Induced travel effects include changes in land development and the location of households and employment, the number of trips made (trip generation), destination choice (trip distribution), mode choice, route choice (traffic assignment), and departure time choice (Transportation Research Board, 1995; Transportation Research Circular, 1998).

The U.S. Environmental Protection Agency (EPA) recently conducted a review of the induced travel literature for their Science Advisory Board and concluded that this research "has not only built a strong case for the existence of induced travel effects, but in some cases suggests that a large fraction of growth in VMT is directly attributed to increases in road capacity" (2000).

One of the difficulties of testing the induced travel hypothesis has been controlling for confounding economic activity variables such as population, income, and other demographic trends. Much of the recent induced travel research, however, has attempted to controll for these variables and has not been able to reject the hypothesis of induced travel (Goodwin, 1996; Hansen and Huang, 1997; Noland and Cowart, 2000; Chu, 2000; Fulton et al., 2000; Noland, 2000). The results of this research have yielded fairly consistent long-term elasticities of VMT with respect to roadway lane miles. See Table 5 below.

Thus, for example, a 10% change in highway lane miles would be expected to result in a 5% to 10% increase in VMT for a metropolitan region in a twenty year time horizon.

As described in the section above, the results of WFRC travel demand model (May 2000) in many cases have negative elasticity of VMT with respect of lane miles and positive elasticity of VMT with respect to travel time. Again, this is inconsistent with travel theory and the evidence for induced travel.

In some cases, however, the elasticity of VMT with respect to lane miles and travel time was consistent with travel theory and the evidence for induced travel. In the cases of the elasticity of VMT with respect to lane miles, the results are consistent with low estimates in the empirical literature. However, the elasticity of VMT with respect to travel time is significantly lower than the figures found in the empirical literature, which reports a range between –0.3 and –1.0.

The latent demand analysis in the FEIS (pg. 1-38) is irrelevant. It is the job of transportation planners to assist in planning for and directing travel demand so that the values and goals of the community can be met in a cost-effective manner. It would be financially impossible and irresponsible to provide unlimited transportation facilities to the community.

The FEIS also suggests that induced demand would not be an important factor in the North Corridor because it is so congestion (pg. 1-38). The theory of induced demand predicts that elasticities will be larger in very congested regions. Thus, the effects of induced demand would be greater in corridors that experience significant congestion.

In sum, the incorrect representation of induced travel in the WFRC travel demand model would result in the underestimation of VMT, emissions, and congestion in the highway alternatives compared to the No build alternative.

Land Use Effects
The National Academy of Sciences/National Research Council (1995) concluded that

Highway capacity additions can improve access to developable land in outlying areas of a metropolitan area. The improved access makes these areas more attractive for future development and influences the location decisions of residents, employers, and shopping facilities. Shifts in the location of residences, jobs, and shopping opportunities affect trip distances and the potential for trips to be made by modes other than the automobile.

A report in the U.K that is equivalent to the above report, Trunk Roads and the Generation of Traffic by the Standing Advisory Committee (SACTRA, 1994), concluded that "the land-use changes consequent on improved access are likely, in turn, to lead to changes in the patterns of travel, car dependence, and the volume of traffic."

The U.S. Environmental Protection Agency (EPA, 2000) recently conducted a review of the literature for their Science Advisory Board. The review cites evidence from Gordon and Richardson (1994):

…various long run effects can have a significant impact on total VMT growth. Long run effects occur due to changes in relative accessibility within an urbanized area and can result in the spatial reallocation of activities. If speeds are higher, many residences and businesses will tend to relocate over time, often resulting in longer distance trips.

All of these reviews cited above were conducted to address the issue of induced travel resulting from new highway capacity expansion. Each review found compelling evidence for induced demand and concluded that the land use and transportation interaction may be a significant contributor to induced travel.

It is also important to note that a U.S. District Court Case in the Chicago, Illinois, region has held that the National Environmental Policy Act requires the consideration of land development changes when a new freeway segment is analyzed in an environmental impact statement (Sierra Club, et al. v. U.S. DOT et al., No. 96 C 4768, U.S. Dist. Ct. for the N. Dist. of Illinois, E. Div., Jan. 16, 1997).

A single set of land use projections is used in the analysis of the alternatives for the FEIS. Because the analysis does not account for the land use and transportation interaction, it would tend to underestimate congestion and VMT in the highway capacity expansion alternative and overestimate congestion and VMT in the no build alternative. The analysis is biased in favor of the highway capacity expansion alternatives. The peer review of the WFRC model recommended that a Delphi Review of land allocations be conducted immediately to develop land use projections specific to transportation alternatives.

In response to our concerns about the land use projections used in the FEIS, UDOT and WFRC officials state that they have spoken to their local planners and that these planners told them that the highway alternatives in the FEIS would not have any land use effects. This position is documented in the FEIS. These officials informed us that they believe that this approximates a Delphi review of land use allocation and is a valid method to assess the land use effects of the proposed highway project.

First, we want to make clear that a Delphi review of land allocations is not the preferred method to represent the land use and transportation interaction. Formal land use models should be applied to evaluate the land use effects of the proposed transportation alternatives. One approach would be to use a land use simulation model, UrbanSim, with a greatly improved WFRC travel demand model. However, this model has not yet been implemented in the region and it is unclear when it will be. Alternatively, an integrated land use and transportation model such as MEPLAN or TRANUS could be used. Both MEPLAN and TRANUS have been implemented in several places in the U.S. and in many places throughout the world. MEPLAN or TRANUS could be calibrated and ready to use in less than six months. Given the magnitude and importance of the decisions at stake, and the amount of time and effort already invested, we believe that it is more than reasonable to apply formal land use models to the analysis of alternatives in the FEIS.

Second, UDOT and WFRC's assertion reflects a misunderstanding of the Delphi review method and of valid scientific methods of inquiry in general. The Delphi review method is described in the National Cooperative Highway Research Program (NCHRP) Report 423A (1999):

Experts provide their judgements about likely future events or the impact of potential transportation investments and programs by responding to several rounds of questionnaires. The Delphi moderator summarizes the results of each round and submits these summaries to the experts for reconsideration of their analysis. In this way the thinking of the experts are shared with each other to either arrive at consensus or clarify differences of opinion.

By definition, the method used to account for the land use effects of the alternatives in the FEIS is not a Delphi review.

Valid scientific inquiry requires the application of systematic methods to the question at issue (in this case, the effect of the transportation alternatives on land use patterns). Such an inquiry should employ valid instruments (e.g., formal interviews or questionnaires) and should allow for replication. The method applied and the results should be well documented. In this way, peers can evaluate the validity of the methods and results based on scientific standards.

The approach used to account for the land use effects of the alternatives in the FEIS fails to meet any of the specified criteria for valid scientific inquiry. The approach employed no replicable instruments (i.e., it used informal questions to officials) and was not documented. Thus, the "results" are invalid.

Officials reviewing the FEIS would not consider the results of informal interviews with local transportation planners as credible evidence of the travel and emissions effects of the proposed alternatives. It is difficult to understand how this could possibly be considered credible evidence of the land use effects of the proposed alternatives.

Methods to Address Land Use Effects and Potential Results
As discussed above, formal land use models should be applied to determine the land use effects of the alternatives in the FEIS. The MEPLAN model has been calibrated in the Sacramento, California, region by researchers at the University of California at Davis and a similar model, TRANUS, has been applied in Oregon, Baltimore, Maryland, and Sacramento. We briefly describe the MEPLAN model and provide the results of a study that suggest the potential significance of the land use effects of highway alternatives.

The basis of the MEPLAN modeling framework is the interaction between two parallel markets, the land market and the transportation market. Behavior in these two markets is a response to price signals that arise from market mechanisms. In the land markets, price and generalized cost (disutility) affect production, consumption, and location decisions by activities. In the transportation markets, money and time costs of travel affect both mode and route selection decisions.

In a recent study, Rodier, Johnston, and Abraham (2000) conduct sensitivity analyses using the Sacramento MEPLAN model to evaluate the potential importance of the land use induced travel effects in the Sacramento, California, region. The model is used to simulate a future base case scenario (low-build) and a beltway scenario for 25- and 50-year time horizons. The study found that the implied elasticity of vehicle miles traveled with respect to lane miles in the Sacramento MEPLAN model compared well to the empirical literature. The study found that the land use effects accounted for approximately half of the elasticity of VMT with respect to lane miles.

As a fair compromise on our part, we suggested that a Delphi review of land allocation be conducted for the alternatives in the FEIS. This is not the preferred option. As stated before, this was also an immediate improvement recommended by the peer review. This is a relatively inexpensive and quick method to account for the land use and transportation interaction.

The NCHRP 423A (1999), Land Use Impacts of Transportation: A Guidebook, provides a good description of the Delphi method and its potential application to evaluate the land use effects of highway alternatives. Our discussion is based largely on this source.

The Delphi review is a systematic method to use expert opinion. The Rand Corporation developed the method for defense purposes in the 1950s. In the 1960s the method was made available to the public, and since then it has been used relatively frequently for the purposes of forecasting and policy analysis.

The Delphi method is generally described as follows:

The Delphi process differs from a panel meeting face-to-face in that participants are anonymous, the process is done iteratively with controlled feedback, and a statistical group response is reported. Anonymity allows participants to focus on the issues, not the personalities of the participants. The repeated rounds with feedback from the moderators allow participants to reconsider their responses in light of new information but prevents lobbying for any point of view. The statistical group response gives the range of opinion as well as the most common response. This helps clarify how strongly people agree or disagree (Linston and Turoff, 1975, Cavelli-Storza, 1982; ctd. NCHRP, 1999).

The Delphi panel should consist of 8 to 12 people who encompass a wide range of opinions on the subject of the land use and transportation interaction (i.e., experts on land use policy and market conditions). More specifically, Bajpai (1990) recommends that the panel include "local government officials, land use and transportation planners, utility company representatives, school district officials, neighborhood and citizen group members, private consultants, academics, and business representatives" (NCHRP, 1999).

The NCHRP report describes many successful applications of the Delphi method by metropolitan planning agencies that are not published in the literature (but contact people are provided). We briefly describe the results of two studies that have been reported in the literature. The first is the a study in San Jose, California, that used the Delphi method to project the land use effects of three transportation alternatives (highway, bus, and rail) (Cavelli-Storza, 1982). The results indicated that the land use effects varied reasonably by alternative (i.e., development was more defused in the highway scenarios compared to the bus and rail scenarios). The second was conducted in Edinburgh, U.K, to examine the land use effects of a road pricing and light rail (Still, 1997). Again, the land use results varied by alternative; however, the variation was small compared to the results of simulations with land use models. This may be explained by the fact that the Delphi method was stopped after the second round because of resource limitations (this was a Ph.D. dissertation project).

Another relatively inexpensive and quick method to address the land use effects of alternatives would be the use of a rule-based urban growth model that runs in geographic information system (or GIS). Such a model can allocate several land uses, according to the user-set rules, and uses commonly available data layers. It permits users to run a variety of scenarios testing policies for new roads or rail lines, urban growth boundaries, and compact growth incentives. Rules could be set using the Delphi method described above. But if they were not, at least the assumptions used in the model to allocate land uses would be applied systematically and described to the public with documentation. This would be a big improvement from WFRC's current method. This type of model has been applied in New Mexico.

The Sierra Club has sponsored the application of such a model (UPLAN) to evaluate the FEIS alternatives. The application of the UPLAN urban growth model in the Salt Lake region showed that future growth in the region tended to locate near existing development and transportation networks. When the proposed Legacy Parkway (South Davis section) was included in the model it showed increased development (approximately 14%) in Southern Davis County (especially near and along the proposed highway alignment).

In sum, there are reasonably available methods to account for the land use and transportation interactions of the proposed alternatives in the FEIS. Moreover, the application of these methods indicates that the land use effects of transportation alternative may be a significant component of travel and emissions analyses.

Auto Ownership and Trip Generation An auto ownership sub-model has been added to the WFRC travel demand model. It was borrowed from Portland, Oregon, and includes an urban design variable, accessibility variables (retail employment within one mile and jobs within 30 minutes of transit access), and other socioeconomic variables (household size, income, and workers per household). However, the accessibility variables are held constant from the No Build to the build alternatives. The revised WFRC model would not represent increased auto trips due to highway project alternatives that reduce the travel time and cost of auto use. As a result, the model would tend to underestimate VMT and emissions and overestimate congestion reduction in the build alternatives.

The peer review committee recommended that the trip generation and trip distribution models for the separate areas of the region be merged. WFRC has merged trip distribution, but it has not merged trip generation. Internal-external trips are not included in feedback from traffic assignment to trip distribution (see discussion below). However, many of these internal-external trips are work trips that should be included in the feedback process. As a result, the model would tend to underestimate VMT, emissions, and congestion from the No Build to the build alternatives.

Trip Distribution
The trip distribution model in the travel demand modeling process requires estimates of travel time in order to estimate the destination of trips. However, travel times depend on the level of congestion on streets in the network. The level of congestion is not known until later model steps are executed (i.e., assignment of traffic on the roadway network). Thus, it is important to feed travel times from the traffic assignment step back to the trip distribution step. This process of recycling travel times once from traffic assignment to trip distribution is commonly called an iteration. Many iterations of a travel model may be required to ensure that the input travel times in trip distribution are consistent with the output travel times from traffic assignment. Modelers must employ criteria to determine when to stop iterating the model. They are called convergence criteria.

The Travel Model Improvement Program's guidance document "Incorporating Feedback in Travel Forecasting" (COMSIS, 1996) recommends and describes several convergence criteria. Some of these include:

• 5% or less change in average system-wide speeds within an iteration
• 95% of the functional classes (or roadway types) by area have a change in speeds of 5% or less within an iteration
• 95% of the links have a change in volumes of 5% or less within an iteration.

The report recommends the use of more than one criterion to determine convergence. Volumes on links and average speeds on links are the two most important variables for determining equilibrium (Stopher, 1993). This is because travel demand models directly relate volume and speed through a functional relationship. Thus, convergence with one of the variables usually implies convergence with another of the variables (Stopher, 1993).

In the U.S., travel modeling studies in the Salt Lake City, Nashville, and Sacramento regions suggest that changes in trip distribution may be a significant effect of induced travel with respect to magnitude of change from the base case and ranking ordering of alternatives (COMSIS, 1996; Johnston and Ceerla, 1996; Rodier, Abraham, and Johnston, 2000). If a travel model is not fully equilibrated on travel times, then its simulations may underestimate VMT and emissions for a transportation plan that includes highway capacity expansions.

The peer review committee recommended that the successive averages approach for convergence be implemented as the feedback mechanism to trip distribution because it is known to consistently converge and it is an efficient feedback method. In this approach, the volumes on each link and each origin destination flow are averaged with previous results.

In the December 1999 WFRC travel demand model, the method of successive averages for feedback was not implemented; instead, the direct method of feedback was used, which is known not always to converge.

The direct method of feedback used in the December 1999 WFRC travel demand model is known not to always converge. We have reported on WFRC's failure to apply formal criteria to determine whether their model has converged (Rodier and Johnston, 1998). We applied several convergence criteria recommended by the Federal Highway Administration's (FHWA) Incorporating Feedback in Travel Forecasting (1996) and found that the model did not meet 2 of 3 convergence criteria (Rodier and Johnston, 1998). FHWA recommends the application of more than one criterion to demonstrate convergence. When a model does not converge, its theoretical validity, and thus its results, are questionable. Convergence of the December 1999 WFRC travel demand model results must be documented to verify their validity.

We have also reported that the direct feedback method has also been incorporated in the WFRC model in an unsystematic fashion (Rodier and Johnston, 1998). The model was programmed to iterate only once; however, this iteration required an input network that contained the results of traffic assignment from a previous model run. This input network could be the product of anywhere between one to four model iterations. If this loaded network is not a product of sufficient runs, then it is likely that the model will not have converged. It is unclear how the feedback mechanism was formalized in the December 1999 WFRC travel demand model. This information must be documented to verify the validity of the results of the December 1999 WFRC travel demand model.

The May 2000 WFRC travel demand model does include the method of successive averages to implement feedback from traffic assignment to trips distribution, as recommended by the peer review. However, the May 2000 WFRC travel demand model does not use AM peak, PM peak, and mid day trip tables to create inputs skims for trip distribution, which was recommended by the peer review committee. To estimate peak period travel times, the revised WFRC model still assigns a daily trip table to the highway network and uses a single peak hour factor, 10% or 12% of average daily traffic, to all links in the region.

In their manual of best modeling practices for metropolitan planning agencies, Harvey and Deakin (1993) state that the use of 12% average daily traffic is a "crude estimate of total bi-directional peak hour travel." Furthermore, "this procedure yields only a rough approximation of link-or corridor-level peaking" and should only be used by smaller MPOs "where the duration and intensity of congestion are limited." Despite this recommendation, WFRC continues to use the 12% factor in a medium-sized region that is using their model for the purpose of addressing growing congestion. Therefore, the use of the 12% factor is clearly not appropriate for use in the revised WFRC model, and AM peak, PM peak, and midday trip tables should be developed.

The use of the crude 10% or 12% factors as congested travel time in trip distribution may be one explanation for the counterintuitive results obtained by the model (described above) and the overly low travel speeds.

Mode Choice
The mode choice model included in the WFRC model set was not used in the evaluation of the alternatives in the FEIS. As a result, average vehicle occupancy is held constant across scenarios. In general, the addition of highway capacity would be expected to decrease average vehicle occupancy because of faster travel times by auto. As a result, the model would tend to underestimate VMT and emissions and overestimate congestion reduction in the build alternatives. The peer review recommended that the mode choice model be included in the WFRC travel demand model for all simulation studies.

Peak-Spreading or Departure-Time-Choice
A peak-spreading or departure-time-choice component is needed in the traffic assignment sub-model to represent how traffic congestion eases and disappears over time. The WFRC travel demand model would currently tend to overestimate congestion in the no build alternative and underestimate it in the build alternative because peak spreading is not represented. A peak-spreading model has been implemented by the San Francisco Bay Area Metropolitan Planning Commission for the Tri-Valley sub-area study. Departure- time-choice models have been implemented in Edmonton, Canada, and in Portland, Oregon.

Summary
In sum, the net effect of the limitations of the WFRC travel demand models would be to underestimate VMT, emissions, and congestion in the highway build scenario compared to no build and transit alternatives. As a result, the analysis of alternatives with the WFRC travel demand model would be biased in favor of the highway alternative and biased against the no build alternative or a transit alternative.

SUMMARY OF TECHNICAL CONCLUSIONS

On July 14, 2000, the Utah Department of Transportation and the Federal Highway Administration approved and issued the Legacy Parkway Final Environmental Impact Statement (FEIS). The FEIS relied on the travel demand forecast made by the Wasatch Regional Council (WFRC) to conclude that the Legacy Parkway highway is needed to meet the future travel demand in the North Corridor.

The authors of this report have reviewed the all the documentation of the Wasatch Front Regional Council (WFRC) travel demand model, the computer files and programs of the WFRC travel demand model, and simulated the FEIS alternatives with the WFRC travel demand model. Based on our reviews and simulations, we have made the following conclusions.

The travel demand forecasts used to determine the need for the proposed Legacy Parkway are critically flawed and unreliable. This conclusion is based on the following findings documented in this report.

• The regional and corridor forecasts for the alternatives evaluated in the FEIS are inconsistent with the basic economic theory used to develop travel demand models (Ben-Akiva and Lerman, 1985, and Small, 1992). For example, when the supply of travel is increased in the model, speeds decrease and less travel is demanded or speeds increase and less travel is demanded. The results of alternatives must be examined to ensure that the model is producing results that are consistent with the theory upon which they are based.

• The sub-models within the WFRC travel demand modeling system produce conflicting results about the effect of the highway alternatives on travel speeds and represent different relationships between travel speeds and vehicle miles traveled.

• The results obtained from the authors' simulations of the FEIS alternatives with the travel demand model that was provided by WFRC are inconsistent with the results documented in the FEIS. It is disturbing that WFRC could not produce a travel demand model capable of replicating the results documented in the FEIS. Moreover, the computer file results from the WFRC simulations that were sent to us are also not consistent with the results documented in the FEIS.

• The simulation of the FEIS alternatives is confounded. Two versions of the WFRC model are used to forecast future travel demand, one version for the No Build alternative and one version for the I-15 and Legacy Parkway alternatives. It is impossible to determine whether the results obtained for the alternatives are due to differences in the travel demand models or differences in the transportation systems represented in the alternatives.

• The travel speeds represented in the WFRC travel demand model are inconsistent with the empirical literature and 20-year forecast of regions that are larger than the Salt Lake City region; that is, they are unreasonably low.

The FEIS incorrectly states that the WFRC travel demand models used in the analysis of alternatives meet the immediate term recommendations of the 1999 peer review. Only approximately half of the immediate term recommendations were implemented in the WFRC travel demand models used in the FEIS. As a result, the WFRC travel demand model cannot be characterized as state-of-the-practice.

The WFRC travel demand model is biased in favor of highway alternatives and biased against transit and no build alternatives. The WFRC travel demand model would tend to underestimate vehicle miles of travel (VMT), emissions, and congestion in the highway alternative compared to the no build or transit alternatives for the following reasons.

• In some cases, the relationships that represent induced travel in the WFRC travel demand model change in the wrong direction; that is, an increase in the supply of transportation facilities decreases the demand for travel.

• In other cases, when this relationship changes in the correct direction, the magnitude of the change can be characterized as low.

• A single set of land use projections is used in the analysis of FEIS alternatives. Formal quantitative and qualitative methods are reasonably available (with respect to both time and cost) to project the land use and transportation interaction of the proposed alternatives. Moreover, research findings included in this report suggest that the land use and transportation interaction may be significant for the alternatives evaluated in the FEIS.

• The WFRC travel demand model does not represent the effect of changes in accessibility among scenarios on auto ownership and thus trip generation rates. As a result, the model would not represent additional non-work trips that may result from the highway alternatives in the FEIS.

• The WFRC travel demand model uses separate trip generation models for different areas of the region and thus the internal and external trips that are work trips will not be included in the feedback process from traffic assignment to trip distribution. As a result, the WFRC travel demand model would tend to underestimate the length of those auto trips in the highway build alternatives.

• One version of the WFRC travel demand model uses an unreliable method of feedback of travel times from traffic assignment to trip distribution. As a result, it is possible that auto trip lengths are underestimated in the highway build alternatives.

• The WFRC travel demand model does not make use of its mode choice model in its evaluation of alternatives and estimation of average vehicle occupancy. As a result, the model would tend to underestimate average vehicle occupancy in the highway build alternatives.

• The WFRC travel demand model does not represent the tendency of travelers to shift the timing of their trips to the shoulders of the peak traffic period when those periods are congestion. As a result, congestion reduction resulting from the highway build scenarios would be overestimated.

As a result of these serious flaws, we can only conclude that the WFRC travel demand model used to evaluate the alternatives in the FEIS is insufficiently accurate for plan or project analysis, or to properly support the analysis and conclusions in the FEIS.

References:

Bajpai, J. N. (1990). Forecasting Basic Inputs to Transportation Planning at the Zone Level. NCHRP Report 328, Washington, DC, TRB, National Research Council.

Ben-Akiva, M.and S. Lerman (1985) Discrete Choice Analysis. The MIT Press. Cambridge, MA.

Cavalli-Storza, V. and L. M. Ortolano (1984). Delphi forecasts of land use: transportation interactions. Journal of Transportation Engineering, Vol. 110, No. 3, May.

Chu, Xuehao (2000). Highway Capacity and Areawide Congestion. Preprint for the 79th Annual Meeting of the Transportation Research Board. National Research Council, Washington, D.C.

COMSIS Corporation (1996). Incorporating Feedback in Travel Forecasting: Methods, Pitfalls and Common Concerns. Office of Environmental Planning, Federal Highway Administration, DOT-T-96-14, Washington, D.C.

Fulton, Lewis M., Daniel J. Meszler, Robert B. Noland, and John V. Thomas (2000). A Statistical Analysis of Induced Effects in the U.S. Mid-Atlantic Region. Preprint for the 79th Annual Meeting of the Transportation Research Board. National Research Council, Washington, D.C.

Goodwin, Phil B. (1996). Empirical evidence of induced traffic, a review and synthesis,Transportation, 23:35-54.

Gordon, Peter, and Harry W. Richardson (1994). Congestion trends in metropolitan areas. In Transportation Research Board's Curbing gridlock: Peak-period fees to relieve traffic congestion volume 2, National Research Council, Washington, DC: National Academy Press: 1-31.

Hansen, Mark and Yuanlin Huang (1997). Road supply and traffic in California urban areas, Transportation Research A, 31:205-218.

Harvey, G. and E. Deakin (1993). A Manual of Regional Transportation Modeling Practice for Air Quality. National Association of Regional Councils.

Johnston, R. A. and R. Ceerla. (1995). Land Use and Transportation Alternatives. Transportation and Energy: Strategies for a Sustainable Transportation System, ed. D. Sperling and S. Shaheen. ACEEE, Washington, D.C.

Linstone, H.A. and M. Turoff (1975). The Delphi Method: Techniques and Applications. Reading, MA, Addison-Wesley Publishing Co.

National Cooperative Highway Research Program (1999). Land Use Impacts of Transportation: A Guidebook. Report 423A. Transportation Research Board, National Research Council, National Academy Press, Washington D.C.

Noland, Robert B. (2000). Relationships between highway capacity and induced vehicle travel. Transportation Research A (forthcoming).

Noland, Robert B. and Cowart, William A. (2000). Analysis of Metropolitan Highway Capacity and the Growth in Vehicle Miles of Travel. Preprint for the 79th Annual Meeting of the Transportation Research Board. National Research Council, Washington, D.C.

National Academy of Sciences/National Research Council (1995). Expanding metropolitan highways: implications for air quality and energy use, Transportation Research Board Special report 245, National Research Council, National Academy Press, Washington, D.C.

Rodier, C. and R. Johnston (1998). Review of the Wasatch Front Regional Council Travel Forecasting Model. Department of Environmental Science and Policy, University of California, Davis, CA.

Rodier, C., J. Abraham, and R. Johnston (2000). Anatomy of Induced Travel Using an Integrated Transportation and Land Use Model in the Sacramento Region. Department of Environmental Science and Policy, University of California, Davis, CA. 1998.

Small, K. A. (1992). Urban Transportation Economics. Harwood Academic Publishers.

Standing Advisory Committee on Trunk Road Assessment (1994). Trunk Roads and the Generation of Traffic. U.K. Department of Transport, December (London).

Still, B. (1997). Transportation impacts on land use: potential methods and their relevance to strategic planning. Pre-print. PRTC European Forum, U.K.

Stopher, P. R. (1993). Predicting TCM Responses with Urban Travel-Demand Models. Transportation Planning and Air Quality II. Proceedings of the National Conference. American Society of Civil Engineers.

USDOT (1997). Our Nation's Travel: 1995 NTPS Early Results.

U.S. Environmental Protection Agency (2000). Induced Travel: A Review of Recent Literature with a Discussion of Policy Issues. Office of Policy. Energy and Transportation Sectors Division, Washington, D.C. http://www.epa.gov/tp/rap.htm

Wasatch Front Regional Council (1999). Travel Demand Model Peer Review. Summary of Proceedings. May 18 and 19.

Wasatch Front Regional Council (2000). Travel Demand Documentation 1999. March.

Wasatch Front Regional Council (2000). Summary of Travel Model Changes. March.