed
countries. A.L. Brown and K. C. Lam estimate that approximately 25% of Australian
urban dwellings are located on roads with over 2, 000 vehicles per day and higher traffic
speeds. Over 12% of dwellings in Australia directly front roadways carrying 8, 000 or
more vehicles per day. In addition, 8% of houses on low volume (<1,000 vehicles per day)
4 Hornberger, Kell and Perkins, Fundamentals of Traffic Engineering, 13th Edition , Institute of
Transportation Studies, UCB (Berkeley), 1992, p.31-3 .
5 From Pearce and Markandya, Environmental Policy Benefits: Monetary Valuation, OECD (Paris), 1989.
6 M. Modra, Cost-Benefit Analysis of the Application ofTraffic Noise Insulation Measures to Existing
Houses, EPA (Melbourne), 1984, cited in Poldy, 1993.
7 Douglass Lee, "Efficient Highway User Charges," USDOT, as cited in MacKenzie.
8 Based on Weatherall1988; Quinet 1990; and Steeting 1990 as cited in BTCE & EPA, "The Costing and
Costs of Transport Externalities: A Review," Victorian Transport Externalities Study, Vol. 1,
Environment Protection Authority (Melbourne), 1994.
Page 3.11-2
Transportation Cost Analysis
are located close enough to a high traffic road to experience traffic noise exceeding 68 dB .
Thus, approximately 1/3 of houses experience significant traffic noise. 9
rt Noise Costs 1o
Table 3.11-2 Selected Estimates of Total T
Country
Percent of GDP
Finland
France
Germany
Norway
United Kingdom
United States,
Japan
OECD, Average
0.3
0.24
0.20
0.23
0.50
0.06- 0.21
0.20
0.15
Table 3.11-2 shows various estimates of total national transportation noise costs as a
percentage of GDP . Some researchers suggest that property value depreciation due to
noise is non-linear, and increases from 0.5 per dB( A) in the range of 50 to 60 dB( A),
rising to 0.8 percent above 65 dB(A).ll
Some researchers point out that hedonic pricing studies only measures a portion of total
noise costs. It does not measure impacts on non-residential environments and ignores
residual noise below a set standard, such as 50 dB. Erik Verhoef estimates that such
estimates of traffic noise represent only l/8th of the total cost 12 and Peter Bein interprets
Srelensrninde's research to imply that hedonic noise surveys identify only about 1/6th of
total motor vehicle noise costs.13 Since most of the estimates cited above are based on
direct hedonic pricing, they are likely to significantly underestimate total noise costs.
9 A. L. Brown and K.C. Lam, "Can I Play on the Road, Mum?- Traffic and Homes in Urban Australia,"
Road and Transport Research, Vol. 3, No. 1, March 1994, p. 12-23 .
10 Based on Bouladon 1991 and Quinet 1990, as cited in BTCE & EPA, "The Costing and Costs of
Transport Externalities: A Review," Victorian Transport Externalities Study, EPA (Melbourne), 1994.
11 BTCE & EPA, "The Costing and Costs of Transport Externalities: A Review," Victorian Transport
Externalities Study, Vol. 1, EPA (Melbourne), 1994, Table 3.4, based on Weatherall, 1988.
12 Erik Verhoef, "External Effects and Social Costs ofRoad Transport," Transportation Research, Vol.
28A, 1994, p. 286.
13 Barnet Hastings Benefit Cost Analysis, BC Ministry of Transportation and Highways (Victoria), 1994.
Page 3.11-3
Transportation Cost Analysis
Estimates:
• Apogee Research estimated noise costs in Boston, MA and Portland, ME for several
modes at high, medium and low densities. Totals are shown in Table 3.11-3.
~
-
-
-
-
- - --
- - - - - - -- -
Expwy
Boston
High
Medium
Low
Portland
High
Medium
Low
· ·-
------
Non-Expwy
0.3
0.1
<0.1
0.6
0.2
<0.1
0.2
0.1
<0.1
0.5
0.1
<0.1
r
.-------o-- ------,
Comm. Rail
Rail Transit
Peak
Off-P
Peak
Off-P
0.4
1.1
n/a
n/a
0.3
0.3
0.4
0.1
0.1
0.1
n/a
n/a
.---
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Bus
Peak
0.5
0.2
<1.0
Off-P
1.3
0.5
0.1
1.1
1.0
0.2
0.1
0.2
0.1
Per Kageson estimates motor vehicle noise costs in Europe at $0.006 per passenger
mile (3 .0 ECU/1,000 km).1s
• Theodore Keeler et al. estimate the marginal noise cost of an added freeway vehicle
mile at $.001-2 in 1975 ($.002-4 current dollars), but offer no estimate for impacts on
local streets, which they state would be considerably higher. 16
• Brian Ketcham estimates average U.S . automobile noise costs at $.001 per vehicle
mile, and noise costs for heavy vehicles average $.04 mile. He also estimates that
ground vibrations by heavy vehicles are responsible for half of urban building
structural damage costs, equal to $.06 per mile. 17
• Peter Miller and John Moffet estimate noise costs at $0.0014 to 0.0023 per automobile
mile and three times higher for buses.l8
• James MacKenzie et al. used estimates developed by Hokanson for the U.S . DOT to
calculate total U.S . noise costs to be $9 billion annually, about $0.004 per VMT. 19
• Transport 2021 estimates noise costs in the Greater Vancouver area equals $0.005
Canadian per km, or about $0.006 U.S. per mile. 20
14
Apogee Research, The Costs ofTransportation, Conservation Law Foundation (Boston), 1994, p. 161.
15 Per Kageson, Getting the Prices Right, European Fed. for Transport & Env. (Bruxelles), 1993, p 102.
16 The Full Cost of Urban Transportation, Institute of Urban and Regional Development (Berkeley),
Monograph #21 1975, p. 52.
17 Making Transportation Choices Based on Real Costs, Konheim & Ketchem (NY), Oct. 1991
18 The Price ofMobility, National Resources Defense Council (Washington DC), Oct. 1993, p.35.
19 James MacKenzie, Roger Dower and Donald Chen, The Going Rate, World Resources Institute
(Washington DC), 1992, p. 21.
2 Cost ofTransporting People in the British Columbia Lower Mainland, GVRD (Vancouver), 1993.
°
Page 3.11-4
Transportation Cost A nalysis
• Saelensminde uses previous studies to estimate noise costs for Norway, resulting in a
range from $88 to $541 per capita annually, or about $0.01 to $0.054 per VMT.21
• Nils Soguel describes a Swiss survey in which residents indicated a willingness to pay
an average of70 francs (about US$55) per month to reduce traffic noise by half. 22
Several statistical strategies were used to minimize survey bias.
• The STAMINA model calculates the relative noise costs oftrucks and automobiles.23
It indicates that one heavy truck produces the same amount of noise as 63 automobiles
at 50 kmlhr, but at 100 kmJhr this decreases to 25 cars per truck noise equivalent.
Medium size trucks produce noise equivalent to 2 to 16 cars, depending on speed.
• The Washington State Department of Transportation uses a formula for calculating
maximum investments in noise reduction that yields values ranging from $5,500 to
$20,000 per exposed household, depending on noise level reduction.24
• The U.S . FHWA estimates noise costs at $.002 per vehicle mile.25
• A U.K. study found a high level of complaint and concern over traffic vibration.26
Along roads with 500 or more vehicles per hour during peak periods, over 50% of
residents are bothered by traffic vibration. However, field studies involving induced
vibration in typical residential structures, and case studies showed only minimal and
superficial structural damage that is likely to be caused by motor vehicle vibration
Variability: Noise impacts vary by vehicle type, vehicle condition, where it is driven, and
when it is driven. Automobiles are generally quieter than either buses or motorcycles.
Electric vehicles produce moderate motor noise at low speeds, and the same level of
wheel noises as a gasoline vehicle, which is the primary source of noise at higher speeds.
Noise costs are higher in urban areas, where there are more human ears, but this difference
is not as great a might be expected, since the impact of a single vehicle in rural areas has a
21 Kjartan Saelensminde, Environmental Costs Caused by Road Traffic In Urban A reas -Results From
Previous Studies, Institute for Transport Economics (Oslo), 1992.
22 Measuring Benefits from Traffic Noise Reduction Using A Contingent Market, Center for Social and
Economic Research on the Global Environment (London), 1994
23 L.R. Rilett, Allocating Pollution Costs Using Noise Equivalency Factors, Transportation Research
Board 1995 Annual Meeting (Washington DC), Paper 950938.
24 Directive 22-22 Noise Evaluation Procedures for Existing State Highways, WSDOT (Olympia), 1987.
Described in March 10, 1994 correspondence from Timothy Coats, WSDOT Traffic Noise Engineer.
25 Douglass Lee, Efficient Highway User Charges- Federal Highway Cost A llocation Study, Appendix E,
FHWA (Washington DC), May 1982.
26 G.R. Watts, Traffic Induced Vibrations in Buildings, TRRL Report #246, (Crowhome), 1990.
Page 3.11-5
Transportation Cost Analysis
greater cost than an additional vehicle added to urban traffic. Noise also impacts wildlife,
which implies additional environmental costs in addition to impacts on humans.
Conclusions: Several studies place average automobile noise costs at $0.001 to $0.02 per
VMT, with higher costs for larger vehicles. Most studies underestimate total costs by
relying on hedonic price surveys without scaling for non-residential and residual impacts.
In other words, if people are willing to pay to avoid the worst traffic noise in their homes,
they should be willing to pay more to completely eliminate traffic noise in all situations.
For this reason these cost estimates can be increased by 2 to 8 times. More research is
needed to better determine true total noise costs.
Automobile and van pool noise costs are estimated here at $0.01 per mile on urban roads
and rural $0.005 on rural roads, based on existing cost estimates increased to take into
account non-residential and residual costs. Electric cars are estimated to produce 30% of
the noise cost of an automobile under urban conditions, and 60% during higher speed rural
driving. Diesel bus noise is estimated to be 5 times greater than an automobile. Electric
bus and trolley noise are estimated to be 3 times greater than an automobile, and
motorcycles are estimated to be 10 times greater than an automobile. Ride share
passengers, bicycling, walking and telecommuting incur no noise costs.
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
Urban Peak
Urban Off-Peak
Rural
Avera2e
0.010
0.010
0.003
0.010
0.00
0.050
0.030
0.100
0.00
0.00
0.00
0.010
0.010
0.003
0.010
0.00
0.050
0.030
0.100
0.00
0.00
0.00
0.005
0.005
0.003
0.005
0.00
0.025
0.015
0.050
0.00
0.00
0.00
0.008
0.008
0.003
0.008
0.00
0.04
0.024
0.08
0.00
0.00
0.00
Page 3.11-6
Transportation Cost Analysis
Automobile Cost Range: These are based on estimates cited above.
Minimum
$0.002
Page 3.11-7
Maximum
$0.06
Transportation Cost Analysis
3.12 External Resource Consumption Costs
Definition: External costs of resources consumed by vehicle production and use.
Description: Automobile construction and use consume approximately 14% ofU.S.
aluminum, 34% of iron, 11% of steel, 71% oflead, 67% ofrubber production, and over
50% of petroleum, which represents more than 20% of all energy consumption.1 Actual
resource consumption is even higher than these figures indicate because the U .S. imports
five vehicles for each one it exports,2 so additional resources are consumed in other
countries to produce our cars.
Resource consumption is not necessarily a problem, but the price consumers pay does not
cover all costs, including environmental impacts and various industrial subsidies. Iron and
steel production involve extensive land use impacts from mining of both ore and coal, and
produce significant air pollution and solid waste. Aluminum production involves mining,
and consumes large amounts of cheap energy. Lead mining and productions produce
hazardous wastes. Petroleum extraction, transport, and processing impose environmental
impacts, dependency on foreign markets resulting in trade imbalances, military costs to
maintain market access, reduction of non-renewable resources available for future
generations, and tax subsidies. The majority of automobile ferrous metals and some other
materials are eventually recycled, 3 but reprocessing still involves substantial energy
consumption and pollution, and has not eliminated the need for mining.
1 Facts
and Figures '93, American Automobile Manufactures Association (Detroit), 1993, p. 49 and 84;
Homburger, Kelland Perkins, Fundamentals of Traffic Engineering, 13th Edition, Institute of
Transportation Studies, UBC (Berkeley), 1992, p. 32-1 .
2 Facts and Figures '93, American Automobile Manufacturers Association (Detroit), 1993, p. 48.
3 Facts and Figures '93, American Automobile Manufacturers Association (Detroit), 1993, p. 50.
Page 3.12-1
Transportation Cost Analysis
Although production, and therefore consumption of these and other resources impose
social costs, the most studied, and probably the greatest overall, are the external costs
associated with energy, especially petroleum use. Energy externalities are therefore the
focus of this section, and are used as a reference for other external resource costs.
Discussion: Motor vehicles are a major consumer of energy. In addition to propulsion
energy, the energy equivalent of 400 gallons of oil or 5,600 kwh in electricity is embodied
in the production of a typical automobile, which represents more than 10% of its typical
lifetime energy use. 4 Petroleum imposes various externalities.5 Peter Miller states:
"money spent at the pump falls significantly short of the true cost of energy used by
the automobile due to the many externalities associated with oil production and use.
Health and environmental costs include the destruction of natural habitat; water
pollution; air pollution; greenhouse gas emissions; and the clean-up and habitat-loss
costs of oil spills that are not directly paid by oil companies. Domestic oil exploration
and extraction are subsidized by means of tax credits and other government
incentives. And imported oil comes at the expense of multi-billion dollar expenditures
of the Naval and Rapid Deployment Forces to protect U.S. shipping and our oil
interests. "6
The external costs of energy consumption are reflected in the variety of publicly supported
efforts to increase national energy efficiency. According to researchers John DeCicco and
Marc Ross energy consumption imposes broad costs to the national economy. They state,
''Money spent on oil imports is mostly lost to the U.S. economy, and gasoline purchases
provide relatively few jobs per dollar spent. ''7 David Greene and K. G. Duleep estimate
petroleum externalities in a benefit/cost analysis of improving U.S. automobile fleet fuel
4 Warner
and Glenys, Canadian Green Consumer Guide , Pollution Probe Foundation (Toronto), 1991.
"Preliminary Estimates of Cumulative Private and External Costs of Energy,"
Contemporary Policy Issues, Vol. VIII, July 1990, pp. 283-307.
6 Peter Miller and John Moffet, The Price ofM obility, NRDC (Washington DC), Oct. 1993, p.19.
7 John DeCicco & Marc Ross, "Improving Automotive Efficiency," Scientific A merican, Dec. 1994, p. 56.
5 Darwin Hall,
Page 3.12-2
Transportation Cost Analysis
economy.8 Their study, and similar analyses by the California Energy Commission, 9
include these external costs of imported oil:
• Oil price benefits: Because North America consumes over 25% of total world oil
production, its demand has a monopsonistic effect. High U.S . demand increases
international oil prices (the elasticity of world oil price with respect to U.S. demand is
estimated at 0.3 to 1.1), imposing a financial cost on all oil consumers.
• Transfer of wealth via monopoly pricing: U.S . demand for imported oil raises the
economic rent paid for oil, transferring wealth to oil producers. This reduces demand
for U.S . goods and services, and lower economic growth. The price of oil over its
competitive market price (estimated at $16/BBL) is considered a cost in their analysis.
• Energy Security: Energy security includes two sets of costs: economic and national
security effects. The economic costs are the effects of sudden oil price changes on
economic growth, inflation, and employment. For example, oil price shocks in 1973
and 1979 are considered to have caused subsequent recessions and extraordinary
inflation. This is the result of the relatively long time needed for the economy to make
price, capital, and technological adjustments to price changes. Until all adjustments are
completed, the economy is inefficient and GNP growth is reduced. The second sets of
costs are associated with strategic costs, especially, military expenditures in the Persian
Gulf region. Although estimates of this cost are controversial, they consider $1 0/BBL
an appropriate value for analysis.
8 David Greene
and K.G. Duleep, "Costs and Benefits of Automotive Fuel Economy Improvement: A
Partial Analysis," Transportation Research A , Vol. 27A, No. 3, pp. 217-235, 1993.
9 California Transportation Analysis Report; Technical Appendices DRAFT, Feb. 1992.
Page 3.12-3
Transportation Cost Analysis
Figure 3.12-1 shows external energy costs estimated by Harold Hubbard.10 The first three
cost categories: corrosion, health impacts, and crop losses are air pollution impacts
considered in chapter 3. 10. Radioactive waste disposal costs do not apply to petroleum
use. The three remaining external motor vehicle energy costs include $12 to $55 billion in
military expenditures to protect petroleum markets (this study was done before the war
with Iraq and so this may be understated), $30 billion in reduced U.S. employment due to
petroleum imports, and $43 to $56 billion in subsidies to the petroleum industry.
Figure 3.12-1
Hubbard's Estimate of Energy Externalities
CP
a.
t!
.!!
0'-
c ftS
0~
II)
c
100
• Range of Disagreements
75
• MinirTUm Estimate
!i)
25
~
o.a......---~
jjj
Corro-
sion
Health
1111l8cts
Crop
Losses
Radioactive
IVIilitary
Ef11>1oy- Subsidies
ment
waste
This graph summarizes external costs of energy consumption, as estimated by Hubbard.
Some of these costs apply to petroleum consumption.
Undertaxing of fuel is another external energy cost. In a study comparing actual taxes on
energy with other classes of consumer products, Joe Loper concludes that fuel taxes
(excluding user fees) are 30% lower than the average state and local taxes on general
commodities, effectively providing a tax exemption to driving. 11 Although economists
often treat taxes as transfer payments rather than costs, Douglass Lee points out that
selected exemptions to broad-based taxes function the same as if all taxpayers paid the tax
10
"The Real Cost Of Energy," Harold M . Hubbard, Scientific American, 264/4, April1991, p. 36.
Joe Loper, State and Local Taxation : Energy Policy by Accident, The Alliance to Save Energy
(Washington DC), 1994.
11
Page 3.12-4
Transportation Cost Analysis
and revenues were then returned as a subsidy payment. 12 This approach also recognizes
that exemptions cause other taxes to increase to meet revenue demands.
Estimates:
(Note, although some estimates below are measured in VMT, this cost is actually based on
fuel consumption rates, not vehicle mileage.)
• Apogee Research estimates external energy costs including government subsidies, tax
breaks, maintenance of the Strategic Petroleum Reserve, and trade effects to total
approximately $0.51 per gallon of gasoline, or about 2.5¢ per vehicle mile.13
• The California Energy Commission estimates energy security costs at $0.31/gallon of
gasoline, or about $0.015 per automobile mile.14
• Greene and Duleep estimate the value ofU.S. motor vehicle fleet energy conservation,
including average savings of $13 .8 billion in energy price reduction, $5 .7 billion in
reduced energy security costs, and $32.4 billion in reduced wealth transfer out ofthe
U.S. (based on their "moderate" parameter values, averaging "high" and "low" oil
prices, using 1993 dollars).l 5 These benefits total about $50 billion for a fuel saving of
about 150 billion gallons over a 30 year period, implying a marginal cost of about
$0.33 per gallon, or about $0.017 per average vehicle mile.
• Harold Hubbard's estimates for energy security, unemployment, and tax subsidies for
petroleum range from $85 to $141 billion.16 Based on 52% consumed by motor
vehicles and 2,300 billion annual miles, this equals $0.02 to $0.03 per vehicle mile.
• Douglass Lee estimates that motor vehicle's share of oil producer tax subsidies is $9
billion a year, Strategic Petroleum Reserve maintenance is $4.4 billion per year, and
local, state and federal sales tax exemptions for fuel total $18.7 billion. This totals
$32.1 billion annually or about $0.013 per vehicle mile.17
• Milton Copulos estimates petroleum import subsidies at $45 billion annually in 1989.
He cites military costs, lost wages, and lost royalties.18
12 Full Cost Pricing ofHighways, National Transportation Systems Center (Cambridge), p. 31.
13 Apogee Research, The Costs of Transportation: Final Report, Conservation Law Foundation
(Boston),
1994, p. 145-147, 158-159.
14
1993-1994 California Transportation Energy A nalysis Report, CEC (Sacramento), Feb. 1994, p. 29.
David Greene and K.G. Duleep, "Costs and Benefits of Automotive Fuel Economy Improvement: A
Partial Analysis," Transportation Research A, Vol. 27 A, No. 3, pp. 217-235, 1993.
16 "The Real Cost Of Energy," Harold M. Hubbard, Scientific American, Vol. 264 No. 4, April1991.
17 Full Cost Pricing of Highways, National Transportation Systems Center (Cambridge), 1995, p. 12.
18 Milton Copulos, Landmarc: National Magazine of Coal and Energy Issues, JIF, 1989.
15
Page 3.12-5
Transportation Cost A nalysis
• Brian Ketcham and Charles Komanoff estimate energy subsidies for driving totals $33
billion a year, which equals about $0.015/mile.
• MacKenzie et al. argue that drivers should pay about half of micro-economic and
security costs of importing petroleum, including maintenance costs for the Strategic
Petroleum Reserve, totaling $25 .3 billion dollars a year, about $0.012 per VMT.l 9
• Peter Miller and John Moffet's estimate of external costs of petroleum includes federal
subsidies provided to the oil industry, micro-economic impacts, and military and other
security costs. Their estimate ranges from $45 (assuming zero military and
microeconomic costs) to $150 billion annually, or $0.015 to $0.05 per vehicle mile. 20
• A study for the Western Regional Biomass Energy Program estimate the annual
military costs of protecting U.S . access to Middle East petroleum supplies is $57
million per year, which averages $9.19 per barrel, or $0.22 per gallon.21
• The Office ofT echnology Assessment indicates average external fuel cost of $0.006 to
$0.025 per vehicle mile based on these energy related cost estimates (billions):22
Low Cost
Monopsony cost of importing oil
$7.5
Military costs related to oil use
5.0
Strategic Petroleum Reserve
0.2
0.0
Tax subsidies
1.0
Oil refineries environmental impacts
Gasoline distribution environmental impacts ___M
$13 .7
Totals
High Cost
$21.6
20.0
0.2
3.0
6.0
_2Q
$55 .8
• A Washington State Energy Office study found that telecommuting incurs a minor
energy cost from increased residential energy use from heating, cooling and office
equipment equal, plus some additional automobile trips. 23
• Energy conservation investments by electrical utilities can be compared with motor
vehicle fuel costs to determine whether petroleum conservation may be justified if
energy policies were consistent across sectors. Utilities currently invest in energy
conservation that is cheaper than a "hurdle rate," which ranges from about $0.03/k:Wh
19 James
MacKenzie, Roger Dower and Donald Chen, The Going Rate, World Resources Institute
(Washington DC), 1992, p. 17.
20 The Price ofMobility, National Resources Defense Council (Washington DC), Oct. 1993, p.16.
21 ENERGETICS & NEOS corporations, The National Security Costs ofPetroleum, Western Regional
Biomass Energy Program (Golden), June 1994.
22 Saving Energy in US. Transportation , U.S. Office of Technology Assessment, 1994, p. 104-108. Also
see pages 123-128 for discussion of energy security threat.
23 Maureen Quaid and Brian Lagerberg, Puget Sound Telecommuting Demonstration; Executive
Summary, Washington State Energy Office (Olympia), Nov. 1992.
Page 3.12-6
Transportation Cost Analysis
to $0.13/kWh.24 These rates are based on marginal costs of energy production, and
often include adders for environmental impacts. Assuming an average energy
conversion heat rate of9,000 BTU/kWh, a gallon of gasoline equals 15 kWh. This
implies that society should be willing to spend $0.45 to $1.95 per gallon of gasoline
conserved, which would justify transportation energy conservation in many cases,
since this often exceeds the $0.75 to $1.00 per gallon per-tax price of petroleum.
Since few of these estimates include environmental or non-use costs such as bequest or
option values, they underestimate total costs. An indication that environmental and nonuse costs of oil production and consumption may be substantial is the fact that the U.S.
has prohibited or limited oil production in several areas where it is considered financially
viable, including in Alaska's tundra, and off California's coast.
Variability: This cost depends on total energy use, including direct fuel consumption and
indirect uses such as vehicle production energy.
Conclusions: Resource use, especially petroleum consumption, incurs external costs
including environmental damage, tax subsidies, energy security, and national economic
impacts. Estimates place these external costs between $25 to $150 billion per year for
petroleum, which averages $0.005 to $0.03 per vehicle mile, based on roadway vehicles
consuming half of total petroleum production.25 Most lower estimates include only a few
ofthe external costs identified. Although the exact value is difficult to determine, the
middle to higher end of this range seems justified to include all external costs, including
embodied energy and non-energy materials used in motor vehicle and road construction.
Therefore, automobile resource consumption is estimated to impose external costs
averaging $0.025 per mile.
This value is used for an average automobile under Urban Off-Peak conditions, with
higher values for Urban Peak driving and lower values for Rural driving to reflect relative
fuel efficiency and vehicle wear. The costs of other vehicles are estimated based on their
24
25
Jim Lazar, Energy Economist, personal conversation, 2/23/94.
Facts and Figures '93, American Automobile Manufactures Association (Detroit), p. 64.
Page 3.12-7
Transportation Cost Analysis
relative fuel consumption. Electric car resource costs are estimated to be half that of an
efficient automobile, to reflect the lower external costs of this energy source.2 6 Rideshare
passengers are estimated to add an incremental cost of 2% each, based on a 20% increase
in fuel use for 10 passengers. Electric buses and trolleys are estimated to impose 50% the
external environmental costs of diesel buses. Telecommuting energy costs are estimated at
10% of an average automobile for the increased energy consumption from residential
heating and increased driving due to automobile availability.
-
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Bus!frolley
Motorcycle
Bicycle
Walk
Telecornmute
Urban Peak
0.029
0.014
0.007
0.039
0.001
0.089
0.045
0.012
0.000
0.000
0.003
-- --
-
~
-
-
-
Urban Off-Peak
Rural
Avera2e
0.025
0.013
0.006
0.033
0.001
0.077
0.038
0.010
0.000
0.000
0.003
0.021
0.011
0.006
0.028
0.000
0.064
0.032
0.009
0.000
0.000
0.002
0.024
0.012
0.006
0.032
0.001
0.074
0.037
0.010
0.000
0.000
0.003
Automobile Cost Range: The minimum is based on the estimate by Douglass Lee. The
maximum estimate is based on a $150 billion total annual cost with an additional10% for
embodied energy and another 10% for non-energy resource costs.
Minimum
$0.008
Maximum
$0.078
26 The main benefit of electric energy over petroleum is that it allows air pollution to occur outside of
urban areas, which is incorporated in Chapter 3.10. At 0.5 kWh/mile, electric cars consume the same total
energy as an 30 mpg car. While not all electric power uses imported petroleum products, it incurs other
external costs, including SOx emissions, hydroelectric facility impacts, and nuclear waste production,
depending on the marginal electrical power source.
Page 3.12-8
Transportation Cost Analysis
3.13 Barrier Effects
Definition: Motor traffic impacts on the mobility, security, and satisfaction of pedestrians
and cyclists, and its effects on their movement and activities.1 Also called severance. 2
Discussion: Roads are typically viewed as transportation links, but they are also barriers,
especially to nonmotorized travel. 3 The barrier effect reduces walking and bicycling, and
increased driving. It represents an increase in accident risk, and a degradation of the
pedestrian and bicyclist environment. Barrier effect costs tend to be inequitable because
they are imposed most on vulnerable and disadvantaged populations, including children,
the elderly, and handicapped people. The UK Environmental Assessment Manual
discusses this problem and provides instruction for measuring (but not monetizing) it. 4
Robert Davis and Mayer Hillman argue that measured reductions in pedestrian and bicycle
accidents may result from reduced travel by these modes rather than increased safety.
Davis reports that the portion of British children walking on their own to school has
decreased from 80% in 1971 to only 9% in 1990, due in part to motor vehicle accident
risk. 5 A study of home-to-school transportation found similar patterns North America.6
School principals cited "volume and speed of vehicular traffic" as the primary barrier to
increased walking and bicycling by students. Hillman states,
1 Swedish National Road Administration
2 Severance often emphasizes the impact of a
new road on access within a community, while the barrier
effect incorporates traffic impacts on all roads, old or new. J. Stanley and A. Rattray, "Social Severance"
in The Valuation ofSocial Cost, Allen and Unwin (Editors), 1978; B.S. Hoyle and RD. Knowles, Modern
Transport Geography, Belhaven Press (London), 1992, p. 62.
3 European Conference of Ministries of Transport, Transport Policy and the Environment, OECD (Paris),
1990, 134; Julian Hine and John Russel "Traffic Barriers and Pedestrian Crossing Behavior," Journal of
Transport Geography, Vol.1 No. 4, 1993, pp. 230-239; J.M. Clark and B.J. Hutton, The Appraisal of
Community Severance, U.K. DoT, TRRL (Crowthorne), Report #135, 1991.
4 Environmental Assessment M anual, HMSO (London), 1993 .
5 Robert Davis, Death in the Streets, Leading Edge (North Yorkshire), 1992, p. 156.
6 University of Florida, Dept. of Urban and Regional Planning, Home-to-School Transportation Study,
Florida Department of Transportation (Tallahassee), 1990.
Page 3.13-1
Transportation Cost Analysis
"Preferred patterns of behavior are altered and an increasing burden of responsibility
is imposed on all road users, especially pedestrians, to reduce their exposure to risk.
This is a social cost which has hardly been acknowledged and which certainly is not
reflected in government transport or road safety policies. "7
Susan Handy attributes reduced walking trips to a commercial district to the barrier effect
created by a major arterial separating it from residential neighborhoods, 8 resulting in half
the walking trip rate of an otherwise comparable community. A study by 1000 Friends of
Oregon concludes that the portion of trips by walking, bicycling and transit in an area
declines as its Pedestrian Environmental Factor (PEF) decreases.9 Automobile oriented
road designs (wide streets, cui de sacs, lack of sidewalk continuity) and high motor vehicle
traffic speeds and volumes reduce the PEF. Traffic calming 10 and neotradtional planning 11
are based, in part, on the desire to improve neighborhood PEF.
Efforts to quantify this cost are currently limited to the Scandinavian literature. Both the
Swedish12 and the Danish 13 roadway investment evaluation models incorporate methods
for quantifying barrier effects on specific lengths of roadway. Both methods involve two
steps. First, a barrier factor is calculated based on traffic volumes, average speed, share of
trucks, number of pedestrian crossings, and length of road way under study. Second, the
demand for crossing is calculated (assuming no barrier existed) based on residential,
commercial, recreation, and municipal destinations within walking and bicycling distance
of the road. The Swedish model also adjusts the number of anticipated trips based on
whether the road is in a city, suburb, or rural area, and the ages oflocal residents.
7 Mayer Hillman,
"Foul Play for Children: A Price of Mobility," Town and Country Planning, Oct. 1988,
pp. 331-332.
8 Susan Handy, Understanding the Link Between Urban Form and Travel Behavior, TRB Annual Meeting
(Washington DC), Paper #950691 , January 1995.
9 The Pedestrian Environment, 1000 Friends of Oregon (Portland), Dec. 1993.
10 Peter Newman, Jeff Kenworthy, Towards a More Sustainable Canberra, Murdoch University, 1991
11 Andres Duany, Anton Nelessen, Chris Duerksen and Walter Kulash, Neotraditonal Town Planning,
Conference Proceedings, American Institute of Certified Planners
12 Swedish National Road Administration, Investment in Roads and Streets, publication 1986: 15E.
13 Danish Road Directorate, Evaluation ofHighway Investment Projects (undersogelse af storre
hovedlandeveejsarbejder. Metode for effektberegninger og okonomisk vurdering), 1992.
Page 3.13-2
Transportation Cost Analysis
Estimates:
• Kjartan Saelensminde estimates that the total cost of the barrier effect in Norway
equals $112 per capita annually (averaging about $0.01 per vehicle mile), which is
greater than the estimated cost of noise, and almost equal to the cost of air pollution.14
• A recent Dutch publication estimates that the barrier effect represents 15% of roadway
costs to be considered in benefit/cost analysis (total costs are 50% economic [travel
time, accident reduction, VOC], 30% noise, 15% barrier effect, 5% air pollution).15
Variability: As described in the Scandinavian literature, this impact depends on traffic
speeds and volumes, and the demand for pedestrian and bicycle crossings.
Conclusions: The barrier effect is implied and described in much planning literature. In
addition to direct costs to pedestrians, bicyclists and residents, it also imposes costs in
terms of increased automobile dependency and use, and increased chauffeuring.
One might argue that there is a symmetry between the impacts of motor vehicles on
pedestrian and bicycle travel and the delays non-motorized modes cause motor vehicles,
resulting in a balance of "costs." However, casual observation indicates that pedestrians
and bicyclists are much more delayed by motor vehicle traffic than vice versa, and that
pedestrians and bicyclists frequently modify or forego trips due to heavy traffic, while
automobile drivers seldom change their trip plans or reduce total travel because of delay,
discomfort and danger imposed by pedestrian or bicycle traffic. It seems safe to estimate
that there is at least an order of magnitude difference in the costs imposed by motor
vehicle traffic on non-motorized travel compared with reciprocal impacts.
14
Kjartan Saelensminde, Environmental Costs Caused by Road Traffic in Urban Areas-Results from
Previous Studies, Institute for Transport Economics (Oslo), 1992.
15 Klaus Gylvar and Leleur Steen, Assessment ofEnvironmental Impacts in the Danish State Highway
Priority Model, 1983
Page 3.13-3
Transportation Cost Analysis
Scandinavian estimates indicate that the barrier effect is a significant cost. There is no
reason to believe that this cost is substantially different in Scandinavian countries than in
North America. The Norwegian estimate of$0.01 per vehicle mile places this cost
comparable to automobile noise, which seems reasonable and is used here to estimate
automobile and motorcycle barrier costs. Transit vehicles are charged $0.025, which
represents an average of the barrier effect cost for trucks in Danish and Swedish models.
Bicycling is estimated to incur 5% of an average automobile's barrier cost. Rideshare
passengers, walking, and telecommuting incur no barrier costs. Although larger urban
traffic volumes are balanced to some degree by higher speeds on rural roads, greater
populations cause this cost to be highest in urban areas, especially during peak periods
when traffic volumes are highest and the greatest demand exists for pedestrian and bicycle
travel. For these reasons, the basic cost is applied to Urban Off-Peak driving, which is
increased 50% for Urban Peak travel and decreased 50% for Rural driving.
Best Guess
Barrier Effect (Dollars per Vehicle MileJ
Vehicle Class
Urban Peak
Urban Off-Peak
Average Car
0.015
0.010
Fuel Efficient Car
0.015
0.010
Electric Car
0.015
0.010
Van
Rideshare Passenger
Diesel Bus
Electric Bus!Trolley
Motorcycle
Bicycle
Walk
Telecommute
0.015
0.00
0.038
0.038
0.015
0.001
0.00
0.00
0.010
0.00
0.025
0.025
0.010
0.00
0.00
0.00
Rural
Average
0.005
0.005
0.005
0.005
0.00
0.013
0.013
0.005
0.00
0.00
0.00
0.009
0.009
0.009
0.009
0.00
0.023
0.023
0.009
0.00
0.00
0.00
Automobile Cost Range: Because of limited research of this cost in North America, the
range is somewhat arbitrarily estimated at 50% and 200% of the estimate developed here.
Minimum
$0.005
Page 3.13-4
Maximum
$0.02
Transportation Cost Analysis
3.14 Land Use Impacts
Definition: External costs ofland use impacts caused by roads and automobile traffic.
Description: Roads and driving impact land use directly, and indirectly by encouraging
low density urban expansion (sprawl). These impose a variety of external costs.
Discussion: Transport and land use patterns are highly interactive. In the short term, land
use patterns affect travel demand.1 In the longer term, land use is affected by transport.2
Travel patterns affect land use, which affects the natural and built environments, which
affects economic, community and individual well being. These links explain why transport
decisions can impose external land use costs. Measuring these costs is difficult because of
the indirect nature of the impacts and the unique character of every area of land. Although
little research has attempted to monetize them, these costs appear to be substantial.
Transportation as a Cause of Sprawl
An important consideration in this discussion is the degree to which roads and automobile
use contribute to land use changes such as sprawl. The proper conceptual measure of such
impacts is the with and without test: the difference in development that would occur with
and without a road project or a certain level of driving.3 Automobile use encourages
sprawl by degrading the urban environment, by demanding large amounts of urban land for
roads 4 and parking, 5 and by accommodating urban fringe development. Low density land
use, in tum, leads to increased automobile use by reducing the viability of walking,
1 Trip
Generation Manual, Institute of Transportation Engineers.
The Geography of Urban Transportation , Guilford Press (NY), 1986, p. 4.
3 C. van Kooten, Land Resource Economics and Sustainable Dev., UBC Press (Vancouver), 1993, p. 86.
4 See Chapter 3.7 for comparisons of the road space requirements of different modes.
5 As described in Chapter 3.4, residential parking requirements tend to significantly reduce the number of
housing units per acre, both by using land and by reducing the profitability of small units.
2 Susan Hanson,
Page 3.14-1
Transportation Cost Analysis
bicycling, and transit service. 6 This self-reinforcing driving/sprawl cycle continues until
other forces, such as travel time, vehicle costs, and congestion become limiting factors.
The Transportation and Traffic Engineering Handbook states, "Although there are other
factors that play a role [in urban sprawl], reliance on the automobile has been most
significant in this trend. 7 Another popular transport engineering text states:
''Automotive transportation allowed and encouraged radical changes in the form of
cities and the use of land Cheap land in the outer parts of cities and beyond became
attractive to developers, much of it being converted from agricultural uses. Most of
the new housing was in the form of single-family homes on generously sized
lots... Automobiles were easily able to serve such residential areas, while walking
became more difficult, given the longer distances involved, and mass transportation
found decreasing numbers ofpossible patrons per mile of route. '18
After studying the relationship between transport and land use patterns, researchers Peter
Newman and JeffKenworthy found strong negative relationships between private vehicle
use and nearly all measures of urban density and provision of automobile facilities (parking
and road space), although causation is not proven.9 Mark Hanson states, "A motorized
means ofpersonal travel is necessarily the dominant transportation technology for
serving highly dispersed trip origins and destinations. "1°
Daniel Solomon argues that the shift from urban to suburban development resulted to a
large degree from the U.S . Federal Housing Administration's Minimum Property
Standards (MPS), established in 1938, which effectively targeted federal housing loans to
automobile oriented suburban developments. He states that, "The MPS was based on the
6
Susan Handy How Land Use Patterns Affect Travel Patterns, CPL Bibliography #279, 1992; Eric D.
Kelley, "The Transportation Land-Use Link," Journal of Planning Literature, 912, Nov. 1994, p. 128-145.
7 John Edwards, Transportation and Traffic Engineering Handbook, Institute of Transportation
Engineers/Prentice Hall (Englewood Cliffs), 1982, p. 401.
8 Hornberger, Kelland Perkings, Fundamentals of Traffic Engineering, 13 Edition , Institute of
Transportation Studies, UCB (Berkeley), 1982 p. 2-8.
9 Peter Newman and Jeff Kenworthy, Cities and A utomobile Dependency, Gower, 1989.
10 "Automobile Subsidies and Land Use," A merican Planning Association Journal, Winter 1992, p. 60.
Page 3.14-2
Transportation Cost Analysis
belief that American gridiron towns could not accommodate the automobile. It imposed a
pattern of enclaves rather than a continuous urban fabric; traffic was restricted to
arterials, and houses stood on curving cui-de-sacs. "11
This low density, automobile oriented land use pattern is still taught to transport planners
and traffic engineers as the preferred and acceptable road system because it best
accommodates motor vehicle travel. 12 Only in 1994 did the Institute of Transportation
Engineers publish a preliminary report on the development of street design standards for
transit and pedestrian oriented communities that provides an alternative road development
model based on nee-traditional street patterns.13
Two arguments are used against treating increased urban sprawl as a cost of transport.
One is that sprawl is a land use management issue not a transport issue. In practice this is
inappropriate because current land use management techniques are not completely
effective. 14 Few governments have the strength to develop and enforce effective land use
controls if strong demand exists, for example, where undeveloped land is easily accessible
to urban areas.15 Even with the best land use management system in place, transportation
improvements have residual impacts that should be considered transportation costs.
Another argument for excluding sprawl as a cost, is that low density development may
offer benefits that offset costs. For example, low density land use allows individuals to buy
more land at a given price, including increased private greenspace. However, since
11 Daniel Solomon, "Fixing Suburbia," in Sustainable
Communities; A New Design Synthesis for Cities,
Sim Vander Ryn and Peter Calthorpe, Sierra Club Books, 1986, p. 22.
12 Hornberger, Kell and Perkings, Fundamentals of Traffic Engineering, 13 Edition , Institute of
Transportation Studies, UCB (Berkeley), 1982, chapter 13.
13 "An Informational Report: Traffic Engineering for Neo-Traditional Neighborhoods," ITE Journal,
March 1994, p. 46.
14 Knaap and Nelson, The Regulated Landscape, Lincoln Institute (Washington DC), 1992, Chapter 5.
15 Harry Dimitriou, Urban Transport Planning, Routledge (NY), 1992, pp. 78-81.
Page 3.14-3
Transportation Cost Analysis
sprawled land use increases the per capita area covered by buildings and pavement, the
total amount of greenspace is reduced.
The benefits of sprawl are almost entirely internalized, so the best test of the hypothesis
that total benefits exceed total costs would be to charge users for all external costs and see
how much they are willing to pay. There is no obvious reason for society to subsidize
these benefits. One justification might be that urban sprawl provides external benefits, but
none have been demonstrated. At one time researchers investigated the possibility that low
density land use reduces social problems such as crime, poverty, depression, and
interpersonal conflict, but most studies find no association between density and crime or
other behavioral problems when income and social class are factored in. 16
This cost varies considerably by mode. Table 3.7-1 shows land use requirements of
various modes as summarized by Emile Quinet. However, this is only one portion of this
cost, since motor vehicle modes also degrade the urban environment and accommodate
low density urban expansion, both of which further encourage sprawl.
Defining Land Use Impact Costs
Automobile oriented land use and sprawl are increasingly recognized as imposing external
costs. One recent study rates "Inefficient Settlement Patterns," "Inefficient Infrastructure,"
and "Loss ofHabitat due to Development" as first, second and third ecological and human
health problems.17 This study used surveys of environmental experts and cited a variety of
other studies supporting the conclusions that urban sprawl is the region's most significant
environmental problem because of its direct and indirect impacts. The California Air
Resource Board also concluded that sprawl imposes an air pollution cost, and therefore
16
Andrew Baum and YakofEpstein, Human Response to Crowding, Hillsdale, 1978.; Newman and
Kenworthy, Cities and Automobile Dependency, Gower, pp. 89-92.
17 Report on Environmental Problem Ranking Process, Capital Regional District (Victoria), Oct. 1994.
Page 3.14-4
Transportation Cost Analysis
recommends development of denser, less automobile oriented communities.18
Transportation land use externalities can be grouped into five categories:19
1.
Environmental Impacts
Biologically active lands such as wetlands, forests, farms, rangelands, and parks
(collectively called greenspace) provide a variety of environmental and social benefits,
including wildlife habitat, air and water regeneration, biological diversity and social
benefits of agricultural production. These external benefits exist in addition to benefits to
the land owner, and are not reflected in the land's market value because they are enjoyed
by society as a whole.20 These benefits are reflected in many ways, for example by
increased value to adjacent real estate, improved community water quality, recreation and
tourism, and in existence, option, and bequest values.21
Roads degrade environmental amenities and agricultural production directly by paving and
clearing land, indirectly by encouraging increased development, sprawl and other
disturbances, and by introducing new species that compete with native plants and animals.
Ecological damage from roads and traffic is well documented.22 Impacts include the loss,
isolation, and disturbance of wildlife habitat, increased paved surfaces, clearing for road
buffers, damage to unique physical features, road kills, and injuries. W . Roley states:
"The net effect on wildlife of automobile-dependent urban sprawl is the fragmentation
of habitat and the isolation of these fragments and their wildlife populations from one
another. The gravest threat to the survival of wildlife in developed areas around the
18
The Land Use-A ir Quality Linkage; How Land Use and Transportation Affect Air Quality, CEPA, Air
Resources Board (Sacramento), 1994.
19 For additional discussion of these costs see Engin lsin and Ray Tomalty, Resettling Cities: Canadian
Residential Intensification Initiatives, Canadian Mortgage and Housing Corporation (Ottawa), Sept. 1993.
20 Knaap and Nelson, The Regulated Landscape, Lincoln Institute (Washington DC), 1992, p. 126.
21 Kopp and Smith, Valuing Natural Assets, Resources for the Future (Washington DC), 1993, pp. 10-19;
and van Kooten, Land Resource Economics, UBC Press (Vancouver), 1993 , pp. 157-187.
22 See for example, Works Consultancy, Land Transportation Externalities, Transit New Zealand
(Wellington), 1993; Environmental Externalities and Social Costs of Transportation Systems, Federal
Railroad Administration (Washington DC) Aug. 1993; H .D . van Bohemen, Habitat Fragmentation and
Roads, TRB Annual Meeting, Paper 950694, January 1995.
Page 3.14-5
Transportation Cost Analysis
world is the reduction of both habitat and mobility of wildlife. The automobile, in
other words, has become the greatest predator of wildlife. ''23
Prime farm land is often located near growing urban areas, making it highly susceptible to
sprawl. Urban development offarrnland is considered semi-irreversible. The total long
term loss of farm production by urban sprawl is often underestimated. Studies estimate
that from 1 to 5 acres are removed from farming for each acre that is actually developed
due to land speculation and other influences of the "urban shadow."24
2. Aesthetic Degradation and Loss of Cultural Sites
Roads and traffic can reduce natural environmental beauty, cause urban blight and destroy
cultural sites.25 The Transportation and Traffic Engineering Handbook, 26 the
Transportation Association of Canada's Environmental Policy and Code ofEthics, 27 the
USDOT's Environmental Assessment Notebook, 28 and a Transit New Zealand study of
transport extemalities29 all cite visual aesthetic degradation as major negative impacts of
roads. Roads and the development they encourage can degrade landscape beauty in many
ways.30 The value of attractive landscapes is indicated by their importance in attracting
tourism and increasing adjacent property values. Aesthetic impacts on the landscape can
be evaluated using public and professional surveys. 31 Such techniques have been used to
evaluate the visual impact of roads and traffic. 32 Ratings consistently became less
favorable as the size of the road construction increased.
23
W. Roley, "No Room To Road," Earthword #4, 1993, p. 35.
Pond and Maurice Yeates, "Rural/Urban Land Conversion I: Estimating Direct and Indirect
Impacts," Urban Geography, Vol14, pp. 323-347.
25 B.S. Hoyle and RD. Knowles, Modern Transport Geography, Belhaven (London), 1992, p. 54-57.
26 John Edwards, "Environmental Considerations," Transportation and Traffic Engineering Handbook,
Second Edition, Institute of Transportation Engineers/Prentice-Hall (Englewood Cliffs), 1982, p. 396.
27 TAC Newsletter, Sept. 15, 1992
28 Environmental A ssessment Notebook Series, USDOT, 1975. In Homburger, 1992, p.29-4.
29 Works Consultancy, Land Transport Externalities, Transit New Zealand, 1993, p.92.
30 British Columbia Scenic Highways Program Study, MoTH (Victoria), October 1994, Chapter 3.3 .
31 Dunne and Leopold, Water in Environmental Planning, Freeman & Co. (NY), 1978, pp. 778-795 .
32 L. Huddart, "Evaluation of the Visual Impacts of Rural Roads and Traffic," TRRL, Report #355 , 1978.
24 Bruce
Page 3.14-6
Transportation Cost Analysis
3. Social Impacts.
Many critics charge that automobile dependent transportation and an overemphasis on
motor vehicle traffic flow in roadway design has negative impacts on society.33 They argue
that automobile oriented land use tends to degrade the public realm and the quality of
residential neighborhoods, disperse activities that support community cohesion (local
schools, stores, and other services), and discourage pedestrian and bicycle travel, reducing
neighborhood interaction. Donald Appleyard reported a negative correlation between
traffic volumes and various measures of neighborly interactions and activities, including
number of friends and acquaintances residents had on their street, and the area that they
consider "home territory."34 he comments:
"The activities in which people engage or desire to engage in may affect their
vulnerability to traffic impact. So many of these activities have been suppressed that
we sometimes forget they exist... Children wanting to play, and people talking, sitting,
strolling, jogging, cycling, gardening, or working at home and on auto maintenance
are all vulnerable to interruption [by traffic] .. . One of the most significant and
discussed aspects of street life is the amount and quality of neighboring. Its
interruption or 'severance' has been identified as one of the primary measures of
transportation impact in Britain. ''35
Richard Untermann and Anne Vemez Moudon perform a more recent study of traffic
impacts on neighborhoods and state,
''A deeper issue than the functional problems caused by road widening and traffic
buildup is the loss of sense of community in many districts. Sense of community
traditionally evolves through easy foot access--people meet and talk on foot which
helps them develop contacts, friendships, trust, and commitment to their community.
When everyone is in cars there can be no social contact between neighbors, and social
contact is essential to developing commitment to neighborhood '136
33
As evidence that roads often sacrifice social goals for the sake of motor vehicle traffic flow, compare
actual road designs with roads optimized for social interaction described in Crowherst Lennard and
Lennard, Livable Cities Observed, Gondolier Press (Carmel), 1995, or Christopher Alexander, et al., A
Pattern Language, Oxford Press, 1977. In most areas, traffic needs have won over other design goals.
34 Donald Appleyard, Livable Streets, University of California Press, 1981.
35 Donald Appleyard, p. 35.
36 Richard Untermann and Anne Vemez Moudon, Street Design: Reassessing the Safety, Sociability, and
Economics of Streets, University of Washington (Seattle), 1989, p.3.
Page 3.14-7
Transportation Cost Analysis
James Kunstler points out that an automobile oriented land use pattern and road system
degrades the public realm (public spaces where people naturally interact) and reduce
community cohesiveness.37 Peter Freund and George Martin criticize the "placelessness"
resulting when urban space is modified for automobile use, and from increased mobility
provided by automobiles. 38 The report Resettling Cities mentions the following possible
social problems associated with low density, sprawled land use (this document includes
arguments both supporting and opposing these concerns): 39
• Reduced choice of housing types suitable for an increasingly diverse population.
• Higher housing costs.
• Increased social alienation.
• Reduced social interaction.
• Decline of central cities and the rise of social problems there.
Merle Mitchell indicates that non-drivers who live in outer suburbs and rural communities
may be "locational disadvantaged" due to relatively poor access to community services.40
David Engwicht describes how automobile traffic reduces neighborhood social
interaction.41 He cites the dispersion of common destinations outside walking and cycling
range from residences, degradation of walk and cycle environment, decline of corner
stores, loss of public spaces suitable for casual social exchange, and increased fear of
crime on city streets that are devoid of pedestrian traffic.
Automobile travel, urban sprawl, and middle-class flight to socially isolated suburbs are
cited as contributors to a reduced sense of community, increased social conflict and
37 James Kunstler,
The Geography of Nowhere, Simon & Schuster, (NY), 1993, Chapter 7.
Peter Freund and George Martin, The Ecology of the Automobile, Black Rose (NY), 1993, p. 104.
39Engin lsin and Ray Tomalty, Resettling Cities: Canadian Residential Intensification Initiatives,
Canadian Mortgage and Housing Corporation (Ottawa), 1993.
40 Merle Mitchell, "Links Between Transport Policy and Social Policy" in Transport Policies for the New
Millennium , Ogden et al. editors, Monash University (Clayton), 1994.
41 David Engwicht, Reclaiming our Cities and Towns, New Society Publishers (Philadelphia), 1993, p. 45.
38
Page 3.14-8
Transportation Cost Analysis
degradation of cities.42 Steven Cochrun cites increased automobile use as a significant
contributor to community design and individual behavior changes that reduce the vigor of
local community.43 Sociologist David Popenoe identifies several negative consequences of
urban sprawl, including segregation by race and social class, economic inequity,
fragmentation of local government, and reduced access for non-drivers, especially children
and non-driving adults.44 A recent Lincoln Institute ofLand Policy newsletter article
describes the impacts of sprawl on the poor:
"Land use patterns that put a premium on mobility actually disadvantage some
segments of the population. Furthermore, a major cause of this poverty, in the opinion
of many scholars and policymakers, is the gap between where these poor people live in
central cities and where job growth is taking place in the suburbs. This transportation
gap can be all but unbridgeable for low-wage workers who do not own cars, especially
when public transit, where it exists, usually focuses on downtown and is often useless
for conveying people to widely dispersed, suburban employment sites. "45
Some critics question whether low density land use is really socially disadvantageous.
Hugh Stretton argues that the higher ratio of recreation sites per capita in Sydney,
Australia compared with Tokyo, Japan indicates that lower density cities provide more
resident benefits, ignoring the fact that preserving openspace often requires increased
densities. 46 He cites survey findings that suburban residents prefer their current housing
over inner-city apartments, but does not consider alternative residential patterns that may
satisfy residents at higher densities. Stretton also claims that in:fill development is more
expensive than building on greenfield (undeveloped) exurban land, but only mentions a
limited number of costs (primarily utility lines) and ignores non-market costs.
42 Douglas Kelbaugh, Housing Affordabili ty and Density, Washington State Department of Community
Development (Olympia), 1992, p. 20; Elmer Johnson, A voiding the Collision of Cities and Cars,
American Academy of Arts and Sciences (Chicago), 1993, p. 7.
43 Steven Cochrun, "Understanding and Enhancing Neighborhood Sense of Community," Journal of
Planning Literature, Vol. 9, No. 1, August 1994, p. 92-99.
44David Popenoe, "Urban Sprawl: Some Neglected Sociological Consideration," Sociology and Social
Research, Vol. 63, 1979, p. 255-268.
45 "Restructuring our Car-Crazy Society," Land Lines 6/2, Lincoln Institute, March 1994, p. 2.
46 Hugh Stretton, "Transport and the Structure of Australian Cities" in Transport Policies for the New
Millennium , Ogden et al. editors, Monash University (Clayton), 1994.
Page 3.14-9
Transportation Cost Analysis
4.
Municipal Service Costs
Several studies have found that low density land use requires significantly higher unit costs
for most public services, such as utilities, roads, schools, and emergency services.47
Results of two studies are shown in Table 3.14-2 and Figure 3.14-1 .
Table 3.14-2 Per H
hold A
Rural Sprawl
Costs
Units/Acre
Schools
Roads
Utilities
__!Qtlll~ ..
t Residential Densit: ...,48
I Municioal Costs for Diffi
Hil.!h Density
Rural Cluster
Medium Density
1:5
$4,526
$154
$992
$5,672
1:1
$4,478
$77
$497
$5,052
4.5 :1
$3,204
$36
$336
$3,576
2.67:1
$3,252
$53
$364
$3t66!_
Per household service costs increase due to sprawl. These are mostly external costs.
Figure 3.14-1 Residential Service Costs49
$125,000
'2
:: $100,000
Ql
II..
0
$75,000
s
·a.
ra
$50,000
0
0
-
Leapfrog, 8 km
-
• Contiguous
• Leapfrog, 0 km
,
•
• • • Contiguous, 0 km
--lnllll
f/1
f/1
,
-Leapfrog, 16.5 km
--Contiguous, 16.5 km
[
'(j
'2
::::J
:::!:
$25,000
$0
30
15
12
10
5
3
0.25
Dwelling Units Per Acre
This illustrates increased capital costs for lower density, non-contiguous development.
47
James Frank, The Costs ofAlternative Development Patterns, Urban Land Institute (Washington DC),
1989; Impact Assessment of the New Jersey Interim State Development and Redevelopment Plan, Office
of State Planning, 1992; Eric D. Kelly, "The Transportation Land-Use Link," Journal ofPlanning
Literature, Vol. 9, No. 2, pp. 128-145, Nov. 1994.
48 Robert Smythe, Density-Related Public Costs, American Farmland Trust (Washington DC), 1986.
Based on prototypical community of 1,000 units housing 3,260 people, 1,200 students.
49 James Frank, The Costs ofAlternative Development Patterns, Urban Land Institute, 1989, summarized from p. 40.
Page 3.14-10
Transportation Cost Analysis
Since these studies focus on capital costs, the total incremental cost of sprawl is higher
than indicated when operating costs are considered. Rural residents traditionally accepted
lower levels of public services, including private water and sewer, and unpaved roads, but
sprawl encourages new residents with higher expectations to move to exurban areas, so
municipal governments face pressure to provide urban services to the urban fringe despite
high unit costs.50 Some communities use impact fees to internalize a portion of these
costs, but in practice these seldom reflect full marginal costs. 51 Since these are fixed costs,
they provide no incentive to use resources efficiently once development costs are paid.
These estimates are limited to residential development. The total costs of suburban sprawl
are probably greater when commercial development costs are also included:
''Because the home and the workplace are entirely separatedfrom each other, often by
a long auto trip, suburban living has grown to mean a complete, well-serviced, selfcontained residential or bedroom community and a complete, well-serviced place of
work such as an office park. In a sense we are building two communities where we
used to have one, known as a town or city. Two communities cost more than one; there
is not only the duplication of infrastructure but also of services, institutions and retail,
not to mention parking and garaging large numbers of cars in both places. "52
5.
Increased Transportation Costs
Numerous studies show a negative correlation between land use density and automobile
use.53 Lower densities increase automobile dependency and mileage, resulting in higher
travel costs. After reviewing current research on the relationship between land use and
travel Duncan McLaren concludes, ''Empirical and modeled evidence supports the
50
Judy Davis, Arthur C. Nelson, and Kenneth Dueker, "The New 'Burbs," Journal of the American
Planning Association, Vo. 60, No. 1, Winter 1994.
51 City of Lancaster (California), Urban Structure Program, 1994.
52 Douglas Kelbaugh, Housing Affordability and Density, Washington Department of Community
Development (Olympia), 1992, p. 17.
53 John Holtzclaw, Explaining Urban Density and Transit Impacts ofAuto Use , Sierra Club (San
Francisco), 1994; Lawrence Frank, Relationships Between Land Use and Travel Behavior in the Puget
Sound Region, WSDOT (Olympia), Report #WA-RD 351.1, 1994; Robert Dunphy and Kimberly Fisher,
Transportation, Congestion and Density: New Insights, Urban Land Institute (Washington DC), 1993 .
Page 3.14-11
Transportation Cost Analysis
hypothesis that higher urban densities can reduce the need to travel. "54 A study of costs
associated with different land use patterns in New Jersey concluded that a development
plan which centralizes a greater portion of future growth would require 83% fewer new
lane miles than continued sprawl that results in a greater amount of generated traffic.55
External Environmental and Social Benefits?
A 1978 report argues that highways provide external environmental and social benefits. 56
Few of these proposed but unsubstantiated benefits seem reasonable based on knowledge
and sensibilities, and some seem outright silly. Here are typical quotations from the report :
Aesthetics: "The freeway can provide open space, reduce or replace displeasing land
uses, enhance visual quality through design standards and controls, reduce headlight
glare, and reduce noise. " and "Regarding the visual quality of the highway and highway
structures, freeways may create a sculptural form of art in their own right. Some
authors note that the undulating ribbons ofpavement possessing both internal and
external harmony are a basic tool of spatial expression. "
Wildlife: ''Freeway rights-of-way may be beneficial to wildlife in both rural and urban
environments... "
Wetlands: "The intersection of an aquifer by a highway cut may interrupt the natura/flow
of groundwater and thus may draw down an aquifer, improving the characteristics of
the land immediately adjacent to the highway. "
Native Vegetation: ''Roadside rights-of-way can be among the last places where native
plants can grow. "
Neighborhood Benefits: ''Highways, if they are concentrated along the boundary of the
neighborhood, can promote neighborhood stability. " and "Old housing of low quality
occupied by poor people often serves as a reason for the destruction of that housing for
freeway rights of way. "
Social Benefits: "Highways can increase the frequency of contact among individuals... "
and "Good highways facilitate church attendance. "
54 Duncan McLaren,
"Compact or Dispersed?" Built Environment, Vol. 18, No. 4, 1993, p. 268-284.
Rutgers University Center for Urban Policy Research, Impact Assessment of the New Jersey Interim
State Development and Redevelopment Plan: Research Finding, Office of State Planning, 1992, p.179.
56 Hays Gamble and Thomas Davinroy, Beneficial Effects Associated with Freeway Construction,
Transportation Research Board (Washington DC), Report 193, 1978
55
Page 3.14-12
Transportation Cost Analysis
External Environmental and Social Benefits?- Continued
Recreation: "Freeways cutting across, through, under, and around the cities afford an
excellent opportunity for innovations in recreation planning and design. "
Additional claimed benefits include improved air quality improvements, energy savings,
and reduce traffic noise. Urban benefits include removal ofblighted housing and slums,
support of mass transit, reduced accidents, greater safety for pedestrians in general and
school children in particular, improved community values, civic pride, increased social
contacts between diverse social groups, increased upward social mobility, in-migration of
better educated families, and increased housing opportunities for racial minorities. Land
use benefits include suburban growth, decentralization, industrial parks, shopping malls,
commercial development at freeway interchanges, and drive-in businesses.
Estimates:
1. Environmental Impacts
A Washington State Governor appointed advisory committee ranked land use impacts
among the state's worst environmental threats, but below air and water pollution.57 A
recent study identified urban sprawl as the highest priority environmental problem in the
Victoria (BC) area and cited numerous documents supporting this conclusion.58 A survey
of Vancouver area residents found 80% of respondents are concerned that not enough
farmland and greenspace are being protected for future generations, and that 65% are
willing to support higher density neighborhoods to protect greenspace.59 These indicate
that sprawl environmental costs are probably greater than zero and less than cost of air
pollution. Using half of air pollution costs as a base, this averages $.025 per VMT.
2. Aesthetic Degradation and Loss of Cultural Sites
Little data is available on monetized roadway aesthetic costs. Segal estimates that a 3/4
mile stretch of Boston's Fitzgerald Expressway reduced downtown property values by as
57 Environment
2010 Survey Results, Washington Department of Ecology (Olympia), March 1990.
Environmental Problems Ranking, CRD Roundtable on the Environment (Victoria), Oct. 1994.
59 Viewpoints Research, Public A ttitudes Survey, GVRD (Vancouver), April18, 1994, p. 1.
58
Page 3.14-13
Transportation Cost Analysis
much as $600 million in current dollars by blocking waterfront views.60 Amortized, this
cost averages $1 .30 to $2.30 per vehicle trip over the Expressway. This is an extreme
case, but indicates that aesthetic degradation from roads probably costs billions of dollars
a year in reduced property values and non-market losses. Overall, aesthetic costs probably
rank with other minor roadway environmental costs such as the barrier effect, water
pollution and waste disposal, so a comparable estimate of$0.005 per average automobile
mile seems appropriate, implying a national total annual cost of $11 .5 billion.
3. Social Costs.
I have found no estimates of this group of costs. They are probably significant in total, and
comparable to the environmental land use impact costs, so an estimate of $0.025 is used.
4. Increased Municipal Costs
Assuming that automobiles induce 50% ofhouseholds to choose one step lower density in
Table 3.14-1 , half the average of the three incremental annual municipal cost increases
([($5,672-$5,052)+ ($5,052-$3 ,669)+($3 ,669-$3 ,576)] x 0.5 = $350), divided by 15,100
annual vehicle miles per household,61 indicates this external cost averages $0.023 per mile.
5. Increased Transportation Costs.
Sprawled land use increases both users and external transport costs, but few studies
attempt to quantify it. One approach is to use estimates of household vehicle ownership
and mileage at different residential densities to calculate expected use travel costs per
household. Applying an estimate developed by John Holtzclaw to the density values in
Table 3 .14-1 , costs can be calculated using figures from Chapter 3 .1. These estimates
understate total sprawl costs because they use a constant transit accessibility index of 10, a
factor that typically increases with density, and because the estimate of $0.10 per mile of
60
61
Segal, The Economic Benefits ofDepressing an Urban Expressway, 1981.
National Personal Transportation Survey: Summary ofTravel Trends, USDOT, 1992, p.18.
Page 3.14-14
Transportation Cost Analysis
external costs is low, as will be discussed in Chapter 4. It also fails to incorporate user
time and accident risk costs, which probably increase with sprawl.
Assuming that sprawl causes 50% of all households to choose a residence one step lower
density in this table, the three incremental increases in household vehicle costs are
averaged and divided by two. Divided by 15, 100 average annual miles this cost averages
$0.092 per mile, as shown in Table 4.14-2. If this is considered entirely a future cost then
this value should be depreciated, but if it is considered a current cost (which seems
appropriate where sprawl is both a current and future problem) no depreciation is needed.
Table 3.14-2 Annual Household Auto Costs Under Four Densities62
2.67:1
units/acre
1:1
1:5
Auto/Household
VMT/Household
Auto Ownership Costs ($2,600/year)
Auto Operating Costs ($0.134/mile)
External Costs ($0 .10/mile)
Total Costs
Incremental cost of reduced density
Average of incremental costs
Average incremental cost per household
Average cost per vehicle mile
4.5:1
1.6
1.77
3.4
2.3
15,100
13,233
28,822
18,603
$4,160
$5,980
$4,602
$8,840
$1 ,773
$2,493
$2,023
$3,862
$1,510
$1,323
$2,882
$1 ,860
$7,256+
$8,135+
$10,333+
$15,584
$879
$2,201
nla
$5,251
($5 ,251 + $2,201 + $879) + 3 = $2,777
$2,777 X 0.5 = $1 ,389
$1 ,389 I 15,100 = 0.092
This table shows estimate of increased automobile costs associated with lower density
land use, with values modified to reflect national vehicle ownership and use.
A study comparing 1976 to 1990 Milwaukee area travel costs for 2,700 and 4,400 average
people per square mile found that user transport costs increased 10% for the lower density
option. These future costs should be depreciated, while including external costs would
significantly increase the total. These factors are assumed to approximately cancel each
other, giving an estimate of 10% of current user vehicle operating costs, or about $0.034
per mile. This represents the reimbursement that future residents would need for increased
62 John Holtzclaw, Using Residential Patterns and Transit to Decrease A uto Dependence and Costs,
National Resources Defense Council (San Francisco), June 1994. Vehicle ownership and annual mileage
data from National Personal Transportation Survey: Summary ofTravel Trends, USDOT, 1992, p . 12, 18.
Page 3.14-15
Transportation Cost Analysis
travel costs resulting from current sprawled development. For a working estimate of
sprawl's transportation costs, the estimates of$0.092 and $0.034 are averaged to $0.062.
Variability: These costs are associated with driving that contributes to the construction
of roads, especially outside of urban areas, or that result in low density urban expansion.
Ideally, this cost should be assessed specifically for each situation. Thus, sprawl costs
would be higher in communities where sprawl impacts are greater, and for specific trips
that accommodate and encourage urban expansion and low density development.
Although most of this cost is assigned to automobile use, some transit services also
contribute to sprawl, indicated by the portion of riders who access bus and trains by car.
Conclusions: Roads and driving cause land use impacts that impose environmental,
aesthetic, social, municipal, and transport costs. If the amount of land devoted to roads
and low density development was significantly reduced, society could be better off due to
the preservation of greenspace, improved views, more interactive neighborhoods and
communities, lower municipal costs, and reduced automobile dependency. This is not to
say that these land use patterns offer no benefits, but most benefits are enjoyed by drivers
and land owners, while costs are borne by society as a whole. Society must therefore be
able to account for these external costs in order to avoid land use changes in which
benefits do not offset total incremental costs.
There are few existing models or studies that measure the total of these costs. Individual
estimates described above can be used to calculate a first-cut estimate of total sprawl
costs, acknowledging that this is preliminary and more research is needed. The cost
charged to drivers should take into account two additional factors. First, automobile use is
not necessarily the only cause of sprawl, other influences such as mortgage and parking
policies also encourage sprawl. Second, not all communities consider urban sprawl to be a
Page 3.14-16
Transportation Cost Analysis
problem. For these reasons, automobile use is only considered responsible for half of total
sprawl costs, calculated below to be $0.07 per vehicle mile. This is charged to urban
driving and telecommuting, because they encourage low density land use. Rural driving is
charged at half this rate, on the assumption that it contributes less to sprawl. Ride sharing,
public transit, bicycling, and walking decrease road building requirements and encourage
higher densities, so incur no land use impact costs, although a sprawl cost should be
assigned to commuter rail services that are primarily accessed by automobile.
Land Use Impact Cost Estimate (dollars per vehicle mile)
Environmental
$0.025
Aesthetic & Cultural
$0.005
Social
$0.025
Municipal
$0.023
Transport demand
$0.062
Total average sprawl cost
$0.140
x 0.5
50% reduction for other contributing factors
Sprawl cost charged to automobile use
$0.07
-
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
~
-
-
---
-
-
~-
- -
--
--
-
Urban Peak
Urban Off-Peak
Rural
Avera2e
0.070
0.070
0.070
0.070
0.00
0.00
0.00
0.070
0.00
0.00
0.070
0.070
0.070
0.070
0.070
0.00
0.00
0.00
0.070
0.00
0.00
0.070
0.035
0.035
0.035
0.035
0.00
0.00
0.00
0.035
0.00
0.00
0.035
0.056
0.056
0.056
0.056
0.00
0.00
0.00
0.056
0.00
0.00
0.056
Automobile Cost Range: This is currently a difficult cost to estimate due to limited
research and data. The minimum estimate is based on just the increased municipal costs
associated with sprawl. The maximum estimate reflects the higher range of each cost.
Minimum
$0.02
Page 3.14-17
Maximum
$0.20
Transportation Cost Analysis
3.15 Water Pollution and Hydrologic Impacts
Definition: Water pollution and hydrologic impacts from vehicles, roads, and parking.
Description: Motor vehicles, roads and parking facilities are a major source of water
pollution and hydrologic disruptions. These include:
•
•
•
•
•
Water Pollution
Crankcase oil drips and disposal.
Road de-icing (salt) damage.
Roadside herbicides.
Leaking underground storage tanks.
Air pollution settlement.
•
•
•
•
•
Hydrologic Impacts
Increased impervious surfaces.
Concentrated runoff, increased flooding.
Loss of wetlands.
Shoreline modifications.
Construction activities along shorelines.
These impacts impose a number of costs including polluted surface and ground water,
contaminated drinking water, increased flooding and flood control costs, wildlife habitat
damage, reduced fish stocks, loss of unique natural features, and aesthetic losses.
Discussion: Roads and motor vehicle use contribute significantly to water pollution and
hydrologic problems. An estimated 46% of vehicles on U.S. roads leak hazardous fluids,
including crankcase oil, transmission, hydraulic, and brake fluid, and antifreeze.1 Between
460 and 600 million gallons of the 1.4 billion gallons of lubricating oils used in cars are
either burned by the car's engine or lost in drips and leaks, and another 180 million gallons
are disposed of improperly onto the ground or into sewers.2 During use, crankcase oil
picks up toxic chemicals and heavy metals. Millions of gallons of petroleum are released
into water bodies from leaks and spills during extraction, processing, and distribution.3
Leaking underground storage tanks, many used for motor vehicle fuel, cause additional
1Christopher
Von Zwehl, Comments at New Jersey Senate Public Safety Committee public hearing on
motor vehicle inspection legislation, Feb. 25, 1991, from Facts and Figures 90, AAMA.
2 Helen Pressley, "Effects of Transportation on Storrnwater Runoff and Receiving Water Quality," internal
agency memo, Washington State Department of Ecology (Olympia), 1991.
3 Peter Miller and John Moffet, The Price ofMobility, NRDC (Washington DC), Oct. 1993, p.50.
Page 3.15-1
Transportation Cost Analysis
groundwater contamination. The oil spots on roads and parking lots, and rainbow sheens
of oil in puddles and drainage ditches are a sign of this problem.
Studies show that runoff from roads and parking lots have high concentrations of toxic
metals, suspended solids, and hydrocarbons, 4 and that automobiles are the primary source
of toxic metals and organics.5 Bioassay tests show mild to acute toxicity of highway runoff
to various aquatic species.6 Decreases in abundance and diversity ofbenthic organisms,
and accelerated eutrophication of lakes has been attributed to urban runoff Road de-icing
salts incur significant environmental and material damage in many areas, 7 and roadside
vegetation control is a major source of herbicide dispersal. An FHWA study indicates that
water pollution is affected by road design, traffic volumes, climate and adjacent land uses
(Table 3.15-1 ), and provides a model for predicting pollution from a particular roadway.8
1a01e .J.l:J-1
.ronuuon Levels m Koad Kunon waters
Pollutant
Urban
Rural
Total suspended solids
Volatile suspended solids
Total organic carbon
Chemical oxygen demand
142.0
39.0
25.0
114.0
41.0
12.0
8.0
49.0
(micro~
Pollutant
Nitrate + Nitrite
Total copper
Total lead
Total zinc
rams per litre}
Urban
Rural
0.76
0.054
0.400
0.329
0.46
0.022
0.080
0.080
Roads and parking lots also have major hydrologic impacts. These include concentration
of stormwater that causes increased flooding, scouring and siltation, increased flood
control costs, reduced surface and groundwater recharge which lowers dry season flows,
4 R T. Bannerman,
D.W. Owens, R.B. Dodds,and N.J. Homewer, "Sources of Pollutants in Wisconsin
Stormwater," Water Science Tech . Vol. 28; No 3-5; pp. 247-259, 1993; Works Consultancy, Land
Transport Externalities, Transit New Zealand (Wellington), 1993, p. 33 .
5 Kevin Weiss, "Water Quality Impacts of Commuting," USEPA Office of Water Quality, 1993.
6 Bioassay is a technique for testing the toxicity of substances by introducing them into the tanks of fish or
other animals in laboratory conditions. Ivan Lorant, Highway Runoff Water Quality, Literature Review,
Ontario Ministry of Transportation, Research and Development Branch, MAT-92-13 , 1992.
7 Field, R. and M. O'Shea, Environmental Impacts of Highway Deicing Salt Pollution., USEPA, Report
No. EPN600/A-92/092 Published in "Chemical Deicers and the Environment" (ed.) F. D'Itri.
8 Eugene Driscoll, et al, Pollution Loadings and Impacts from Highway Stormwater Runoff, Publication
Number FHWA-RD-88-007, FHWA (Washington DC), April1990.
Page 3.15-2
Transportation Cost Analysis
and constriction of streams into culverts that increase physical barriers to fish. A 1992
survey of726 culverts in Washington State found that 36.4% interfere with fish passage at
least sometimes, of which 17.4 were total blockages. 9 Reduced flows and plant canopy
along roads can increase water temperatures. These impacts reduce wetlands and other
wildlife habitat, degrade of surface water quality, and contaminate drinking water. In many
cases the hydrologic impacts of road and urban runoff are more harmful to receiving
waters than the effects of toxic pollutants. to
Quantifying these costs is challenging. First, it is difficult to determine exactly how much
motor vehicles and roads contribute to water pollution problems since impacts are diffuse
and cumulative. Although pollutants measured in roadway runoff are usually well below
water quality standards, some build up in stream sediments where they can be toxic.
Second, it is difficult to place a dollar value on water quality and flow. Even if we know
the quantity of pollutants originating from roads and motor vehicle traffic and their general
environmental impacts, we face the problem of monetizing costs such as loss of wildlife
habitat, reduced wild fish reproduction, and contaminated groundwater.
New laws and policies designed to reduce pollution, prevent fuel tank leaks, and
internalize cleanup expenses may reduce costs of some impacts, so it could be argued that
current motor vehicle use imposes lower costs than has occurred in the past. However,
growth in population density, total driving, and public concern about water quality will
probably increase total costs, even if impacts per automobile mile decrease.
9 Tom
Burns, Greg Johnson, Tanja Lehr, Fish Passage Program; Progress Performance Report for the
Biennium 1991-1993, Washington Dept. of Fisheries, WSDOT (Olympia), Dec. 1992.
10 Waste Management Group, Urban Runoff Quality Control Guidelines for the Province ofBritish
Columbia, BC Ministry of Environment, Environmental Protection Division (Victoria), June 1992.
Page 3.15-3
Transportation Cost Analysis
Estimates:
• The California Energy Commission estimates major petroleum oil spill costs at $0.004
per gallon of gasoline, or about $0.0002 per mile, based on the calculated risk of a
major oil spill such as the Exxon Valdez.11
•
Paul Chernick and Emily Caverhill estimate average petroleum marine oil spill costs by
multiplying the minimum Exxon Valdez cleanup cost estimate of$1 .28 billion times 5
(because the cleanup only collected 20% of total oil released), for an estimated cost of
$6.4 billion, or $582 per gallon spilled.12 They consider this estimate conservative:
"While Exxon has been criticizedfor doing too little, and spending too little, we are
not aware of any criticism of Exxon spending too much. If cleaning up 20% of the
spill was worth $1.28 billion, cleaning up all the oil must have been worth more than
$6.4 billion. The first barrel in the environment probably has greater impact than the
last 20% (After all, each animal can only be killed once. The practical difference
between pristine water and slighly polluted water is almost certainly greater than the
difference between very polluted water and slightly more polluted water), so the value
of cleaning up all the oil would probably be much higher than $6.4 billion. The value
of avoiding the spill in the first place must be greater than the value of cleaning it up,
because returning the environment to its pre-spill pristine condition is desirable but
impossible."
This report cites estimates that oil tankers spill from %0.02 and %0.11 of their
contents, for an estimated cost of $0.10 to $0.4 7 per gallon of imported crude oil,
based on $582 per gallon. However, because ofuncertainty concerning the application
of Alaskan oil spills to other situations, the authors use a lower value of$0.026 per
gallon to represent this cost in their own analysis of electrical generation impacts.
•
In September 1994 an Alaska jury awarded $5 billion in damages to businesses and
individuals harmed by the Valdez oil spill, which in addition to the $3 billion Exxon
claims to have spent on cleanup implies a total cost greater than $8 billion, since the
legal judgment does not compensate for all non-market damages. This estimate implies
a cost greater than $728 per gallon of spilled oil.
•
Douglass Lee estimates annual uncompensated oil spills average $2 billion, totaling
about $0.00 1 per VMT. 13
•
Peter Miller and John Moffet cite leaking underground storage tanks, oil spill cleanup
and road deicing costs, to estimate annual automobile water pollution costs at $3 .8
billion, or $.0013 per VMT.14
111993-1994 California Transportation Energy Analysis Report, CEC, (Sacramento), Feb. 1994, p. 31.
12 Paul Chernick and Emily Caverhill, Valuation ofExternalities from Energy Production, Delivery and
Use, Boston Gas Company (Boston), Dec. 1989, p. 85.
13 Full Cost Pricing of Highways, USDOT, National Transportation Systems Center (Cambridge), p. 21.
14 The Price ofMobility, National Resources Defense Council (Washington DC), Oct. 1993, p.50.
Page 3.15-4
Transportation Cost Analysis
• Murray and Ernst estimate road salting costs at $4.7 billion (in 1993 dollars). 15
• The Office of Technology Assessment study estimates that leaking fuel tanks and oil
spills associated with motor vehicle use costs $1 to $3 billion per year in the U.S . 16
• Transport 2021 estimates external water pollution costs from automobile use to be
$0.002 Canadian per km, or $0.0025 U.S . per VMT, based on a review of studies.
• The Washington Department of Transportation (WSDOT) estimates that meeting its
stormwater runoff water quality and flood control requirements will cost $75 to $220
million a year in increased capital and operating costs, or $0.002 to $0.005 per VMT.
Variability: Hydrologic impacts ofstormwater depend on the amount of paved surface,
so impacts are generally proportional to lane mileage. Water quality impacts are more
closely related to vehicle mileage and maintenance.
Conclusion: Motor vehicles and roads impose a number of water quality and hydrologic
costs, including roadway and parking lot stormwater runoff pollution, flooding and other
hydrologic impacts, petroleum spills, road salting, and habitat loss (especially for fish and
other aquatic animals). Available estimates of these costs range from $0.001-$0.005 .
However, no existing estimate incorporates all identified impacts, so they understate total
costs. The WSDOT's cost estimate for meeting water quality standards for state highway
runoff is notable because it alone exceeds most other estimates despite its limited scope,
implying that water quality and hydrologic costs of roads and motor vehicle traffic are
substantially higher than usually considered.
Here is an estimate of total water pollution costs from roads and motor vehicles:
15
16
Murray & Ernst, Economic A ssessment of the Environmental Impact ofHighway Deicing, EPA 1976.
Saving Energy in U.S. Transportation, U. S. Office of Technology Assessment, 1994, p. 108.
Page 3.15-5
Transportation Cost Analysis
1. State highways account for approximately 5% ofU.S. road miles and 10% of lane
miles, and carry about 50% ofVMTP An estimated 100 million commercial and 200
million residential parking spaces add approximately 30% to total road surface area,
and more than 50% to urban road surface.18 These figures indicate that total water
pollution and hydrologic impacts are significantly greater than just state highway
impacts. State highway runoff impacts are conservatively estimated here to represent
one-third of total roadway runoff impacts, so the middle value ofWSDOT's estimated
cost of meeting its highway runoff mitigation requirements ($75 + $220 I 2 = $147.5)
is tripled to include non-highway roads, parking spaces, and residual impacts ($147.5
x 3 = $442.5 million), and scaled to the entire U.S. road system ($442.5 x 50) for a
total annual national runoff cost of $22.1 billion.
2. Add Douglass Lee's estimate of oil spills ($2 billion).
3. Add Murray and Ernst's estimate road salting costs ($4.7 billion)
This totals $28.8 billion per year, or about $0.013 per automobile mile. Note that this
estimate does not include costs of residual runoff impacts, shoreline damage, leaking
underground storage tanks, reduced groundwater recharge and increased flooding due to
pavement, so it should be considered a conservative value. This cost is applied equally to
all petroleum powered motor vehicles. Although it could be argued that larger vehicles
require slightly more road surface and consume more petroleum products per mile, private
vehicle owners are more likely to allow their vehicles to drip and to dispose of used fluids
17 FHWA
Annual Statistics, 1992, assuming that interstates, freeways and principal arterials represent
state facilities, and other roads are locally owned.
18 Commercial parking estimate from Douglass Lee, Full Cost Pricing ofHighways, National
Transportation Systems Center (Cambridge), 1993, p.21. Residential parking spaces assume that there are
slightly more parking spaces than registered automobiles. Parking lot area is calculated based on 250
parking spaces equal one lane mile.
Page 3.15-6
Transportation Cost Analysis
incorrectly, so overall impacts are considered equal. Electric cars and trolleys are
estimated to have water pollution cost half of an average automobile because they use few
petroleum products, but still require roads and parking spaces. Bicycling, walking and
telecommuting are not considered to impose any significant water pollution cost.
Best Guess
Water Pollution Costs Dollars per Vehicle Mile)
Vehicle Class
Rural
Urban Peak
Urban Off-Peak
Average Car
Fuel Efficient Car
Electric Car
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
0.013
0.013
0.007
0.013
0.00
0.013
0.007
0.013
0.00
0.00
0.00
0.013
0.013
0.007
0.013
0.00
0.013
0.007
0.013
0.00
0.00
0.00
Average
0.013
0.013
0.007
0.013
0.00
0.013
0.007
0.013
0.00
0.00
0.00
0.013
0.013
0.007
0.013
0.00
0.013
0.007
0.013
0.00
0.00
0.00
Automobile Cost Range: The Minimum is based on literature cited. The Maximum is the
estimate developed above doubled to reflect costs not included in this estimate.
Minimum
$0.001
Page 3.15-7
Maximum
$0.026
Transportation Cost Analysis
3.16 Waste Disposal
Definition: External costs of automobile waste disposal.
Description: Disposal of used tires, batteries, junked cars, oil and other semi-hazardous
materials resulting from motor vehicle production and maintenance.
Discussion: Over 70% of Washington's moderate risk waste stream is from automobiles,
and there is no reason to consider this atypical. 1
Moderate Risk Waste
Used Oil (Primarily Automobile)
Batteries (Primarily Automobile)
Antifreeze (Primarily Automobile)
Cleaners, Paints, Adhesives
Pesticides, Other
Percent
50%
15%
7%
21%
7%
Used tires and junked cars also create significant disposal problems.2 Tire piles create
environmental and health hazards, especially when they catch fire. Although efforts are
underway to find uses for waste tires, none have created enough demand to eliminate land
fill disposal. Many junked cars sit for years before they are recycled. Some are simply
abandoned and must be disposed of at public expense. Junked cars impose aesthetic
impacts, and are sources of pollution from fluids, lead batteries, and metals.
These wastes impose a variety of environmental, human health, aesthetic, and financial
costs, through improper disposal, residual impact even when proper disposal is observed,
and because some disposal efforts are subsidized by general taxes. A number of recent
laws and policies are intended to internalize these costs. Crankcase oil recycling networks
1
Problem Waste Study (Moderate Waste), Washington Department of Ecology (Olympia), 1990, p. 12.
1992 Washington State Waste Characterization Study, Washington Department of Ecology (Olympia)
July 1993.
2
Page 3.16-1
Transportation Cost Analysis
have been established, vendors are required to recycle used car batteries, and in some
states a tire tax is dedicated to tire disposal. It is uncertain to what degree these policies
will reduce external disposal costs.
There is the potential of overlap between water quality costs described in the previous
chapter and waste costs described here, since both include waste crankcase oil. Water
quality costs cover impacts of oil and other fluids that drip during vehicle use. Waste costs
address impacts of oil and other fluids after their useful life, during disposal. A review of
the previous chapter will show that there is no overlap in calculating these costs.
Estimates:
• Douglass Lee estimates the following external disposal costs:
Table 3.16-1
Aut
--- -
----
--
-
--
- ·- ·- -
bile
Ext
·- - -- - -----I Waste
-- - n·
---
..
-
-
Product
Annual Volume
Unit Costs
Waste Oil
Scrapped cars
Used tires
960 million quarts
2.82 million
3 billion
$0.50
$25
$1
Total
~
~
I Cost Estimat..3
Total Annual Cost
$0. 5 billion
$0.7 billion
3.0 billion
$4.2 billion, $0.002 per VMT
Variability: Impacts depend on vehicle design, construction and user waste management.
Conclusions: Waste disposal has been a significant problem of automobile production and
use. Lee's estimate that U.S . external motor vehicle waste costs total $4.2 billion per year
seems reasonable. Although it may overstate some waste costs if new management efforts
are successful, it excludes other wastes altogether. This cost is applied equally to all motor
vehicles. Although electric vehicles do not create waste oil, they do produce used
3 Full
Cost Pricing ofHighways, National Transportation Systems Center (Cambridge), p. 31.
Page 3.16-2
Transportation Cost Analysis
batteries, hulks and tires. As described in chapter 3.15 (Water Pollution), although buses
and trams may produce more waste per vehicle, their waste tends to be managed better
than those of private vehicles, so costs are considered equal.
___
,...
___
...
..........
.. ---- ---r----
Vehicle Class
Average Car
Fuel Efficient Car
Electric Car
- ------ r--
. -------
Urban Off-Peak
0.002
0.002
0.002
0.002
0.00
0.002
0.002
0.002
0.00
0.00
0.00
0.002
0.002
0.002
0.002
0.00
0.002
0.002
0.002
0.00
0.00
0.00
Van
Rideshare Passenger
Diesel Bus
Electric Busffrolley
Motorcycle
Bicycle
Walk
Telecommute
~----
Urban Peak
- - - --- - --- -- -
-·----.r
Rural
Average
0.002
0.002
0.002
0.002
0.00
0.002
0.002
0.002
0.00
0.00
0.00
0.002
0.002
0.002
0.002
0.00
0.002
0.002
0.002
0.00
0.00
0.00
Automobile Cost Range: Due to the uncertainty of this cost and its relatively small
magnitude, the minimum cost is zero. The maximum is 2.5 times the estimate used here.
Minimum
$0.00
Page 3.16-3
Maximum
$0.005
Transportation Cost Analysis
4.0 Costs Totals
Chapters 3 .1 through 3. 16 provide Best Guesses of 20 costs for 11 modes under three
travel conditions, totaling 660 individual estimates. These were put into a spreadsheet for
calculating statistics. This chapter summarizes the results.
When reviewing these estimates it is important to remember:
• They include non-market costs such as users' travel time, accident risk, and
environmental impacts, which is why they are higher than most travel cost estimates.
• Estimates are based on average vehicles and conditions. Costs may differ in specific
situations.
•
Some estimates describe costs per passenger mile not per vehicle mile, assuming
average vehicle occupancy.
4.1
Summary Graphs
The following graphs summarize the results. Figure 4-1 shows average automobile costs
per vehicle mile. Figure 4-2 shows costs for an Average Automobile traveling under
Urban Peak, Urban Off-Peak, and Rural conditions. Figure 4-3 compares total average
costs for each of the eleven modes.
Page 4-1
Transportation Cost Analysis
Costs Per Vehicle Mile for Average Automobile
lUI
D Internal Costs
• External Costs
a
l1t211
J
~
l
11.00
"'
<:::
"'e
8.
0
~
0
:E
~
a.
Q)
:E
~
E
i=
~
~
;:)
Q)
" "
Q)
Q)
"0
"0
·o
~
m
E
Q)
0
:E
~
E
~
·o
~
m
E
~
"'
<:::
~
~
~"'
"'
Q_
'E" '"
Q)
"
-
E
<:::
0
Kl
;g
~
·o
<:::
lJ._
"'
8
~
"'
"0
~
Q)
"
~
~
·~
"0
<:::
c?l
-'
!·o
"'
IX
.~
cr
"
w
<:::
0
'§
'5
Q_
il!
z
·o
~
"~
Q)
IX
~
~
w
Iii
.E
~
·c
{!
~
E
-
~
s:"'
*
s:"'
il!
;:)
"0
<:::
~
"
::::;:
1his graph shows average costs per vehicle mile.
Costs for Average Automobile Under Three Travel Conditions
10.30
•Urban Peak
• Urban Off-Peak
ORural
10.25
:!
:IE 10.20
~
i
>
10.15
.!
10.10
:e
8
10.05
10.00
. .,c. .,c. "' "'
:f.
:f..
u
. ~X.
a..
a..
"'
~.
.
8=
=
u
u
j
:c :c
- ~ ...
:c
I!!
"'c
c
0
~
~
E
i=
~
c
·o;
u
u
C(
C(
;;;
;;;
E
E
.$
c
c
;;;
E
.$
c
;;;
c
0
c
=:
:£!
i...
i
..
::0
.,~c
j
&
w
=
u
·~
Jl
iu
c::0
~
::0
a[
c
0
~
0
a..
...
c
=
0
z
..I!. 'E....
::0
J
.3
.,=
c
~
..
j
I
..J
::E
Some costs are significantly higher under urban and peak-period travel conditions. 1
1 Note that time and internal accident costs are higher per vehicle for off-peak than for peak travel. This
is because urban peak driving tends to have lower vehicle occupancy than driving under other conditions.
Page 4-2
Transportation Cost Analysis
Total Costs Per Passenger Mile for Eleven Modes
12.110
$1.75
•uman Peak
J!
• Urban Off-Peak
DR ural
$1.50
i
t$1.~
••
: $1.110
l
SG.75
~
!
8
$8.50
·~
$0.110
Average
Car
Fuel
Efficient
Car
Electric
Car
Van
Rldeshare Diesel Bus
Passenger
Electric
Bus/
Trolley
Motorcycle
Bike
Walk
Telecommut
This graph compares total costs of each travel mode under the three travel conditions.
There is an important difference between public transit rider costs and typical private
vehicle. Public transit ridership tends to experience increasing economies of scale, since
many costs are fixed and most transit systems have excess capacity. Therefore, the cost
estimates for Diesel Bus and Electric Bus/Trolley overstate marginal costs. The marginal
cost ofbus and trolley riders is best reflected in the Rideshare Passenger cost estimate. 2
Private vehicle costs tend to experience diseconomies of scale due to increasing
congestion and other external costs, and so may understate long-run marginal costs.
4.2 Cost Distribution
The twenty costs can be grouped into categories defined in Table 4-1 .
2 Note
that Rideshare cost estimates are based car pooling costs and do not include a user fare.
Page 4-3
Transportation Cost Analysis
Table 4-1
•
T ranst~ortauon Coste
Cost
Vehicle Ownership
Vehicle Operating
Operating Subsidies
User Time
Internal Accident
External Accident
Internal Parking
External Parking
Congestion
Road Facilities
Land Value
Municipal Services
Equity and Option
Air Pollution
Noise
Resources
Barrier Effect
Land Use Impacts
Water Pollution
Waste
Internal/
External
Internal
Internal
External
Internal
Internal
External
Internal
External
External
External
External
External
External
External
External
External
External
External
External
External
Fixed/
Variable
Fixed
Variable
Fixed
Variable
Variable
Variable
Fixed
Fixed
Variable
Variable
Variable
Variable
Variable
Variable
Variable
Variable
Variable
Variable
Variable
Variable
3
Market/
Non-market
Market
Market
Market
Non-Market
Non-Market
Non-Market
Market
Market
Non-Market
Market
Non-Market
Market
Non-Market
Non-Market
Non-Market
Non-Market
Non-Market
Non-Market
Non-Market
Non-Market
Tax Based
Social/
Environmental
Tax Based
20% Tax Based
Tax Based
Tax Based
Tax Based
Social
Environmental
Environmental
Env. & Soc.
Social
Environmental
Environmental
Environmental
This table indicates how costs are classified for analysis. Since internal variable costs,
have the greatest effect on users' decisions, the distinction between internal and external,
fixed and variable indicates the effect a cost is likely to have on travel demand
Table 4-2 summarizes the cost distribution between five major cost categories for average
automobile use. Figure 4-4 illustrates this same information in graph form.
Table 4-2
Cost Distribuf
Variable Vehicle
Fixed Vehicle
User Time & Risk
External Market
External Non-Market
Total
Urban Peak
$/Mile
Percent
$0.16
12%
$0.25
19%
$0.31
23%
$0.18
13%
$0.43
33%
$1.33
100%
Vehicle Mile for A
Urban Off-Peak
$/Mile
Percent
13%
$0.14
$0.25
23%
$0.33
31%
$0.09
9%
$0.25
24%
$1.06
100%
Aut -------bil -
Rural
$/Mile
Percent
$0.11
14%
27%
$0.23
36%
$0.30
$0.06
7%
16%
$0.14
100%
$0.84
Avera2e
$/Mile
Percent
13%
$0.13
23%
$0.24
32%
$0.34
9%
$0.10
23%
$0.24
$1.05
100%
Costs can be divided into five major categories. Variable Vehicle costs are one of the
smallest categories, yet this is the cost category that most affects vehicle use. External
costs are significant in each of the three driving conditions.
3 These
categories indicate general tendencies and are not absolute. There are exceptions, but they are
considered minor.
Page 4-4
Transportation Cost Analysis
Figure 4-4
Average Automobile Cost Distrr-:i:..:::b.. :;u:..:.t•=-=·0="=-------,
• External Non-Market
• External Market
D User Time & Risk
$1 .40
$1 .20
.!!
i
$1 .00
Ql
:g
~ $0.80
£ $0.80
~
~ $0.40
8
$0.20
$0.00
Urban Peak
Average
Rural
Urban Off-Peak
This graph illustrates the costs in Table 4-2.
Figure 4-5 shows costs for each mode divided into major categories. Travel time and
accident risk is the largest cost category for most modes. Automobilefc operating costs
alone are lower than those of transit riders, indicating that car owners usually find it
cheaper to drive than to ride a bus. The ratio between internal and external costs varies
significantly between vehicles. Non-market external costs as a percentage of total costs
rank from motorcycles, vans, average automobiles, fuel efficient cars, electric cars, transit
vehicles, telecommuting, bicycling, rideshare passengers, and walking. These differences in
ratio between internal and external costs increase for Urban Peak travel.
Cost Distribution for Various Modes
Figure 4-5
~
:!!:
tt.50
• External Non-Market
$1.25
DTravel Time & Risk
• External Market
...
• Fixed Vehicle Costs
II
~ tt.OO
.,
= 10.75
II
Q.
...
II
Q.
10.50
~
.!
~
10.25
10.00
Average
car
Fuel
Efficient
Car
Electric
Car
Van
Rldeshare Diesel Bus
Passenger
Electric
Bus/
Trolley
Motorcycle
Bike
Walk
This graph shows the average values for five major cost categories.
Page 4-5
Telecommut
Transportation Cost Analysis
Figure 4-6 shows the internal costs for each vehicle type further disaggregated. Travel
time is the largest cost for most modes, and especially dominates bicycling and walking, so
their total costs are extremely sensitive to the value used. Internal accident risk is the
largest cost for motorcycling.
Figure 4-6
Internal Costs
$1.26
•lnt Parking
•lnt Accident
$1.00
.!
DUserTime
i...
• Vehicle Ownership
II
~ $0.75
• Vehicle Operating
..=
:.
: $0.50
~
.!
8$0.26
$0.00
Average car Fuel Efficient Electric car
car
Van
Rldeshare
Paoeenger
OleMI Buo
Electric Bu8l
Trolley
Motorcycle
Bike
walk
Telecomm,_.
This graph shows the internal costs of each travel mode.
Figure 4.7 shows external costs disaggregated. Environmental and Social is the largest
category for automobiles and motorcycles, but is relatively low for other travel modes.
This is true even of electric cars, although this cost is lower than for gasoline vehicles.
External parking costs are another major category. During peak periods congestion is a
significant cost, but becomes a relatively small cost when averaged over all mileage. Bus
and trolley average passenger costs are dominated by tax based operating subsidies but
the marginal cost of an additional rider, indicated by Rideshare Passenger, incurs virtually
Page 4-6
Transportation Cost Analysis
no external costs. Since transit passenger fares exceed marginal operating costs, the
marginal cost to the system per additional rider is actually negative: the more paying
passengers the bus system carries the lower the systems' average operating cost.
Figure 4-7
10.50
Average External Costs
• Environmental & Social
•Tax Based
o Congestion
• Ext Parking
• Ext Accident
10M
!
i
&10.30
i
f
r
•~
10.20
8
10.10
10.00
Average
Car
Fuel
Efficient
Car
Electric
Car
Van
Rldeshare Diesel Bus
Passenger
Electric
Bus/
Trolley
Motorcycle
Walk
Bike
Tetecommute
This graph shows the external costs of each travel mode.
4.3
Total Transportation Costs
Table 4-3 shows estimated U.S . motor vehicle travel costs based on this analysis.
Table 4-3
U.S. Motor Vehicle Costs.' bv Mil
Mileage
Units
Urban Peak
Urban Off-Peak
Rural
Total
billions
460
920
920
2,300
Internal Costs
Total
per mile
billions
$0.71
$0.71
$0.64
$327
$653
$559
$1,539
d Total
External Costs
Total
billions
per mile
$0.61
$0.34
$0.20
$281
$313
$184
$778
Total Costs
Total
billions
per mile
$1.32
$1.05
$0.84
$607
$966
$773
$2346
This table summarizes total motor vehicle costs based on the estimates from this report.
Page 4-7
Transportation Cost Analysis
That internal costs are the same for Urban Peak and Urban Off-Peak travel ($0. 71 per
mile) is unexpected, since the stop-and-go driving of peak period traffic increases travel
time, stress and vehicle operating costs. However, this is offset by the lower automobile
occupancy rates (approximately 1.2 average) compared with off-peak travel (greater than
1.5 average), resulting in comparable total travel time and higher accident risk costs.
This total of$2.3 trillion, which equals over 40% the U.S . Gross National Product (GNP),
may seem unreasonably high, especially since this does not include other transportation
costs such as air travel and rail. The explanation of this apparent anomaly is that a major
portion of these costs are non-market, including personal time, accident risk, and
environmental degradation. Only six of the twenty cost categories in this study are direct
market costs (although these non-market costs impose significant indirect market costs
such as medical care and disability costs from accidents and air pollution). Thus, the
majority of these costs are not incorporated into GNP calculations. The market economy
can be imagined as the tip of a pyramid that consists of the larger non-market human
economy (unpaid travel time, housework, childcare, and volunteer activities that
contribute to society), atop an even larger non-market natural economy that provides
clean air, water, and beauty, in addition to marketed natural resources (Figure 4-8).
Figure 4-8
Market, Human Non-Market, and Natural Non-Market Goods
Natural Non-Market Goods
Markets represent only a small portion of all human benefits. A larger portion of benefits
are provided by human non-market and natural non-market goods. Many of the costs of
transportation represent losses of these goods.
Page 4-8
Transportation Cost Analysis
Our total endowment ofwealth, including non-market goods, is much larger than just
market activities. We are richer than indicated by just our financial assets due to these
non-market goods. Non-market costs, such as accident risk, lost time, and environmental
degradation, represent the loss of non-market resources to ourselves, to other members of
society, and to future generations. This estimate indicates what it would cost ifwe paid for
the non-market goods consumed by transport.
4.4
Cost Ranges
As discussed in Chapter 1, section 1. 10, the cost estimates developed in this report
incorporate various degrees of uncertainty. Although point estimates are necessary for the
analysis in this chapter, such estimates actually represent a range of possible costs that
depend on particular circumstances and uncertain input data. The Minimum and Maximum
estimates represent the widest reasonable range of average automobile costs. They can be
used for sensitivity analysis. Figure 4-9 illustrates these ranges.
Ranges of Average Automobile Costs
Figure 4-9
$OM
..
i
DMaxlmum
• Point Estimates
•Minlrn.Jm
$0.31
.!
2i
0
E
-s0
c...
$0.21
:.
I!!
.!!
8
$0•••
$0.H
1
u::
)(
i
:::>
..
.!
.1:1
·c:
~
;::
.. •!!
>...
>
..,'.C
·u
u
<(
;;;
..,'•C
uu
:e"."' :e"."' 1"
... ... "'
0
.. .. ...
<(
E
•
;Ji
E
~
E
"0
u
..,•
!
0:
....,•
:I
.
>
..."
I!
rll
u
c
~
:I
c:r
w
:I
ll
;Ji
This graph shows the most likely range for each cost.
Page 4-9
"
.2
;;
......
0
:
·c;
z
~
:I
i
0:
....
'E
.:1
I
.
..,:::>
..."
i
i
•
tl
~
Transportation Cost Analysis
Figure 4-10 shows Minimum, Point Estimate and Maximum costs by major category. The
greatest variation occurs among travel time, internal accident risk, and external costs.
Note that even the Minimum estimate shows external costs to be approximately equal to
the Maximum estimate of variable vehicle costs (fuel, oil, tires, maintenance, and short
tenn parking). Thus, even if the lowest reasonable values are used for each cost,
externalities are still significant relative to the costs most often considered in private and
public transportation decisions.
Figure 4-10
Ranges of Average Automobile Costs by Major Category
$1 .75
•external Market
$1 .50
CJExternal Non-Market
•rime & Internal Risk
$1 .25
•veh . Ownership
I
t
l
•veh. Operating
$1 .00
$0.75
~
$0.50
$0.25
$0.00
Minimum
Point Estimate
Maximum
This graph compares cost range by major categories. Even the lowest estimate indicates
that external costs are significant.
4.5
Survey Test of Cost Estimates
Since limited data is available on some of the costs included in this study, a preliminary
survey was performed to determine whether the public's ranking of transportation costs is
consistent with this study's results. One hundred eleven (Ill) surveys were distributed to
Page 4-10
Transportation Cost Analysis
households randomly selected across North America. Ofthose, 11 were returned as
undeliverable.4 Thirty-eight completed survey forms were received.
The survey asked respondents to identify how serious they consider various transportation
problems. The responses were numbered from 1 (Very Serious) to 4 (Not At All Serious).
Table 4-4 shows the survey results. These indicate that the public considers social and
environmental transportation costs significant. Even the lowest ranking cost, Ugliness of
roads, has a value indicating that respondents, on average, considers it between "Not Very
Serious" and "Somewhat Serious." Costs ranked according to the survey show a strong
correlation to the ranking of average automobile costs in this study, with the exception of
urban sprawl. This may be explained by the technical nature of many sprawl costs.
Table 4-4
----
.
Public Survey and This Study's Cost Estimate Ranking Compared
-- -- -- - - -
Survey
Rank
-------
----------
- ----
...
------
---
- --
Rankin This
Study
Transportation Problems
1
1
2
3
4
5
6
3
5
4
7
8
9
6
2
10
Traffic accidents
Air Pollution
Excessive energy consumption.
Traffic congestion
Barrier Effect
Traffic Noise
Mobility problems for non-drivers
Harm to wildlife caused by roads and traffic.
Urban sprawl
Ugliness of roads
7
8
9
10
Survey
Averaee
Variance
1.53
1.59
1.74
1.89
2.19
2.20
2.24
2.27
2.37
2.79
0.72
0.45
0.37
0.84
0.50
0.87
0.50
0.64
0.86
0.90
This table shows that survey respondents gave similar rankings to transportation
problems as the cost estimates in this survey. Note that the lower the average value, the
more serious respondents consider the problem.
A second question asked respondents to identify how important they consider various
transportation goals (Table 4-5). "Very Important" counted as a 1, while "Not At All
Important" counted as a 4. Although these questions are more difficult to compare
4 Additional
letters were probably undelivered but not returned across the international border
Page 4-11
Transportation Cost Analysis
directly, the results are consistent with this study's cost estimates. The top ranking of
Develop a more diverse transportation system, and Provide better transport to poor,
handicapped, and elderly, support the concept of transportation equity and option value,
and indicate that these costs are, if anything, underestimated in this study. Similarly, the
high ranking of Reduce environmental impacts and Reduce urban impacts indicate that
the public perceives environmental degradation and negative social impacts to be
significant external costs of our current transport system.
Table 4-5
Survey Ranking of Transportation Goals
-
Rank
Question
Averaee
Variance
1
2
3
Develop more diverse transportation system.
Provide better transport to poor, handicapped, and elderlv.
Reduce environmental impacts.
Reduce urban impacts.
Reduce/avoid urban sprawl
Accommodate increased driving
1.26
1.58
1.74
1.77
2.21
2.34
0.20
0.44
0.81
0.39
0.92
0.81
4
5
6
While its small size and methodological limitations prevent this survey from providing
conclusive results, it supports this study's estimates oftransportation costs and
demonstrates that surveys can be useful for this research. Survey results indicate that the
magnitudes of this study's cost estimates are appropriate, and that costs which have
previously been ignored in transport planning, such as Equity, Option Value, the Barrier
Effect, and Land Use Impacts, may be even greater than estimated here.
4.6
Summary
If you ask people what it costs to drive they typically mention vehicle operating expenses,
which average approximately 12¢ per mile for a typical car. Some may also include a
portion of vehicle ownership costs, which averages about 21¢ per mile. A few may also
Page 4-12
Transportation Cost Analysis
mention the value of travel time and accident risk. These however are only a portion of
total costs. The full cost of driving includes all of these internal costs plus a significant
number of external costs. Total costs actually range from about $0.84 per vehicle mile for
rural driving to $1 .3 3 for urban peak driving. Of course there is considerable variation in
the cost of any specific trip, but these estimates, and variations for different travel modes
and specific conditions, provide a reasonable basis for analyzing true transport costs.
Some specific cost estimates used here are of uncertain precision, but this does not change
the analysis conclusions. The existence of each cost has been demonstrated, double
counting is avoided, and the best available data are used. Even using the lowest reasonable
cost estimates, total external costs are significant.
Page 4-13
Transportation Cost Analysis
5.0
Travel Elasticities and Generated Traffic
Transportation improvements and cost reductions can encourage more and longer trips,
changes in travel patterns, and land use changes which require special consideration when
assessing benefits and costs. Current transportation planning often fails to do this,
resulting in incorrect conclusions. This chapter describes how increased travel and related
impacts should be assessed, and provides analysis tools for doing this.
5.1
Introduction
Transportation improvements that reduce user costs tend to increase travel and divert trips
from other routes, times and modes. This is called generated traffic. Economic analysis
requires that all incremental (net) benefits and costs be including in the evaluation of
policies, programs and projects. Incremental costs and benefits to existing trips are
relatively easy to determine, but special consideration is needed to determine net benefits
and costs of generated traffic.
Generated traffic has three important implications on transportation decision making.
First, as discussed in greater detail later in this chapter, generated travel tends to provide
relatively little user benefit, since these are trips that users chose to forego when traffic
conditions are suboptimal.
Second, it erodes a portion of the congestion reduction benefits that are often predicted
for transportation improvements.1 A recent report by the UK Standing Advisory
Committee on Trunk Road Assessment concludes, "These studies demonstrate
convincingly that the economic value of a [road] scheme can be overestimated by the
omission of even a small amount of induced traffic. We consider this matter ofprofound
1 This
is the same as the "take back" or "snap back" effect found by energy planners, in which consumers
increase their energy use a result of conservation efforts that reduced their unit energy costs.
Page 5-l
Transportation Cost Analysis
importance to the value-for-money assessment of the road programme. '12 A recent study
found that the ranking of preferred projects changed significantly when generated traffic
"feedback" is incorporated into conventional project assessment analysis.3 Specifically,
capacity expansion options were found to provide less congestion reduction benefit and
negative air emission effects, while demand management and No Build options have more
relative benefits.
The third implication is that generated travel increases total transport costs, including the
external costs identified earlier in this report. While users' marginal benefits exceed their
marginal costs of increased travel (if not, users would not take the additional trips), these
benefits do not necessarily exceed total incremental costs. Generated traffic, therefore,
may create more costs than benefits.
Traffic models used in some large urban areas incorporate generated traffic feedback, and
U.S. Clean Air laws require increased use of such models. But models used in medium
and small communities usually omit this step, and generated traffic costs are usually
ignored in the economic analysis of specific projects.4 Motor vehicle use is frequently
assumed to grow at a constant rate, unaffected by road improvements. Ignoring the effects
of generated traffic in economic analysis tends to overstate benefits and understate costs
of roadway improvements, leading to non-optimal transport investments. A common
excuse for this omission is that no tools exist to predict how much traffic will be generated
or to determine resulting net costs. These excuses are no longer justified.
2 SACTRA,
Trunk Roads and the Generation of Traffic, UKDoT, HMSO (London), 1994.
Johnston and Raju Ceerla, A Comparison ofModeling Travel Demand and Emissions With
and Without Assigned Travel Times Fed Back to Trip Distribution, Institute of Transportation Studies,
University of California at Davis, 1994. Submitted to the Journal ofTransportation Engineering.
3 Pro. Robert
4 H.C.W.L.
Williams and W.M. Lam, "Transport Policy Appraisal with Equilibrium Models I: Generated
Traffic and Highway Investment Benefits," Transportation Research B , Vo. 25, No 5, pp. 253-279, 1991.
Page 5-2
Transportation Cost Analysis
5.2
Transportation Elasticities
Abasic rule of economics (and common sense) states that products which are cheaper or
more convenient will be used more. This applies to transportation, not only for financial
costs, but also for improvements in travel time, convenience and comfort. Economists
measure the sensitivity of this effect using elasticities, which is defined as the percentage
change in consumption caused by a percentage change in user costs.5
Let's consider how a reduction in price or an increase in travel speed increases driving.
First, consider the elasticity of all travel. For example, rank all of the trips that you might
consider making during a certain time period from highest to lowest value, as illustrated in
Figure 5-1 . There are typically some high value trips (such as urgent medical services,
commuting, major shopping trips, special social and recreational events), some medium
value trips (such as less important errands and less enjoyable social and recreation
activities), and some low value trips (such as frivolous errands and the least enjoyable
social and recreation activities). If travel costs increase you will forego the lower value
trips, but take them if prices decline.
Figure 5-l
Individual's Travel Ranked by Value
=
::;)
.
s
::J
~
Enwgoncy
-
Sc:OOd
Mojor
Shopping
Spodll
E-
Soc:illi>lng
Dl'*>g Ola
lrrp&o
Rocrootlon
Jot-
Shopping
TypeofTr1p
The trips you take vary in importance. Some trips you will take only ifyour cost
(including financial, time and discomfort) is low.
5 For
example, a price elasticity of driving with respect to fuel of -0.5 means that a 1% increase in fuel
induces a 0.5% reduction in driving.
Page 5-3
Transportation Cost Analysis
If you create the same type of graph for an entire community it would include thousands
or millions of potential trips, as shown in Figure 5-2. This is a travel demand curve. It
would typically be a more-or-less straight line, indicating a relatively consistent sensitivity
of consumption (travel) to users' costs.
Figure 5-2
Travel Demand Curve (All Travel)
a.
'1:
.....
:
~
..
""
'tl
c
~
Number of Trips
Considering all trips made in a community, some have higher value to users than others.
A demand curve for just automobile trips will look somewhat different. Rather than being
a straight line it is typically concave, indicating a higher sensitivity to price, as shown in
Figure 5-3 . Why? Because if the price of driving increased, users can either forego the trip
or take another mode. If the price of driving decreased, it would increase total trips and
attract trips from other modes. These two effects are additive. In general, the more
narrowly a good is defined the more concave its demand curve because consumers have
more alternatives. For example, the demand curve for peak period automobile trips along
a certain corridor would be even more sensitive to price changes (more convex) since
users can change their amount of travel and shift trips to other routes, times or modes.
Page 5-4
Transportation Cost Analysis
Figure 5-3
Travel Demand Curve (All Travel and Auto Travel)
· · · · · All Travel
-Auto Travel
~
=:
;::)
Number of Trips
Automobile travel is more sensitive to cost changes than travel taken as a whole, because
users can shift to andfrom other travel modes. This results in a concave demand curve.
Elasticities depend on several factors. Some trips are price inelastic because they are
highly valued and users may have few travel choices, while others are quite sensitive to
price either because the trip itself is discretionary or because there are substitutes, which
can include alternative destinations, times and modes. Transport overall is a major portion
of most household budgets, which implies high elasticity, but many costs are fixed, so the
user's marginal cost of any one trip may be small, reducing their elasticity.
Elasticities are affected by time. 6 In the short term about the only way consumers can
reduce their fuel consumption is by eliminating trips and shifting destinations when
possible, or using existing travel alternatives. In the medium term they can also buy more
efficient cars and choose housing and job locations that require less driving. In the long
term additional land use and transport changes can occur that reduce automobile
dependency. Thus, it may take many years for the full effect of a price increase to be felt.
Short term is typically less than two years, medium term is two to 15 years, and long term
is 15 years or more, although definitions vary. Large price changes tend to be less elastic
than small price changes, since consumers make the easiest accommodations first.
6 J.M.
Dargay and P. B. Goodwin, "Estimation of Consumer Surplus with Dynamic Demand Changes," in
Proceedings of European Transport Forum, PTRC, Sept. 1994.
Page 5-5
Transportation Cost Analysis
s) p
Country
Canada
USA
Austria
Belgium
Denmark
Finland
France
Germany
Greece
Ireland
Italy
Price
-2.0
-1.2
-1.2
-1.5
-0.8
-1.2
-0.4
0.1
0.2
-1.0
-0.7
Income
0.5
1
1.2
1.3
0.7
1.3
1.2
0.5
2.0
0.9
1.3
1992)7
Country
Netherlands
Norway
Portugal
Spain
Sweden
Switzerland
UK
Australia
Japan
Turkey
Mean
Price
-3 .2
-2.5
-0.7
-1.2
-0.1
0.2
-1.4
-0.2
-0.3
-1.1
-1.0
_L
Income
0.6
1.3
1.9
2.1
1.2
1.5
1.5
0.7
0.8
1.3
1.2
In the long run fuel consumption is quite sensitive to changes in price and user income.
Goodwin estimates the price elasticity of gasoline at -0.27 in the short term and -0.7 in the
long term, meaning that a 10% increase in fuel price reduces fuel consumption by 2.7% in
the short term and 7% in the long term. 8 Dargay reports higher values averaging -0.67
when price increases and decreases are calculated separately. 9 Sterner et al. find long run
North American fuel elasticities to be greater than 1.0, as shown in Table 5-1. Kageson
cites studies indicating that the elasticity of fuel is -0.2 to -0.3 in the short run, and -0.8 to
-1.0 in the long run, most of which results from increased fuel efficiency. 10 Schipper and
Johansson estimate vehicle ownership, use and fuel consumption elasticities.11 They
conclude that the long run elasticity of driving with respect to fuel price is -0.3.
John DeCicco and Deborah Gordon conclude that the medium-term elasticity of motor
vehicle fuel in the U.S. is probably -0.3-0.5.12 They point out that CAFE standards have
artificially increased U .S. fleet vehicle efficiency enough that consumers are unlikely to
7
Sterner, Dahl, Frazen, "Gasoline Tax Policy, Carbon Emissions and the Global Environment," Journal
ofTransport Economics and Policy, 26/2, p. 109-119, Cited in Works Consultancy, 1993.
8 Goodwin, "Review of New Demand Elasticities," Journal ofTransport Economics, May 1992, p.157.
9 Joyce Dargay, "Demand Elasticities," Journal ofTransport Economics," January 1992, p. 89.
10 PerKageson, Getting the Prices Right, European Fed. for Transport & Env., 1993, p.175.
11 Lee Schipper and Olof Johansson, Measuring Long-Run Automobile Fuel Demand, TRB Annual
Meeting (Washington DC), Paper #950168, January 1995.
12 DeCicco and Gordon, Steering with Prices: Fuel and Vehicle Taxation and Market Incentives for
Higher Fuel Economy, American Council for an Energy Efficient Economy (Washington DC), Dec. 1993.
Page 5-6
Transportation Cost Analysis
respond significantly to small increases in fuel prices, such as an additional $0.05 per
gallon tax. They refer to this as a "rebound" effect, because increasing fuel efficiency
reduces users' marginal costs, encouraging more driving and reducing the effect of fuel
price hikes on consumption.
forV -- · -
Table 5-2
Trip Type
-....
....
13
Elasticity of Road Travel with Respect to
Out of Pocket Expenses
Urban shopping
Urban commuting
Inter-urban business
Inter-urban leisure
Freight
-2.7 to -3.2
-0.3 to- 2.9
-0.7 to -2.9
-0.6 to -2.1
-0.6 to -2.0
These are elasticities offuel use with respect to fuel price. Although the major variable
financial cost of driving, fuel accounts for only about 15% of users' financial costs. It is
therefore not surprising that vehicle use does not decrease significantly in response to
moderate change in fuel prices, since this only represents a very small change in total user
costs. Increased fuel prices will cause a combination of reduced driving and increased fuel
efficiency (especially in the long run).
Our concern is with motor vehicle use, not just fuel consumption. Kenneth Button's
elasticity estimates of driving with respect to user out ofpocket expenses for various types
of trips show relatively high values, Table 5-2. Oum, et al. estimate the elasticity of
automobile use with respect to overall price is -0.23 in the short run and -0.28 in the long
run, with a wide range ofvariations.14 Although most discussions oftravel elasticities
focus on financial costs, reductions in congestion and other transport improvements that
save travel time typically increase travel and attract shifts from other modes, which
13
14
Button, Market and Government Failures in Environmental Management, OECD (Paris), 1992, p.53.
Tae Hoon Oum, W.G. Waters II, and Jong-Say Yong, "Concepts of Price Elasticities of Transport
Demand and Recent Empirical Estimates, Journal of Transport Economics, May 1992, pp. 139-154.
Page 5-7
Transportation Cost Analysis
indicates the elasticity of driving with respect to time costs. Similar elasticities exist for
comfort and probably prestige, although there is little quantified data on these.
Table 5-3
Estimated Elasticit ·
Cost Component
Out-of-Pocket Price
Fuel (work)
Fuel (non-work)
Highway tolls
Parking fees
Time Costs
Riding time
Parking search
Congestion
Cost of Alternatives
Transit fare
Transit access time
Elasticities: Low= 0 to 0.5;
fVMTwithR'•es pee t to User Costl5
Short Run Effect
Lon2 Run Effect
-Low
-Medium
-Medium
-Low
- Low to Medium
- Medium to High
-High
-High
-Low
-Low
-Low
-Medium
-High
-High
+Low
+Low
+Low
+Low
Medium = 0.5 to 1.0;
High= 1.0+
Terry Moore and Paul Thorsnes indicate that driving and other transportation activities are
relatively elastic when properly measured, especially in the long run, shown in Table 5-3 .
The Australian Road Research Board publishes travel elasticity estimates in shown in
Table 5-4.
Table 5-4
Australian Travel Demand Elasticities 16
Elasticity Type
Petrol consumption and petrol Price
Travel level and petrol price
Bus demand and fare
Rail demand and fare
Mode shift to transit and petrol price
Mode shift to car and rail fare increase
Road freight demand and road/rail cost ratio
Short-Run
-0.12
-.10
-0.29
-0.35
+0.07
+0.09
-0.39
15
Long-Run
-0.58
-0.80
Terry Moore and Paul Thorsnes, The Transportation/Land Use Connection, American Planning
Association (Chicago), Report #448/449, Washington DC, 1994, Appendix B.
16 James Luk and Stephen Hepburn, New Review ofA ustralian Travel Demand Elasticities, Australian
Road Research Board (Victoria), December 1993.
Page 5-8
Transportation Cost Analysis
Transportation modelers have developed elasticity coefficients for various cities and trip
types that include vehicle access time (both walking and waiting), vehicle travel time,
vehicle costs, and parking costs.17 Greig Harvey summarizes a variety of transport
elasticity estimates, including toll prices, fuel taxes, transit fares, and parking pricing. 18
Since fuel represents about 15% of total vehicle costs, a -0.2 elasticity of driving with
respect to fuel prices represents an elasticity of -1 .4 with respect to the total financial costs
of driving. In other words, if all user costs were converted into a single variable charge,
current fuel price elasticities imply that a 1% increase in this user charge would reduce
driving by -1.4%. This implies a relatively high degree of elasticity.
The hypothesis that driving is actually relatively elastic with respect to total user charges is
further supported by elasticity estimates of driving with respect to parking price. Shoup
and Willson found that charging employees for parking tends to reduce solo commuting
by 20-40%. They estimate the employee parking elasticity of demand at -.16, 19 which
means that a 10% increase in parking charges reduces employee SOV commuting by
1.6%. Assuming a $30 average monthly parking fee and average monthly user costs of per
automobile of $3 80, a -.16 elasticity of employee parking implies a total price elasticity of
about -2.0 with respect to total user financial costs of driving.
Research indicates that cross elasticities between driving and other travel modes are highly
sensitive to land use patterns, transit service quality, and the ease of walking and
bicycling.20 Increased urban density, improved pedestrian facilities and transit service, and
increased prices for driving are estimated to have a greater effect on reducing VMT when
17 Travel
Model Improvement Program, Short-Term Travel Mode/Improvements, Technology Sharing
Program, USDOT (Washington DC), 1994, Table 7.1 & 7.2.
18 Greig Harvey, "Transportation Pricing and Travel Behavior," Curbing Gridlock, National Academy
Press, 1994.
19 Donald Shoup, "Employer-Paid Parking," Transportation Quarterly, April1992, 46/2, p. 172.
20 See research publications by LUTRAQ program, 1000 Friends of Oregon (Portland).
Page 5-9
Transportation Cost Analysis
implemented together than each could have individually. According to one estimate,
automobile user costs would have to rise 300% to reduce VMT by 33%, but if
accompanied by density increases near transit, better transit speeds, and traffic congestion,
pricing would have a much greater effect. 21 Robert Johnson and Raju Ceerla state,
"Since the work trip is so unresponsive to price increases (demand is inelastic), good
transit service to work centers was found to be needed ..Increased auto operating costs
per se was found to increase transit travel to work in the various regions, especially if
good radial service (to the urban center) was simulated 22
The actual synergetic effects of these factors are currently uncertain, but elasticity
estimates in Table 5-5 are probably the lower bound where effective land use and TDM
programs are implemented, and an upper bound where no such efforts are made.
Table
ces
Time Period
< 1 year
1-15 year
> 15 year
Fuel Prices
Total User
Financial Costs
-0.1
-0.4
-0.5
-0.5
-1.0
-2.0
Elasticity estimates are useful for analyzing and predicting potential impacts ofprice
changes. Actual effects depend on the availability of alternatives and other factors.
An understanding of travel elasticities is useful because pricing can be used to manage
transport. Elmer Johnson states, ''As people begin to pay the full social costs of driving,
they would take greater care in deciding when and how to move from place to place. Solo
travel would decrease substantially. ''2 3 There is considerable interest now concerning the
potential of pricing to encourage more efficient travel patterns, the economic and social
impacts of increased prices, and the best way to implement such strategies.
21
Webster, Bly and Paulley, Urban Land-Use and Transportation Interaction: Policies and Models,
Avebury (Brookfield, MA), 1988, cited in "Effects of Land Use Intensification and Auto Pricing Policies
on Regional Travel, Emissions and Fuel Use" draft report by Robert Johnston and Raju Ceerla, 1994.
22 "Effects of Land Use Intensification and Auto Pricing Policies on Regional Travel, Emissions and Fuel
Use," draft report by Robert Johnston and Raju Ceerla, 1994, p.6.
23 Avoiding the Collision of Cities and Cars, American Academy of Arts and Sciences, 1993, p. 43 .
Page 5-10
Transportation Cost Analysis
5.3
Defining Generated Traffic
Generated traffic is the additional travel resulting from a transport improvement which
would not otherwise occur (Figure 5-4). The existence of generated traffic is proven both
theoretically and empirically. 24 It is recognized by economists, urban planners, and traffic
modelers (who call it feedback). Generated traffic results from latent travel demand
constrained by user costs (vehicle expenses, travel time, discomfort and risk) caused by
poor roads and traffic congestion. A reduction in these costs can induce more travel.
If roads are congestion residents tend to defer trips that are not urgent, and forego trips
that can be avoided. For example, when choosing where to go for dinner or shopping, you
may consider a wide range of destination when traffic flows freely, but limit yourself to
nearby destinations if roads are congested. In some situations you may even choose to
walk, bicycle or take transit rather than fight traffic to reach your destination.
Affect of Road Capacity on Traffic Volumes
Figure 5-4
--Trame Growth With Add•d Capaci ty
,•
--Trame Growt h
~
~"
.
"'
c:
Projected
Traffle
Growth~ '
~ .
[
1:1
"
~
Time --->
Traffic grows quickly after a road is built, then the growth rate declines as congestion
develops. A demand projection made during the high growth period indicates the need
for more capacity, but this need declines as congestion becomes self-limiting. If capacity
is added, traffic growth continues until it is filled This is called ''generated traffic. "
24 Mark Hansen, et al. , A ir Quality Impacts of Urban Highway Capacity Expansion: Traffic Generation
and Land Use Changes, Institute of Transport Studies, University of California (Berkeley), Research
Report UCB-ITS-RR-93-5, 1993; SACTRA 1994; Terry Moore and Paul Thorsnes, The
Transportation/Land Use Connection, American Planning Association (Chicago), #448/449, 1994.
Page 5-11
Transportation Cost Analysis
Generated Traffic Example
Aperson must deliver a package 10 kilometers across town (20 km round trip). Her
driving time is worth $10 per hour. Her marginal vehicle costs are $0.10 per km under free
flowing conditions and $0.15 under congested conditions. When the roads are congested
the trip takes 60 minutes. When the roads are uncongested the trip takes 30 minutes. Her
alternative is to mail the package at the local post office, which takes 15 minutes on
average walking and waiting in line, and costs $5 .00, or a total cost of $7.50. As long as
the cost of mailing is greater than the cost of driving she will make the cross town trip. In
this case she would choose to deliver it herself if the road is uncongested, saving $0.50,
but would mail it if the road is congested, saving $5 .50 in total costs.
Total User Costs
Time cost @ $10.00/hr
Postage cost
Vehicle cost
Congested
Trip
Uncongested
Trip
Mail
Package
$10.00
0.00
3.00
$13 .00
$5 .00
0.00
2.00
$7.00
$2.50
5.00
0.00
$7.50
Acongestion reduction project could generate this trip by making personal delivery
cheaper than mail. Users' potential net benefits range from $0 to $6.00, with an average of
$3.00. This reflects the Rule-of-Half, which states that net benefits of generated travel are
approximately half of total time saving benefits.
The generated trip also incurs external costs, including congestion on roads other than the
one being considered for improvement, air and noise pollution, parking requirements,
increased energy consumption, road wear, accident risk and reduced travel options for
non-drivers. Assume that the average total external cost of driving is $0.50 per mile under
congested condition and $0.25 per automobile mile under uncongested conditions, and
that the mail truck carries an average of 1, 000 packages, imposes costs double an
automobile, and requires 20 miles of travel to deliver the package.
Total costs
Internal Costs
External Costs
Total Costs
Congested
Trip
Uncongested
Trip
Mail
Package
$13.00
10.00
$23.00
$7.00
5.00
$12.00
$7.50
0.02
$7.52
The total cost of driving is higher than the total cost of mailing the package under either
level of congestion. While it is possible that some generated trips substitute for activities
that have even greater external costs than driving, this is unusual since driving has higher
external costs than most other activities. Generated traffic can also have long term effects.
For example, if enough residents shift from mailing to delivering packages by car the local
post office may close due to low use, increasing mailing costs, which imposes extra
burdens on non-drivers who depend on it.
Page 5-12
Transportation Cost Analysis
The term generated traffic has two somewhat different meanings depending on the
perspective. Traffic planners and engineers are primarily concerned with traffic generated
on a particular road or corridor since it affects their efforts to improve traffic flow. Policy
makers and economists are concerned about increases in total vehicle travel because it
affects total costs. As discussed in the next section, some of the traffic that appears on an
improved roadway is actually diverted from other routes and times and does not reflect
increased driving. The emphasis in this chapter is on overall motor vehicle travel, but
generated traffic on improved roadways is also considered.
In the short term, most of the increase in traffic that occurs on an improved road results
from trips that are diverted from other routes, travel times and modes. Downs calls this
Triple Convergence. 2 5 Over the long term structural changes including land use changes
(sprawl, and agglomeration of services), increased automobile ownership, and reduced
transport choices that result in more and longer automobile trips. These individual changes
can have synergetic effects called automobile dependency.26
5.4
Types of Generated Traffic
Road improvements produce several different effects which increase traffic on a particular
road, as described in Table 5-6. These distinctions are important because different types of
generated traffic tend to have different impacts and costs. Although all increase traffic on
an improved road, some have less net cost than others. In general, diverted automobile
trips have little incremental costs because they simply change the route or time of an
existing automobile trip. Longer trips have moderate incremental costs. Increased
automobile trips have the largest incremental costs. Some types of generated traffic also
impose secondary costs by encouraging land use changes (such as sprawl) and transport
system changes (such as reduced public transit service and pedestrian facilities) that
25
Anthony Downs, Stuck in Traffic, Brookings Institute (Washington DC), 1992.
and Jeff Kenworthy, Cities and Automobile Dependency, Gower (Aldershot), 1989.
26 Peter Newman
Page 5-13
Transportation Cost Analysis
increase travel needs, increase automobile ownership, and reduce affordable travel
alternatives. These increase automobile dependency and have negative equity impacts by
making non-drivers and poor people relatively worse off, as discussed in chapters 3. 9 and
3.14.
Table 5-6
Name
Traffic Effects of Road
I
Shorter Route
Description
Improved road allows drivers to
use more direct route.
Longer Route
Improved road attracts traffic
from other, more direct routes.
t
Type of
Chan2e
Time
Frame
Diverted trip. Short
term
Longer trip.
Short
term
Reduced peak period congestion
reduces the need to defer trips to
off-peak periods.
Diverted trip. Short
term
Mode Shift; No Improved traffic flow makes
Diverted trip,
Capital
driving relatively more attractive generated
Short
Changes
than other modes.
auto trip.
term
Less demand leads to reduced
Mode Shift;
rail and bus service, reductions in Diverted trip,
With Capital
bicycle and pedestrian facilities, generated
Long term
Changes
and more automobile ownership. auto trip.
Destination
Reduced travel costs allow
Change;
drivers to choose farther
Current
destinations. No change in
Short
Land Use
destination locations.
Longer trip.
term
Time Change
Destination
Change; Land
Use Changes
New Trip; No
Capital
Changes
Improved access allows land use
changes, especially urban fringe
development.
Improved travel time allows
driving to substitute for nontravel activities.
New Trip; With Improved access increases
Capital
activities that require driving,
Changes
reduced alternatives to driving.
Synergetic effects of increased
Automobile
automobile oriented land use and
Dependency
transportation system.
Longer trip.
Generated
trip.
Generated
trip.
Generated
trip.
Travel
Imoacts
Reduction
Reduction
Small increase
Slight increase
None
Moderate
increase
Increase
Moderate to
large increase
Increased
driving,
reduced
alternatives.
Large increase,
with equity
costs.
Increase
Moderate to
large increase
Increased driving, automobile
dependent
Long term land use
Short
term
External Cost
Imoact
Increase
Increased driving, automobile
Long term dependency.
Increased
driving, few
Long
alternatives.
term.
Moderate to
large increase,
with equity
costs.
Large increase
Large increase,
with equity
costs.
Large increase,
with equity
costs.
This table describes various traffic effects of road improvements. Some are diverted trips
while others are generated travel. All increase traffic on the improved roadway.
Page 5-14
Transportation Cost Analysis
5.5
Predicting Generated Traffic
Arecent University of California study calculated elasticities of vehicle traffic with respect
to road capacity to be 0.15-0.3, 0.3-0.4, and 0.4-0.6 for 4, 10, and 16 years respectively, 27
meaning that up to 60 of each 100 additional roadway spaces are filled with generated
traffic within 16 years. The researchers state that higher rates of traffic generation are
likely in many urban areas due to increased latent demand. Kenneth Small cites research
indicating that 50% to 80% of increase in highway capacity is soon filled with generated
traffic.28 Figure 5-5 shows the likely range of generated traffic on an improved highway.
Elasticity of Traffic Volume With Respect to Road Capacity29
Figure 5-5
. - - - - - - - - - - -------·
100 ...
....
.. ....
j ,.
i ....
i ....
I ....
I ....
1 ....
'
"'
"' "'
'll
•
~
u
'll
L
- Extreme Latent Demand
- H i gh Latent Demand
---Medium Latent Demand
- - L o w Latent Demand
,..
...
10
Years After Capacity
12
,.
10
Expansion
10
"
This shows expected generated traffic on a road after its capacity increases. About half
of added capacity is filled with new traffic within a decade of construction, and even
more traffic can be generated on extremely congested roads.
This research indicates that about half of new road capacity is typically filled with new
trips that would otherwise not have otherwise occured within a decade of construction. In
27 Mark Hansen,
et al. , A ir Quality Impacts of Urban Highway Capacity Expansion: Traffic Generation
and Land Use Changes, Institute of Transport Studies, University of California (Berkeley), Research
Report UCB-ITS-RR-93-5, 1993.
28 Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992, p. 113.
29 Based on Hansen, et al. and comments by Professor Robert Johnston.
Page 5-15
Transportation Cost Analysis
areas with extreme latent demand half of the added capacity can be filled with generated
traffic within two years, and nearly all of the added capacity may be filled after 20 years.
These elasticity estimates refer only to traffic increases on the improved road. Some of this
traffic increase results from diverted trips and does not represent a total increase in vehicle
travel. One type of generated traffic (Shorter Route in Table 5-6) actually reduces vehicle
travel and external costs. But such savings are overwhelmed by generated travel, causing
net increased vehicle travel and external costs. The University of California team found
that total vehicle travel increased 1% for every 2% to 3% increase in highway lane miles.
This supports the hypothesis by Newman and Kenworthy that automobile oriented
transport and land use policies increase automobile dependency and use.
5.6
Calculating Internal Benefits of Generated Traffic
Most transport benefit/cost manuals specify how to calculate net user benefits of
generated traffic using the Rule-of-Half. 30 This states that generated traffic consists of
relatively low value travel because they are trips that users choose to forego under more
congested conditions. Their net benefit is assumed to be half of the benefit for existing
trips. The Rule-of-Half is illustrated using a demand curve in Figure 5-6.
30
See for example, Manual on User Benefit Analysis ofHighway and Bus Transit Improvements,
AASHTO, 1977, p.26; COBA Manual, British Dept. of Transport (London), 1989 Reprint, p. 1-15.
Page 5-16
Transportation Cost Analysis
Figure 5-6
Vehicle Travel Demand Curve Dlustrating the Rule-of-Half
-D•mona .-;urvo
• Orlglnol User Coat
-
-
-Reduced User Coat
§
1
1-
~
I
A
::J
Vehicle Travel
A reduction in user costs (downward shift on Y axis) increases vehicle travel (rightward
shift on X axis). Rectangle A shows the benefits of reduced user costs for existing trips.
Triangle B shows the benefits of generated traffic.
For example, if there are 100 possible peak period automobile trips on a road but only
room for 75, travelers must forego 25 trips due to congestion delays. The foregone trips
are those users consider less valuable than the trips they take. If roadway capacity
increases, the 25 new trips are relatively low value. Economists estimate that net benefits
of generated traffic average half the benefit to existing travelers (illustrated in Figure 5-6
by the fact that B is a triangle rather than a rectangle) and call this the Rule-of-Half.
5.7
Calculating External Costs of Generated Traffic
Driving imposes a number of external costs, as described earlier. Urban peak period
driving, the type of driving most likely to be generated by increased road capacity, usually
has the highest external costs. Net external costs should be charged as costs of a project
that induces generated traffic, with the exception of congestion on a route being
considered for capacity expansion to avoid double counting delay costs. The contribution
of generated traffic to congestion on other roads should be considered a cost of the
project that creates it.
Page 5-17
Transportation Cost Analysis
To illustrate this, consider the effects of building or expanding a highway into a city's
downtown. If the number of automobile trips don't change the improved highway would
simply benefit travelers without increasing external costs. But if the highway generates
new automobile trips, surface street congestion, pollution and parking problems (costs) in
downtown will increase. Alternative investments (transit service, bicycle lanes or a TDM
program) can provide mobility to downtown without incurring these costs. In order to
accurately assess and compare these potential investments, the additional external costs of
the generated traffic must be included as a cost of the highway project.
Determining net external costs of driving requires subtracting the external cost of the trip
alternative (the activity that would occur without the policy, program or project under
consideration). In the case of diverted traffic this is the difference in external costs
between the two trips. For longer trips this is the increase in externalities over the shorter
trip. For generated travel this is the external cost difference between the trip and the nontravel activity that it replaces. There is no specific research on the external costs of
activities that generated traffic substitutes for, but since driving has greater social and
environmental impacts than most other activities people typically engage in, we can
assume that such costs are overall significantly lower than driving.
Roadway improvements that induce mode shifts may also result in diseconomies in transit
service, which should also be considered costs.31 Adding road capacity is increasingly
expensive in most urban areas due to rising land acquisition costs and community
resistance. Failure to consider these increasing marginal cost (for example, by estimating
congestion costs based on previous rather than future road construction costs or ignoring
increases in transit service costs) further understates the total costs of generated traffic.
31 H.C.W.L. Williams, et al. "Transport Policy Appraisal with Equilibrium Models III: Investment
Benefits in Multi-Modal Systems," Transportation Research B, Vol. 25, No 5, pp. 293-316, 1991. Also
see John Kain, "Impacts of Congestion Pricing on Transit and Car Pool Demand and Supply," in Curbing
Gridlock, TRB, National Academy Press (Washington DC), 1994, pp. 502-553 .
Page 5-18
Transportation Cost Analysis
Short and Long Term Effects
As described in Table 5-6, short term generated traffic consists primarily of diverted trips.
Longer trips, generated trips and increased automobile dependency tend to be long term
effects. Most short term changes probably occur within about a year of capacity expansion
since they only require users to change their habits and do not involve significant changes
in land use or transit service provisions. Eventually generated trips will crowd out some of
the diverted traffic, since the improved road offers increasingly less relative advantage
over other routes. For example, if one year after a road project is completed there are 400
additional peak period trips, these can be assumed to be primarily diverted. If, after 10
years there are 1,200 additional peak period trips, somewhat less than 400 of these (say
200) result from diverted traffic and the rest are new trips resulting from land use changes
and increased automobile dependency.
Figure 5-7
Increased Travel as a Percentage of Generated Traffic on a Roadway
-
100'11
Gi
>
...!
80%
10%
'i
•
"
!
&0%
.5
~
..
'0
~"'
•.f!
40%
30'JI.
Q,
20%
10%
0%
0
2
4
8
8
10
12
14
18
18
20
Yeal'8 Altar Capacity Expansion
This graph estimates the portion of travel that is generated rather than diverted
The distribution of different types of generated traffic over time is an important subject
that has received little research. One recent survey estimates that short term generated
Page 5-19
Transportation Cost Analysis
traffic increases travel only 3-5%, indicating that it consists mostly of diverted trips.32
Figure 5-7 shows the likely distribution of generated traffic between diverted and
generated trips, based on the assumption that diverted trips are initially the major source
of increased traffic on an improved road, but over a 20 year time period generated travel
(new and longer trips) increase and eventually dominate.
There is currently no standard procedure for calculating the net cost of generated traffic.
In order to help develop useful working estimates, what we do know about generated
traffic costs is summarized below:
I.
Net costs are calculated by subtracting the cost of the trip alternative, which may be
a different trip or a non-travel activity.
2.
Net costs depend on the type of generated traffic, as described in Table 5-6. This
shows that only one type (shorter routes) reduces costs. Of the rest, two divert
automobile trips and tend to have small net costs, three lengthened trips and have
moderate net costs, and four generate new automobile trips and have large net costs.
Total costs of generated traffic can be greater than the direct external costs of
individual trips as a result of indirect effects on alternative modes and land use.33 This
can be illustrated by an example in which a highway improvement results in the
abandonment of parallel rail service due to reduced ridership. The net external costs
of the project include the external costs of the generated traffic plus the incremental
external costs of any reduction in rail service resulting from reduced demand,
including driving on other roads. In this way the road improvement may increase
overall regional traffic congestion. Such an situation is unlikely to occur in North
America now, simply because few communities still have viable rail service, but a
similar effect may result from other impacts of increased automobile dependency,
including reduced bus service, sprawled development, and loss of local services such
as neighborhood shops and schools, all of which result in increased overall
automobile use.
3.
Net costs as a portion of total costs tend to increase over time because to the
increasing portion of generated tips. Although the effects of automobile dependency
are largely long term (such as increased sprawl and reductions in the availability of
32 Richard Dowling and Steven Colman, Effects ofIncreased Highway Capacity: Results of a Household
Travel Behavior Survey, TRB Annual Meeting (Washington DC), #950409, January 1995.
with Equilibrium Models III: Investment
Benefits in Multi-Modal Systems," Transp ortation Research B , Vol. 25, No 5, pp. 293-316, 1991.
33 H.C.W.L. Williams, et al. "Transport Policy Appraisal
Page 5-20
Transportation Cost A nalysis
alternative modes), the causes occur as soon as land and transportation investment
decisions are influenced.
Figure 5-8
,....
Estimated Net Costs as Percentage of Generated Traffic Total Costs
10%
.! .....
0
'""
1
;ll .....
~
....
• .....
:::1
11
lS
1-
'6
~
•
...
-
- -Higher Range
--Dofaull
- - - Lower Range
2...
'""
""
10
12
14
11
11
20
Years After Capacity Expansion
Since generated traffic tends to result more from diverted trips in the short term and
generated travel over the long term, net costs as a portion of total costs increase over
time. This graph is a multiplier for estimating net external costs.
Based on these assumptions, net generated traffic costs are estimated to be 40-60% of
total external costs in the short term, and increase to 80-100% over the long term, using a
20 year planning horizon. Figure 5-8 illustrates a proposed default multiplier that can be
used to estimate net external costs for planning purposes.
5.8
Land Value Impacts
Road improvements can lead to land use changes that affect real estate values. This occurs
because transport is used to compete for desirable locations. For example, people often
face tradeoffs between travel and location costs: lower priced urban fringe land requires
more driving. Road improvement benefits are captured as increased vehicle operating
costs (due to longer trip distances), consumer surplus (commuters can purchase more land
at a lower cost than would be possible closer to the urban center) and producer surplus
(profits to urban fringe land owners). Sometimes communities compete for economic
Page 5-21
Transportation Cost Analysis
development by offering transport facility subsidies. Such subsidies can also capture a
portion oftransport improvement benefits.
Competition for access can create a self-perpetuating cycle of increasing costs, since
increased motor vehicle traffic degrades the urban environment, thereby increasing the
desire by individuals for exurban residences. Both increased driving and increased
development at the urban fringe impose external costs. This creates a "social dilemma" in
which individuals' short term interests conflict with their long-term interest. 34 The
tendency of increased travel speeds to result in dispersed destinations, no reduction in
travel time, and increased overall travel costs, is described as space pollution and time
pollution by Geographer John Whitelegg.35 He and other researchers find that the amount
of time people spend on travel varies little, regardless of speed or mode of travel.
Increased travel speeds often result in increased travel, not more free time. He writes,
"Those who use technology to travel at greater speeds still have to make the same
amount of contacts--still work, eat, sleep and play in the same proportions as always.
They simply do these further apart from each other. "
In economic analysis it is important to include the cost of all public subsidies and to
exclude any benefits associated with increased real estate values in order to avoid double
counting. Some transportation investment economic analysis models such as MEPLAN
and TRANUS track land value benefits, but exclude travel time savings for this reason.
5.9
Applying Generated Traffic External Cost Estimates
There has been little research on the subject of calculating net costs of generated traffic
and even less to develop practical tools for incorporating these costs into common
transportation decision making such as evaluation of particular transport plans, projects or
34 Emin
Tengstrom, Use of the A utomobile, Swedish Transport Research Board (Stockholm), 1992, p. 59.
Also see Garret Hardin, "Tragedy of the Commons," Science Magazine , Dec. 1968, pp. 1243-1247.
35 John Whitelegg, "Time Pollution," The Ecologist, Vol. 23 , No. 4, July/Aug. 1993, p. 131-134.
Page 5-22
Transportation Cost Analysis
policies. This is an unfortunate omission because this cost has significant implications in
such decisions. The current practice of ignoring generated traffic overstates the benefits of
increasing road capacity and understates total costs, which tends to skew decisions toward
automobile dependency and away from other transportation options.36
Generated traffic can be incorporated into transport planning in three ways:
1.
Roadway capacity.
Total increased motor vehicle travel that would result from expanding capacity on a
roadway can be calculated using the elasticity estimates illustrated in Figure 5-5 .
Incorporating generated traffic is important to accurately determine travel time
savings, user benefits, and external costs. These curves can be used to calculate the
effects of generated traffic for each project year to determine accurate net present
value. Ideally, each type of traffic effect described in Table 5-6 would be assessed
separately, since each has different economic impacts. Figure 5-8 provides a default
estimate of net external costs as a portion of total external costs.
2.
Lane mileage.
Increasing lane miles increases overall vehicle travel. The travel generated by
roadway projects can be estimated using the University of California's research
results, indicating that a 1% increase in lane miles increases total vehicle travel by
0.33% to 0.50%.
36 H.C.W.L. Williams and W.M. Lam, "Transport Policy Appraisal
with Equilibrium Models I: Generated
Traffic and Highway Investment Benefits," Transportation Research B , Vol. 25, No 5, pp. 253-279, 1991.
Page 5-23
Transportation Cost Analysis
3.
Land Use Factors.
Some land use policies, programs and projects can also generate traffic.37 Estimates
of the relationship between land use density and vehicle travel, such as those by
Holtzclaw, 38 and the LUTRAQ project39 can be use to predict the transportation
effect of land use decisions. The conceptual measure of this impact is the with-and-
without test: the type and amount of development that would occur with and without
the transport project. 40
Since these three approaches for estimating generated traffic overlap in their effects (for
example, the elasticity of vehicle travel with respect to road capacity typically incorporates
some traffic generated by sprawl), only one of the three should normally be used for
evaluating a specific project to avoid double counting.
37 Eric Kelley, "The Transportation Land-Use Link," Journal ofPlanning Literature, Vol. 9, No.2, Nov.
1994, p.l28-145 ; Lawrence Frank, "Impacts of Mixed Use and Density on the Utilization of Three Modes
of Travel" paper presented at the Transportation Research Board Annual Meeting, January, 1994;
Reducing Transport Emissions Through Planning, Dept. ofEnv. & DoT, HMSO (London), 1993 .
38 John Holtzclaw, Explaining Urban Density and Transit Impacts ofAuto Use, Sierra Club and NRDC,
San Francisco, 1994
39 Sam Seskin, The LUTRAQ Project; Travel Behavior, 1000 Friends of Oregon (Portland), 1994
40 van Kooten, Land Resource Economics and Sustainable Development, UBC Press (Vancouver), 1993.
Page 5-24
Transportation Cost Analysis
6.0
Transportation Cost Implications
Consumers and industry face tradeoffs between transport and other costs or benefits. The
relative price difference between transport and other goods affects countless decisions
such as where to buy a house, where to locate a business, how to distribute the goods that
a business produces, and where to go for vacation. These tradeoffs have many indirect
impacts. In this chapter the cost estimates developed in this report are used to analyze the
implications of transportation costs and underpricing on economic efficiency, economic
development, land use, stakeholder perspectives, and travel patterns.
6.1
Economic Efficiency Impacts
A basic tenet of market theory is that economic efficiency is maximized when marginal
user prices (defined as perceived variable internal costs) reflect total marginal costs.
Mispricing causes inefficient use of resources because it prevents users from accurately
incorporating costs into their consumption decisions. As one pricing study describes,
''Price is the mechanism by which scarce resources are allocated efficiently between
competing uses. For consumers, price encourages a purchase if the benefits of making
the purchase exceed the benefits of alternatives. For producers, prices provide
incentives for resources to move to those uses which people value most highly by
informing firms how to produce, which products to produce, when and where to sell
the products, and when, where and how to invest. "1
According to this study, motor vehicle use is significantly underpriced compared with the
costs it imposes on society. External costs are estimated to average 33% of total costs,
with a range from 41% for Urban Peak driving to 23% for Rural driving, as shown in
Table 6-1. In other words, user costs would need to increase 35% to 72% to incorporate
all costs. Other studies described in Chapter 2 reach similar conclusions.
1 Halcrow
Fox and Associates, Land Transport Pricing for New Zealand, Transit New Zealand
(Wellington), 1993, p. 47.
Page 6-1
Transportation Cost Analysis
p
A vera e A
bile C
fTotal C
External Costs
Total Costs
Internal Costs
Units
per mile
per mile %of Total per mile %of Total
Table 6-1
Urban Peak
Urban Off-Peak
Rural
Weighted Average
$1.32
$1.05
$0.84
$0.99
$0.71
$0.71
$0.64
$0.67
54%
68%
76%
68%
$0.61
$0.34
$0.20
$0.32
46%
32%
24%
32%
On average, about one-third of the costs of driving are external.
Externalized costs are not the only cause of underpricing. Many vehicle costs are fixed,
which further reduces the ratio between prices and total costs. Operating costs (variable
financial costs) are only about 35% ofusers' financial costs, and only about 13% of the
total costs of driving. External costs equal48% to 130% of user marginal costs (vehicle
operating costs plus, user time and accident risk), and 181% to 3 81% of vehicle operating
costs. Each dollar spent on gas, maintenance, and short term parking incurs on average
$2.61 worth of external costs, including congestion, accident risk, parking subsidies, and
environmental degradation. Car owners are "captive" to most costs of driving. We pay
fixed and external costs no matter how much or little we drive, which reduces the
incentive to limit driving to high value trips. Automobile owners receive only a small
portion of the total savings they produce by driving less or more efficiently.
Although underpricing of such a common consumer good may appear beneficial from a
narrow perspective (and indeed benefits many individuals in the short term), mispricing
reduces overall economic e:fficiency. 2 Underpriced automobile use increases purchases of
transport over other consumer goods, and driving over other travel modes. Low price
causes per capita mileage to increase, forcing other constraints, such as congestion,
pollution, and resource depletion to limit growth. As Elmer Johnson states,
"When a good as central to American life as the automobile remains underpricedfor
several decades, that good tends to be used more than it otherwise would be. Habits
2 Terry Moore and Paul Thorsnes, The Transportation/Land Use Connection, American Planning
Association, Report #448/449 (Chicago), Washington DC, 1994.
Page 6-2
Transportation Cost Analysis
become ingrained and are hard to break. They are reinforced by the present urban
infrastructure designed to exploit the full possibilities ofprivate mobility. '13
External costs are not eliminated, they simply show up elsewhere, for example as higher
prices for commercial goods (for parking), increased local taxes (to pay for road services),
higher insurance premiums (from automobile accidents), illnesses (from pollution), and
lower residential property values (from urban traffic). Another effect ofunderpriced
driving is that non-automotive modes decline. Walking, bicycling, transit, and rail
transport receive little capital investment, land use patterns and social habits develop
which conflict with these travel options, and they develop a social stigma.
This is not to say that driving would cease if more costs were internalized. Consumers
would be willing to pay more for some trips. However, some driving has relatively low
value to the user, either because the trip itself provides little net benefit or because
reasonable alternatives exist. Increasing prices to reflect a greater portion of total costs
would reduce low value driving, improving the transportation system's overall efficiency.
A variety of strategies have been proposed for internalizing costs. A common suggestion
is to increasing fuel taxes.4 This, however, does not reduce all external costs, since fuel
prices do not directly affect when or where driving takes place, or provide incentives to
buy a low polluting car. Over the long run drivers would buy more fuel efficient cars,
which does not reduce congestion, accidents, parking costs, noise or most other
environmental impacts. Many countries with fuel prices two or three times higher than in
North America experience comparable or greater per capita external costs of driving. 5
3 Elmer
Johnson, Avoiding the Collision of Cities and Cars, American Academy of Arts and Sciences
(Chicago), 1993, p.11.
4 Steve Nadis and James MacKenzie, Car Trouble, WRI, Beacon Hill Press (Boston), 1993, p. 160.
5 For a comparison see Victorian Transport Externalities Study; Summary Report, Environment
Protection Authority (Melbourne), Table 3, p.8.
Page 6-3
Transportation Cost Analysis
Other analysts emphasize peak hour pricing as a method of internalizing costs.6 This could
reduce congestion but not other externalities such as pollution and parking subsidies.
Underpriced Driving Reduces Transportation Diversity and Efficiency
Underpricing increases automobile dependency and reduces travel choices. These are
disadvantages to non-drivers, and may reduce the overall efficiency of the transport
system. Harry Dimitriou identifies a "transportation gap" between the 1/2-mile average
range of walking, and automobile travel that is inefficient for trips less than about 4 miles
due to high per mile pollution, congestion and parking costs.7 Most urban trips are in this
range, which is efficiently handled by bicycles, small low powered vehicles, and local bus
service along busy corridors. Dimitriou argues that using automobiles for the relatively
short trips common in cities is a "sub-optimal" use of technology that exacerbates urban
problems. Underpricing encourages automobile use for trips when more efficient
alternatives may be more appropriate.
User charges should be applied as closely as possible to the source of an externality to
optimize economic efficiency, although economists and policy makers recognize that in
practice a "second best" solution is often necessary.8 Because of the diverse nature of
transport costs no single mechanism can capture all external costs. Charles Komanoff
identifies several price changes needed: weight-distance charges, fuel taxes, congestion
pricing, smog fees, parking fees, marginalized insurance, and higher fines for violators for
optimal efficiency.9 He estimates that no charge should raise more than 33% of total user
revenue for maximum economic efficiency. The Office of Technology Assessment study
6 Anthony Downs,
Stuck In Traffic , Brookings Institute (Washington DC), 1992.
Urban Transport Planning: A Developmental Approach, Routledge (NY), 1992, p. 245.
8 This assumes that charges are intended to achieve economic efficiency. If the concern is horizontal
equity (its not fair to charge people for goods that somebody else consumes) then a broader range of
pricing mechanisms can be considered.
9 Charles Komanoff, Pollution Taxes f or Roadway Transportation, KomanoffEnergy (NY), 1994
7 H. Dimitriou,
Page 6-4
Transportation Cost Analysis
Saving Energy in US. Transportation and the European Federation for Transport and
Environment reach similar conclusions and recommendations for internalizing costs. 10
Douglass Lee excludes indirect costs such as pollution embodied in resources consumed in
vehicle production and use, arguing that one sector should not be charged for the sins of
another sector. 11 Although charging automobile drivers for air pollution produced during
steel production or water pollution produced during petroleum distribution may marginally
reduce these impacts, it is more effective to apply such charges directly to the steel or
petroleum firms in proportion to their external costs as an incentive to reduce such
impacts. If pollution costs are simply passed on as surcharges to end users, individual
firms have little incentive to minimize the damage they produce. However, until marginal
charges are effectively applied it is still legitimate to consider residual impacts an
externality of consumption and to charge end users as a second best solution.
It would be inefficient to compensate for externalities so generously that individuals have
no incentive to avoid such impacts, as described in the Coase Theorem.12 For example, it
would be a mistake to reward accident victims so well that road users become careless or
intentionally cause crashes. Although this argument is sometimes used to justify
government inaction on external costs, it seldom actually applies since charging for an
externality and compensating victims are separate activities. 13 Rather, it simply means that
victim compensation must use common sense to avoid perverse incentives.
10
Saving Energy in U. S. Transportation , Office of Technology Assessment, 1994. Per Kageson, Getting
the Prices Right, European Fed. for Transport & Environment (Bruxelles), 1993, p.154-161.
11 Douglass Lee, Full Cost Pricing ofHighways, USDOT National Transportation Systems Center
(Cambridge), January 1995, p. 25.
12 R. Coase, "The Problem of Social Cost," Journal of Law Economics, Vol. 3, Oct. 1960, pp. 1-44.
13 The Coase Theorem states that it may be most efficient to allow producers and victims to negotiate
directly for compensation of externalities, provided negotiations are possible and "property rights" are
established. In practice, governments must often represent victims' interests and define property rights.
Page 6-5
Transportation Cost Analysis
Many costs associated with increasing transportation prices, such as unemployment and
reduced profits in motor vehicle dependent industries, are transition costs which decrease
over time and eventually disappear. Transition costs are the economic inefficiencies that
result when capital equipment, policies, contracts and employment become non-optimal
due to changes in prices or demand. For example, a business may invest in a market that
declines due to higher costs, or choose to purchase a cheaper but less efficient vehicle only
to incur losses when fuel prices increase. Similarly, an individual may spend more than
expected on commuting if fuel prices increase unexpectedly after they purchase a house
that is far from their work site. These result in reduced employment, productivity, and
profits in existing sectors (such as new automobile production) until an opportunity occurs
to change to more optimal market, equipment or location.
It is tempting to say that these are ''just" transition costs, but that would be unfair. These
are very real costs, both financially and in emotional stress to individuals who lose jobs or
receive less pay. But this does not justify underpricing and subsidies. Although automobile
industries have traditionally provided economic development and good jobs, it would be a
mistake to assume that these benefits are unique to that sector. At best, automobile
industries might continue providing current profits and employment, but little growth can
be expected due to international competition for markets and increased productivity.
Protecting automobile industries reduces the viability of other industries. Labor and other
resources that are currently employed in automobile industries can be more productive in
other sectors. Transitions provide benefits by allowing new businesses opportunities and
jobs to develop. The economic inefficiency resulting from efforts to protect the North
American automobile industry may explain, for example, why the U.S . has failed to
succeed in many new industries, such as consumer electronics and manufacturing. A better
strategy would be to accept and plan for economic transitions to minimize costs.
Page 6-6
Transportation Cost Analysis
6.2
Economic Development Implications of Underpricing
Transport is essential to any economy. It is needed for most economic activities, from
production of raw materials, to manufacturing, distribution, supplying labor and
professional services. The transport industry is itself a major economic sector. Motor
vehicle production, servicing, and use are all major industries, accounting for
approximately 17% of GNP, 11% of personal consumption expenditures, and over 10% of
total employment in the U .S.14 As a result, it is often argued that transport investments
and low prices encourage economic productivity and development. 15 Increased transport
prices are claimed to have a "multiplier effect" that raises overall costs and reduces
productivity.16 Lobbying organizations use these arguments to support underpricing.17
Piet Rietveld calls these claims scientific mythification. He states that " ... the direct
contribution of infrastructure improvement to a reduction in transport costs is in general
small in industrialized countries. "18 These analyses often ignore total costs and
distributional effects. Objective research indicates that transport improvements provide
only marginal economic productivity or development benefits.19 Each dollar of increased
revenue to one company resulting from underpricing means a dollar or more of revenue
lost elsewhere in the economy.
14
American Automobile Manufacturers Association, Facts and Figures 93, p. 58; Transportation In
America, 1 Jth Edition, Supplements, ENO Foundation (Lansdowne). Sept. 1993, p. 4
15 D. Aschauer, "Is Public Expenditure Productive?" Journal ofMonetary Economics, Vol. 23, pp. 177200, 1989; Alicia Mannen, "How Does Public Infrastructure Affect Regional Economic Performance?"
New England Economic Review, Sept./Oct. 1990, p.ll-33 ; Theresa Smith, "The Impact of Highway
Infrastructure on Economic Performance," Public Roads, Spring 1994.
16 Campbell Anderson, "A Business Perspective on Transport Reform" in Transport Policies For the New
Millennium , Ogden et al. editors, Monash University (Clayton), 1994.
17 The Allen Consulting Group, Land Transport Infrastructure: Maximizing the Contribution to
Economic Growth , Australian Automobile Association (Canberra), Nov. 1993.
18 Piet Rietveld, "Spatial Economic Impacts of Transport Infrastructure Supply," Transportation
Research, Vol. 28A, No.4, p. 339.
19 Joseph Berechman, "Urban and Regional Economic Impacts of Transportation Investment: A Critical
Assessment and Proposed Methodology," Transportation Research, Vol. 28A, No.4, pp. 351-362.
Page 6-7
Transportation Cost Analysis
Recent macroeconomic research indicates that transport infrastructure investments may
have high economic returns.20 However, this does not prove that increased driving
provides economic benefits. Rather, it implies that increased transport efficiency provides
benefits. Public transit expenditures provide twice the return as highway improvements, 21
indicating that mobility rather than just driving that provides economic benefits.
Economies and Diseconomies of Scale
The price of a single envelope at my local stationary store is 10¢, but 100 cost only $2. 50,
offering a 75% per unit savings. This is not surprising because many costs of distributing
envelopes are fixed; the same amount of checkout time is taken to sell me one as the box
of 100. This is an example of economies of scale. Economies of scale also exist with
respect to industrial production. Henry Ford's use of mass production to reduce
automobile prices, and the recent development of low-cost personal computers are two
examples. You benefit if your neighbors consume more of certain goods.
While human production and market activities tend to experience economies of scale,
most natural costs tend to experience diseconomies of scale. For example, the most
suitable land for a given use such as farming will be used first. Providing additional land
for the same production requires using less suitable land. Similarly, increasing extraction
of natural resources (minerals, fish, timber, etc.) requires using less accessible supplies or
more expensive extraction techniques, again incurring diseconomies of scale. Waste
disposal can also show diseconomies of scale. Ecosystems can absorb a certain amount of
some wastes, but as the volume or rate increases so does the negative impact per unit.
Our economy is increasingly limited by environmental and social constraints that
experience diseconomies of scale, while improved logistics and computer controls reduce
many of the advantages of large scale production. We cannot expect reduced costs if our
neighbors buy more cars and drive more each year, but you will experience increased
congestion, pollution, and land use impacts. It is no longer appropriate to encouraging
increased production and consumption, especially with respect to transport.
20 David Aschauer, Transportation Spending and Economic Growth, American Public Transit Association
(Washington DC), 1991
1991, p. 10.
21
Page 6-8
Transportation Cost Analysis
Peter Nijkamp and Eddy Blaas state that transport facility investments only contribute to
development if other conditions are ripe and high transport costs are a constraint. 22
Christine Kessides reaches a similar conclusion.23 She cites studies indicating that
infrastructure investments can provide high returns in fast developing countries but only
normal returns in developed economies such as North America and Western Europe.
These researchers emphasize that regions which already have paved roads, rail, and ports
will only enjoy small economic development from marginal improvements in such
infrastructure. Since underpricing reduces overall economic and transport efficiency by
incurring indirect costs and increasing traffic congestion, it may reduce rather than
stimulate overall economic development. Macroeconomic modeling by Arie Bleijenberg
indicates that a significant increase in fuel taxes (intended to internalize external costs and
reduce travel demand) with revenues used to reduce income taxes would have little overall
effect on the Dutch economy, and would slightly increase employment. 24
Current policies that underprice transportation may have been justified when they were
established. The automobile market and road system probably experienced significant
economies of scale during the first half of this century.25 Increased automobile sales
allowed manufacturers to reduce production costs, and by distributing construction
expenses over more automobile travel, facility costs per VMT declined. This was a unique
historical event, however, that does not apply to mature markets. There are probably few,
if any, further economies of scale in automobile, petroleum and roadway industries.
22
Nijkamp and Blaas, Impact Assessment and Evaluation in Transport Planning, Kluwer, 1993, p. 45-49.
Christine Kessides, The Contributions ofInfrastructure to Economic Development, World Bank
Discussion Paper #213 (Washington DC), 1993.
24 "The Art oflntemalising," inlnternalising the Social Costs ofTransport, OECD (Paris), 1994.
25 Clay McShane, Down the Asphalt Path , Columbia University Press, 1994, p. 105. Also see Stephen
Goddard, Getting There , Basic Books, (NY), 1994
23
Page 6-9
Transportation Cost Analysis
The Effects of Underpricing: Two Industry Example
Consider an economy with two industries: Heavy and Light. Transport is a relatively large
portion ofHeavy's production costs, and a small portion ofLight's.
1. These industries initially face a transport price structure which underprices shipping
buy using a property tax to fund roads and other transport costs. Since both industries
pay the same tax, Light is effectively subsidizing Heavy.
Score:
Heavy +
Light -
Economic Efficiency -
2. A $0.10 per mile road use fee is implemented to replace the property tax. Heavy
industry pays more taxes, while Light pays less.
Score:
Heavy Light +
Economic Efficiency +
3. At worst, Heavy pays the full road charge. But, Heavy may change shipping practices
to reduce costs, so its net cost increase is minimal.
Score:
Heavy =
Light +
Economic Efficiency ++
4. FairPrice also reduces traffic congestion and accidents. Assume that Heavy industry's
shipping time is reduced by 1/2 hour for every 100 miles of trucking and its trucks cost
$50 per hour to operate, this means the $10 per hundred mile road user charge saves
$25 in operating expenses, for an overall saving to Heavy.
Score:
Heavy +
Light +
Economic Efficiency +++
Although this example is simplistic, it emphasizes two important points:
1.
Underpricing does not eliminate costs it simply transfers them.
2.
Full Cost pricing provides economic incentives for business to use resources more
efficiently, which provides benefits, especially to transportation dependent industries.
The overall negative economic effects of transport price increases are relatively small, and
appear to be declining in most industrial sectors. Transport, especially freight transport has
decreased as a percentage of GNP, Industrial Production, and national employment in
recent decades due to increased efficiency and more high-value, low-bulk products.26
Transport accounts for only about 5-6% of most manufactured product prices. 27 Fuel
26
Transportation In America, 11th Edition, ENO Foundation, 1993, p. 23, and Supplement, p. 4.
Diamond and Spence, 1989, quoted in Piet Rietveld, "Spatial Economic Impacts of Transport
Infrastructure Supply," Transportation Research, Vol. 28A, 1994, p. 337. National Transportation
Agency, A n Integrated and Competitive Transportation Sy stem, (Ottawa), March 1992, from
Transportation, Taxation and Competitiveness, Transport Association of Canada (Ottawa), 1993, p. 56.
27
Page 6-10
Transportation Cost Analysis
taxes are only about 4% oftrucking industry gross revenues, and 0.5% of railroad gross
revenues, 28 so increased fuel taxes would only have a slight effect on end prices.
Higher vehicle taxes can benefit the economy by reducing traffic congestion, encouraging
more efficient shipping patterns, and reducing other taxes, resulting in constant or even
declining prices. Because fuel represents a relatively small portion oftotal industrial costs,
Kageson concludes that, "The effects from internalizing the social costs of transport on
the ability of European industry to compete on the world market will be almost
negligible. The total impact will amount to less than 0.5% of the annual turnover ofmost
industries. " and "GDP costs of a carbon tax may be fully offset by taking advantage of its
efficiency value and using the revenues to cut existing taxes that discourage capital
formation. '129 The Office of Technology Assessment considers the possible effects of
higher fuel taxes on the U .S. economic development and concludes,
"... if a gasoline tax were coupled with an equal-revenue increase in investment
tax credits, short-run macroeconomic losses resulting from motor fuel tax
increases could be more than offset by the short-run macroeconomic gains". 30
A $43 .50 per tonne of C02 fuel tax (increasing fuel prices 20-30%) on freight transport in
Norway is estimated to reduce long-haul rail and truck shipping by 0.26% and 1.82%
respectively, local for-hire truck transport would decline only 0.11 %, fleet truck shipping
would decline 2.08%, and fleet van shipping would decline by 0.36%.31 The greater
reduction in fleet trucks results from their relative inefficiency, since they often travel with
part loads, especially on return trips. Over several years these changes would be virtually
insignificant compared with normal turnover in these industries.
28
Transportation, Taxation and Competitiveness, TAC (Ottawa), 1993, p. 25, 29.
Per Kageson, Getting the Prices Right, European Fed. for Transport & Env. (Bruxelles), 1993, p. 183.
30 Saving Energy in U. S. Transportation , Office of Technology Assessment, US Congress, 1994, p.225 .
31 Trond Jensen and Knut Eriksen, "A General Equilibrium Model for Freight Transport in Norway,"
published in European Transport Forum Conference Proceedings, PTRC, Sept. 1994.
29
Page 6-11
Transportation Cost Analysis
Table 6-2
I
ts ofT
Better Off
Motor vehicle production, sales and service.
Bulk commodities.
Low-value manufacturing.
Imports.
Isolated companies.
Isolated retail and recreation.
Und
Worse Off
Alternative forms of transportation.
High-value products.
Communications and information industries.
Domestic and local production.
Centrally located companies.
Local oriented retail and recreation.
Internalizing transport costs benefits some companies and harms others. Overall, winners
should exceed losers due to overall increased economic efficiency.
Table 6-2 identifies the types of companies that are likely to be better or worse off from
underpricing. Only if companies in the "Better OfP' category provide higher profits or
employment than those in "Worse OfP' might underpricing provide overall economic
benefit. In fact, the automobile industry has experienced relatively low profits in recent
years and its portion of total U.S. employment and industrial production has declined. 32
Similarly, there is no evidence that bulk commodities, low value manufacturing, and
isolated commercial and recreation centers are economically more beneficial than the
competing industries that are less transportation dependent.
Automobile expenditures provide minimal or negative economic benefits at the regional
and local level. Most money spent on new vehicles and parts leaves a community. Fuel
purchases make even less contribution to local economies since petroleum products
provide minimal local economic activity. A study by Montgomery County, Maryland
found that only 15% of gasoline expenditures remained in the regional economy.33 Even in
Los Angeles County, where petroleum is produced and processed, only about 50% of
petroleum expenditures stay in the local economy, resulting in an economic multiplier of
32
American Automobile Manufacturers Association, Facts and Figures 93, p. 76. Transportation In
America, 1 Jth Edition, Supplements, ENO Foundation (Lansdowne), Sept. 1993, p. 4
33 Montgomery County Energy Study, Montgomery Dept. ofEnv. Protection (Rockville), 1985.
Page 6-12
Transportation Cost Analysis
1.8, compared with 2.7 for general goods and services.34 Thus, reduced driving and
increased use of other travel modes can provide local economic benefits.
The relative strength of Asian and European country economies where private automobile
use and road investments are much lower further indicates that road building and driving
do not directly contribute to economic growth. Walter Hook argues that excessive
automobile dependency and use associated with underpriced driving gives the U.S . an
economic disadvantage relative to Japan.35 He points out that transport accounts for only
9% of Japanese GNP, about half of the percentage in the U.S . This reduces Japanese
industrial and employment costs, increases productivity and frees funds for capital
investment. Harry Dimitriou recommends treating intra- and inter-community transport
needs separately.36 It is inter-community (long distance) transport that he believes
supports economic development. Marginal improvements to intra-community travel
provide little economic benefit in developed regions.
International competition is often used to justify underpricing, but North American fuel
prices and user fees are among the lowest among developed nations (Figure 6-1). Our
taxes could increase significantly and still be competitive with world markets. Since export
industries use only a small portion ofU.S. transport, broad underpricing is an inefficient
way to support a single sector. A better export stimulation strategy may be to increase
domestic prices and use a portion of the revenue to directly support export industries.
34
L.A. County Transportation Comm., Transportation Energy Conservation in Los A ngeles, Nov. 1979.
Walter Hook, "Are Bicycles Making Japan More Competitive?," Sustainable Transport, Summer 1993;
Walter Hook, "The Evolution of Japanese Urban Transportation and Non-Motorized Transport," paper
presented at the TRB Annual Meeting, January 9-13 , 1994.
36 Harry Dirnitriou, Urban Transport Planning: A Developmental Approach, Routledge, 1992, p. 144.
35
Page 6-13
Transportation Cost Analysis
Figure 6-1
1992 Gasoline Taxes and Prices in Selected Countries37
•Gas Taxes
•Pretax Gasoline Price
$5 .00
.§
$4 .00
ii
Cl
~
$3 .00
8..
$2 .00
1!!
uj
::;;
$1.00
$0.00
Italy
Sweden
Prance
Spain
Oermany
CB
Japan
Canada
USA
North America has the lowest fuel taxes and prices among the developed nations.
Transaction costs can make it expensive and therefore inefficient to actually charge
marginal prices, which could justify continued underpricing. Direct parking fees and
traditional road tolls are a bother because users must deliver money in the correct form,
and administrating such fees impose additional costs. People often prefer to reduce
transaction costs by paying lump sums or by including such fees in other purchases. This is
a legitimate concern, but new technologies significantly reduce transaction costs, lowering
the threshold under which marginal pricing is justified. Although transaction costs prevent
society from ever achieving pricing that perfectly reflects costs, this is not a barrier to
significant progress in marginalizing transport costs.
6.3
External Benefits of Transportation?
Transportation underpricing and subsidies would be justified if there are significant
external marginal benefits from automobile use. Some lobbying organizations argue that
such benefits exist. The Highway Users Federation, 38 the International Road Union, the
Deutsche Strassenliga (a German freight organization), and the German Club of
37 Pucher
and Horschrnan, "Public Transportation in the United States," Public Transport International,
Vol. 41, No. 3, Sept. 1993.
38 Eric Beshers, External Costs ofAutomobile Travel and Appropriate Policy Responses, Highway Users
Federation (Washington DC), March 1993 .
Page 6-14
Transportation Cost Analysis
Automobilists have each published reports arguing that driving provides significant
external benefits.39 Supposed benefits include improved personal mobility, improved
economic productivity, and general regional economic development.
These studies have been criticized for methodological problems, such as failure to
distinguish between internal and external benefits, counting distributional changes as
benefits, and non-marginal analysis.40 As discussed in the previous section, most
indications of economic benefits of driving actually indicate much higher productivity and
economic development benefits from more efficient travel modes. That driving provides
benefits does not mean that these are external, that more driving is necessarily good for
society (marginal benefits), that more driving is the best way to improve transportation if
other options are available, or that driving should be underpriced.
Significant true external benefits are unlikely because rational consumers tend to
internalize benefits and externalize costs, and because external benefits, when they do
exist, are quickly captured in economic competition. 41 A 1982 US DOT study concluded,
"the preponderance of expert opinion probably lies on the side of saying that there are
no external benefits of highway consumption beyond the benefits to the users. "42 Two
recent studies also conclude that transport benefits, while of great value, are almost
entirely internalized and no market failure exists to justify underpricing.43
39
Cited in Werner Rothengatter, "Do External Benefits Compensate for External Costs of Transport?",
Transportation Research, Vol. 28A, 1991, p.325.
40 Per Kageson, Getting the Prices Right, European Fed. for Transport & Env. (Bruxelles), 1993, p. 37.
Also see Werner Rothengatter, "Obstacles to the Use of Economic Instnunents in Transport Policy," in
Internalising the Social Costs ofTransport, OECD, 1994.
41 Kenneth Button, Internalising the Social Costs ofTransport, OECD, 1994, p.12.
42 Final Report on the Federal Highway Cost Allocation Study, USDOT, FHWA, 1982, p. E-9.
43 Werner Rothengatter, "Do External Benefits Compensate for External Costs of Transport?",
Transportation Research, Vol. 28A, 1991, p.321-328; Dr. Heini Sommer, Felix Walter, Rene
Neuenschwander, External Benefits of Transport?, ECOPLAN (Bern), March 1993.
Page 6-15
Transportation Cost Analysis
A recent Office of Technology Assessment report also explores the possibility that driving
offers external benefits to society, which might include economies of scale in retail, out-ofhome social activities, and spontaneous trip making. 44 Whether any ofthe benefits
described (such as improved access to evening education and social activities) are unique
to automobile travel and cannot be provided by other transportation systems is not
discussed. That report demonstrates no significant external benefits that could approach
the magnitude of external costs, nor does this report indicate external marginal benefits to
justify increased automobile use or underpricing.
Most claimed external benefits of driving are actually external benefits of access. Focusing
on just one mode would fail to obtain the full potential benefits of transportation
improvements. Even proponents of increased infrastructure investment such as Dr. David
Aschauer point out that optimum future infrastructure investments are likely to be less
emphasis on roads and motor vehicle traffic than in the past, and that transit investments
provide twice the rate of return as highway investments. 45
External benefits, when they do exist, are usually captured in market competition. If
driving provides external benefits, businesses or communities will compete for it, offering
incentives until benefits are internalized. For example, communities often subsidize roads
and parking facilities to attract development. This may provide a benefit to the first
communities to use this approach, but other communities will then be forced to provide
comparable subsidies until most benefits are captured by developers or new residents.
44 Saving Energy in U.S. Transportation , Office of Technology Assessment, 1994, p. 97.
45 Dr. David Aschauer, Public Investment and Private Sector Growth , Economic Policy Institute
(Washington DC), 1990; Dr. David Aschauer, Transportation Spending and Economic Growth , American
Public Transit Association (Washington DC), 1991.
Page 6-16
Transportation Cost Analysis
Since user fees do not cover the full cost of increasing road capacity to eliminate growing
traffic congestion, automobile use appears to experience diseconomies of scale. In other
words each driver benefits if others drive less, making road space and parking available.
This implies that the overall benefits of driving would probably increase if user prices were
raised to internalize a greater portion of costs, to discourage driving. In this way, low
value travel would decrease and reduced congestion would make travel more efficient.
Just as patriotism is called the "last refuge of a scoundrel" because it can be used to justify
careless and selfish motives, so the potential of external transportation benefits is often
invoked to avoid scrutiny which might illuminate poor analysis, selfish interests and
strategies that increase social costs. Although automobile use provides significant benefits,
that is no reason to underprice or otherwise encourage increased driving. Virtually all
objective studies conclude that motor vehicle users should bear the full costs of driving.
6.4
Land Use Impacts of Underpricing
Transportation has direct and indirect impacts on land values and use.46 These result in
benefits, which are largely capitalized into the value of land, 47 and external costs, as
described in Chapter 3 .14.
The tendency of land values to reflect access costs and capture transportation
improvement benefits is the basis for location theory, a seminal concept in economics.48 If
highways improve access between urban and rural areas, rural property values increase as
urban home buyers compete with other rural land uses. This provides profits to current
(often speculating) land owners, increases the supply ofland available for urban type
46 Terry Moore
and Paul Thorsnes, The Transportation/Land Use Connection, American Planning
Association, Report #448/449 (Chicago), 1994; "The Transportation Land-Use Link," Journal of Planning
Literature, Vol. 9, No. 2, Nov. 1994, p. 128-145.
47 That is, they result in higher property values to current land owners and lower real estate prices to
buyers for a given combination of land amenity and access.
48 Donald McCloskey, The Applied Theory of Price , MacMillan, 1985.
Page 6-17
Transportation Cost Analysis
development, and reduces urban land prices. It can also allow rural residents to compete
for urban jobs, benefiting rural residents and employers, and access urban services.
An automobile-oriented transportation system requires a relatively large portion of land,
leaving less for other uses. About 50% of land area in automobile-oriented cities is
devoted to roads and parking, compared with about 10% in pre-automobile cities. 49 This
increase in transport facility land requirements can be considered a cost of automobile use.
Devoting land to roads and parking imposes of external costs, including increased
competition for the remaining land, so environmental impacts and automobile dependency.
Transport is used by individuals to compete for desirable locations. People often face
tradeoff's between travel and location costs: lower priced urban fringe land requires more
driving. This has important implications on land use. In the short term transportation
improvements allow people to get to the same destinations more quickly, but in the long
term people tend to take more or longer trips, increasing urban sprawl. At one time this
was considered good for society.51 Exurb an development allows individuals to purchase
more land, privacy, and rural amenities than they can in a city. Melvin Webber states,
"Today [ 1985] people are moving into outlying areas because technological
improvements in transportation and communications have reduced the real cost of
traveling and communicating... current transportation and communication systems are
generating new forms of urbanization that are highly efficient, yet spread over
thousands of square miles. I suggest that this calls for celebration, not commiseration.
It promises unprecedently amiable living and working arrangements in pleasant
surroundings and increasingly intimate contact with friends and associates, many of
whom may be located miles away. When combined with high automobility, the exurbs
49 World Bank, cited in Harry Dimitriou, Urban Transport Planning, Routledge, 1992, p. 136. Note this is
larger than the estimated 30% of land devoted just to roads in automobile dependent cities.
50 lfthe supply of exurban land is unlimited, land used for urban roads is more than offset by increased
access to this cheap land, resulting in reduced overall land prices. But as competition for exurban land
increases, consumption of land for roads and parking reduces total available land, raising prices.
51 James Kunstler, The Geography ofNowhere, Simon & Schuster, 1993, p. 39; Peter Muller,
"Transportation and Urban Form," in Geography of Urban Transportation, Guilford Press (NY), 1986.
Page 6-18
Transportation Cost Analysis
promise spacious residential sites, temporal proximity to numerous employment sites,
and relatively easy access to recreational resources and culturally rich activities. "52
Webber is only half correct. While increased driving allows individuals access to land that
is less impacted by urban problems, it expands the range and scale of those impacts, so net
benefits are reduced. Webber ignores the external costs of exurban development, increased
internal and external costs of driving, and the problems faced by non-drivers, which call
into question his claim of "efficiency."
Competition for exurban locations creates a self-perpetuating cycle of increased costs,
since increased motor vehicle traffic degrades the urban environment, thereby increasing
the desire by individuals for exurban residences. This creates a "social dilemma" in which
individuals' short term interests are in conflict with society's long-term interest. 53 The
stakes continuously rise, so an increasing portion of user benefits are dissipated on
transport costs and capitalized in land values. Researchers find that the time people spend
on transport varies little, regardless of speed or travel mode.54 As a result, it is possible
that little or no overall benefit is derived from a substantial increase in driving. The
tendency of increased travel speeds to result in dispersed destinations, no reduction in
travel time, and increased overall travel costs, is described as space pollution and time
pollution.55 Whitelegg writes,
"Those who use technology to travel at greater speeds still have to make the same
amount of contacts--still work, eat, sleep and play in the same proportions as always.
They simply do these further apart from each other. "
52
Melvin Webber, "The Emerging Metropolis: Trends and Trepidations." In: Metropolitan Growth
Centers: A New Challenge for Public-Private Cooperation , UMTA-CA-06-0196-1 , Nov. 1985, p.9,
quoted in Hornberger, Kelland Perkins, Fundamentals of Traffic Engineering, 13th Edition, Institute of
Transportation Studies, UCB (Berkeley), 1992, p. 2-12.
53 Emin Tengstrom, Use of the A utomobile, Swedish Transport Research Board (Stockholm), 1992, p. 59.
Also see Garret Hardin, "Tragedy of the Commons," Science Magazine , Dec. 1968, pp. 1243-1247.
54 Gordon Stokes, "Travel Time Budgets and Their Relevance for Forecasting the Future Amount of
Travel, in Proceedings ofEuropean Transport Forum , PTRC, Sept. 1994, p. 25-36.
55 John Whitelegg, "Time Pollution," The Ecologist, Vol. 23, No. 4, July/Aug. 1993, p. 131-134.
Page 6-19
Transportation Cost Analysis
As discussed in Chapter 3.14, sprawl has significant environmental and social costs,
including degradation of natural habitat, ecosystems, and aesthetic amenities, reduced
community cohesion, increased per household municipal service costs and increased long
term transportation costs. Sprawled land use tends to be highly automobile dependent,
leading to more per capita driving and increased external costs. Automobile dependency
also increases dispersion of destinations within urban areas, leading to longer average trip
distances and a larger gap between drivers and non-drivers. Many urban planners now
argue that land uses have become unnecessarily separated and recommend more mixing of
compatible land uses.56 These land use impacts are exacerbated by underpriced driving,
and to a lesser degree by underpricing other long distance travel modes such as van and
car pooling, commuter rail, and telecommuting.
Growth Control or Traffic Control?
Although increased driving is justified by Webber and others to allow residents to avoid
urban problems, many such problems result from automobile use. The concerns associated
with increased urban density and growth are typically dominated by traffic and parking
congestion, traffic noise and air pollution, reduced open space, and higher taxes required
to provide new infrastructure. Rather than focusing on limiting population growth,
communities may benefit more from reducing automobile use.
Exactly how benefits are distributed between land sellers and land buyers depends on
specific market circumstances. Since the benefits of increased travel and reduced
development density are largely internal, while many costs are external, underpriced
transportation increases conflicts between individual and societal interests. Market
efficiency depends on prices reflecting marginal costs, so underpricing increases the
likelihood that transportation improvements and increased travel may result in no net
benefit when both internal and external costs are balanced against benefits.
56 For example see
The Land Use-Air Quality Linkage, California Air Resources Board, 1994, p. 9.
Page 6-20
Transportation Cost Analysis
Douglass Lee argues that transport and land use patterns are significantly distorted by
underpricing (including externalized environmental and social costs), overinvestment in
roads, and distortions in land markets such as overly restrictive zoning, all of which tend
to reinforce each other, increasing automobile use and sprawl. 57 He concludes that an
efficient transportation/land use system would increase economic efficiency, reduce total
transport expenses, reduce subsidies and tax burdens on non-users, improve urban
environmental quality, reduce urban sprawl, and increase the use of efficient travel modes.
6.5
Transportation Decision Making and Underpricing
From society's perspective, all costs and benefits must be considered in each decision, but
the perspective of individual decision makers is often more limited. People tend to have
different, often conflicting perspectives of transport costs and benefits, depending on their
role (Table 6-3). Even definitions of cost, benefit, equity, and efficiency can differ. These
differing perspectives and definitions create conflicts over goals, objectives and strategies,
and can result in decisions that increase overall costs and reduce overall efficiency.
Table 6-3Stakeholder P
Perspective
-
~~
Society
Driver
Non-Driver
Politician
Highway Planner
Urban Planner
Energy Planner
Environmentalist
f
Costs
fT
All costs
Time, vehicle costs, risk
Time, fares, discomfort, risk
Political jurisdiction costs
Roadway and drivers' costs
Facility costs, traffic impacts
Fuel consumption
Environmental impacts
Benefit
d Cost
Benefits
All benefits
Mileage
Access
Political jurisdiction benefits
Vehicle mileage, road capacity
Mobility/Access
Mobility/Access
MobilityI Access
The definition of transportation costs and benefits varies depending on a person's
perspective. These differences lead to conflicts between stakeholders.
57 Douglass Lee, "A Market-Oriented Transportation and Land Use System: How Different Would it Be?"
in Privatization and Deregulation in Passenger Transport: Selected Proceedings of the 2nd International
Conference, Espoo, Finland, Viatek, Ltd., June 1992, pp. 219-238.
Page 6-21
Transportation Cost Analysis
Another way to view the conflicts in our transport system is to consider average cost
curves of driving from three perspectives:
1.
2.
3.
Users, who decide how much to travel.
Transportation agencies, which supply road facilities.
Society, which bears environmental and social costs.
Each perceives different automobile use cost curves, as illustrated in Figure 6-2.
Individuals face incentives to maximize and increase their total driving, "to get their
money's worth" from large fixed investments in automobiles. The transport agency's Ushaped average cost curve implies economies of scale when roadway development is a
goal giving politicians and transport professionals an incentive for underpricing.58 Once
congestion develops there is little or no economic justification for underpricing, but there
are frequently institutional benefits since public agency funding depends on dedicated fuel
taxes. The upward sloping cost curves associated with congestion, and with social and
environmental costs, means that society increasingly benefits from reduced driving. As a
result of these different price signals, the perspectives of individual drivers and automobile
oriented planners conflict with progressive transportation planners and society in general.
58 Decision makers perceive benefits that exceed investment costs and conclude, "This road improvement
program is good for the community.", or "A small tax expenditure offers many benefits." See Stephen
Goddard's Getting There for discussion of the arguments used by road advocates to gain support.
Page 6-22
Transportation Cost Analysis
Motor Vehicle Use Conflicting Cost Curves
Figure 6-2
~User
Costs
-Transport Agency Costs
. . . . . .Environmental and
Social Costs
.!!
i
:.
~
Annual Mileage
Since most motor vehicle costs are fixed, average costs decrease with increase mileage.
Automobile owners face incentives to maximize their driving. Facility development has a
downward sloping cost curve (economies of scale) when traffic is low, since increased
driving allows costs to be divided among more miles of use. Once the system becomes
congested average costs increase. Most environmental, and social costs of driving are
minimal when use is low, but slope upward, especially once congestion develops.
The emphasis on increasing automobile capacity, rather than broader community
development and environmental goals, has become deeply ingrained in transportation
planning and financing . The existence of dedicated roadway agencies and funding skews
planning decisions toward roadway development. As described by Harry Dimitriou,
"... the conventional bias in traditional [planning] methodologies with their concern for
transport systems efficiency above all else, exists because those most intimately
involved in such approaches are well equipped with tools and techniques to design
and plan 'operational efficient' networks, whereas the equivalent expertise in the
planning and management of more 'developmentally effective' transport systems is
much less advanced "59
By articulating the differing perspectives held by various stakeholders in the transportation
planning process, participants can begin to understand and resolve the conflicts that exist.
59
Harry Dimitriou, Urban Transport Planning, Routledge (NY), 1992, p. 220.
Page 6-23
Transportation Cost Analysis
6.6
Summary: Implications of Underpricing on Households and Individuals
You may be wondering, "What does this mean to me?" Underpriced driving provides
various benefits. More households are able to own a car, live in suburban and e:xurban
areas, and travel farther than if prices were higher. It gives drivers a competitive advantage
in obtaining jobs, education, housing, services, safety and status over non-drivers. To what
degree these effects increase overall wealth and happiness is difficult to determine.60 At
least some, and perhaps most, of these benefits are captured. In addition, underpricing
imposes costs that show up elsewhere in household budgets, such as higher consumer
prices to pay for "free" parking, higher taxes to pay for roads and related services, and
increased health costs.
The effects of transport underpricing on commercial activities and employment are
uncertain. It benefits some industries and firms but burdens others. At one time,
underpricing may have provided significant external benefits by reducing average roadway
and industrial development costs. These are historical benefits. There is no evidence that
current driving provides external marginal benefits, which is to say that you benefit overall
ifyour neighbors drive more. Documented negative effects of underpriced driving include:
• Increased overall transportation costs. Low marginal prices for driving encourage
individuals to spend a greater portion of their budget on driving and incur greater
external costs. U.S. residents spend a greater portion of household expenditures on
transportation than people in other countries, and the U.S. devotes a greater portion of
GDP on transport than most other nations.
• Increased automobile dependency and reduced transportation choices, since fewer
people (especially middle- and upper-class people who have resources) walk, bicycle,
°
6
For an inquire into whether increased wealth and driving provides increased happiness see Richard
Douthwaite, The Growth Jllusion , Council Oak Books (Tulsa), 1993 . This book also explores many of the
overall negative impacts of increased automobile dependency and use.
Page 6-24
Transportation Cost A nalysis
or ride public transit and trains, so services and facilities have received significantly
less investment over the last century than would otherwise occur.
• More pollution (air, noise, water) and resource consumption, especially petroleum,
and increased motor vehicle accidents.
• Increased resources devoted to roads, parking facilities and automobile oriented
services. This means that more land is removed from other productive uses, higher
taxes, or reductions in other government services, and higher costs for many services
to pay for "free" parking.
• Automobile oriented land use, economic and social patterns. Increased centralization
and scale of services and activities, and less emphasis on neighborhood activities,
services and relationships. The use and usefulness of streets for non-driving purposes,
including walking, playing and other socializing has declined. There is evidence that
fewer pedestrians on streets reduces the safety of those who do walk.
How these affect you, or any specific individual or household depends on many factors
including driving ability and automobile ownership, income, residence and job location,
and future goals. The benefits of underpriced driving are highly skewed toward those who
drive the most, which includes people who are relatively wealthy, exurban and rural
residents, and long distance commuters. Children and teenagers, the elderly, the very poor
and the handicapped tend to use automobiles relatively little, receive the least benefits of
underpricing and suffer the most disbenefits. The effects of underpricing on middle- and
lower-middle class families appears mixed. Although they enjoy benefits from driving, they
are forced to spend more resources on transport than would be necessary with a less
automobile dependent transportation system, which strains many household budgets.
Page 6-25
Transportation Cost A nalysis
The costs of underpriced driving depend on how much you value environmental
protection, much you enjoy walking, bicycling and interacting with neighbors, and what
importance you place on providing benefits for economically, physically and socially
disadvantage people, and future generations.
The effects of underpriced transport on individuals and families depends significantly on
whether the analysis is individual or social, short or long term. Many benefits of
underpricing give individuals competitive advantages, but provide little or no overall
benefit to society. From a social rather than an individual perspective, all costs, including
external costs, must be considered, which further reduces the net benefits expected from
underpricing. Many of the benefits of underpricing (and disbenefits of increased prices)
decrease over time as individuals and communities respond by changing expectations
(such as the size and number of automobiles that are expected to maintain a household at a
certain status level), changes in land use, and investments in alternative travel modes. A
short term, individualistic perspective will estimate much greater benefits and fewer costs
of underpricing than will a long term, social perspective.
The effects of increased transport prices depends how new prices are structured, how
quickly and predictably changes occur, whether alternative travel (walking, bicycling, bus
and train) provisions are improved, and how revenues are distributed. Many ofthe costs
ascribed to transportation price increases, such as unemployment in automobile oriented
sectors, are temporary and avoidable transition costs.
Page 6-26
Transportation Cost Analysis
7.0 Evaluating Transportation Equity
The benefits and costs of transportation are not allocated equally. Is that fair? This chapter
explores the concept of transportation equity and suggests better ways to incorporate
fairness into transportation decisions.
7.1
Defining Transportation Equity
Although equity (fairness) is often cited as a concern in transportation decisions, this
subject has received little research among transportation professionals. 1 What is meant by
transportation equity is often unclear. There are three common definitions:
1.
Horizontal Equity.
This is concerned with the fairness of cost and benefit allocation between groups who
have comparable wealth and ability. Horizontal equity is cited when communities battle
for transport funding, and when charges are distributed among transport system users.
2.
Vertical Equity, By Income
This focuses on the allocation of costs between income classes. According to this
definition transport is most equitable if it provides the greatest benefit at the least cost
to the poor, therefore compensating for overall social inequity. This definition is often
used to support transport subsidies and oppose price increases.
1 Overviews
of transport equity include Hank Dittmar, "Isn't It Time We Talked About Equity" Progress,
Vol. IV, No. 5, Surface Transportation Policy Project, June 1994; Schaeffer and Sclar, A ccess for A ll,
Columbia University Press (New York), 1980; David Hodge, "Social Impacts of Urban Transportation
Decisions: Equity Impacts," in The Geography of Urban Transportation , Susan Hanson (Ed.) Guilford
Press (New York), 1986; Rosenbloom and Altshuler in "Equity Issues in Urban Transportation", Policy
Studies Journal~ Autumn 1977, p. 29-39. The Highway Cost & Pricing Study, by Cambridge Systematics
for the Wisconsin Dept. of Transportation Translinks 21 project includes comprehensive equity analysis.
Page 7-1
Transportation Cost Analysis
3.
Vertical Equity, By Need/Ability
This is a measure of whether an individual is relatively transportation disadvantaged
compared with others in their community. It assumes that everyone should enjoy at
least a basic level of access, even if people with special needs require more resources
per mile, per trip or per person. Applying this concept is difficult because there are
currently no standards for transport need, nor a consistent way to measure access.
Quantifying Transportation by Access
Access depends on the time, expense and effort required to reach destinations and
services. It varies with individual and community circumstances, making it difficult to
quantify. For example, who has better access, a low income driver in an automobile
dependent city, or a low income non-driver in a multi-modal community? Does a nondriver who can afford taxi fares have mobility problems? Are children in auto dependent
communities disadvantaged because they must be chauffeured to any destination?
7.2
Current Transportation Equity Analysis
Most current analyses of transportation equity focus on only one type of equity and
consider only market effects. Examples are described below.
Horizontal Equity
Horizontal equity is often an issue in the allocation of transport funds to geographic
jurisdictions because transportation projects provide short term economic stimulation
Gobs and contracts) and long term economic development. Political representatives often
fight for a fair share of this money and various formulas and decision making frameworks
have been developed to distribute these resources fairly. Some studies examine the ratio
between the state or federal transportation tax contributions from a jurisdiction and the
funding it receives back, on the assumption that a low ratio would be unfair.
Page 7-2
Transportation Cost Analysis
The equitable distribution of costs and benefits between modes and vehicle classes has
received a moderate amount of research among transportation economists. A number of
cost allocation studies have examined whether different vehicle classes (automobiles,
medium trucks and heavy trucks) pay a fair share of the costs they impose on the roadway
in taxes.2 One study compares overall automobile financial costs and revenues.3 A few
studies have also compared the costs of road and rail transport.4
An increasing concern is the equity of disbenefits to a community caused by a transport
project that provides few local benefits. 5 Urban neighborhoods are negatively impacted by
freeways or other road improvements that primarily benefit suburban commuters. For
example, despite vigorous opposition, freeway construction in the late 1950's leveled 750
African American homes and businesses in Nashville, Tennessee, virtually destroyed that
community.6 In addition to being unfair in terms ofhorizontal equity, this frequently
imposes vertical inequity since the people who are impacted tend to be less aftluent than
the drivers who use the facilities. Such unfair treatment of local communities lead to
"freeway revolts" which often pit poorer urban residents against suburban development
interests. This resistance has stopped many planned urban freeway projects.
Income Equity
Underpricing driving is often justified for the sake of income equity. A few studies have
examined the distribution of transport financial costs and benefits by income class, and the
impacts of price changes. One study examined the effects of sudden oil price increases by
2 Kenneth Small, et al, Road Work, Brookings Institute (Washington DC), 1989.
3 Cora Roelofs and Charles Komanoff, Subsidies for Traffic: How Taxpayer Dollars Underwrite Driving
in New York State , Tri-State Transportation Campaign (NY), 1994.
Concepts, External Costs ofTruck and Train , Brotherhood of Maintenance of Way Employees
(Ottawa), Nov. 1994.
5 Helen Leavitt, Superhighway-Superhoax, Ballantine, 1970; Ben Kelley, The Pavers and the Paved,
Brown, 1971 ; Stephen Goddard, Getting There , Basic Books (New York), 1994, especially chapter 13.
6 David Hodge, "Social Impacts of Urban Transportation Decisions: Equity Impacts," in The Geography
of Urban Transportation , Susan Hansen (Ed.) Guilford Press (New York), 1986, p. 302.
4 Transport
Page 7-3
Transportation Cost A nalysis
income class and concluded that, "The greatest beneficiaries of lower energy prices would
be the poor. "7 Merle Mitchell argues that fuel taxes are regressive since the lowest income
households spend a larger portion of income on motor vehicle fuel than the highest income
class. 8 Mark French concludes that a $0.15 per gallon fuel tax increase would be regressive
with respect to income, 9 although his analysis significantly overstates this factor by
assuming incorrectly that driving and fuel consumption rates are the same for all income
classes. The poor, especially the very poor, own fewer automobiles, drive less and rely
more on alternatives to driving (Figure 7-1) than the rich. 10
Figure 7-1
Annual Vehicle Travel By Income 11
40000
.,
36000
.!!
30000
Gl
"'
:E
Gl
u
:c
~
iii
:I
c
~
I:..:..:: ::ookjl
25000
20000
15000
10000
"'
. - .. - - .
.
. .
"' "'
"'
"' "'
"'
"' "' "'
"' "'
5000
0
<$10,000
$10,000$19,999
$20,000$29,999
$30,000$39,999
$40,000+
Annual Household Income
Higher income households and individuals use motor vehicles more than those with lower
incomes, and thus enjoy a greater share of benefits and user subsidies. This implies that
reducing subsidies to driving may be progressive if savings benefit poor households.
7
Steven Rock, "Distributional Changes in Consumer Transportation Expenditures: 1972-1985,"
Transportation Research Record 1197.
8 See for example Merle Mitchell, "Links Between Transport Policy and Social Policy," in Transport
Policies for the New Millennium , Ogden et al. editors, Monash University (Clayton), 1994.
9 Mark French, "Efficiency and Equity of a Gasoline Tax Increase," Energy Systems and Policy, Vol. 13,
1989, pp. 141-155. It incorrectly assumes that average fuel consumption is the same for all income levels.
10 Golob, "Casual Influence oflncome and Car Ownership," Transport Economics, May 1989, p. 149.
11 Hu and Young, 1990 NPTS Databook, Vo/.1, FHWA (Washington DC), Nov. 1993, Table 3.14.
Page 7-4
Transportation Cost Analysis
The actual burden imposed on poor households by increased automobile taxes is less
certain than these studies imply. James Poterba has demonstrated that the poorest
households actually spend a smaller portion of their annual expenditures (which he
considers a more appropriate reference than income) on gasoline than middle class
families, and concludes that, "The gasoline tax thus appears far less regressive than
conventional analyses suggest. "12 His work indicates that fuel taxes have little or no
regressivity relative to lifetime expenditures.13
Werner Rothengatter argues that wealthier European drivers misrepresents equity
concerns to justify low vehicle use taxes that are actually regressive due to higher
automobile use by wealthier households. 14 Similarly, David Banister indicates that a 26%
increase in British petroleum prices would be a progressive tax with respect to income
since many poor households do not own a car. 15 Shifting fixed vehicle ownership taxes to
variable vehicle use taxes would be even more progressive since poor car-owning
households tend to drive less annually per automobile than wealthier households. He
concludes that the overall equity impacts of increased vehicle taxes depends in part on
whether automobile use is considered a necessity or a luxury, since poor households that
do own cars are disadvantaged by such tax changes. 16
The equity impacts of congestion pricing and other road fees has been studied by several
researchers. These typically conclude that equity impacts depend on how revenues are
distributed.17 Genevieve Giuliano summarizes current research and the fairness of current
12
James Poterba, "Is the Gasoline Tax Regressive?", Tax Policy and the Economy, MIT Press, 1991.
13 James Poterba, "Reexaminations of Tax Incidence: Lifetime Incidence and the Distributional Burden of
Excise Taxes," The American Economic Review, Vol. 79, No. 2, May 1989, p. pp. 325-330.
14 Werner Rothengatter, "Obstacles to the Use of Economic Instruments in Transport Policy," in
Internalising the Social Costs ofTransport, OECD (Paris), 1994.
15 David Banister, "Equity and Acceptability Questions in Intemalising the Social Costs of Transport," in
Internalising the Social Costs ofTransport, OECD (Paris), 1994.
16 Banister's analysis of the equity impacts of urban road pricing appear to assume that automobile use is a
necessity, even by city dwellers, despite evidence he presents to the contrary.
17 Ken Small, "Using the Revenues from Congestion Pricing," Transportation , 19/4, pp. 359-381.
Page 7-5
Transportation Cost Analysis
road funding mechanisms, which she concludes are overall regressive (Table 7-1 ) . 18 She
emphasizes the problems facing women commuters in shifting from SOV travel due to
family responsibilities and inflexible employment conditions, implying an inequitable
burden, and that at least some low income drivers would be worse off overall from
congestion fees. She emphasizes the need to analyze impacts by gender, employment type,
location, commute distance, and other criteria in addition to income. John Kain identifies
significant potential benefits to poor commuters (and non-drivers) from congestion pricing
by incorporating transit and ride sharing service improvements, including travel time
savings and increased bus frequency, plus revenue rebates. 19
Table 7-1
-
-
Incid
------ -- - fT
- ------ Used
~ --
to S
Hi2h
Tax
s
-------
20
Incidence
Regressive
Regressive
Regressive
Regressive
Progressive
Regressive
Federal, state fuel gasoline tax
State use fees
State sales tax
Local sales tax
Federal, state income tax
Property tax
Most current taxes used for roadway funding are regressive.
A recent study indicates that Pay-As-You-Drive insurance increases income equity by
eliminating the high premiums often required for residents of low income communities,
reducing costs for low annual mileage drivers, and providing overall insurance system
savings.21 Low income households would pay 30 to 80% lower premiums than under the
current system, in part because low income households drive less than wealthier
households. This analysis understates total potential benefits to the poor by considering
18 Genevieve Giuliano, "Equity and Fairness Considerations of Congestion Pricing," in Curbing Gridlock,
TRB, National Academy Press (Washington DC), 1994, p. 250-279.
19 John Kain, "Impacts of Congestion Pricing on Transit and Carpool Demand and Supply," in Curbing
Gridlock, TRB, National Academy Press (Washington DC), 1994, p. 502-553 .
20 Genevieve Giuliano, in Curbing Gridlock, National Academy Press, (Washington DC), 1994, p. 260.
21 Jeff Allen, Roland Hwang, and Jane Kelly, An Equity A nalysis of "Pay-A s-You-Drive" Automobile
Insurance in California, Union of Concerned Scientists (Washington DC), Nov. 1994.
Page 7-6
Transportation Cost Analysis
only market costs. Reduced pollution, traffic congestion and automobile dependency
would also benefit lower income households.
Michael Cameron concludes that a $0.05 per mile road user charge in Southern California
would not necessarily be regressive, since all income quintiles benefit from reduced
congestion and air pollution including the poorest residents.22 Cameron suggests that this
estimate probably understates benefits to the poor, since low income people tend to be
exposed to higher than average pollution, 23 a factor not incorporated into his model.
Although air pollution reduction benefits would be lower in other regions, Cameron
considers air pollution a proxy for other environmental impacts not priced in his study,
which would provide additional benefits in all regions.
Robert Johnston, et al. consider the income equity of congestion management strategies,
including HOV facilities, metering, pricing, and rationing, and conclude that pricing can be
equitable if revenues are appropriately spent. 24 Similarly, Ken Small states that, "... when
central cities are the recipients of the toll revenue, the toll causes a monetary transfer
from rich to poor plus a uniform time saving enjoyed by all. '125
After examining impacts on the poor of internalizing transport costs, Per Kageson
concludes, "... the reform (without any refund of the revenues) will do little to change the
existing differences between income groups, and the strain on low income groups can be
offset by refunding a portion of revenues, in equal amounts, to all citizens. '126 Douglass
22
Efficiency and Fairness on the Road, Environmental Defense Fund (Oakland), 1994.
David Hodge in The Geography of Urban Transportation, Susan Hansen (Ed.) Guilford Press (New
York), 1986, p. 303 . This issue has received increasing attention as an issue of environmental justice.
24 Robert Johnston, Jay Lund and Paul Craig, "Capacity-Allocation Methods for Reducing Urban Traffic
Congestion," Journal ofTransportation Engineering, Vol. 121, No. 1, pp. 27-39, January 1995.
25 "The Incidence of Congestion Tolls on Urban Highways", Ken Small, Journal of Urban Economics,
December 1983, p. 90-111.
26 Per Kageson, Getting the Prices Right, European Fed. for Transport & Env., 1993, p.185.
23
Page 7-7
Transportation Cost Analysis
Lee concludes that congestion charges are unlikely to be overall regressive because peak
period drivers tend to be wealthier than average. He states,
"...peak tolls (in the peak direction during peak hours) would be a progressive source
of revenue. All existing user and non-user funding sources (such as property and sales
tax) are less progressive or are regressive. "27
Regulation is sometimes seen as more equitable than pricing to manage markets and
reduce impacts, because with pricing, "motoring will become the prerogative of the
wealthy. "28 This conclusion is debatable since regulations increase costs (for example,
increasing the price of automobiles), and penalties for failing to comply are often fines, all
of which are a greater burden to low income drivers. Elizabeth Deakin finds that nearly
half of all vehicles owned by poor families (annual income less than $25,000) in the San
Francisco Bay area are older, high polluting models, but such households produce only
12% of mileage and 15% of trips made in these older vehicles.29 The majority of older
vehicles are owned by wealthier families, which account for 3/4 of their mileage.
Congestion is a progressive cost with respect to income because wealthier people have
higher opportunity costs for their time.30 This is tempered by the fact that wealthy drivers
can afford more comfortable cars, mobile communications, a greater choice of housing
locations, more flexible schedules, and in some cases alternative modes.
27
Highway Pricing as a Tool for Congestion Management, Douglass Lee, Principal Investigator,
Transportation Systems Center (Cambridge), October 1989, p. 13.
28 David Banister and Kenneth Button, "Environmental Policy and Transport: An Overview," in
Transport, the Environment and Sustainable Development, E&FN SPON (NY), 1993, p.7.
29 Elizabeth Deakin, "Policy Response in the USA," Transport, the Environment and Sustainable
Development, E&FN SPON (NY), 1993, p. 95.
30 Herbert Mohring and David Anderson, Congestion Pricing for the Twin Cities Metropolitan Area,
Department of Economics, University ofMinnesota, January 1994.
Page 7-8
Transportation Cost Analysis
Need/Ability Based Equity
Several factors can make an individual transportation disadvantaged, including age,
physical disability and poverty.31 Approximately 26% of the U.S . population is under 18,
and about 12% is over 65 years of age.32 Since young children couldn't travel
independently even if a vehicle was available, it is the approximately 15% in the 7 to 17
range that can be considered most disadvantaged.33 About 7% ofthe U.S . population is
estimated to be mobility impaired due to mental or physical disability. 34 Over 13% live in
poverty (defined as a family offour earning less than $13,359 annually in 1990).
The 1990 National Personal Transportation Survey indicates that 9.2% of households with
6.4% of the population do not own an automobile, but since this survey is based on
telephone interviews it is believed to underrepresent poor households.35 The 1989
American Housing Survey and the 1990 Census indicate that 15.9% and 11.5% of
households own no automobile respectively, so the portion of the population living in a
household without a car is probably about 10%. The number of households without
vehicles may be a poor indicator of the total number of people who are transportation
disadvantaged. Simply because an individual lives in a household that has at least one
automobile is no proof they are not transportation disadvantaged due to automobile
dependency. Increasing travel demands and reduced travel choice mean that a personal
automobile is required for full participation in society. Automobile transportation is
increasingly required for jobs, schooling, recreation and participation in civic activities.
3l "Mobility or, Rather the Lack of It" in Transport Sociology, by Hillman, Koutsopoulos, Schmidt,
Henerrson, and Whalley.
32 The World Almanac, 1993, Pharos Books (New York), 1993.
33 Although 16 and 17 year olds can obtain drivers licenses, and some people over 65 continue to drive,
young driver's freedom is often limited by access to vehicles and liability insurance costs, and many people
younger than 65 must curtail their driving due to age related constraints, so the under 18 and over 65 age
ranges seem reasonable to represent portions of the population that are transport disadvantaged by age.
34 John Meyer and Jose Gomez-Ibanez, Autos Transit and Cities, Harvard U. Press (Cambridge), 1981.
35 Richard Crepeau, Zero Vehicle Households: Issues ofTransport and Housing, 1995 TRB General
Meeting (Washington DC).
Page 7-9
Transportation Cost Analysis
There is considerable overlap between the transportation disadvantaged groups described
above. In their 1981 book Autos Transit and Cities, John Meyer and Jose Gomez-Ibanez
estimate that 23% ofthe U.S . population is transportation disadvantaged due to some
combination ofpoverty, disability or age over 65 . Including young people aged 7 to 17
would indicate that 1/3 or more of the population is transportation disadvantaged.
According to some studies, women tend to be represented in transportation disadvantaged
categories more than men.36
Need based transportation equity concerns tend to focus on two issues. One is access for
physically disabled people to public facilities, which has resulted in the passage of various
handicapped access requirements including the Americans with Disabilities Act. The
second area of attention is support for transit and special mobility services to provide a
basic level of mobility for all residents. Although there is no doubt that many transport
disadvantaged people rely on public transit, there is little research on the degree to which
they depend on transit, and even less information on how much these groups use other
modes, such as bicycling and walking.
Few current analyses consider both income and need/ability equity, 37 and none effectively
incorporate non-market effects such as the distribution of pollution, accident costs or
comfort by income or ability. Most equity analysis focus on the short term, failing to
consider long term effects on land use and automobile dependency. Failing to consider
these impacts ignores important equity considerations and tends to skew results toward
justification of underpricing and continued overemphasis on automobile transport.
36 Julia
Walton, "Gender and Urban Form," The Urban Ecologist, Fall 1993; Karen Overton, "AutoDependence: A Driving Force for Gender Inequality," The Urban Ecologist, 1995, No. 1, p. 16.
37 An exception is Access to Opportunity: Cooperative Planning to Improve Mobility for Residents of
Inner-City Communities, East-West Gateway Coordination Council (St. Louis, MO), 1995.
Page 7-10
Transportation Cost Analysis
7.3 Automobile Dependency as an Equity Issue
Due to its importance in economic and personal development mobility is frequently
considered a necessity and even a right. 38 Some trips are considered more important to
society than others, and can be defined as "basic mobility." This basic level of mobility
includes access to services and employment, and to a lesser degree social and recreational
activities. Whether increased prices for driving is inequitable to low income drivers
depends in part on whether driving is a necessity or a luxury. If poor people must drive,
any increase in user prices is an inescapable and unfair burden. If usable transport
alternatives exist then increased prices can be considered acceptable. However, if driving
is a necessity, then non-drivers are disadvantaged even if driving is cheap, and their
relative disadvantage is exacerbated by anything which decreases their transport options.
This issue is explored in this section.
The degree to which transport, land use and social patterns emphasize driving relative to
other modes is called Automobile Dependency.39 Automobile dependency can create a
self-perpetuating cycle. As discussed in chapters 3.9 (Equity), 3.13 (Barrier Effect), and
3.14 (Land Use Impacts), automobile use increases barriers to pedestrian and bicycle
travel, reduces the viability of transport alternatives, and increases the amount of travel
between destinations, all of which further increases driving.
The equity impacts of automobile dependency can be defined in formal economic
language. Consider all trips you would like to take during a time period ranked from
highest to lowest user cost (including financial costs, time, accident risk, and comfort).
Economists call this a supply curve. This transport supply curve varies from one individual
and community to another. For many trips, non-drivers have significantly higher costs than
drivers. This difference increases as communities become automobile dependent, as travel
38
39
John Hamburg, Larry Blair and David Albright, Mobility as a Right, TRB Annual Meeting, Jan. 1995.
Peter Newman and Jeff Kenworthy, Cities and Automobile Dependency, Gower (Aldershot), 1989.
Page 7-11
Transportation Cost Analysis
needs increase, land use becomes more dispersed and other travel modes decline. This
average cost premium for non-drivers varies from one community to another depending
on the degree of automobile dependency. This is illustrated in Figure 7-2.
Figure 7-2
Transportation Supply Curve for Drivers and Non-Drivers
-Average Driver
--Average Non-Dr1ver
~
~
Trlpo
Each individual faces a different transportation supply curve (the costs of trips ranked
from lowest to highest). In most communities non-drivers have significantly higher total
costs, which reflects the degree of automobile dependency.
Economists measure benefits of consumer expenditures net their costs, called consumer
surplus. If the average cost of travel is higher for non-drivers than drivers, non-drivers
enjoy less consumer surplus. This is illustrated in Figure 7-3 . This difference in consumer
surplus between drivers and non-drivers is one way to measure transportation inequity.
Automobile dependency increases this difference in consumer surplus between drivers and
non-drivers. It is also unfair to low income drivers who must spend a greater portion of
their income on transport than they would otherwise, leaving fewer resources for other
expenditures and less consumer surplus.
Page 7-12
Transportation Cost Analy sis
Figure 7-3
Consumer Surplus of Drivers and Non-Drivers
1
1il
Driver
Consumer'
Surplus •
8
~
1-
Driver Average Trip Cost
Trips
Costs per trip are higher on average for non-drivers compared with drivers. Non-drivers
take fewer trips and enjoy less consumer surplus than drivers.
Automobile Dependency Equity Cost Example40
The Chimawa Indian Health Clinic provides health care to Native Americans in Western
Oregon. It has over 18,000 regular clients. When opened in 1970 the clinic had no public
transit service. Patients who relied on transit, including those who were sick, pregnant,
disabled, elderly and children, had to walk a mile on a muddy trail from the nearest bus
stop. After years of political pressure, protests and legal challenges the local transit agency
extended the bus route to the clinic, but this service is in jeopardy due to low ridership.
This illustrates the equity costs of automobile dependency. If the community was less
automobile dependent, transit access would have been a requirement when originally siting
the facility. A larger portion of non-driver employees and patients would create demand
for better transit service, pedestrian and bicycle facilities. Automobile dependency forces
households to own automobiles whether or not they can afford the costs to access critical
services such as health care. Since these costs are most significant for people who are
already disadvantaged (the poor, handicapped, and elderly), it is inequitable.
40 Jacky Grimshaw, Impacts of Siting Non-Transportation Public Facilities, Center for Neighborhood
Technology (Chicago), November 1994.
Page 7-13
Transportation Cost Analysis
7.4
Comprehensive Transportation Equity Analysis
The total cost perspective presented in this report allows a more comprehensive analysis
of transport equity. Equity analysis depends on how transportation system user classes are
defined. Major variables include location (urban, suburban, rural), income (low, medium,
high), physical ability (disabled non-driver, able non-driver, driver), and lifecycle stage
(child, adolescent, adult, parent, elderly). For the sake of simplicity, users are divided into
four major classes which capture many of the variables just listed:
•
Non-drivers. This includes people who cannot drive due to age or disability, or
poverty. Non-drivers use automobiles as passengers, but except for those who can
afford unlimited chauffeuring, their use is typically much less than drivers.
•
Low income drivers. This includes people who can drive and have access to an
automobile but whose travel decisions are significantly affected by financial costs.
Vehicle user prices have a major effect on their travel habits.
• Middle income drivers. This includes drivers who normally have unrestricted access to
an automobile and who are only moderately burdened by their automobile financial
costs. Vehicle user prices have a moderate effect on their travel habits.
•
Upper income drivers. This includes drivers who normally have unrestricted access to
an automobile and are not burdened by their automobile financial costs. Vehicle user
prices have little effect on their travel habits.
For equity analysis, transportation costs are divided into six categories:
l . User market. These include vehicle ownership and operating costs, out-of-pocket
parking expenses, and transit fares . These are the focus of most transportation
planning and equity analysis.
Page 7-14
Transportation Cost Analy sis
2. User non-market. These include user travel time, accident risk and comfort. These are
often recognized in planning but are frequently ignored in equity analysis. For
example, few analysis formally identify how the speed, safety and comfort benefits of a
transportation decision are distributed by income or user class, although this is
sometimes considered informally when allocating transport resources between
different geographic locations, or between automobile and transit investments.
3. External market. These include roadway facility costs, parking subsidies, and transit
subsidies that originate as general taxes or increased consumer prices. Equity is
sometimes a factor in the allocation of these costs and benefits (for example, in the
allocation of road tax burdens between different motor vehicle classes) but there is
often disagreement as to how equity should be measured.
4. External environmental. This includes noise, air and water pollution, the barrier effect,
aesthetic degradation, and habitat loss. These are sometimes considered during
transportation planning but are usually ignored in equity analysis. For example, there is
little data on the impacts of air pollution or the barrier effect by income class. Some
analyses assume environmental benefits are valued most highly by wealthy residents,
while others point out that pollution costs tend to be borne most by lower income
residents, but little quantitative research has been done.
5. Automobile dependency. This includes the costs of reduced transportation choices and
generated traffic that result from automobile dependent transportation system and land
use patterns. These are seldom incorporated in transport planning or equity analysis,
although some recent discussions of sustainable community planning consider them.
6. Economic. This includes changes in consumer prices, economic development,
employment, and productivity, as discussed in sections 5.1 and 5.2. These costs are
Page 7-15
Transportation Cost A nalysis
frequently cited in general discussions of transportation policy, but little quantitative
research has been done. Many claimed economic benefits of transport improvements
are distributional, representing gain in one community that is offset by a loss
elsewhere, which is a horizontally inequitable. Except for a few studies of the
employment benefits of specific transportation projects, few studies have examined
how such costs and benefits are distributed by income or class. Transaction costs, such
as economic changes that result in unemployment among traditional industries tend to
be vertically inequitable since disadvantaged people have fewer resources to fall back.
7.5
Comprehensive Equity Analysis Applications
These three user classes and six major cost categories can be used for comprehensive
analyses of transportation equity impacts. Below are three examples.
Example 1.
Equity Effects of Automobile User Prices.
As described in Chapter 4, driving is significantly underpriced. Is this equitable? Would
increasing prices (higher fuel taxes or road user fees), to better reflect marginal costs and
to encourage more efficient travel habits, be more or less equitable? For example, a $0.50
per gallon fuel tax increase would increase average marginal automobile costs by about
20% and total internal financial automobile costs by about 6%. It would raise about $500
billion annually in the U.S . and would decrease total driving by approximately 4%. What
would be the equity impacts of such a change? Specific impacts by cost category and
transportation system user class are evaluated in Table 7-2.
Table 7-3 summarizes these impacts of transportation underpricing. This analysis shows
that increased prices has complex equity impacts that are ignored most analyses, including
indirect benefits that are usually unrecognized. Although higher fuel taxes or a road user
fees increase market costs for all drivers and force drivers (especially low income drivers)
Page 7-16
Transportation Cost Analysis
to use less desirable (to the user) travel modes for some trips, there are significant nonmarket benefits to all residents, including reduced congestion, reduced road and parking
subsidy costs, reduced environmental impacts, and increased transport choices. These
offer significant potential benefits to non-drivers and low income people.
User Class
Cost
Non-Driver
User Market
Low Income
User Market
Drivers
Middle and Upper
Income Drivers
Non-Drivers
Low Income
Drivers
Middle and Upper
Income Drivers
User Market
User
Non-Market
User
Non-Market
All Road Users
User
Non-Market
External
Market
All Residents
Environmental
All Residents
Automobile
Dependency
All Residents
Economic
Expected Effect
No short term change. Small long term benefits due to economies of
scale in transit service.
Higher short term costs, with mixed long term effects since drivers
will sometimes shift to cheaper (but less desirable to the user)
modes such as public transit, bicycling, and walking, and other
times will pay the higher cost. Whether the net result is increased
market costs or savings depends on the availability of substitutes. If
alternatives are viable, higher driving costs will motivate a shift to
these modes, resulting in overall financial saving. If few alternatives
exist, low income households will spend more overall on transport.
Higher costs since they will usually drive despite higher prices.
No short term change is likely. A moderate long term benefit is
likely due to improved transit, bicycling and pedestrian service.
Significant increased costs since these users will be priced out of
driving for some trips, and forced to use less desirable (to the user)
modes such as transit, bicycling and walking. The size of this cost
increase depends on the quality of alternative modes. This cost is
offset somewhat by reduced congestion delay and accident risk.
Moderate benefit due to reduced congestion delay and accident risk.
Moderate benefit, including reductions in other taxes and reduced
parking subsidy costs due to less driving.
Moderate benefit from reduced pollution, energy consumption, and
urban sprawl due to less driving.
Various benefits, especially for non-drivers and low-income drivers
who are most likely to use the increased transport choices. These
benefits increase in the long term.
Slight short term consumer price and employment transition costs,
and slight long term benefits due to productivity gains, as discussed
earlier. Automobile sector job losses may be slightly regressive, but
these could be offset by increased employment in transit and other
sectors that substitute for driving.
The overall equity impacts of price changes depend largely on how revenues are
distributed. If each income class receives revenues comparable to what they pay, the
Page 7-17
Transportation Cost Analysis
overall effect is probably slightly progressive (increases vertical equity) due to benefits to
non-drivers. If some revenues are specifically targeted at disadvantaged people (the poor
and non-drivers) the overall effect could be strongly progressive. Spending revenues only
on new highways is probably regressive (since these are used primarily by higher income
travelers), but typical expenditures on transit, bicycle and pedestrian improvements are
probably progressive. Analysis of the distributional effects of increased prices and benefits
by geographic area or subgroups may indicate additional horizontal inequities.
blj
ffi
User Market
User Non-Market
External Market
Environmental
Auto Dependency
Economic
Example 2
s
NonDrivers
Short Term
Low
Middle-High
Income
Income
Drivers
Drivers
NonDrivers
None
Slight
Benefit
Moderate
Benefit
Moderate
Benefit
Large
Benefit
Small
Cost
Moderate
Cost
Large
Cost
Moderate
Benefit
Moderate
Benefit
Moderate
Benefit
Slight
Cost
Slight
Benefit
Moderate
Benefit
Moderate
Benefit
Moderate
Benefit
Large
Benefit
Slight
Benefit
Moderate
Cost
Small
Benefit
Moderate
Benefit
Moderate
Benefit
Slight
Benefit
Slight
Cost
Lon2 Term
Middle-High
Low
Income
Income
Drivers
Drivers
Mixed
Mixed
Moderate
Benefit
Moderate
Benefit
Large
Benefit
Slight
Benefit
Moderate
Cost
Small
Benefit
Moderate
Benefit
Moderate
Benefit
Small
Benefit
Slight
Benefit
Equity of Transit Subsidies
Public transit service receives significant financial subsidies. Is this fair? An analysis that
only considers market costs may conclude that transit subsidies are inequitable, at least in
the narrow terms of horizontal equity. A comprehensive equity analysis can better identify
how costs and benefits are distributed.
If all costs are considered, the difference in external costs between public transit and other
travel modes is small. Figures 4.3 to 4.5 show that the external costs per passenger mile of
Page 7-18
Transportation Cost Analysis
a Diesel Bus rider are comparable to an Average Automobile during Urban Peak travel,
but average bus external costs are higher under Urban Off-Peak and Rural conditions. This
implies that some transit riders receive more subsidy than average drivers. However:
•
Although transit riders receive a higher subsidy than drivers per mile, drivers travel
much more than non-drivers per year, so transit dependent users receive a much
lower annual subsidy.
•
Due to unused capacity and economies of scale, a typical additional bus rider
increases the transit system's efficiency, reducing the average unit subsidy. The
marginal cost per rider using existing capacity is therefore low or negative.
•
Transit satisfies the definition of "basic mobility" and thus should be evaluated
differently from driving, which can be considered a relative luxury. It is therefore
reasonable to subsidize transit use at a higher rate per mile than driving just as public
health services provide subsidies for childhood immunization than for plastic surgery.
•
Transit service incurs many costs such as wheelchair lifts and service to special
destinations to meet community needs that should not be charged to general riders.
Taking these factors into account, the average annual subsidies received by transit riders is
significantly lower than that received by drivers, indicating that drivers as a class are
unfairly subsidized compared with transit riders in terms of horizontal equity. This inequity
is greater if vertical equity is considered, since bus riders tend to be disadvantaged
compared with average drivers. Even greater vertical inequity exists for individual transit
services or routes serving low income urban riders since these are often more cost
effective than those targeting wealthier suburban riders. For example, in the Los Angeles
area, bus riders who are primarily lower income receive average subsidies of $1 .17 per
Page 7-19
Transportation Cost Analysis
trip, while suburban rail riders who tend to be wealthier receive subsidies averaging $11 to
$21 per trip. 41 BART benefits are similarly inequitable with respect to income.42
The user and cost categories described earlier in this section can be used to identify the
benefits of transit subsidies. Table 7-4 shows how transit subsidy costs and benefits are
probably distributed in a typical urban area.
User Class
Non-Driver
Low Income
Drivers
Middle and Upper
Income Drivers
Cost
User Market
User Market
All Residents
All Residents
User Market
User
Non-Market
User
Non-Market
User
Non-Market
External
Market
Environmental
All Residents
All Residents
Automobile
Dependency
Economic
Non-Drivers
Low Income
Drivers
Middle and Upper
Income Drivers
Expected Effect
Large benefits. Transit service would not exist in most communities
without subsidies. Transit is much cheaper than alternatives such as
taxies. Subsidies provide financial savings to users and allow access
to jobs and a wider selection of commercial services.
Moderate benefits. Transit is used regularly by some low income
drivers, and it provides a backup when a car is unavailable.
Small benefit. A few afiluent drivers regularly use transit such as
suburban rail, and it provides a backup when a car is unavailable.
Large benefit. Transit service provides many non-market benefits to
non-drivers, and reduces traffic congestion and accident risk.
Moderate benefit, including reduced congestion and accident risk.
Moderate benefit, including reduced congestion and accident risk.
Moderate cost. Transit subsidies require additional taxes that are
offset slightly by reduced automobile parking and road subsidies.
Moderate benefit, including reduced air pollution and urban sprawl.
Moderate to large benefits, including more short term mobility
choices, and less car dependent transport and land use in the long
term. Non-drivers benefit most.
Mixed overall, with significant distributional effects in some areas.
41
"LA Bus Riders Union Sues MTA," Progress, Vol. V, No. 1, Surface Transportation Policy Project
(Washington DC), February 1995, p. 7.
42 David Hodge, "Social Impacts of Urban Transportation Decisions: Equity Impacts," in The Geography
of Urban Transportation , Susan Hanson Editor, Guilford Press (New York), 1986, p. 307.
Page 7-20
Transportation Cost Analysis
Table 7-5 summarizes the impacts of transit underpricing.
Table 7-5
T
·· Subsidv Benefits S
User Market
User Non-Market
External Market
Environmental
Auto Dependency
Economic
Non-Drivers
Low Income
Drivers
Middle-High
Income Drivers
Large Benefit
Large Benefit
Moderate Cost
Moderate Benefit
Large Benefit
Mixed
Moderate Benefit
Moderate Benefit
Moderate Cost
Moderate Benefit
Moderate Benefit
Mixed
Small Benefit
Moderate Benefit
Moderate Cost
Moderate Benefit
Moderate Benefit
Mixed
This analysis shows a variety of transit subsidy benefits that are greatest for non-drivers,
and to a smaller degree low income drivers, indicating vertical equity benefits. Analysis of
the distribution of specific transit subsidy costs and benefits by route, type of service (rail
vs. bus) and geographic area may indicate additional equity impacts, including horizontal
equity.
The equity effects of transit service and subsidies vary considerably. As mentioned above,
bus service used most by low income riders often have the highest farebox recovery, while
commuter bus and rail service to higher income suburbs, airports and other special
destinations are less cost effective. Funding structures that favor suburban over urban
service, and therefore fail to deliver transit service where the need and system efficiency
are greatest, are considered inequitable by Brian Taylor.43 Equity analysis by type of
service or route is likely to show that some transit subsidies (such as rail service to
wealthy suburbs) are regressive but general bus service is highly progressive.44
43
Brian Taylor, "Unjust Equity: An Examination of California's Transportation Development Act,"
Transportation Research Record 1297, Transportation Research Board (Washington DC), pp. 85-92.
44 The Los Angeles area Bus Riders Union, for example, concludes that rail transit intended for afiluent
riders is inequitable because it diverts funding from basic bus transit.
Page 7-21
Transportation Cost Analysis
Example 3
Traffic Management Equity
Different users often have conflicting interests in roadway design. Automobile users
benefit from streets designed to maximize traffic capacity and speeds, with minimal
variations or distractions. Cyclists benefit from streets designed for moderate traffic
speeds and volumes, with special provisions for bicycles. Pedestrians benefit from streets
designed for minimal traffic speeds and volumes, and special provisions for walking and
sitting. 45 Transit riders benefit from street designs that both facilitate transit movement and
that enhance the pedestrian environment. Nearby residents, visitors (such as diners at a
restaurant) and property owners benefit from streets designed for minimal through traffic
and which accommodate other activities (sitting, playing, and community interactions). 46
Because road space and funding are usually limited, and because motor vehicle traffic has
disbenefits to other users, the interests of different user groups often conflict. Roadway
design decisions therefore have equity implications.
Current street design and funding practices tend to emphasize motor vehicle needs at the
expense of other users. Traffic engineers refer to projects that increase motor vehicle
traffic speeds and capacity as "upgrading," although this may result in reduced safety and
comfort for other user classes. Specific design features that benefit drivers at the expense
of other users include hierarchical street networks, wide lanes, straight alignments, smooth
surfaces, large turning radii, synchronized traffic signals, and maximum surface parking.
Because of limited resources, accommodating motor vehicle traffic often reduces the size
and quality of sidewalks, bike paths and other facilities for non-motorized users.
45 Tim Pharoah, Less Traffic, Better Towns, Friends of the Earth (London), 1992.
46
D. Gordon Bagby, "Effects of Traffic Flow on Residential Property Values," Journal of the American
Planning Association, Vol. 46, No. 1, January 1980, p. 88-94. W. Hughes, Jr. and C.F. Sirmans, "Traffic
Externalities and Single-Family House Prices," Journal ofRegional Science, 32/4, 1992, pp. 487-500.
Donald Appleyard, Livable Streets, University of California Press, 1981.
Page 7-22
Transportation Cost Analysis
Tabl1
User Class
Non-Driver
Low Income
Drivers
Middle and Upper
Income Drivers
Cost
User Market
Expected Effect
Moderate benefit. Reduced traffic allows more walking and
bicycling, reducing transit and taxi fare expenses. It can also
increase the market value of residences adjacent to streets.
User Market
Small cost. Traffic restrictions increase automobile operating costs.
Non-Drivers
Low Income
Drivers
Middle and Upper
Income Drivers
User Market
User
Non-Market
User
Non-Market
User
Non-Market
All Residents
External
Market
Small cost. Traffic restrictions increase automobile operating costs.
Large benefit. Traffic restrictions and pedestrian/bicycle improvements increase safety, comfort and mobility under all conditions.
Small cost. Motor vehicle traffic restrictions increase driving time
but reduce accidents, and increase comfort when not driving.
Small cost. Motor vehicle traffic restrictions increase driving time
but reduce accidents, and increase comfort when not driving.
Moderate benefit. Short term expenses required to implement traffic
management projects. These are offset by savings from reduced
vehicle accidents and increased property values.
Large benefits, including reduced traffic noise, severance and
sprawl. Overall energy saving and air aualitv benefits likelv.47
Large benefits. Significantly increases the viability of nonautomotive modes.
Mixed; location specific. Constrains some economic activities (such
as freight delivery) but enhances others (such as tourism).
All Residents
All Residents
Environmental
Automobile
Dependency
All Residents
Economic
In recent years urban design, bicycle and pedestrian advocates, and neighborhood groups
have argued that roadways are a public realm that should accommodate all users, and that
overemphasizing automobile benefits in design is unfair. Advocates recommend alternative
standards that reverse many current priorities in order to discourage traffic, reduce motor
vehicle speeds, and emphasize other street functions, especially in urban commercial and
residential areas. Called neo-traditional streets (when applied to new streets) and traffic
calming (when applied to existing streets), these concepts include modified grid street
networks, smaller scale streets and blocks, narrow lanes, tight comers, textured road
surfaces, and greater integration of street users.48 The ultimate goal of these efforts is to
improve local and neighborhood environmental quality, and to encourage alternatives to
47
Peter Newman and Jeff Kenworthy, Cities and A utomobile Dependency, Gower, Aldershot, 1989.
Gordon Shaw, "Impact ofResidential Street Standards on Neo-Traditional Neighborhood Concepts",
ITE Journal, July 1994, p. 30-32; D.T. Brennan, "Evaluation of Residential Traffic Calming: A New
Multi-Criteria Approach," Traffic Engineering and Control, January 1994, p. 19-24. Tim Pharoah, Traffic
Calming Guidelines, Devon County Council, UK, 1992.
48
Page 7-23
I
Transportation Cost Analysis
driving. Table 7-6 summarizes the expected costs and benefits of traffic management
based on the equity criteria described earlier (of course, actual impacts are situation
specific and may differ significantly from these general estimates).
Table 7-7 summarizes these impacts of traffic management.
Table 7-7
Traffic M
User Market
User Non-Market
External Market
Environmental
Auto Dependency
Economic
Benefits S
Non-Drivers
Moderate Benefit
Large Benefit
Moderate Benefit
Large Benefit
Large Benefit
Mixed
-
Low Income
Drivers
Small Cost
Small Cost
Moderate Benefit
Large Benefit
Large Benefit
Mixed
Middle-High
Income Drivers
Small Cost
Small Cost
Moderate Benefit
Lar_g_e Benefit
Large Benefit
Mixed
~------ -------- ------ - --~--- -----------------
This analysis indicates that traffic management provides equity benefits. Horizontal equity
increases because the external impacts that drivers impose on others are reduced. Vertical
equity increases because non-drivers (who tend to be socially disadvantaged) receive extra
benefits. In many cities, low income urban neighborhoods could benefit most from traffic
calming, indicating further potential vertical equity.49 Although the financial impact of
increased automobile operating costs is greater per mile for low income drivers than more
affluent drivers as a percentage of income, low income households drive significantly less
than wealthier households (Figure 7-1) and benefit more from reduced automobile
dependency, so the vertical equity impacts among drivers is probably neutral.
7.6
Equity Analysis Conclusions
Most current analysis of transportation equity focus on a limited number of costs and
benefits, consider only one equity variable, and ignore most long term effects. The
significant external costs identified in this report indicates the potential for inequity. It is
likely that people who drive a lot receive an unjustified subsidy from those who drive little
49
David Engwicht Reclaiming Our Cities and Towns, New Society Publishers (Philadelphia), 1993.
Page 7-24
Transportation Cost Analysis
or not at all. Since driving tends to increase with wealth and physical ability, subsidies to
driving appear to be regressive. Underpriced driving is probably moderately unfair in terms
of horizontal equity, neutral to moderately unfair in terms of income, and highly unfair in
terms of need. People who are economically, physically, and socially disadvantaged are
harmed by an automobile dependent transport system that does not meet their travel
needs, and they tend to suffer a disproportionate share of external non-market costs use
since they can afford fewer protections against traffic impacts.
The overall equity effects of price changes are largely dependent on how revenues are
distributed, and could be highly progressive with respect to income while still providing
overall benefits to all income groups. The income inequity impacts of price increases can
be reduced by providing exemptions to low income households, increasing the quality and
quantity of low cost alternatives, and returning revenue to low income households in the
form of reduced regressive taxes, improved social services (including subsidies for transit
and other low-cost travel modes), and rebates. Equity impact of specific regulations and
pricing options must be evaluated individually, and should include analysis ofhow costs
and revenues impact both low income and transportation disadvantaged people.
Transaction costs, such as economic changes that result in unemployment among
traditional industries, tend to be inequitable with respect to income and social position
since low income people have fewer resources to fall back on. These problems can be
minimized or avoided by implementing changes gradually and predictably, through good
planning, job development and retraining programs that improve employment
opportunities for displaced workers.
Transit subsidies appear to be highly progressive, especially for services used by transit
dependent riders. A test of the equity benefits of a particular transit program or project is
the number of transportation disadvantaged and low income people who can and do use it.
Page 7-25
Transportation Cost Analysis
Indications of transit system equity are whether fares are affordable, proximity of service
to affordable housing, and the system's ability to transport people with disabilities. A
commuter rail transit system that is accessed primarily by drivers (park-and-ride or kissand-ride) may provide no equity benefits and may be overall inequitable if it is subsidized
with general taxes. However, this could change if affordable housing is developed within
walking and bicycling distance of stations. Improvements for other transportation
alternatives, including bicycling, walking and car pooling are probably also progressive.
Traffic management strategies such as traffic calming and nee-traditional street designs
that restrict automobile traffic and enhance bicycle, pedestrian and transit travel are
probably moderately to highly progressive in most circumstances, although actual impacts
depend on specific conditions.
Although this analysis does not attempt to quantify each cost and benefit, the estimates in
this report make such an undertaking possible by determining the portion of a community's
residents who are in each of the user classes, and estimating how both market and nonmarket costs are distributed among them. Some effects, such as automobile dependency,
cannot currently be quantified and should be incorporated qualitatively.
Page 7-26
Transportation Cost Analysis
8.0
Applications and Case Studies
The cost and elasticity estimates developed in this study are applied in this chapter to
representative examples of transportation decision making.
8.1
Evaluating Transportation Demand Management (TDM) Savings
Many North American communities have TDM programs that encourage residents to use
alternative travel options and reduce their driving. What are the benefits of such programs,
and what resources should they receive relative to other transportation investments? The
Oil-Smart Commute Performance Test offers an example for cost analysis.
-
~
-
-
- -
---
--
Mode
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Walk
Bike
Bike
Van Pool Driver
Van Pool Passenger
Van Pool Passenger, Walk
Van Pool Passenger, Walk
Van Pool Passenger, Walk
Van Pool Passenger, Walk
Bus Rider, Walk
Bus Rider, Walk
Car Pool Driver
Car Pool Passenger, Walk
Car Pool Passenger, Walk
SOVDriver
Totals
-
- -
- -----
--
-
-
-
--
-
--
Dist- Travel Internal External
Total Cost
ance Time
Cost
Cost
per trip
miles minutes per trip per trip
$0.01
$2.77
2.2
41
$2.76
$0.06
$1.02
2.75
10
$0.96
$0.08
$1.51
3.5
16
$1.43
$0.28
$3 .27
2.7
18
$2.99
$2.24
$1 .95
$0.29
2.7
18
$2.58
2.8
24
$2.30
$0.29
$2.30
$0.29
$2.58
2.8
24
$2.58
2.8
24
$2.30
$0.29
$2.58
$2.30
$0.29
2.8
24
$0.96
$4.88
3.2
35
$3 .92
$0.85
$4.50
2.5
30
$3 .65
$3.42
3.4
15
$2.81
$0.61
$2.19
$0.62
$2.81
3.3
20
$0.62
$2.81
3.3
20
$2.19
$5 .06
3.4
10
$3 .00
$2.06
$44.61
$37.05 $7.56
44.15 329
Savings Over
sov
per day
$4.58
$8.07
$7.10
$3 .57
$5 .63
$6.01
$6.01
$6.01
$6.01
$0.35
$1.12
$3.28
$4.49
$4.49
$0.00
$66.72
Savings for each trip are calculated based on cost estimates in this report. Savings
compared with an SOV trip are doubled to estimate savings per day. This illustrates one
offour Commute Performance Test days.
The Oil-Smart campaign, an annual program spearheaded by the Seattle based Bullitt
Foundation, encourages residents to use efficient travel modes. Dozens of community
organizations participate in the campaign. During four days in March, 1994, 62 trips were
Page 8-1
Transportation Cost A naly sis
monitored for a Commute Performance Test to determine the benefits of changing travel
patterns. Of these trips, about half consisted of two links, such as walking to a park-andride lot to catch a van pool, so a total of92links were analyzed. Table 8-1 summarizes the
distances, times, costs and savings for one day's trips. User (internal) and social (external)
cost and saving estimates are derived from the estimates in this report.
The external cost reduction of these alternative modes is especially significant. For
example, the calculated external cost of a trip from Capitol Hill to Pioneer Square is about
$2.00 for an SOV driver, but averages only $0.39 per for other modes. Ifall15 round
trips that day were made by SOV the total external cost would have increased from about
$15 to $60. Table 8-2 summarizes total costs and cost per mile for all trips in the test.
Table 8-2
Cost
-
- -
--
-
-
c
Perf1
TestS
s
Public
Van
Car
Totals Walk Bicycle Pool
Pool Transit
8
6
93
20
7
31
Number of Trips
Total miles
489.9
269.1
52.9
40.2
14.4
45
User Costs
$255.35 $19.44 $17.47 $103 .86 $26.33 $26.15
User Costs Per Pass. Mile
$0.65
$0.50
$1.35
$0.39
$0.39
Social Costs
$100.10 $0.05
$0.99 $31.00 $13 .16 $12.19
$0.00
$0.02
$0.12
$0.25
$0.30
Social Costs per Pass. Mile
Total Costs
$355.44 $19.49 $18.46 $134.86 $39.50 $38.34
Total Costs per Pass. Mile
$0.95
$1.35
$0.41
$0.50
$0.75
This table summarizes costs per travel mode for the Commute Performance Test.
Figure 8-1 shows these costs broken down into major cost categories.
Page 8-2
sov
21
68.3
$57.98
$0.85
$41.39
$0.61
$99.38
$1.46
-
-- -
··· - - - -
Transportation Cost Analysis
Figure 8-1
Major Cost Categories per Mile by Mode
•
•
0
•
•
$1 .60
$1.40
.! $1 .20
i
&
$1 .00
c
:.=
External Non-Market
External Market
User Time and Risk
Vehicle Ownership Costs
Vehicle Operating Cost
$0.80
~
$0.60
~
$0.40
$0.20
$0.00
Walk
Bicycle
Van Pool
CarPool
Transit
sov
This graph compares average travel costs per passenger mile for six modes used in the
1994 Oil Smart Commute Performance Test.
Significant Findings:
•
Total savings by the 55 Oil Smart participants who used alternative modes were $467
compared with the same trips made entirely by SOV. This averages about $8.50 per
person per day in savings.
•
External costs of SOV travel average about 5 times greater per trip than other modes.
•
The greatest savings per trip resulted from van pool riders who did not drive to their
van pool stop. Total costs of van pool, car pool, and transit trips were sensitive to how
the traveler got to their transit stop or rideshare meeting place.
•
The greatest savings per mile resulted from bicyclists, since they had low operating
and external costs but travel faster than pedestrians. The costs of bicycle and
pedestrian trips are sensitive to the time value assigned to travel.
These findings indicate that significant investments in Transportation Demand
Management programs are justified for programs that encourage use of alternative modes
and reduce automobile use. They also indicate which modes and trip combinations offer
the greatest total savings and the greatest potential for reducing external costs.
Page 8-3
Transportation Cost Analysis
8.2
Price Impacts on User Travel Decisions
Current travel trends indicate continued growth in automobile use and automobile
dependency. Motor vehicle driving has increased both absolutely and as a portion of total
land travel in recent years. Figure 8-2 illustrates these trends.
Figure 8-2
U.S. Vehicle Travel Trends, 1977-1990 1
111%
1U'X.
141%
UI'X.
.
f
121%
6
111%
l
UO%
~Vehicles
Per Licensed
Driver
- D a l l y VMT Por Household
-w-Porcont Transit Trips
.....
.....
i<:::::::::
~
..
71%
.....
1969
1983
1977
1990
Indicators show increasing automobile use and automobile dependency.
Although increased driving may result in part from demographic trends such as growth in
female employment which increases commuting travel, other trends such as increased
urbanization and improved communication and logistics could have compensated,
resulting in no or negative growth in per capita driving. This increase in motor vehicle use
is sometimes cited as proof of 11 America's love affair with the automobile. 11 But an
alternative explanation is that low user prices simply make driving too attractive for other
modes to compete. As shown in Section 4.2, the out of pocket cost of driving is typically
lower per mile than the cost of a bus fare. Studies described in Section 5.2 show that
transport prices significantly affect travel patterns. Low priced driving supports a cycle of
increased automobile use, automobile ownership and automobile dependency.
1 Nationwide
Personal Transportation Survey 1990: Summary ofTravel Trends, USDOT, 1993.
Page 8-4
Transportation Cost Analy sis
Consider the impacts of different transport prices (defined as the perceived variable
internal cost, which includes user non-market costs such as travel time and risk) on typical
user travel decisions. Assume that a resident has three shopping options: a local store
accessible by a 1/2-mile walk, a small supermarket 2 miles away where prices average
15% lower than the local store, and a megastore 7. 5 miles away where prices average
30% lower than the local store. Below is a comparison of the size of the shopping that
would justify traveling to the farther stores between current and full-cost pricing.
The current price of Urban Off-Peak driving is $0.47 per vehicle mile. This includes
vehicle operating costs, travel time, and internal risk. The total cost of driving (including
fixed vehicle ownership and external costs) averages $1 .06 per mile. Since walking has
virtually no external costs, both price and total cost are $1 .09 per mile under the same
conditions. Table 8-3 summarizes the three trip options.
Table 8-3
c
d Total-Cost T
Round Trip.
Savings over Local Store.
Current trip price.
Current travel price premium over Local Store.
Current shopping total to justify longer trip.
Full trip cost.
Full-cost travel price premium over Local Store
Full-cost shopping total to justify longer trip.
Selecf
I Price I m l)act
St
Local Supermarket
Megastore
Local Store
4 mile drive
15 mile drive
1 mile walk
15%
30%
$0
15 X 0.47 = $7.05
4 X 0.47 = $1.88
1 X 1.09 = $1.09
7.05-1.09=$5.96
$0
1.88-1.09= 0.79
5.96/30% = $19.77
0.79/15% = $5.27
$0
15 X 1.06 = $15.90
4 X $1.06 = $4.24
1 X $1.09 = $1.09
$4.24-1.09=$3 .90 $15.90-1.09=$14.81
$0
$3 .90/15% = $26.00 14.81/30o/o=$49.37
$0
This table shows how underpricing discourages use of local services.
This analysis indicates that current underpricing gives users little economic incentive to
walk 1/2 mile to a local store or shop at a local supermarket. At $0.47 per mile, the price
of driving to a store 2 miles away appears almost the same as the price of walking to a
store 1/2 mile away, and even a purchase under $20 justifies the 15 mile round trip to the
Megastore. But when all costs are considered the shorter trips become more attractive,
Page 8-5
Transportation Cost Analysis
and the Megastore is only justified for a large shopping. This shows how prices that are
below total costs skew user decisions to make longer and more frequent automobile trips.
Of course, other factors affect shopping habits. It can be difficult to carry big shopping
loads without a car (although easy with a wagon or bicycle trailer), and large stores have a
wider selection of goods. On the other hand, walking and shopping at local stores offers
health, enjoyment and community contact benefits. Shopping is often part of linked trips,
which reduces per trip costs, but linked trips tend to occur during peak periods when
congestion and travel time values are high. This analysis indicates that much of the savings
that individuals enjoy by shopping at a large, central store may be offset by incremental
external transport costs, and the discrepancy between user price and total costs affects
many travel decisions. Table 8-4 shows a similar analysis for home location decisions.
Table 8-4
Current and Total-Cost Travel Price Im act on Home Selection2
Exurban Home
Saving_s
Cars owned.
Annual Household VMT.
Annual user costs.3
Annual external costs. 4
Total costs.
I
I
2
25,000
$9,000
$8,513
$17,513
I
I
1
12,500
$4,500
$4,792
$9,292
I
I
1
12,500
$4,500
$3,721
$8,221
I
I
I
I
I
I
I
I
I
Many trip decisions involve a tradeoff between travel costs and potential benefits. The
more travel is underpriced the more marginal trips can be expected.
The Central Home reduces external costs by $3,721 annually compared with the Exurban
Home, with a capitalized value of approximately $40,000 (the additional housing value
that could be purchased if savings were invested in the mortgage). This implies that
underpriced driving underprices exurban housing by at least this amount per unit. The
2
John Holtzclaw, "Explaining Urban Density and Transit Impacts on Auto Use," NRDC, 1990.
Based on Facts and Figures '93 , American Automobile Manufacturers Association (Detroit), p. 54.
4 This assumes that Central vehicles are driven 33% Urban Peak, 33% Urban Off-Peak, and 33% Rural,
Exurban vehicles are driven 23% Urban Peak, 33% Urban Off-Peak and 44% rural, and this driving
incurs external costs of$0.61 , $0.34 and $0.20 per mile respectively, as calculated in this report.
3
Page 8-6
.....t
Transportation Cost Analysis
Central Home saves $8,221 annually in total driving costs over an Exurban Home, worth
over $80,000 or more in capital value if used for mortgage payments.
Some economists argue that transport costs should be considered when calculating
maximum mortgage payments. 5 Currently, the increased travel expenses associated with
an automobile dependent home are not considered a cost by most lending agencies. As a
result of underpriced driving and the omission of transportation expenses in mortgage
budget analysis, home selection decisions are skewed toward automobile dependent, high
travel cost houses, resulting in greater internal and external costs.
8.3
Marginalizing User Costs
Automobile owners typically pay approximately $0.21 per mile in fixed costs and $0.13
per mile in variable costs to drive. Fixed costs include about $0.08 per mile in vehicle
insurance, licenses, registration, and vehicle ownership taxes, totaling about $1,000 per
year. 6 Marginalizing these costs by paying them through additional fuel taxes or a mileage
charge instead of fixed annual payments would allow automobile owners who reduce their
driving to enjoy savings not currently available.7 Various versions of this concept have
been advocated by environmental and consumer organizations for years.8 Table 8-5 shows
the effect of an $0.08 per mile increase in vehicle operating costs.
5 Patrick Hare, Making Housing Affordable by Reducing Second Car Ownership, Patrick Hare
Planning
and Design (Washington DC), April, 1993. John Holtzclaw, Using Residential Patterns and Transit to
Decrease A uto Dependence and Cost, National Resources Defense Council (San Francisco), 1994.
6 Jack Faucett Asso., Costs of Owning and Operating A utomobiles, Vans and Light Trucks, FHWA, 1992.
7 Vehicle owners who currently reduce their driving by 100 miles only save about $13 .00. By
marginalizing these costs the same 100 mile reduction in driving would save $21.00.
8 See for example M. El-Gasseir, Potential Benefits and Workability of Pay-As-You-Drive Automobile
Insurance , for the California Energy Resources Conservation and Development Commission, June 1990.
Page 8-7
Transportation Cost Analysis
Table 8-5
Estimated A
Current Vehicle Operating Cost
Current VMT
Revised Price (+$0.08/mile)
1-10 Year Elasticity
1-10 Year Revised VMT
-
IVMTI----.---- tofM
---
-
Units
$per mile
billions
per mile
billions
. --
Urban Peak
0.15
460
0.23
-0.2
400
lizin2 User Cost"9
Urban Off-Peak
0.13
920
0.21
-0.2
806
Rural
0.11
920
0.19
-0.2
813
Totals
2,300
2,019
Changing insurance, registration, licensing, and taxes into variable costs would reduce
overall driving at no extra cost to users, increasing overall transportation efficiency.
The estimated 281 billion miles per year that would be eliminated represent low value
driving that users would forgo rather than pay an extra $0.08 per mile. Marginalizing these
costs provides benefits to users (who enjoy savings not currently available) and society
from reduced external costs. Table 8-6 shows the potential savings from this price change.
Table 8-6
s
Travel Reduction
Internal Saving
Total Internal Saving
External Savings
Total External Savings
Total Savings
f Reduced Drivin2 f1
Units
billion VMT
$/mile
$billions
$/mile
$billions
$billions
Urban Peak
60
0.71
$43
0.61
$37
$80
M
lizin2 Sel
Urban Off-Peak
114
0.71
$81
0.34
$39
$120
d User C
Rural
107
0.64
$69
0.20
$21
$90
10
Totals
281
$193
$97
$290
Marginalizing costs that are currently fixed could save over $290 billion annually.
The analysis in Table 8-6 oversimplifies actual travel cost changes. In practice, some
reduced automobile costs would be offset by increases in other types of travel. Table 8-7
recalculate the savings assuming that VMT reductions result 1/3 from reduced trips, 1/3
from reduced trip length, and 1/3 from mode shifts that are distributed equally among van
pools, car pools, bus, bicycling, walking, and telecommuting. This more accurate analysis
shows lower savings than in Table 8-6, but still worth over $200 billion annually.
9
Data from Chapter 3.1, and tables 4-3 and 5-5 .
°Cost estimates from Table 4-3 . This assumes that users enjoy savings proportional to their reduced
1
driving. For example, depreciation costs would decline since they need to buy new cars less frequently.
Page 8-8
Transportation Cost Analysis
s
M -a VIOl so f Reduced Drivin2 fi
Units
Urban Peak Urban Off- Peak
billion VMT
20
38
Eliminated Trips
$0.71
$0.71
Internal Savings
$/mile
Total Internal Savings
$billions
$14
$27
$0.34
External Savings
$0.61
$/mile
Total External Savings
$billions
$12
$13
$40
Total Savings
$billions
$26
Shortened Trips11
20
38
billion VMT
$0.47
$0.47
Internal Savings
$/mile
$18
Total Internal Savings
$billions
$9
$0.30
$/mile
$0.49
External Savings
$11
Total External Savings
$billions
$10
$31
$billions
$21
Total Savings
6
billion VMT
3
Shift to each of Six Modes
Internal & External Savings 12 $/mile
Varies
Varies
$billions
$5
$11
Total Internal Savings
$8
Total External Savings
$billions
$10
$19
$billions
$15
Total Savings
billions
60
114
Total VMT Reduction
Total Internal Saving
$billions
$28
$56
Total External Savings
$billions
$32
$32
$billions
$60
$88
Total Savings
Table 8-7
M
-
. ---
A
- ----
---
----
--
lizin2 C
Rural
36
$0.64
$23
$0.20
$7
$30
36
$0.41
$15
$0.18
$6
$22
6
Varies
$8
$1
$9
107
$46
$14
$60
Totals
192
$64
$31
$96
94
$42
$27
$74
$24
$19
$43
281
$130
$78
$208
This analysis, more accurate than Table 8-6, shows annual savings over $200 billion.
Another way to let marginal user prices reflect a greater portion of automobile costs is to
"Cash Out" free parking.13 This means giving employees who currently receive free
automobile parking the option of receiving its comparable cash value instead. Thus,
employees who currently get a $30 to $60 per month subsidy for driving could receive an
equal cash incentive for commuting by transit, ride share, bicycle or foot. 14 This offers
potential benefits to employers from reduced parking supply costs, to employees who have
the option of a financial bonus not currently available, and society due to reduced external
costs such as congestion, pollution, and energy consumption. It is also equitable, since
non-drivers are currently excluded from a valuable subsidy enjoyed by drivers. Donald
11 Internal savings include user variable costs. External savings are all external costs except parking.
12 Calculated in a separate spreadsheet.
13 Donald Shoup, Cashing Out Employer-Paid Parking, Federal Transit Administration, December 1992.
14 Non-drivers would actually receive a somewhat lower benefit because the cash is taxable while parking
is exempt under current US income tax rules.
Page 8-9
Transportation Cost Analysis
Shoup estimates that this measure alone could reduce solo commuting by 20%, and total
vehicle travel by 3.3%, increase federal tax revenue by $1.2 billion annually. 15 This 76
billion VMT reduction would save an estimated $46 billion per year in external costs.16
The combination of maginalizing automobile insurance and registration, and Cashing Out
employee parking could reduce current driving by an estimated 357 billion miles a year, or
about 15% of total automobile travel, providing many billions of dollars in savings to users
and society. These strategies involve neither increased costs nor coercion. The foregone
trips represent low value travel that automobile users are willing to eliminate given greater
choice. The only class of drivers likely to be disadvantaged are those who are currently
uninsured, which is illegal in most states. Additional reductions in low value travel could
be expected if user prices were increased to incorporate external costs. These estimates
indicate that a significant portion of driving has negative net value (total benefits minus
total costs), and that society would benefit significantly from these price changes.
8.4
Critiquing Transportation Investment Models
A number ofbenefit cost models are used for the economic analysis of transport
investments.17 These models compare project benefits (travel time savings, accident
reductions, and vehicle operating savings) with financial costs (land acquisition,
construction, and maintenance). They specify how the benefits of generated traffic should
be measured, and emphasize that all costs must be considered.
15 Donald Shoup, "Cashing-Out Employer-Paid Parking; An Opportunity to Reduce Minimum Parking
Requirements," Journal of the American Planning Association, Forthcoming, June 1994.
16 Based on Urban Peak external costs of$0.61 per mile, from Table 4-2. Although not all commuting is
Urban Peak, mode shifts are most likely under these conditions because the most options are available.
17 Examples include the MicroBENCOST computer program developed for the USDOT, and the COBA
(COst Benefit Analysis) model developed by the British Department of Transport.
Page 8-10
Transportation Cost Analysis
As discussed previously, generated traffic and external costs are often ignored in
transportation planning. Both omissions skew the results in the same direction, making
road expansion projects appear more attractive and other options such as demand
management and public transit investments appear less attractive. Another significant
omission in most current modeling is the use of short term point elasticity values rather
than dynamic values that vary over time.18 Long term elasticities are usually much greater
than short term elasticities. The justification often used for this exclusion is that the
necessary data are not available, but that is untrue.
To test the effect of omitting generated traffic in transportation decisions, researchers
Robert Johnston and Raju Ceerla used a conventional four-step traffic model to evaluate
transport investments in Sacramento, California, and then evaluated the same investments
with a newer model that incorporates feedback from generated traffic. 19 The model that
incorporates generated traffic changed the ranking of options compared with the standard
model. The ranking ofNo Build, Light Rail, and Road Pricing increased with feedback,
while building additional highway capacity becomes less attractive. Johnston and Ceerla
did not incorporate estimates of external costs in their analysis, but doing so would
certainly increase the calculated costs and decrease the benefits associated with projects
that add roadway capacity. Williams and Lam reached similar conclusions concerning the
impacts of ignoring generated traffic, 20 and also point out that highway investments can
impose external costs in terms of reduced transit service efficiency.21
18
J.M. Dargay and P. B. Goodwin, "Estimation of Consumer Surplus with Dynamic Demand Changes,"
in Proceedings of European Transport Forum, PTRC, Sept. 1994.
19 Robert Johnston and Raju Ceerla, Institute of Transportation Studies, University of California (Davis),
"A Comparison of Modeling Travel Demand and Emission with and Without Assigned Travel Times Fed
Back to Trip Distribution" Submitted to the Journal of Transportation Engineering, Aprilll , 1994.
20 H.C.W.L. Williams and W.M. Lam, "Transport Policy Appraisal With Equilibrium Models I:
Generated Traffic and Highway Investment Benefits," Transport. Research B , 28/5, pp. 253-279, 1991.
21 H.C.W.L. Williams and W.M. Lam, "Transport Policy Appraisal With Equilibrium Models III:
Investment Benefits in Multi-Modal Systems" Transportation Research B , 28/5, pp. 293-316, 1991.
Page 8-11
Transportation Cost Analysis
Generated traffic increases virtually all external costs, including air pollution, nose, energy
consumption, parking demand, congestion on local streets, urban sprawl, and automobile
dependency. Projects that reduce overall vehicle use provide benefits in terms of external
cost savings that are not recognized in current models. Only if economic analysis
incorporates total external costs including impacts of generated traffic can society be sure
that projects being funded actually provide overall benefit.
Evaluating Congestion Reduction as a Transport Improvement Priority
8. 5
The cost estimates in this report can be used to compare the relative significance of
transport costs. Figure 8-3 shows automobile costs ranked by magnitude.
Figure 8-3
Average Automobile Costs Ranked by Magnitude
SI .JI
u .u
i
u .u
I
u .u
j
u .u
~
u .u
.....
!
l
f.
f
i
>
f~ Il
::.
~
!
3
I lf lf I Il
~
!l
u
J:
Jj
!
~
J
j
i
I I
i
I
i j
i.
i
Since user travel time is the highest ranking cost, it could be argued that projects which
increase travel speeds offers significant potential benefits. However, as discussed in
Section 5.4, individuals tend to maintain a constant travel time budget, so in practice the
benefits of most investments that increase travel speeds translate into shifts in activity
locations, and are capitalized into real estate values as residents travel longer distances to
work, school, shopping, and recreation centers.
Page 8-12
1
Transportation Cost A nalysis
Traffic congestion is the additional travel time imposed on society above optimum traffic
volumes. Although congestion is often assumed to be our greatest transportation problem,
the reduction of which consumes most transport investment funds, this analysis indicates
that overall it is only a middle-range cost. More importantly, traffic congestion is a
relatively small cost compared with the total of costs that typically increase in response to
efforts to accommodate more vehicle traffic. Of 17 transport costs, two (travel time and
congestion) tend to be reduced by increased traffic capacity and speeds, while 15 tend to
increase, as shown in Table 8-8.22
Table 8-8
Trans ortation Costs Affected b Increased Roadwa Ca
Costs Typically Reduced by
Costs Typically Increased by
Increased Road Capacity
Increased Road Capacity23
User Travel Time
Congestion
I
Vehicle Costs
Road Facilities
Municipal Services
Air Pollution
Waste generation
Land Use Impacts
Resource consump_tion
Parking
Accidents
Equity & Option
Barrier Effect
Noise
Water Pollution
RoadwllY_ Land Use
A number of urban economists have concluded that current roadway investment policies
and failure to internalize costs leads to overinvestment in roads both in terms of financial
costs and as a percentage of urban land.24 These perspectives imply, but do not prove, that
traffic congestion is an overrated problem. To prove this it is necessary to compare
potential benefits of congestion relief with other possible investments, which is not
22
Some projects are intended to improve roadway safety, but drivers frequently respond by driving faster,
resulting in no change in accidents. See Robert Davis, Death on the Streets, Leading Edge (London),
1992. Recent research on highway safety improvements by Mike Kawczynski P.Eng. of MAC Engineering
for the B.C. Ministry of Transportation and Highways found the same results.
23 Although reduced traffic congestion and stop-and-go driving decrease per mile energy consumption, air
pollution, and vehicle operating costs, current estimates indicate that these savings are overwhelmed by
overall increases in driving. See Hansen, et al., The A ir Quality Impacts of Urban Highway Capacity
Expansion: Traffic Generation and Land Use Change, Institute of Transportation Studies, UCB
(Berkeley), 1993; Newman and Kenworthy, Cities and A utomobile Dependency, Gower Press, 1989.
24 Takahiro Miyao and Yoshitsugu Kanemoto, Urban Dynamics and Urban Externalities, Harwood
Academic Publishers (NY), 1987, pp. 77-87.
Page 8-13
.
~
Transportation Cost Analysis
possible based on data in this study. However, additional factors also indicate that current
congestion reduction benefits may be lower than is often assumed, which further implies
that transport planning and investments overemphasize congestion relief As discussed in
the previous section, most economic estimates of traffic congestion overstate the potential
of reducing this cost because traffic congestion maintains a self-limiting equilibrium.
Efforts to accommodate more trips leading to generated traffic. Increased congestion
forces individuals and communities to limit their driving and use substitutes, which include
shifts in route, mode, time destination, and alternatives such as faxes and delivery services.
Traffic congestion does not stop economic activity, it simply causes individuals to choose
a marginally more expensive alternative.
Efforts to reduce traffic congestion can have perverse effects. Kenneth Small suggests that
efforts to reduce traffic congestion by increasing road capacity can incur external costs by
diverting travel from public transit to automobiles, therefore reducing the efficiency of the
transit system (reduced economies of scale) while the generated traffic reduces or
eliminates any long term congestion reduction benefits.25 In other cases, road
improvements that create more direct routes can divert traffic from circuitous but less
congested roads, resulting in slower travel for everyone.
Several studies (some mentioned in Chapter 3.5) estimate current and future congestion
based on trend analysis, often with alarming conclusions. The US Congress's Office of
Technology Assessment found significant problems with some of these studies, including
methodological mistakes in calculating roadway capacity, and the failure to recognize that
traffic congestion tends to be self-limiting. 26 Although some indicators show growing
traffic congestion, others, such as declining urban commute travel times and increased
25
26
Kenneth Small, Urban Transportation Economics, Harwood (Chur), 1992, p. 115.
Saving Energy in U.S. Transportation, OTA (Washington DC), July 1994, p. 111-122.
Page 8-14
Transportation Cost Analysis
average highway speeds in many areas, show contradicting trends. Rather than assuming
that a community's primary transport problem is congestion, it would be better to define
the problem as: Transportation is too expensive in terms of all costs. This includes travel
time (which incorporates congestion delays), other user costs, plus all costs to society.
Does Traffic Congestion Significantly Burden the Local Economy?
Empirical evidence supports the hypothesis that predictable traffic congestion imposes a
relatively minor constraint to economic activity, provided that other transport options are
available. Cities such as Hong Kong, Tokyo, New York, London and Paris, have extreme
levels of traffic congestion. Similarly, traffic congestion is inescapable in fast growing
suburban and exurban communities due to high levels of automobile dependency and
use.27 These indicate that a positive correlation exists between traffic congestion and local
economic activity. Of course, this does not mean that congestion contributes to economic
growth, but it indicates that congestion does not significantly limit economic activity,
economic development, or property values. Although traffic congestion is clearly an
economic cost, it does not appear to be a significant burden, especially if alternative access
options, such as telecommunication, subways and local shops and services are available.
This study's cost estimates can help identify the overall optimal congestion reduction
strategies. Pricing appears most cost effective because it reduces congestion, reduces total
vehicle travel thereby reducing total external costs, and raises revenue. Programs to
reduce traffic congestion by increasing road capacity appear to be least cost effective,
because in addition to their direct financial, social and environmental costs they
accommodate additional vehicle use that increases total external costs, and as described
earlier the improvement they provide degrades over time due to generated traffic. Nonpricing TDM programs (promotion and support of alternative modes, and land use
27
Robert Cervero, America's Suburban Centers: The Land Use-Transportation Link, Unwin Hyman,
1992. Joel Gerreau, Edge City, Doubleday, 1992.
Page 8-15
Transportation Cost Analysis
management) is probably intermediate between pricing and road capacity programs, and
appears to be highly variable depending on specific circumstances.28 Although the
existence of generated traffic and external costs does not exclude the possibility that some
capacity expansion projects are still cost effective, it is a more rigorous standard which
would probably indicate that costs exceed benefits of many approved projects.
As discussed in Chapter 5, traffic tends to fill available road capacity due to generated
traffic, but grows less or not at all if new capacity is not added. Some people argue that
road capacity must increase to accommodate population growth.29 It is true that traffic
increases with urban sprawl and poorly planned development. On the other hand, more
population within a given area (increased density) can increase the accessibility of services,
such as shops and schools, and increase mode choices, reducing per household automobile
use and dependency, and per capita road requirements.30 Development practices that take
advantage oftravel reduction opportunities can avoid the need to increase road capacity.31
This analysis returns us to consideration of the meaning of transportation. If society
defines transport simply as vehicle travel it is easier to conclude that costs decline with
road building, and conventional planning and investment programs are justified. If
transportation is defined as access, then roadway expansion projects may actually increase
total transport costs by increasing urban sprawl, automobile dependency and use, and
associated social and environmental impacts. Investment projects must be evaluated
28 In his book Stuck in Traffic (Brookings Institute, 1992) Anthony Downs argues that such programs are
popular because they are relatively ineffective. However, they do seem to have a marginal effect by
reducing some market and social barriers to travel pattern changes.
29 Myths and Facts about Transportation and Growth, Urban Land Institute (Washington DC), 1989.
30 See Chapter 3.14 for discussion of the effects of sprawl on transportation.
31 For information on designing communities to reduce automobile dependency see John Holtzclaw's
"Using Residential Patterns and Transit to Decrease Automobile Dependence and Costs," National
Resources Defense Council (San Francisco), 1994; Steve Weissman; Judy Corbett, Land Use Strategies
for More Livable Places, California Environmental Protection Agency, 1992; Rebecca Ocken Site Design
& Travel Behavior: A Bibliography, 1000 Friends of Oregon (Portland), 1993.
Page 8-16
Transportation Cost Analysis
according to total costs, including long-term impacts on land use and automobile
dependency, to insure that they provide net benefits.
Framing the Congestion Cost Question
If you ask people, ''Do you think that traffic congestion is a significant problem that
deserves significant investment?" most would probably answer yes. Ifyou ask them,
"Would you rather invest in road capacity expansion or use lifestyle changes, such as
increased urban density and more use of walking, bicycling, car pooling and public
transit to solve congestion problems?" a smaller majority would probably choose the road
improvement option. These are essentially how choices are framed by conventional
transportation plans. But if you presented a more realistic description of our choices by
asking, "Would you rather spend a lot of money increasing road capacity that will
provide only moderate and temporary reductions in traffic congestion and will increase
personal, municipal, social and environmental costs, encourage urban sprawl, raise rural
property values, and leave a legacy of automobile dependency to future generations, or
would you rather modify and accelerate lifestyle changes that will occur anyway
(increased urban density and multimodalism) over the next few decades to avoid these
problems?" a majority would probably choose alternatives to more road building.
8.6
Evaluating Traffic Management Benefits
As discussed in section 6.5, a conflict exists in roadway design between maximizing traffic
flow and protecting local environmental quality. Increased traffic volume and speed:
•
Require more land for streets and parking at the expense of pedestrian and bicycle
facilities, and other public spaces.
•
Increase risk of accidents between automobiles and other road users.
•
Increase barriers to pedestrian and bicycle movement.
•
Increase noise, air pollution and dust.
•
Increased automobile dependency
•
Increase urban sprawl.
Page 8-17
Transportation Cost Analysis
Transport cost analysis can help determine the optimal allocation of resources between
motor vehicle traffic and local environmental quality. Current local traffic planning tends
to overestimate the benefits of increasing road capacity, and underestimate external costs
as previously discussed. Figure 8-4 illustrates estimated costs that are likely to decline due
to traffic calming and nee-traditional streets, assuming that the same amount of driving
takes place but at lower speeds.32 These costs average $0.19 out of $1.3 7 total cost per
vehicle mile, and may be much higher in many urban areas.
Costs Reduced by Traffic Calming
Figure 8-4
$0.30
i•
.!!
:g
E
$0.25
$0.20
•Affected by Traffic Calming
~
D Not Affected by Traffic Calming
II
"
$0.15
~
:.
$0.10
~
8 $0.05
$0.00
I'
'I'
a.
a>E
0 ~
~!
1
'I'
0>
c:
Ill
~
E
i=
0
~
8.
~
:::>
1
cIll
<(
c
8<(
Iii
E
Iii
E
"0
8
~
Ill
"0
Ill
0>
c:
~
n.
Iii
"'
i-
0>
c:
~
n.
Iii
"'
~
c:
0
8l
0
c:
8
-~
~
·o
"'
LL.
"0
2!
Ill
~"
"0
c:
"'
....J
0:
~~
-~ -~
~/Jl
~
c:
CT
'S
"
w
0
'C5
n.
<
5l
·a
z
~
"
~
0:
~
w
Q;
.E
&l
"'
~
E
0
Q;
~
Ill
1ii
"'
~
5l
:::>
"0
c:
a\
a\
"'
Automobile costs that traffic calming programs are likely to reduce.
:E
~
....J
This analysis indicates that local environmental and social costs can be substantial
compared with other transportation costs.33 Current traffic planning and funding ignore
many of these costs, so automobile traffic improvements tend to occur at the expense of
local environmental and social benefits. Since motor vehicle traffic imposes these costs, it
32
Based on an average of Urban Peak and Off-Peak costs, with noise and barrier effect costs doubled to
represent higher impacts on neighborhood streets.
33 If anything, this underestimates potential benefits, since neo-traditional neighborhoods can also provide
traffic congestion and user costs savings by reducing automobile dependency.
Page 8-18
Transportation Cost Analysis
is both equitable and efficient to use motor vehicle user funds to implement traffic calming
and related projects to reduce impacts and improve neighborhood environmental quality.
8. 7
Least-Cost Transportation Planning
Least-Cost planning is a concept used in utility planning that is now being applied to
transportation investment decisions. Despite being relatively new, Least-Cost transport
planning is required in California at the state level, 34 in Washington State at the regional
level, 35 and is recommended by several analysts.36 Least-Cost planning includes:
•
Consideration of supply and demand management options on an equal basis.
•
Use of standard measurements of costs and benefits for evaluating investments.
•
Selection of projects and programs according to cost effectiveness.
Least-Cost transportation planning means, for example, that Transportation Demand
Management (TDM) programs are compared equally with investments that increase
facility capacity. This represents a change because TDM programs currently receive
limited consideration and less funding than general roadway improvements. Researchers
Caroline Rodier and Robert Johnston point out that current transport funding formulas
tend to reward regions that demonstrate increased travel demand and tend to give fewer
resources to communities that successfully reduce demand.37 They describe a method for
34
California Transportation Energy Analy sis Report (Draft), California Energy Comm. , Feb. 1994, p. 1
Cy Ulberg, Jane Meseck Yeager and Matthew Hansen, Least-Cost Planning Implementation,
Washington State Transportation Center (Seattle), 1995.
36 Ruth Steiner, Least Cost Planning for Transportation; What We Can Learn, TRB 1993 Annual
Meeting; Sheets and Watson, Least Cost Transportation Planning: Lessons from the Northwest Power
Planning Council, Bullitt Foundation (Seattle), January 1994; Nelson and Shakow, Applying Least Cost
Planning to Puget Sound Regional Transportation , Bullitt Foundation, (Seattle), February 1994; Michael
Grant, A Methodology for Evauating Cross-Modal Transportation Alternatives, Office of Intermodalism,
USDOT (Washington DC), 4 August 1994.
37 Caroline Rodier and Robert Johnston, Incentives for Local Governments to Implement Travel Demand
Management Measures, Institute of Transportation Studies, UCD (Davis), October 1994.
35
Page 8-19
Transportation Cost Analysis
calculating the financial benefits of deferring a highway capacity project and apply it in a
case study in the Sacramento, California region. They estimate that local governments
there could justify spending $3 7 million per year in TDM programs if it delayed future
anticipated roadway expenditures for seven years. Even greater demand management
expenditures could be justified if external costs are incorporated into the analysis.
As an example ofLeast-Cost planning, consider a transport problem facing Olympia,
Washington. Access between the city's downtown and the Westside is limited by a
bottleneck at the Fourth and Fifth avenue bridges. 38 Together the two bridges can carry
approximately 1, 800 vehicles per hour in each direction. A $1 0. 4 million widening project
has been proposed to increase peak bridge capacity by 1, 149 vehicles. This would require
annual payments of$838,100, for a cost of$1.40 per additional peak period automobile
one-way trip.39 As discussed in Chapter 5, increased road capacity can:
1. Shorten some trips by allowing more direct routes. Although shorter trips usually
reduce external costs, downtown Olympia is very sensitive to traffic impacts
(congestion, noise, barrier effect, etc.) so this is unlikely to provide overall savings.
2. Encourages some longer trips which increase external costs.
3. Generates some new trips. This increases external costs, especially due to downtown
Olympia's sensitivity to increased traffic.
4. Increase peak periods trips. This increases costs, including congestion on other roads.
5. Encourages mode shifts from transit to driving.
The additional external costs of these effects can be calculated. Assume that the 1, 149
additional peak period trips are equally divided among effects 1-4, (effect 5 is not
significant in this case due to low transit use) times two peak periods per day, times 260
38
Gilbert McCoy, Kristine Growdon and Brian Lagerberg, Apply ing Electrical Utility Lease Cost
Planning Approaches to the Transportation Sector, Washington State Energy Office (Olympia), 1993.
39 Assuming 7 percent interest on 30-year bonds, and 260 annual work days.
Page 8-20
Transportation Cost Analysis
annual work days, equals approximately 150,000 annual trips changed per effect ( 1, 149 +
4 x 2 x 260 = 149,370), as summarized in Table 8-9. The total cost to society ofthis
proposed project includes $838,100 in construction costs and $1,434,000 in external
costs, totaling over $2.2 million per year, $3 .77 per additional trip, or over $7.50 per
additional round trip commute. The specific values used are for illustrative purposes only,
but they indicate the significant costs imposed on society from increased urban driving.
---------- - - ------
Effect
1. Average trip length
reduced by 4 miles.
2. Average trip length
increase by 4 miles.
3. Generated trips,
average 8 miles.
4. Shift from off-peak
5. Mode change.
Totals
- - - - ----
-----
- - - --- - - - - - - - - - - - - - - -
-
---
-----
Peak Mileage
Chan2e
External Cost Per
Mile
Total Additional
External Cost
-600,000
None 40
$0
+600,000
$0.61 41
$366,000
+1 ,200,000
+1 ,200,000
None
2,400,000
$0.61
$0.28 42
None
$732,000
$336,000
None
$1,434,000
Increasing road capacity provides user benefits and external costs from increased total
motor vehicle use. Additional costs should be considered in investment evaluation.
The city could consider demand management options rather than invest in increased
capacity. The components of a TDM program depend on specific geographic and
demographic conditions, but might include improvements to transit, bicycle and pedestrian
facilities, and incentives to reduce peak period driving. Once the costs and effectiveness of
specific TDM options are estimated, a supply curve is developed based on the cost per
vehicle trip reduced across the bridge. The most cost effective options would be chosen
until a goal (such as 1, 149 peak hour trips reduced) or budget constraint is reached. The
$83 8, 100 annual cost of increasing bridge capacity could fund a respectable TDM
program. The $2.2 million total annual cost could fund an outstanding TDM program. The
40
41
42
Due to the sensitivity of downtown Olympia to traffic impacts, described above.
Average automobile external Urban Peak external costs from Table 4-3 .
Difference between average automobile Urban Peak and Urban Off-Peak external costs.
Page 8-21
Transportation Cost Analysis
TDM option should be chosen instead of the capacity construction option if expenditures
less than the $2.2 million total would reduce peak period driving along that corridor. Note
that this estimate only considers costs on one corridor. Reduced traffic on other roads
would provide additional benefits that could justify even greater TDM expenditures.
8.8
Evaluating Electric Vehicle Benefits
There is considerable interest in alternative automobile engines and fuels to reduce
environmental costs. Alternative fuels, especially electric vehicles, are often cited as
solutions to the environmental impacts of our current transportation system. The costs
developed in this report can be used to evaluate these options from an overall economic
perspective. For simplicity sake this analysis focuses on electric vehicles, although a
similar analysis could be performed for a variety of fuels.
As discussed in specific cost chapters, electric vehicles reduce but do not eliminate several
costs. For example, local air pollution may be avoided but a portion of electric generation
capacity comes from fossil fuels which produce global air pollutants. Similarly, although
engine noise is greatly reduced, electric vehicles still emit road-tire noise. Cost effects are
summarized in Table 8-10.
Table 8-10
Fuel T
Effect
Costs Typically Reduced in
Electric Vehicles
Air pollution
Noise
Water pollution
Waste
Resource (energy) consumption
externalities
L _ _ _ __ _ _
T
rtation Cost
Costs Unaffected by
Electric Vehicles
User travel time
Congestion
Accidents
Parking
Land value
Barrier effect
Land use impacts
Equity & Option
Municipal services
Costs Typically Increased in
Electric Vehicles
Vehicle ownership
Vehicle operating
Road facilities 43
This table shows how costs typically differ between gasoline and electric vehicles.
43
Although road facility costs do not actually increase, electric vehicle use does not contribute to
dedicated fuel taxes so their subsidy is greater based on the cost analysis framework used in this report.
Page 8-22
Transportation Cost A nalysis
Three types of electric vehicles are considered:
1.
Standard Electric. This is based on current electric car ownership and operating
costs, which are higher than a standard automobile. This uses the electric vehicle
costs defined earlier in this report.
2.
Cheaper Electric Car. This is based on the assumption that these costs will decline
in the near future due to increased production. Ownership and operating costs are
equal that of an average automobile, and other costs are as defined earlier for an
electric vehicle.
3.
Neighborhood vehicle. These are small, inexpensive, low power, low speed electric
vehicles intended for local urban travel. 44 These are estimated to reduce all costs
except travel time, congestion, and road services (policing, planning, etc.) by 50%.45
Figure 8-5 shows the total costs of these four vehicles by major category. Although
Standard Electric Cars reduce some non-market externalities, their current high ownership
and operating costs make them slighly more expensive overall. Of course, these average
values underestimate the cost differential in urban areas with high noise and local air
pollution costs. 46 Assuming that reduced future production costs will make Cheaper
Electric Cars available, overall savings are possible. However, electric cars do not reduce
many external costs of driving, including parking subsidies, accident risk, urban sprawl, or
inequity. To significantly reduce total costs requires an inexpensive, efficient, safe, small
vehicle that does not encourage urban sprawl, such as the Neighborhood Car.
44
Daniel Sperling, "Prospects for Neighbohood Electric Vehicles," Transportation Research Record
1444, 1995, p. 16-22.
45 This estimate is somewhat arbitrary since specific performance and cost data are not available.
46 Roland Hwang, et al. , Driving Out Pollution: The Benefits ofElectric Vehicles, Union of Concerned
Scientists (Berkeley), 1994 estimate lifetime benefits of electric vehicles at $17,570 in So. California.
Page 8-23
Transportation Cost Analysis
Electric Vehicle Cost Comparison by Major Category
Figure 8-5
•External Market
•External Non-Market
CJTim e & R lsk
tt.U
.1 .00
iJ!
f
U .TI
>
l
j ....
to .U
to .oo
Average
Au tom oblla
Standard
Electric
Cheaper
Electric
N alghborhood
Car
This graph compares cost categories of three electric vehicles and an average
automobile based on the assumptions stated above.
8.9
Critiquing Taxation Report
The study Transportation Taxation and Competitiveness, published by the Transportation
Association of Canada (T AC) in September 1993 examines the economic impacts of
Canadian transport taxes.47 It concludes that road transportation is overtaxed because
motor vehicle tax revenues are not spent entirely on roadway facilities. This conclusion
has been widely publicized by industry lobbying groups to justify lower fuel taxes and
increased expenditures on driving.
This argument can be analyzed using the cost estimates from this report. Current Canadian
automobile fuel taxes average $0.263 per litre Canadian, or about U.S . $0.84 per gallon.
Based on average automobile fuel efficiency of21 mpg, this equals about $0.04 per mile.
Automobile owners also pay annual registration fees and taxes that might increase average
user payments to as much as $0.05 per mile. Assuming that Canadian average external
costs of driving are comparable to the $0.32 per mile estimated for the U.S. (Table 4-3),
automobile user taxes cover only about 16% of external costs.
47
T AC membership includes transport industries and agencies.
Page 8-24
Transportation Cost Analysis
Not all fuel taxes should be considered user fees. A portion are general taxes (Goods and
Services Tax, or GST, and Provincial Sales Tax, or PST). As discussed in Chapter 1,
broad based taxes such as these should not be considered user fees, or the tax system
would become absurd. If all sales taxes were limited to providing services for the sector
from which they originated there would be little funding for essential general public
services such as education, planning, and law enforcement. Taxes paid on hats would be
targeted for public hatracks, and theater taxes would be dedicated to popcorn subsidies.
Excluding this revenue, fuel taxes cover only about 13% of estimated external costs.
These estimates suggest that motor vehicle user payments are low compared to the costs
motor vehicles impose on society. This contradicts the TAC report's conclusions and
recommendations. Although the report acknowledges the potential of significant external
costs, the authors make no effort to incorporate them in the analysis. Their justification is,
"... most analyses to date readily acknowledge that data quality problems and
theoretical/imitations (e.g., pavement deterioration models) make it difficult to
accurately quantify the extent of cost recovery. "48
The T AC report argues that minimizing prices will improve national productivity and
competitiveness. However, as discussed in Section 6.1, economic efficiency is optimized
when prices reflect marginal costs. Current low taxes reduce the nation's overall economic
efficiency and competitiveness by diverting resources from other sectors. Like other
lobbying organizations, TAC uses data selectively to support arguments for increased
subsidies without consideration of economic efficiency or fairness.
48
Denis Lacroix, Jan Bowland and Frank Collins, Transportation Taxation and Competitiveness, TAC
(Ottawa), Sept. 1993, p. 44.
Page 8-25
Transportation Cost Analysis
9.0 Conclusions and Recommendations
This chapter summarizes this report's major conclusions and provides recommendations
for improving the efficiency and equity of our transportation system. These conclusions
won't surprise readers familiar with recent literature on transportation economics or policy
issues. Other books, reports and articles make similar points.1 This study augments
previous documents by providing specific cost estimates and an analysis framework that
can be applied to analyze policies and programs.
The first conclusion of this study is that the high levels of automobile use found in North
America and other high consuming countries are not essential, and probably reduce overall
economic success or personal happiness compared with transport systems that provide
more travel options, reduce external costs (accidents, pollution, congestion), avoid urban
sprawl, minimize financial costs, and increased overall economic efficiency. Another
important conclusion of this study is that driving is significantly underpriced compared
with total costs. Two factors contribute to this: some costs are fixed and others are
external. Variable user costs (including vehicle operating expenses, travel time, and
accident risk) comprise less than half of total costs. About a quarter of total costs are
fixed, and a third are external. These ratios vary depending on travel conditions and
vehicle types, but the basic relationships are consistent for virtually all driving.
The benefits of motor vehicle travel are substantial, but these are squandered when society
succumbs to the temptation to underprice. Automobile owners have no incentive to limit
driving to trips in which benefits exceed total costs. This results in wasteful travel in which
1 Examples include Elmer Johnson's Avoiding the Collision of Cities and Cars, American Academy of
Arts and Sciences (Chicago) 1993 ; Deborah Gordon, Steering a New Course, Union of Concerned
Scientists (Cambridge) 1991; S. Nadis and J. Mackenzie, Car Trouble , Beacon Hill (Boston) 1993.
Page 9-1
Transportation Cost Analysis
a dollar is often spent to provide fifty cents worth of benefit. Other problems such as
congestion, pollution, and community degradation become constraints to traffic growth.
According to conventional wisdom traffic congestion is our greatest transportation
problem. This justifies current planning and funding practices that emphasize projects to
increase road capacity. This study provides a different perspective. According to estimates
developed here, congestion is a moderate problem (cost). Other costs which increase
when automobile use grows are greater overall. Underpriced driving encourages overuse,
forcing congestion to be self limiting. As expressed by Moore and Thorsnes,
''No rational concert promoter would decide how big to build a stadium based on
the number ofpeople who would come to see the Greatful Dead if the tickets were
free. But that is often how transportation planners decide highway capacity: they
estimate how many trips would be make on an unpriced facility, then try to build
a facility big enough to accommodate that number of trips. ''2
Efforts to reduce congestion by increasing road capacity creates more traffic and increases
automobile dependency in the long run. Transport programs should be evaluated
according to how well they improve access at the lowest total cost to users, society, and
the environment. Only by considering all costs can society be confident that transport
investments really provide net benefits.
There is no single solution to our current transportation problems. Neither improved
transit service, increased bicycling and walking facilities, "smart" highways, nor less
polluting vehicles alone can reduce the inefficiencies of our transport system while the
price of driving is so low. Solutions to congestion that increase road capacity and traffic
speeds exacerbate transportation problems. Such solutions accommodate existing
inefficiencies and inequities, and increase the total amount of driving in the long term.
2
Moore & Thorsnes, Transportation/Land Use Connection, American Planning Association, 1994, p. 57.
Page 9-2
Transportation Cost Analysis
Making user prices reflect marginal user costs is the key to encouraging more efficient
transport, but increasing prices alone is only half the solution. Changes in land use
patterns, planning, investment policy, and personal habits are also needed. Our current
transportation system encourages every driver to own a "personal" car. A more efficient
and equitable transportation system would provide users with more travel choices, and
provide incentives to use each mode for what it does best.
9.1
Costs and User Pricing
As stated above, a primary conclusion of this study is that transport, especially automobile
use, is underpriced with respect to total costs. External costs that tend to be ignored
during transportation planning and investment decisions include parking, congestion,
accidents, municipal service costs, land use impacts, roadway land value, environmental
degradation, and social impacts. Urban peak driving incurs the greatest external costs per
mile, but external costs of driving under other conditions are also significant.
The problem is not only that costs are externalized. Although automobile ownership costs
are a major portion of most household budgets, automobile operating costs are typically
lower per mile than a local bus fare . Automobile owners have little financial incentive to
limit their driving or use other modes. This price structure provides an incentive to
maximize driving in order to "get your money's worth" from high fixed costs.
Transportation costs affect economic productivity and development. As discussed in
Section 6.2, this does not justify underpricing or subsidies. Although underpriced driving
provides many visible benefits, these are mostly transfers. Each dollar of benefit from
underpricing creates at least a dollar's worth of economic loss somewhere. Underpricing
encourages inefficient use of resources that reduces economic efficiency. Although a
Page 9-3
Transportation Cost Analysis
particular transport improvement may contribute to development in a region or
community, there is no economic reason that the facility users shouldn't pay the facility's
cost. In practice, most transport improvements provide only marginal benefits in countries
that already have extensive road, railroad, and shipping networks.
"Raise My Prices, Please!"
There is a vivid and highly emotional vocabulary to describe overpricing. A person who
paid too much is said to have been "gouged," "gypped," or "fleeced." It is easy to
demonstrate that overpricing reduces economic efficiency, and tends to be inequitable, so
overpricing is a favorite issue for economists and policy analysts. Countless political
campaigns, debates, policies, and programs focus on eliminating overpricing.
Underpricing has a similar negative impact. Underpricing leads to economic inefficiency
and unfairly imposes costs on individuals and society. It can have significant negative
social and environmental impacts. But we are unlikely to hear a popular cry, ''Raise my
prices, please. " Low prices may be acknowledged intellectually as an economic problem,
but because impacts are dispersed and nearly invisible, it seldom creates emotional fervor.
Educating policy makers, planners, and the public about problems created by underpricing
is a key challenge to developing an efficient and equitable transportation system.
Although it is often claimed that Americans have a love affair with the automobile, high
levels of automobile dependency and use are more accurately explained by decades of low
prices and skewed investments. At one time society may have benefited from increased
motor vehicle use due to economies of scale in vehicle and roadway production, but no
longer. Increased driving probably incurs diseconomies of scale in most areas.
Page 9-4
Transportation Cost Analysis
Economic efficiency, equity, and long term development are optimized if user prices
incorporate total costs. Increasing prices to better reflect costs encourages more efficient
use of our transportation system. In the long term this can reduce the need for subsidies to
transit and other special programs, due to economies of scale. In the short term, however,
increased transit subsidies may be required to overcome decades of under investment.
Even with increased use, targeted subsidies may still be justified for modes which serve
low income and disabled people for the sake of equity.
Pricing recommendations
Ideally, drivers should pay variable prices exactly equal to all marginal costs. Although it is
impossible to create an absolutely "ideal" price structure, a number of practical measures
could greatly improve current pricing:
1. Increased fuel taxes are an easy way to internalize some costs, since administrative
mechanisms exists, but as discussed in Section 6.1, this is not the best instrument since
it does not affect when and where driving occurs. A variety of charges are needed to
capture external costs, including weight-mileage changes, congestion fees, pollution
taxes, and parking charges. Payments should be frequent so drivers perceive them as
marginal costs.
2. Congestion fees can improve traffic efficiency. Several charging methods are available,
ranging from highway tolls, to electronic monitoring that automatically charges for
mileage and time spent on congested roads. It is important to consider all effects of
charges to avoid undesirable consequences. For example, freeway tolls increase
congestion on parallel surface roads, and city center tolls can encourage urban sprawl
if improperly applied.
Page 9-5
Transportation Cost A naly sis
3. One of the easiest ways to marginalize costs is to make insurance, registration,
licensing, and vehicle taxes proportional to mileage, as discussed in Section 8.3.
4. Another relatively easy and effective strategy for internalizing and marginalizing costs
is to require employers to cash out parking subsidies, also discussed in Section 8.3.
This means that employees who receive free off-street parking could choose to receive
the same value in cash. Parking should be charged daily rather than monthly, so
commuters who drive part-time only pay for what they use.
5. As much as possible, commercial parking should also be short term user paid. Parking
must be managed at the regional level to prevent communities from using free parking
to compete for business, and to prevent spill-over parking problems.
6. Pricing should be used to encourage individuals to buy fuel efficient and low emitting
vehicles, and to scrap dangerous, inefficient, and high polluting ones.
9.2
Equity
As discussed in chapters 3.9 and 7, transportation equity is a multi-dimensional problem
that depends on individual needs, ability and community circumstances. In automobile
dependent communities, anybody who does not have a personal car is disadvantaged.
Survey results (described in Section 4.5), transit funding referenda, and handicapped
access efforts indicate public support for increased transport equity and diversity.
Underpriced driving cannot be considered equitable. Underpricing forces non-drivers to
subsidize automobile use, reduces travel options, and imposes land use and social patterns
that increase travel requirements. This would be unfair even if drivers and non-drivers had
Page 9-6
Transportation Cost Analysis
comparable incomes and abilities (horizontal inequity), and is especially inequitable since
non-drivers tend to be economically, physically, and socially disadvantaged (vertical
inequity). Efforts to mitigate this inequity by providing direct subsidies for transit service
do not eliminate the imbalance between drivers and non-drivers. The relatively low
external costs of walking, bicycling, ride sharing, and telecommuting indicate that people
who use these modes subsidize SOV drivers.
The equity of increasing motor vehicle user prices depends on how revenue is used. Price
increases can be progressive if revenue is used for expenditures that significantly benefit
low income people. Using road pricing revenue only for roadway transportation
improvements is not necessarily fair or efficient since driving incurs external costs borne
by all of society. It is important for both equity and efficiency that society provide
affordable and effective travel alternatives before significantly increasing driving costs.
These typically include improved public transit service, bicycle and pedestrian facilities,
and affordable housing that does not require a car for access to jobs and public services.
Equity recommendations:
More research is needed to define transportation equity, determine ways to measure it,
and identify how it is affected by various policies. Many strategies to develop a more
efficient transport system also contribute to equity, but some efficiency strategies conflict
with equity goals. Here are specific ways to support transportation equity:
7. A basic level of transport should be defined in each community. This might include, for
example, freedom to walk safely, access to public services, employment, schools,
recreation, and social activities.
Page 9-7
Transportation Cost Analysis
8. All transportation policies and programs should be evaluated in terms of how they
affect disadvantaged groups.
9. Transportation equity and option value costs should be borne by all of society, not just
by users of particular modes. For example, the incremental costs of handicapped
access for transit systems should not incorporated into the base price of all transit
riders.
10. Transport user price increases should be predictable and gradual to allow individuals
to adjust travel patterns (housing and job locations, vehicle purchases, etc.).
11 . A significant portion of revenue from increased automobile user charges should be
targeted at refunds, tax reductions, and services that benefit disadvantaged people.
12. Transportation equity mitigation should not focus only on transit. Other modes,
including walking, bicycling, ride sharing, taxies, delivery services, telecommuting, and
land use pattern changes can provide access for non-drivers and the poor.
13 . Transition costs associated with reduced automobile dependency and use, such as
unemployment in automobile industries, should be anticipated and minimized.
9.3
Land Use Patterns
As discussed in sections 3.14 and 6.4, transportation and land use are interrelated and
must be considered together. Low transport prices reduce rent and density gradients. This
creates both benefits, which are primarily internalized and capitalized in land values, and
costs, which are primarily externalized. Driving contributes to urban sprawl, reduces
Page 9-8
Transportation Cost Analysis
neighborhood interactions and services, and increases per capita road and parking
requirements, causing substantial social and environmental costs.
The impacts of traffic on the public realm and neighborhood environmental quality deserve
special attention. The street system (including sidewalks) is the most valuable publicly
owned physical asset in most jurisdictions. In addition to accommodating vehicle
movement, streets define a community's character, accommodate pedestrian and bicycle
travel, and are an important community meeting place. These functions are degraded by
automobile traffic. While it is possible to walk, bicycle, and socialize on streets with
moderate to high vehicle traffic volumes and speeds, doing so is less efficient, pleasant and
safe than on low traffic streets. Subjugating street designs to motor vehicle traffic needs
increases automobile dependency and use, exacerbates urban sprawl, and reduces mobility
for non-drivers.
Nee-traditional neighborhood design and traffic calming programs described in section 8.6
are increasingly popular strategies to reduce traffic impacts and return streets to multifunction use. Implementing these improvements requires changes in transport planning and
funding patterns. Many benefits of increased travel are lost due to competition for
location, as discussed in section 6.4. Increased driving ability allows individuals to buy
relatively low priced homes at the urban fringe . This perpetuates urban sprawl, increased
automobile dependency, and degradation of urban environments. As a result, land prices
escalate, more driving occurs, and traffic impacts increase with little or no net benefit.
Land use goals such as greenspace preservation, improved neighborhood environments,
and reduced automobile dependency should preempt short term transport objectives.
Specific projects and policies should be evaluated in terms of how they support or
contradict a community's land use goals. Various land use design factors have been shown
Page 9-9
Transportation Cost Analysis
to affect automobile dependency and use, including residential and employment density,
land use mix, transit service quality, pedestrian and bicycle network quality, and building
site design. Effective consideration of these factors can significantly reduce travel demand.
Land Use recommendations:
14. Transportation and land use planning should be integrated so policies and projects are
mutually supportive.
15. Prior to developing a transportation plan, communities should establish land use and
environmental goals and plans.
16. Transport improvements that contradict land use plans and neighborhood
environmental quality goals should be avoided even if they are otherwise justified. In
economic evaluation this could be accomplished by assigning "Strategic Goal" cost
and benefit factors to factors that support or contradict these goals.
17. Full-cost pricing ofland development, utilities and other public services help reduce
subsidies that fuel urban sprawl.
18. Parking standards should specify maximum rather than just minimum capacities.
Parking requirements should be flexible, and should be reduced where they are not
needed due to low automobile ownership (for residential developments) or use (for
commercial developments).
Page 9-10
Transportation Cost Analysis
19. Zoning laws, development standards, home buyer programs, and other land use and
land development policies should be modified as needed to conform with and support
community transport goals.
20. Parking management programs are needed to avoid conflicts (such as spill-over
parking into residential neighborhoods) and to internalize parking costs.
21 . Communities should insure that at least a portion of housing units in all price ranges
are accessible to stores, employment, and other public services without driving.
22. Zoning laws and development policies should encourage increased diversity ofhousing
types, infilling and appropriate land use mixing.
23. Greater attention should be paid to streetscape design and development of local
activity centers to encourage walking, bicycling and neighborhood interaction.
24. It is especially important to develop moderate density, mixed use communities near rail
stations and bus routes.
25 . Local services, such as neighborhood stores, local schools, and small parks should be
encouraged to reduce travel needs.
26. Zoning and development policies that preserve greenspace and discourage urban
sprawl should be implemented.
27. Traffic Calming and other traffic manage-ment strategies should be used to reduce
traffic impacts and better accommodate pedestrians and bicyclists on existing
Page 9-11
Transportation Cost Analy sis
residential and commercial streets. Cities should establish mechanisms to implement
traffic management and calming.
9.4
Transportation Decision Making
Transportation decisions have tremendous impacts, many of which tend to be ignored
during policy making, planning and budgeting. Planning is often uncoordinated, resulting
in a "tyranny of small decisions." Decision makers often treat driving as an end in itself,
rather than a means for achieving access, and focus on the benefits of accommodating
motor vehicle traffic without assessing total costs. This leads to increase automobile
dependency and use, exacerbate environmental and social problems, reduce access for
mobility disadvantaged people, reinforce social inequity, and provide less benefit than
predicted. Current transport planning practices have five major problems:
•
Limited scope . Current planning and funding practices do not provide equal
consideration to all options for meeting access needs. Demand management tends to
be considered and implemented only where traffic congestion or air quality problems
are significant, and ignored in other situations. Funding allocation tends to favor
roadway improvements over other modes.
•
External costs ignored. Current planning practices tend to ignore many costs,
especially environmental and social impacts. Economic evaluation models can, but
usually do not, incorporate monetized estimates of these non-market costs. Even costs
such as parking demand and public service demands of increased motor vehicle use are
seldom considered in transportation planning and project evaluation.
Page 9-12
Transportation Cost Analysis
•
Poor public involvement. Although transportation decisions impact many aspects of
individual and community life, transport planning is considered a technical field and the
public is excluded from many critical decisions. Residents are seldom involved early
enough in the planning process to place constraints or establish broad goals that reflect
community values, and even with citizen involvement transportation decisions are
highly influenced by professional biases and preferences.
•
Missing link between transportation and land use planning. Although transportation
and land use patterns are highly interrelated, they are seldom planned together.
Transportation planning should be considered a subset of land use and community
development planning. Since transport to a large degree determines long term land use
patterns, transportation decisions should be based on long term land use goals.
•
Generated traffic effects ignored. Research described in Section 8.4 shows that
increasing road capacity increases total driving, especially in congested areas. As a
result, projects that expand urban roadway capacity usually provide significantly less
congestion reduction than predicted because latent demand fills much of the new
capacity, and automobile use increases throughout the region.
Since most urban trips are relatively short (less than 5 miles), there is a "transportation
gap" caused by overemphasis on long-distance travel and too little attention to bicycling,
local transit, and low powered vehicles. This creates a self-fulfilling prophecy of increased
driving, automobile dependency, inequity and sprawl. Electric cars and other alternative
fuels reduce some external costs, particularly urban air pollution, noise, and petroleum
externalities, but do not affect others such as accident risk, congestion, and parking
Page 9-13
Transportation Cost Analysis
subsidies. Introduction of relatively inexpensive, small, low power, low speed vehicles for
local travel may offer greater overall savings.
Transportation decision making recommendations:
28. Measures of transport system effectiveness should be based on access rather than
traffic volumes and speeds. Policies and programs that reduce the need to travel
should be compared equally with measures that increase mobility in planning and
funding.
29. Transportation economic analysis must consider all costs. Non-market and indirect
costs should be given the same weight as market costs. Non-market costs should be
quantified and monetized as much as possible for use in economic evaluation.
30. Least-cost planning should be used as a model for transportation decision making.
This means that a broad range of options are considered, including both supply and
demand management, and evaluated based on a standard set of criteria that takes into
account all costs and benefits. A "no-build" option with land use management
strategies should be considered in transport planning and receive equal funding to road
building.
3 1. Research is needed to understand how public policies and land use patterns affect
travel decisions, and to develop practical strategies and programs that achieve
transportation demand reduction goals.
Page 9-14
Transportation Cost Analysis
32. Transport planners should become familiar with the environmental and social impacts
of their decisions. Environmentalists, urban planners, and social policy analysts need to
learn more about transport issues.
33. Transportation equity and diversity should be recognized as important goals in
planning and policy making. Achieving these goals requires development of a diverse
and integrated transport system that accommodates non-drivers.
34. Non-motorized transportation modes (walking, bicycling, and telecommuting) deserve
increased emphasis in planning and funding. Special attention is needed for intermodal
connections, such as the integration ofbicycling with transit.
35. Traffic analysis must consider the effects of generated traffic. Generated traffic should
be assessed using the "rule-of-half' which recognizes that these trips tend to have
relatively low value, since they are trips that users forego if roads are congested.
36. The incremental external costs of generated traffic should be treated as a cost of
projects that increase roadway capacity.
37. Many of the resources that are currently targeted at large scale regional transport
projects may provide greater benefit if used for local and neighborhood improvements.
For example, improvements to local shopping districts, pedestrian and bicycle
facilities, and neighborhood services (parks, schools, etc.) may communities to become
more self-sufficient and thus reduce motor vehicle traffic and automobile dependence.
Page 9-15
Transportation Cost Analysis
38. Transportation professionals and decision makers should make a habit of not using an
automobile for at least two consecutive weeks each year in order to experience the
practical problems facing non-drivers.
39. Impacts on human life and health, and irreversible environmental damage should be
assessed with a low discount rate for the sake of intergenerational equity.
40. Neighborhood car rentals and ownership co-ops should be encouraged to help reduce
the need for residents to own cars and trucks that are seldom used to capacity.
9.5
Research Recommendations
Further research is needed on transportation external costs including air pollution, noise,
accident expenses, parking subsidies, and municipal service costs to give representative
values for different communities and driving conditions. For example, several estimates of
air pollution exist, but most are either for areas with extreme pollution problems, such as
Southern California, or they are nationwide totals.
Transport equity and diversity appear to be significant values and which deserve much
more research. The concept of automobile dependency deserves more analysis. Research
is needed to identify factors that contribute toward automobile dependency, ways to
quantify its costs, and strategies for reducing it.
Transportation land use impacts need more research to understanding of how transport
decisions affect land use, and methods to measure and monetize these effects. This
research should cover impacts to both natural environments, such as the loss of wildlife
habitat and landscapes, and impacts on the built environment, such as the degradation of
Page 9-16
Transportation Cost Analysis
neighborhood life from high traffic volumes. These appear to be significant costs with
major implications to many transport decisions.
The barrier effect (severance) has been well studied and measured in Scandinavian
countries, but their quantification techniques have not been applied in North America.
Research is needed to test the Scandinavian formulas here and develop estimates of this
cost per vehicle mile under a variety of typical conditions.
Latent demand has tremendous implications on transport decisions. Some progress has
been made to develop tools for predicting generated traffic. We need more information to
help predict generated traffic on highly congested roads and other typical conditions. Most
current studies focus on traffic generated on single roads. Of equal or greater importance
is the overall increase in regional automobile use that results from increased road capacity.
Page 9-17
Transportation Cost Analysis
Bibliography
This is just a portion of the total literature reviewed for this document. A complete and
current annotated bibliographic computer database is available from the author.
Alexander, Christopher, et al., A Pattern Language, Oxford University Press, 1977.
American Association of State Highway and Transportation Officials, A Policy on Geometric
Design ofHighways and Streets, AASHTO, Washington DC, 1990.
American Public Transit Association, Transit Fact Book, APTA, Oct. 1992.
Apogee Research, Inc., Richard Mudge, The Costs of Transportation: Final Report, Conservation
Law Foundation, March, 1994.
Bartholomew, Keith, "Making the Land Use, Transportation, Air Quality Connection," American
Planning Association, PAS Memo, May, 1993.
Bhatt, Kiran, "Implementing Congestion Pricing: Winners and Losers," Institute ofTransportation
Engineers Journal, Dec. 1993 .
BTCE & EPA (Franzi Poldy and Mark Dess), Costing and Costs ofTransport Externalities; A
Review, Victorian Transport Externalities Study, Vol. 1, Environment Protection Authority
(Melbourne), 1994.
Bureau ofTransport Statistics, Transportation Statistics Annual Report; 1994, USDOT,
(Washington DC), January 1994.
California Energy Commission, California Transportation Energy Analsysis Report (Draft) ,
1993-1994, California Energy Commission (Sacramento) Feb. 1994.
Cameron, Michael, Efficiency and Fairness on the Road, Environmental Defense Fund, Oakland,
CA, (510)658-8008, Feb. 1994.
Cameron, Michael, Transportation Efficiency, Environmental Defense Fund, March 1991.
Cannon, James : for American Lung Association, The Health Costs ofAir Pollution, 1990.
Cervero, Robert, America's Suburban Centers: The Land Use-Transportation Link, Unwin
Hyman (Boston), 1989.
Citizens Against Route Twenty, Traffic Calming, Solution to Route 20, CART, Melbourne, 1989.
Clough, Peter, Land Transport Pricing: Digest Report, Transit New Zealand, Report #20, 1993 .
Covil P.E., James; Richard Taylor; Michael Sexton P .E. , Will Multimodal Planning Result in
Multimodal Plans, Transportation Research Board Annual Meeting, January 1994.
de Boer, Enne (ed.), Transport Sociology: Social Aspects of Transport Planning,Pregamon, 1986.
DeCorla-Souza, Patrick; Ronald Jensen-Fisher, Comparing Multi-Modal Alternatives in Major
Travel Corridors, Transportation Research Board Annual Meeting January 1994.
Dess, Mark; Godfrey Lubulwa; Sophia Schyschow, Externalities- The Concept and Their Control
in the Transport Context, 16th Australian Road Research Board Conference, 1992.
Dimitriou, Harry, Urban Transport Planning; A Developmental Approach, Routledge, 1992.
Bib.-1
Transportation Cost Analysis
Dodds, Daniel; Jonathan Lesser, Monetization and Quantification ofEnvironmental
Impacts,Washington State Energy Office (Olympia), June 1992.
Downs, Anthony, Stuck in Traffic; Coping with Peak-Hour Traffic Congestion, Brookings
Institute (Washington DC), 1992.
Dunphy, Robert; Kimberly Fisher, Transportation, Congestion, and Density: New Insights,
Urban Land Institute (Washington DC), Jan. 1994.
Engwicht, David, Reclaiming Our Cities and Towns: Better Living With Less Traffic, New
Society Publishers (Philadelphia), 1993 .
Eurpoean Conference of Ministers of Transport, Internalising The Social Costs of Transport,
OECD (Paris), 1994.
Ewing, Reid, Transportation Services Standards -As If People Matter, Transportation Research
Board Annual Meeting, January 1993.
Ewing, Reid; Padma Haliyur; G. William Page, Getting Around a Traditional City, A Suburban
PUD, and Everything In-Between, Transportation Research Board, January 1994.
Federal Railroad Administration; Office of Policy; USDOT, Environmental Externalities and
Social Costs ofTransportation Systems-Measuring, Mitigation & Costs, USDOT; Federal
Highway Administration, August 1993 .
Fisher, Peter; Philip Viton, The Full Costs of Urban Transport: Part 1: Economic Efficiency in
Bus Operations, Institute ofUrban and Regional Development, UC Berkeley, #19, 1974.
Frank, Lawrence, Impacts ofMixed Use and Density on the Utilization ofThree Modes of Travel,
Transportation Research Board, January 1994.
Giuliano, Genevieve; Keith Hwang; Martin Wach, "Employee Trip Reduction in Southern
California: First Year Results," Transportation Research; 27:, #2, 1993 .
Goddard, Stephen, Getting There, Basic Books (New York), 1994.
Goodland, RJA, The Urgent Need for Environmental Sustainability in the Transport Sector; An
Informal Personal Holistic View with Emphasis on Developing Countries, Wold Bank,
Environment Department, Jan. 12, 1994.
Grant, Michael, A Methodolgy for Evaluating Cross-Modal Transportation Alternatives, Office
of Intermodalism, US DOT (Washington DC), 4 August 1994.
Hansen, Mark; David Gillen; Allison Dobbins; Yuanlin Huang; Mohnish Puvathingal, The Air
Quality Impacts of Urban Highway Capacity Expansion: Traffic Generation and Land Use
Changes, Institute ofTransportation Studies, University of California at Berkeley, April, 1993.
Hanson, Mark E., "Automobile Subsidies and Land Use: Estimates and Policy Review," Journal of
the American Planning Association, Winter 1992.
Hanson, Mark; Resource Management Associates ofMadison, The Nature and Magnitude of
Social Costs of Urban Roadway Use; Literature Survey and Summary ofFindings, Federal
Highway Administration, USDOT, Jul-92 .
Hart, Stanley, An Assessment of the Municipal Costs ofAutomobile Use, Self published, 1985 .
Hart, Stanley; Alvin Spivak, The Elephant in the Bedroom; Automobile Dependency and Denial,
New Paradigm Books (Pasadena), 1993 .
Bib.-2
Transportation Cost Analysis
Holtzclaw, John, Using Residential Patterns and Transit to Decrease Auto Dependence and
Costs, National Resources Defense Council, San Francisco; funded by the California Home
Energy Efficiency Rating Systems, June, 1994.
Hook, Walter "Are Bicycles Making Japan More Competitive?, Sustainable Transport~ Institute
for Transportation Development Policy (Washington DC), Summer 1993 .
Hook, Walter, Counting on Cars, Counting Out People, Institute for Transportation and
Development Policy (NY), Paper# 1-0194, Winter 1994.
Hughes, William; C.F. Sirmans, "Traffic Externalities and Single-Family House Prices," Journal
ofRegional Science, 32/4, 1992.
Jack Faucett Associates, Cost of Owning and Operating Automobiles, Vans and Light Trucks,
1991, USDOT, FHWA (Washington DC), April1992 .
Johnson, Elmer, Avoiding the Collision of Cities and Cars; Urban Transportation Policy for the
Twenty-First Century, American Academy of Arts and Sciences (Chicago), Sept. 1993.
Johnston, Robert; Raju Ceerla, "A Comparision of Modeling Travel Demand and Emissions With
and Without Assiged Travel Times Fed Back to Trip Distribution," Submitted to the Journal of
Transportation Engineering, 11 April, 1994.
Kageson, Per, Getting the Prices Right: A European Scheme for Making Transport Pay Its True
Costs, European Federation for Transport and Environment (Bruxelles, Belgium), May, 1994.
Keeler, Theodore; Kenneth Small, The Full Costs of Urban Transport; Part IlL Intermodal Cost
Comparisons, Institute of Urban and Regional Development (Berkeley), Monograph #21, 1975 .
Kelbaugh, Douglass; Mark Hinshaw; David Wright, Housing Affordability and Density:
Regulatory Reform and Design Recommendations, Department of Architecture, University of
Washington (Seattle), 1992.
Ketcham, Brian, Making Transportation Choices based on Real Costs, Konheim & Ketcham Inc.
(New York), October 1991.
Komanoff, Charles, Pollution Taxes for Roadway Transportation, Kamanoff Energy Associates
(New York), January, 1994.
Komanoff, Charles; Brian Ketcham, Win-Win Transportation: A No-Losers Approach to
Financing Transport in NYC and the Region, Transportation Alternatives (New York), 1992.
Kunstler, James Howard, Geography ofNowhere, Simon & Schuster, 1993 .
Laube, Felix; Michael Lynch, Costs and Benefits ofMotor Vehicle Traffic in Western Australia,
Institute for Science and Technology Policy, Murdoch University, March, 1994.
Lee, Douglass, Recent Advances in Highway Cost Allocation Analysis, Transportation Research
Record #791 , 1981 .
Lee, Douglass, Full Cost Pricing ofHighways , Volpe National Transportation Systems Center
(Cambridge), January 1995.
Lee, Douglass, "A Market-Oriented Transportation and Land Use System: How Different Would it
Be?" in Privatization and Deregulation in Passenger Transport: Selected Proceedings of the
2nd International Conference, Espoo, Finland, Viatek, Ltd., June 1992, pp. 219-238 .
Bib.-3
Transportation Cost Analysis
Levinson, Herbert; Robert Weant editors, Urban Transportation; Perspectives and Prospects,
Eno Foundation (Westport, CT), 1982.
Loder & Bayly Consulting; RJ Nairn; Sustainable Solutions, Greenhouse Neighborhood Project;
A Low Energy Suburb- Summary Report, Victorian (Australia) Government, July, 1993 .
Lowe, Marcia, Alternatives to the Automobile: Transport for Livable Cities, Worldwatch Institute
(Washington DC), 1990.
Luk, James; Stephen Hepburn, "New Review of Australian Travel Demand Elasticities,"
Australian Road Research Board #249, (Victoria, Australia) 1994.
MacKenzie, James; Roger Dower, Donald Chen, The Going Rate, World Resources Institute
(Washington DC), June 1992.
Masser, Ian; Ove Sviden; Michael Wegener, "Transport Planning for Equity and Sustainability,"
Transportation Planning and Technology, 1993, Vol. 17.
McElhaney, David, Highway Funding Bulletin, USDOT, FHWA (Washington DC), April, 1994.
Miles, John, The Costs of Congestion : A Preliminary Assessment ofMelbourne's Road Network,
(Victorian, Australia) Transport Externalities Study, Feb. 1994.
Miller, Peter; John Moffett, The Price ofMobility: Uncovering Hidden Costs ofTransportation,
Natural Resources Defense Council (Washington DC), Oct. 1993.
Moore, Terry; Paul Thorsnes, The Transportation/Land Use Connection, American Planning
Association, Report #448/449, January 1994.
National Highway Traffic Safety Administration, Traffic Safety Facts, 1992, USDOT, NHTSA,
(Washington DC), May 1993 .
Nelson, Dick; Don Shakow, Applying Least Cost Planning to Puget Sound Regional
Transportation, Institute for Transportation and the Environment (Seattle) Feb. 1994.
Newman, Peter; Jeffrey Kenworthy, Cities and Automobile Dependency, Gower Publishing, 1989.
Nijkamp, Peter, "Roads Toward Environmentally Sustainable Transport," Transportation
Research, Vol. 28A, No.4, 1994.
Environmental Imapct Assessment of Roads, OECD (Paris), IRRD No. 859799, 1994.
Saving Energy in U.S. Transportation, Office ofTechnology Assessment, July 1994.
Pearce, David, Economic Values and the Natural World, MIT Press (Cambridge), 1993.
Pisarski, Alan, New Perspectives in Commuting, USDOT; FHWA (Washington DC), July 1992.
Pucher, John, "Urban Travel Behavior as the Outcome of Public Policy," American Planning
Association Journal, Fall1988 .
Pucher, John; Ira Hirschman, Path to Balanced Transportation: Expand Public Transport and
Require Auto Users to Pay Full Costs , American Public Transit Association (Washington DC),
Oct. 1993.
Replogle, Michale, Transportation Conformity and Demand Management: Vital Strategies for
Clean Air Attainment, Environmental Defense Fund (Washington DC), April1993.
Rodriguez-Bachiller, Agustin, "Discontiguous Urban Growth in the New Urban Economics,"
Urban Studies, February 1989.
Bib.-4
Transportation Cost Analysis
Roelofs, Cora; Charles Komanoff, Subsidies for Traffic: How Taxpayer Dollars Underwrite
Driving in New York State, Tri-State Transportatation Campaign (New York) March 1994.
Schaeffer, K.H.; Elliott Sclar, Access for All; Transportation and Urban Growth, Columbia
University Press (New York), 1980.
Sheets, Edward; Richard Watson, Least Cost Transportation Planning: Leassons From the
Northwest Power Planning Council, Bullitt Foundation (Seattle), Dec. 1993 .
Shoup, Donald, Cashing Out Free Parking, USDOT, Federal Transit Administration (Washington
DC), FTA-CA-11-0035-92-1, December 1993.
Slater, Rodney, Highway Statistics 1992, USDOT, FHWA (Washington DC), 1992.
Small; Winston and Evans, Road Work, Brookings Institute (Washington DC), 1989.
Steiner; Ruth, "Least Cost Planning for Transportation?; What We Can Learn," Transportation
Research Board Annual Meeting, January 1992.
Stevens, Paula, Costs Associated with Passenger Vehicle Use, California Air Resources Board
(Sacramento), June, 1993 .
Tatineni, Maya; Mary Lupa; Dean Englund; David Boyce, Transportation Policy Analysis Using
a Combined Model ofTravel Choice, TRB Annual Meeting, January 1994.
Tengstrom, Emin, Use ofthe Automobile; Its Implications for Man, Society and the Environment,
Swedish Transport Research Board (Stockholm), 1992.
Tomazinis, A., Productivity, Efficiency & Quality in Urban Transportation, Lexington, 1975.
Transport Concepts, External Costs ofTruck and Train, Ottawa, October 1994.
Back to the Future; Re-Designing Our Landscapes with Form, Place & Density, Urban
Development Institute (Vancouver), 1993 .
Van Seters, Levelton, Pammenter, Powell, Paul, Litman, Cost ofTranporting People in the
British Columbia Lower Mainland, Greater Vancouver Regional District (Vancouver), 1993 .
Verhoef, Erik, External Effects and Social Costs of Road Transport, Transportation Research,
Vol. 28A, No.4, 1994.
Walter, Felix; Dr. Heini Sommer; Rene Neuenschwander, External Benefits ofTransport?,
Ecoplan (Bern, Switzerland), March 1993 .
Walter, Felix; Rene Neuenschwander, External Costs ofTransport-An Overview, Ecop1an (Bern,
Switzerland), October 1992.
Whitelegg, John, "Time Pollution," The Ecologist, Vol23, No. 4, Jul-93.
Wilson, Tay; Charlotte Neff, Social Dimension in Transportation Assessment, Gower, 1983 .
Winston, Clifford, "Efficient Transportation Infrastructure Policy," Journal ofEconomic
Perspectives, 5: 1, Winter 1991.
Works Consultancy Services Ltd., Land Transport Externalities, Transit New Zealand
(Wellington), Report #19, 1993.
Wright, Charles, Fast Wheels, Slow Traffic, Temple University Press (Philadelphia), 1993 .
Wright, Charles, Characteristics Analysis ofNon-Motorized Transport, UMTRI Research
Review, March 1990.
Bib.-5