Monday, May 24, 2010

How much is that car really costing? You? The rest of us?

Editor's introductory note: I have long maintained that the cost of driving one more car in a city is far greater than normally understood, with the result that the benefits to the city of getting one car off the street are very very considerable. My own working rule of thumb, admittedly crude and entirely unscientific, is that every time a mayor or her team figure out how to remove one car from the traffic stream -- without decreasing the quality of the overall mobility system - brings about a benefit for all equal to at least one dollar a car/km. But let's hear what Todd Litman of the Victoria Transport Policy Institute has to say about it.

Transportation Cost Analysis

- by Todd Litman, VTPI, Victoria, BC, Canada. 24 May 2009;

Mobility provides many benefits, but it also incurs significant costs. It is therefore important to optimize our transport system to maximize net benefits, considering all impacts.

What does it cost to commute by car? Would you save by shifting to bicycling or public transportation? How would this affect your neighbors? What are the optimal fees for using roads, parking facilities, vehicle insurance, and fuel?

These and many other questions can be answered with comprehensive analysis of transportation costs. This column explores this issue and discusses the implications of current, based on information from my report, Transportation Cost and Benefit Analysis, available free on our website at

By “costs” I don’t just mean monetary expenses, it also includes impacts on non-market resources such as personal time, health and environmental quality.

The table below lists the 23 categories of transport costs considered in my analysis. Some costs, such as parking and accidents, are divided into internal costs, which are borne directly by users, and external costs, borne by other people. My report provides detailed analysis of each cost category.

Table 1 Transport Cost Categories (Click tables and figures for clearer image.) This table defines the 23 transport cost categories evaluated in this study.

Some of these costs are relatively easy to measure. For example, vehicle ownership and operating costs are relatively easy to estimate. Transportation economists have developed good estimates of the costs of building and maintaining roads and parking facilities, and the values of travel time, congestion and accidents. Other costs have received less research, but their values can still be estimated using various data sources and quantification methods.

The figure below illustrates the estimated monetized value of these costs, averaged per vehicle-mile for a typical automobile driven in the U.S.

Figure 1 Costs Ranked by Magnitude

This figure shows costs per vehicle-mile for an average North American automobile.

The results are, for the most part, unsurprising. Vehicle ownership and operation, crash damages, travel time, and the costs of building and maintaining road and parking facilities are among the largest costs. Congestion and air pollution, two costs that receive considerable attention in transport policy debates, are actually moderate in magnitude. Many of the largest costs are internal, borne directly by users, external costs are individually relatively small, but numerous.

Figure 2 summarizes the overall distribution of these costs. About a third are external, a quarter are internal-fixed (motorists pay the same amount, regardless of how much they drive) and less than half of these costs are internal-variable (borne by users according to the amount a vehicle is driven, and therefore directly affecting short-term travel decisions).

Figure 2 Average Car Cost Distribution

This figure illustrates the aggregate distribution of costs for an average car. About 60% of total vehicle costs are either External or Internal-Fixed, and so do not directly affect short-term travel decisions.

These costs vary depending on travel conditions, with higher costs (particularly congestion, pollution, and infrastructure costs) under urban-peak conditions, as illustrated in Figure 3.

Figure 3 Average Car Cost Distribution By Driving Conditions

This figure illustrates how aggregate costs vary by travel conditions. Urban-Peak travel has the largest total costs largely due to higher external costs.

My analysis also estimates these costs for different travel modes, as summarized in Figure 4. Of course, these costs vary significantly depending on travel conditions and user needs and preferences.

For example, public transit travel costs are much lower than automobile costs under urban-peak conditions, and under favorable conditions walking and cycling can have very low costs because users enjoy these activities, but under unfavorable conditions their costs per mile can be very high, which can justify investments in sidewalks, crosswalks and traffic calming to improve non-motorized travel conditions, and therefore encourage shifts from motorized to non-motorized modes.

Figure 4 Cost Distribution by Mode

This graph shows the cost distribution of each mode. These costs are measured per passenger-mile, not per vehicle-mile, as in previous graphs. Note that transit costs are based on average U.S. ridership levels and would be lower in areas with higher ridership rates.

Conventional planning tends to consider some of these costs but ignores others, as summarized in Table 2. For example, conventional planning generally ignores parking costs and vehicle ownership costs, it assumes that businesses must supply the same number of employee and customer parking spaces, and households will own the same number of vehicles, regardless of transport policies or planning decisions, and so ignores the savings to businesses and consumers of improvements to alternative modes that allow consumers to own fewer vehicles and drive fewer annual miles.

This tends to bias planning decisions toward highway projects, and away from improvements to walking, cycling and public transit, and away from mobility management policies that encourage use of alternative modes.

Table 2 Scope of Conventional Planning Analysis

If you ask people what it costs to drive they typically mention vehicle operating expenses, which average approximately 16¢ per mile for a typical automobile. Some may include vehicle ownership costs, which average about 27¢ per mile. A few may also mention travel time and crash risk.

These, however, are only a portion of total costs. The full cost of driving includes these direct, internal costs, plus various indirect and external costs. Total estimated costs range from about $0.94 per vehicle mile for rural driving to $1.64 for urban-peak driving. Of course there is considerable variation in the costs 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.

The largest categories of transport costs tend to be internal, including vehicle ownership and operation, travel time, and crash risk borne by motorists. External costs tend to be smaller, and so are easy to overlook, but numerous, so their aggregate value tends to be significant. About half of transport costs are either external or internal-fixed, and therefore do not directly affect individual travel decisions. This represents underpricing, which results in economically excessive automobile travel (more vehicle travel than would occur in a more efficient market). Other transport modes have different cost profiles, some having much smaller external costs under certain circumstances.

For More Information

European Transport Pricing Initiatives ( includes various efforts to calculate transportation costs and optimal pricing.

Todd Litman (2008), Transportation Cost and Benefit Analysis; Techniques, Estimates and Implications, Victoria Transport Policy Institute (

Todd Litman (2009), Socially Optimal Transport Prices and Markets, VTPI (; at

M. Maibach, et al. (February 2008), Handbook on Estimation of External Cost in the Transport Sector: Produced within the study Internalisation Measures and Policies for All external Cost of Transport (IMPACT) Version 1.1, CE Delft, for the European Commission DG TREN; at

Nariida C. Smith, Daniel W. Veryard and Russell P. Kilvington (2009), Relative Costs And Benefits Of Modal Transport Solutions, Research Report 393, NZ Transport Agency (; at

Swiss ARE (2005), External Cost of Transport In Switzerland, Swiss Federal Office of Spatial Development (; at English summary in
Externe Kosten des Verkehrs in der Schweiz; Aktualisierung für das Jahr 2005 mit Bandbreiten.

TC (2005-08), The Full Cost Investigation of Transportation in Canada, Transport Canada (

# # #

About the author:

Todd Litman is founder and executive director of the Victoria Transport Policy Institute, an independent research organization dedicated to developing innovative solutions to transport problems. His work helps to expand the range of impacts and options considered in transportation decision-making, improve evaluation techniques, and make specialized technical concepts accessible to a larger audience. He can be reached at: 1250 Rudlin Street, Victoria, BC, V8V 3R7, Canada. Email: Phone & Fax: +1 250-360-1560

Print this article


  1. Simon Norton, Cambridge UKMonday, 24 May, 2010

    Does the calculation of external costs include those costs which have longer causation chains ?

    Example. If more people in an area drive, there are fewer passengers on the local buses, which also take longer because of increased congestion. So the service becomes less attractive -- and more expensive unless the subsidy is increased. This means that bus users may switch to cars, causing further external costs.

    Similarly the area might become less attractive for cycling, leading to a switch from bike to car and more external costs.

    My contention is that motorists should pay not only those external costs attributable to their own cars, but also those attributable to those extra cars that take to the road as a result of their decision to go by car.

    Simon Norton, Cambridge UK

  2. David Levinger | The Mobility Education Foundation | USATuesday, 25 May, 2010

    There is a new study by McCubbin and Delucchi (2010) that I have found quite useful and I am attaching here.
    Download link:

    Delucchi pubs:

    David Levinger | The Mobility Education Foundation | USA

  3. John Warren, Toronto ON CanadaWednesday, 26 May, 2010

    This is dear to my heart

    As a member of the Canada transportation climate change table prior to signing the Kyoto agreement I advocated including healthcare costs as an externality to fund public transit. – Public transit much safer than cars hence use cost savings health death injury lost time productivity avoided by lack of car use to public transit. My guess is that this would make public transit cost positive.

    I was informed in one session that “this will not be included in the analysis’

    The process included oil and other lobby groups which might have been a factor in this decision.

    Have you / anyone else specifically researched this question? – more appropriate in countries with public health care but USA also moving in that direction

    John Warren, Toronto ON Canada

    ph (1) 416-444-6099
    mobile (1) 647-899-7163


Thank you for your comment. You may wish to check back to the original entry from time to time to see if there are reactions to this. If you have questions, send an email to: