Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration
Office of Planning, Environment, & Realty (HEP)

A Least Total Cost Approach to Compare Infrastructure Alternatives

Patrick DeCorla-Souza, Brian Gardner, Jerry Everett & Michael Culp
Federal Highway Administration
400 Seventh St SW
Washington DC 20590

Why a Least Total Cost Approach?

The new environment for transportation planning in the 1990s presents a challenge to planners and decision makers in evaluating multimodal alternatives. The Intermodal Surface Transportation Efficiency Act (ISTEA) of 1991 provides new intermodal funding flexibility. Also, ISTEA requires consideration of efficiency, social, economic and environmental factors in the evaluation process. The Act's emphasis on "management" calls for development of procedures that allow comparisons across a variety of alternatives including new services, land use and demand management as well as high capital investment-type solutions. Additionally, the Clean Air Act Amendments (CAAA) of 1990 emphasize vehicular demand management as an important strategy to reduce air pollutant emissions. Future evaluation procedures will thus need to: (a) give adequate consideration to economic efficiency and social and environmental impacts; and (b) be capable of allowing comparisons across modes as well as across a variety of high capital and low capital or management strategies.

In the past, Metropolitan Planning Organizations (MPOs) have usually compared transportation projects using measures of effectiveness which are uniquely applicable to a specific mode. For example, measures of highway project effectiveness commonly used are improvement in highway level of service (LOS) or highway speed, reduction of highway accidents or savings in highway user costs. Transit project effectiveness, on the other hand, is usually measured by transit ridership or public capital and operating costs per new rider. It is likely that Intelligent Vehicle Highway System (IVHS) projects will also use different measures of effectiveness, depending on their modal orientation. If IVHS projects or programs benefiting different modes (e.g. highway solo-driver, highway shared ride or transit) are to be compared with one another or with other types of investment or management strategies, common measures of effectiveness will have to be used i.e. measures applicable across modes, and across supply-enhancing and demand-reducing strategies.

The least total cost approach uses a common measure (i.e. total cost) which is applicable across all types of alternatives. It attempts to account for the full costs of each alternative. The main advantages of this approach are: (1) It allows comparisons of transportation investments across modes; (2) It allows comparisons of major investment alternatives (e.g. new highway or transit capacity) with management alternatives such as new or improved services (e.g. using IVHS technology), pricing strategies, land use strategies and other strategies which moderate travel demand.

The least total cost approach facilitates accounting for costs of competing highway-oriented and transit-oriented IVHS projects in a comparable manner. For example, in current practice, when computing costs for transit alternatives, analysts include vehicle capital and operating costs and costs for garaging the vehicle. On the other hand, analysts computing the costs for highway travel may include the variable portion of vehicle operating costs such as costs for gas and oil, maintenance and tires, but exclude the fixed costs such as vehicle ownership costs and parking or garaging costs at each end of the trip. (Note that, in the long range, vehicle fixed costs and parking fixed costs are avoidable costs i.e. they are not sunk costs to be ignored). For valid comparisons across modes, the full avoidable future costs of each alternative will have to be taken into account, not just costs incurred by transportation agencies for capital investment and operation. Public costs incurred by non- transportation public agencies (e.g. police, fire, court systems, etc.), fixed private costs (e.g. auto ownership costs), and external social and environmental costs cannot be ignored. From a societal point of view, it is irrelevant whether costs are borne privately, publicly or socially.

In a least total cost approach, user benefits other than satisfaction of the basic need for access, for example comfort and convenience advantages of a particular modal alternative, need not be excluded. User benefits or "amenities" can be included in the cost totals as negative costs if they are quantifiable and can be converted to monetary terms. Some user benefits and dis-benefits, as well as some external costs and benefits, cannot easily be converted to monetary terms. They may be listed with some measure of their magnitudes for use in trade-off analysis. For example, a break even analysis could be done to determine how much additional benefits from a higher total cost alternative would have to be worth in dollars in order to make decision makers indifferent between the higher cost alternative and the one with the least total cost.

The base to which alternatives are compared in current practice also poses a problem. In current practice, the base used for comparison is usually a future year "do-nothing", or "no- build plus Transportation System Management (TSM)" alternative. Benefits of the alternatives are calculated based on savings with respect to the base. However, the savings estimates will not be real if the base itself could never exist in reality, which is often the case. For example, before the large delays forecasted under base conditions could ever occur, it is probable that travelers would change their travel patterns (either traveling at different times of the day, by different modes, to different destinations, or by different routes); or they may even decide not to make the trip. It is therefore probable that benefits claimed for alternatives by comparing them to the base are inflated to some extent. (Note that travelers do suffer losses in overall utility when they are compelled to shift their travel patterns; however, the increase in travel times modeled under the typical base year scenario probably overestimates their utility losses.)

The least total cost approach as applied in this paper embodies the following major features:

Applying the Least Cost Approach

The approach is demonstrated in this paper through application to a case study using a simplified microcomputer-based spreadsheet (LOTUS 123). The focus of the case study is on comparison of land use and IVHS strategies. A previous paper presented a case study application of the approach focusing on evaluation of major transportation investments (8).

Unit costs of travel differ depending primarily on two variables: (1) time of day e.g., peak or off-peak; and (2) type of trip e.g., personal travel for work, personal travel for non-work purposes, or freight travel. These two variables can be used to categorize travel demand into six travel markets. The case study application focuses on the peak period work (person) travel market.

All costs for providing access are included in the evaluation of costs for accommodating future trips, whether or not the tripmaker bears them directly. Costs may be categorized based on whether or not they have market prices. Market-priced costs include dollar costs borne privately by system users and publicly by transportation or other agencies. Market- priced costs may be categorized as private vehicle costs, public transportation system costs, highway facility costs and safety and security costs. Costs which have no market prices include travel time costs, environmental costs, pain and suffering components of accident costs, and other social costs such as community disruption. They may be borne by system users (e.g., travel time costs) or externally (e.g., environmental costs).

Dollar value estimates of many of these costs may be found in the literature, as indicated in Table 1. However, there are other social costs for which it is unlikely that dollar values can be developed -- they will simply have to be listed with estimates of their orders of magnitude for consideration in trade-off evaluation in the decision-making process. Examples of these impacts are: national defense implications for protection of oil sources, community cohesion or disruption, community pride, aesthetics, accessibility of disadvantaged segments of the population, loss of cultural, historic, recreational and natural resources, loss of open space and depletion of non-renewable energy resources.

Cost parameters used in the application example presented in this paper are based on values shown in Table 1, with appropriate adjustments as presented in Table 3 for IVHS alternatives.

table 1. Example unit costs

The adjustments account for cost increases due to IVHS technology (both publicly and privately borne) and cost savings from reduced accidents and reduced needs for new highway lanes. More detailed methods for calculation of costs could certainly provide more accurate estimates of costs. The purpose of the example is simply to demonstrate how the approach may be used in real world situations, and not to provide definitive answers about the cost-effectiveness of the alternatives evaluated.

The basic process for computation of costs is indicated in Figure 1. The process relies heavily on output from the four-step travel demand modeling process (9) , both for the base year condition as well as for future year alternatives. As Figure 1 indicates, the outputs from the travel models needed for input into the costing procedures are the following, for each person travel market: (1) person trips by mode (from mode choice); (2) travel miles (from trip assignment) by mode -- person miles of travel (PMT) on transit line-haul and transit access modes, as well as vehicle miles of travel (VMT) on the highway system; and (3) travel minutes (also from trip assignment) by mode. As Figure 1 indicates, the travel measures output from the travel models are input into cost models which provide unit cost parameters for the various cost components. Unit costs may be costs per trip, per PMT, per VMT or per minute of travel time, as indicated in Table 1.

figure 1

The case study urban area was Washington, DC. A previous study (10) provided model output data. In cases where needed travel parameters were not available from the study report, national averages from the Nationwide Personal Transportation Study (NPTS) were used (11) . The Washington, DC study involved analysis of the systemwide travel and transportation system impacts of two alternative urban development patterns for the year 2010. The first alternative (BAL) promoted a closer balance between housing and employment growth, both regionwide and within individual "employment growth" subareas within the region. The second alternative (CONC) maintained regionwide balance between housing and employment as in the first alternative, but concentrated employment in areas with good transit service and significant levels of transit use at the job end of the work trip. The study also provided a base model run for 1995. To demonstrate the application of the least total cost approach with IVHS alternatives, two new alternatives were developed by the author. Both built upon the concentrated (CONC) alternative. The first alternative, IVHS(S), assumed use of only supply-enhancing IVHS technologies such as technologies which smooth the flow of highway traffic, provide priority to transit vehicles, provide real- time information to highway and transit users, provide new services e.g. single-trip carpooling, and enhance highway and transit safety. The second alternative, IVHS(D), added to IVHS(S) by managing demand through pricing mechanisms for peak use of highways.

The travel data and results of the cost analysis are presented in Table 2. A comparison of total costs which were calculated by the spreadsheet suggests that the concentrated (CONC) alternative has lower total costs than the balanced alternative (BAL). Based on the liberal use of cost and travel demand assumptions for IVHS by the author, the IVHS(S) scenario could save about $400,000 daily in aggregate mobility costs relative to the concentrated (CONC) scenario. For the IVHS(D) scenario, the savings would be about $5.7 million daily. Public agency costs (for highways and for public transportation deficits assuming a 40 % farebox recovery rate) would be about $244,000 lower daily under IVHS(S) and $3.4 million lower daily under IVHS(D). As indicated earlier, the cost totals include only those cost items included in Table 1, and exclude some non-monetizable environmental and social costs. Many of these costs are primarily related to auto travel. Since the IVHS(D) scenario involves much less auto travel than the other scenarios, additional savings in non- monetizable environmental and social costs may be expected.

Table 2 also indicates that, while providing mobility currently costs about $5.90 per work trip (including all cost items listed in Table 1), the cost per new trip added by 2010 will be significantly higher under all future alternatives except for IVHS(D). Average cost per added trip amounts to $10.35 under the balanced scenario, $10.03 under the concentrated scenario and $9.54 under the IVHS(S) scenario, but only $3.00 under the IVHS(D) scenario.

Note that Table 2 includes a line item for "negative costs". These are the additional user benefits for the BAL, CONC and IVHS(S) alternatives relative to the IVHS(D) alternative, reflecting primarily the consumer surplus enjoyed by single occupant vehicle (SOV) drivers who are tolled off by the IVHS(D) alternative. This consumer surplus is calculated by multiplying the number of SOV drivers tolled off by half the tolls they would have had to pay. The IVHS(D) alternative is assumed to cause shifts of SOV drivers to other modes only, since work trips are not very likely to shift out of the peak periods during which tolls apply due to limited flexibility of work start and end times. (Note that there may be some debate as to whether the consumer surplus losses suffered by tolled off SOV drivers have already been accounted for through the higher travel times on the HOV and transit modes which are included in the "positive" cost totals. The excess travel time costs incurred by SOV drivers who shift modes may need to be subtracted if their consumer surplus losses are included as negative costs. The spreadsheet has not been set up to do these calculations at this time.

table 2. unit cost changes for ITS

Least Cost Conclusions

This paper has explained the theory in support of a least total cost approach to compare transportation investment alternatives across modes, and to compare significant changes in management and land use policies. The approach is based on assessing the relative economic efficiency of alternatives by determining which alternative involves the least total cost for providing access for various travel markets. The approach has been demonstrated through application of a simplified analysis technique using a LOTUS 123 spreadsheet. Results from the analysis have been presented for demonstration purposes only. The application of the approach to the case study suggests that the approach can be a useful tool for comparison of multimodal investment, IVHS, management and land use policy alternatives.

table 3. total weekday travel demand

table 4. total travel costs per week day

table 5. peek period work trip costs per week day

table 6. peek period truck trip costs per week day


The author would like to acknowledge the valuable comments on an earlier draft of this paper by James Saklas and James March of the Federal Highway Administration, and Ronald Fisher of the Federal Transit Administration. However, the views expressed in this paper are those of the author alone, and do not necessarily represent the policies of the Federal Highway Administration or the U.S. Department of Transportation.


  1. Characteristics of Urban Transportation Systems, US DOT, Pub. no. DOT-T-07, 1992.
  2. DeCorla-Souza, Patrick and Anthony Kane, Peak Period Tolls: Precepts and Prospects, Transportation, 19: 293-311, 1992.
  3. Charles River Associates, Transit Deficits: Peak and Off-Peak Comparisons, CRA Report no. 784-30C, prepared for UMTA, April 1989.
  4. FHWA, Federal Highway Cost Allocation Study , Appendix E, May 1982.
  5. Litman, Todd. Transportation Cost Survey, February 2, 1992. 113 Decatur, Olympia, WA 98502. (206) 943-9025.
  6. MacKenzie, James, Roger Dower and Donald Chen. The Going Rate: What it Really Costs to Drive. World Resources Institute. June 1992.
  7. FHWA. The Costs of Highway Crashes. Pub. no. FHWA-RD-91-055. October, 1991.
  8. DeCorla-Souza, Patrick and Ronald Jensen-Fisher. Comparing Multimodal Alternatives in Major Travel Corridors. In process of publication by the Transportation Research Board. Paper No.940643.
  9. FHWA, National Highway Institute. Introduction to Travel Demand Forecasting. Course Notebook. Text for NHI course no.15254.
  10. Metropolitan Washington Council of Governments. Transportation Demand Impacts of Alternative Land Use Scenarios. May 31, 1991.
  11. FHWA. Summary of Travel Trends. Pub. no. FHWA-PL-92-027. March 1992.
  12. Hanson, Mark. Automobile Subsidies, Land Use and Transportation Policy. Occasional Paper no. 32, Department of Urban and Regional Planning, University of Wisconsin. July, 1989.
Updated: 3/25/2014
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000