“Induced travel demand” is the additional travel and traffic generated by improvements in transportation capacity and service levels. Induced travel demand results in new travel above and beyond the increase in travel that was originally forecast in the planning process, as a result of changes in travel route, travel mode, time of travel, amount of travel, and location. Transportation planners have developed ways of predicting the induced demand that results from all of these changes. Governments should take induced travel into account in transportation planning.
REALTORS® are interested in existing property, of course, but there is also good reason for them to be interested in the development of new property and in community growth generally. New transportation infrastructure benefits communities, among other ways, by providing opportunities for real-estate development. Real estate development, in turn, often leads to an increased demand for transportation--whether in the form of new roads, expanded roads, or various kinds of mass transit.
The creation of new transportation infrastructure, however, often has the paradoxical effect of creating further transportation demand above and beyond that predicted in the planning process. This phenomenon is known as “induced travel demand”: the additional travel and traffic generated by improvements in transportation capacity and service levels.
Induced demand is a multimodal concept; the phenomenon occurs with motor vehicle traffic, public transit ridership, and walking and bicycling. The idea is that when roadways are widened (or faster transit is implemented) people choose to make more trips or travel farther than they did before.
But the fact that induced travel demand tends to work against the desired efficiency of newly developed roads or mass transit does not mean that the creation of new transportation infrastructure should be avoided. Demand for new transportation cannot simply be ignored. Instead, planners must take induced travel demand into account as they decide what kinds of new transportation are necessary. At the same time, they should take advantage of transportation demand management (see “Managing Transportation Demand Puts Livable Communities Within Reach”) to make sure that shifts in the use of one kind of transportation are properly balanced by accompanying shifts in other kinds; and they should adjust land use plans to alter traffic as desired.
What Causes Induced Demand?
Induced travel results from two primary factors. First, in most places today there is a significant amount of latent demand for travel--trips that people do not make because they will take too long or cost too much. For example, a skier might decide not to ski on a certain Saturday knowing that the highway leading to the ski resort will be crowded. Or a family might choose a less desirable daycare facility because it is close to their home whereas a better daycare would require too much driving and increase the cost of gasoline. If transportation improvements are made (widening the highway to the ski resort, or building a light rail line that provides access to the more distant day care facility) people may decide to make trips they previously avoided, or to make longer trips, both of which would increase total travel.
Second, in a growth environment where new homes and businesses are being built, “live and work” location choices take into account current and future travel times, congestion and similar factors. One effect of a new roadway or transit line is reduced travel times in specific corridors, encouraging the development of new homes and businesses farther from existing city centers.
Induced travel is a public policy concern for several reasons. By generating additional traffic, a widened highway might provide much less congestion relief for existing travelers than was expected, raising questions about whether the investment was beneficial. By providing high-quality, high-speed access to areas outside a city, a commuter rail line might encourage suburban growth far from jobs and services (paradoxically increasing auto traffic). For many years, as settlement moved westward across the country and later as we built our cities and suburbs, increased travel was seen in positive terms--a sign of progress. But today, with rampant highway congestion, high cost of travel in family budgets, a national dependence on imported petroleum and emerging issues associated with climate change, we are more likely to question the wisdom of inducing new travel. Serving new travel demand is one thing; causing it might be something else entirely.
Sources of Induced Demand
There are five primary sources of induced travel demand. These apply to all modes (auto, transit, pedestrian, bicycle), but only auto and transit induced travel have been studied extensively. They are presented here in the context of auto travel for purposes of simplicity.
Changes in travel route. The roads and streets of an area comprise a network. We tend to ignore this fact and focus only on specific corridors, or even just on specific segments of corridors, but every street is connected into a larger network. Some networks are more complete than others, but in most networks there will be alternative parallel routes for any specific corridor we might consider widening or otherwise “improving.” When average speeds (or travel flow) are increased (by reducing congestion, or by eliminating intersections, etc.) daily traffic on the local network will readjust to the new condition. Part of that adjustment process includes the shifting of travel from alternative parallel routes to the newly improved route. This occurs very quickly, often in a matter of days.
Changes in travel mode. The relationship between the travel times of different modes is an important factor in determining the “mode share” in an area or in a corridor. (Mode share is a term used to describe the percentage of trips made by each mode.) One sure way to increase transit mode share is to speed up transit connections between a set of origins and destinations by implementing express bus service, installing special bus lanes, or building a rail line. Conversely, when investments are made in a street or highway to increase traffic flow and speeds, the result can be a shift from transit back to driving. This is referred to as an induced mode shift. Similarly, one reason people carpool is to alleviate the difficulty of making long trips in congested corridors by driving alone. However, the extension of a new freeway without HOV lanes can reduce the incentive to carpool, raising the single occupant vehicle (SOV) mode share and increasing traffic.
For walking and bicycling, trip length is an important factor in potential mode shifts. People will walk or bicycle only relatively short distances for most trips (generally less than a mile for walking and less than five miles for bicycling). So the tendency for travelers to shift travel mode in response to changes in traffic corridors is limited to shorter trips for the pedestrian and bicycle modes. However, pronounced mode shifts away from walking and biking can occur where the impact of road construction is to make walking and bicycling less safe and convenient--a common outcome.
Changes in time of travel. In small cities and young suburban areas, traffic flow peaks during daily “rush hours”--usually 7:30 a.m. to 8:30 a.m. and 4:30 p.m. to 5:30 p.m. However, in larger cities and metropolitan regions, this peaking effect is spread over longer periods of time. This occurs because as traffic grows and congestion increases, some travel during peak periods is discouraged. Popular perception is that rush hour traffic is made up primarily of commuters going to work or going back home from work. Actually, however, in many cities commuting represents less than 25 percent of daily travel and less than 50 percent of traffic in the afternoon peak period.
The remaining travel--shopping, social and recreational travel, deliveries, school-related trips and so forth--represents “discretionary” travel for which drivers have more flexibility to choose their time of travel to avoid congestion. In congested areas, some of these trips shift to “shoulder periods” earlier in the afternoon or later, in the early evening. As a result, in larger urban areas, the afternoon “rush hour” actually lasts as long as three to four hours. When highways are widened or new freeways are built, some of that shoulder travel shifts back into the peak period. Because most traffic engineering is focused on providing capacity for peak travel periods, this time shift is an important form of induced travel that causes the public and policy makers to seek further street and roadway expansion. This is troubling to traffic engineers who know that concentrating traffic in peak periods represents an inherently inefficient use of the public’s roadway investment.
Changes in amount of travel. Humans are inherently mobile beings--a result of shared DNA reflecting our heritage as hunter-gatherers who roamed over large areas daily and seasonally. And, our society and economy require high levels of mobility and transportation service to run the way we want them to. As a consequence, there is almost always a significant amount of latent demand for travel, even in rural places. What prevents latent travel demand from becoming actual travel is primarily time and cost. So when travel times are reduced by construction of new routes or delay is (temporarily) reduced by widening or otherwise improving existing routes, some of the latent demand for travel becomes actual travel.
This conversion of latent demand to new traffic is a major source of induced travel. To get some idea of how significant it is, the graphic below shows the growth in U.S. population and in total U.S. daily vehicle miles of travel (VMT) over the past half century. This substantial increase in per-capita VMT is in part a result of roadway investments and induced travel.
Changes in location. Every day in the United States millions of “location choices” are made: where to live, where to work, where to send the kids to school, where to shop, where to invest in commercial real estate, where to lease office space, and so forth. Since at least the 17th century economists have known that access, connectivity, travel distances, and travel costs play major roles in determining land values and development potential. When new transportation facilities are built that improve connectivity or reduce travel times, they “open up” new areas for development. When freeways or commuter rail lines are extended out from cities, new possibilities are made available to people for places to live and work. As we now know, one result of the past hundred years of transportation investment in this country has been the expansion of suburban and exurban development, including new residential subdivisions and commercial sites farther from existing city centers. For most of our history, this has been seen as a good thing--as progress. However, the irony that many people are beginning to see is that much of this investment was originally justified to “alleviate congestion” in existing corridors for existing travelers. In practice, however, the induced travel associated with location decision-making has consumed much or all of the new capacity in many if not most of these corridors.
Measuring and Forecasting Induced Demand
In order for cities and regions to grow according to plan, it is important for planners and policy makers not only to understand induced travel demand but to be able to quantify and forecast it. In fact, most transportation planning and most urban land use planning has ignored induced demand. This has occurred in part due to the structure of the planning process and how transportation investment decisions are made. It has also in part been due to technical shortcomings, which are described below.
Changes in travel route. The redistribution of traffic over a network in response to changes in capacity and travel speed in certain corridors is something we know how to estimate. In fact, almost any city or urban region has the capability of doing this using “traffic models” that are readily available and in general use. (See text box.) However, many roadway projects are planned and designed as corridor improvements, ignoring their impacts on the larger network. Federal transportation policy has discouraged this myopic approach for major projects since at least 1991, but many smaller projects--especially the widening of arterial intersections--escape the planning process without ever being subjected to network-level analysis or traffic modeling.
Changes in travel mode. It is common for large cities and metropolitan regions to develop more robust transportation models that forecast not only motor vehicle traffic, but also transit ridership. These more complex models include equations that predict how much travel demand will be captured by transit. (Usually such models are developed in connection with planning for rail transit systems.) Theoretically, these models could be used to forecast the reverse change—the increased capture of travel demand by the automobile mode in response to improved traffic speeds and reduced traffic delay. In practice, this is almost never done. So mode share analysis is done in connection with evaluation of future transit investments (where federal policy requires it), but not in connection with evaluation of future roadway investments (where federal policy does not require it).
How Traffic Models Work
Transportation planners forecast travel demand in response to land uses using computer software that we refer to as “traffic models.” Planners input population forecasts and other demographic data into the models, assigning the numbers to “traffic analysis zones,” and then estimate the travel demand that will result. Algorithms are used to assign these trips to a network, resulting in forecasts of future traffic in specific corridors.
Changes in time of travel. Most urban transportation demand analysis (traffic modeling) focuses on “peak travel periods.” Thus, the evaluation of traffic demand in relation to roadway capacity addresses a PM peak hour or a PM peak period (e.g., 4 p.m. to 6 p.m.). In virtually all cases, the percentage of travel occurring in this peak travel period is static in the traffic model: once estimated, it is not later adjusted. So although we know that people adjust their time of travel in response to congestion and delay, we generally do not take that fact into account in designing our traffic models and forecasting travel demand.
Changes in amount of travel. An early step in the transportation planning process involves the estimation of “trip generation” associated with specific land uses. This is a significant characteristic of traffic models because it determines how much traffic the model predicts will be produced in the future based on forecast land uses. For example, a traffic model may be designed to predict that each single-family residential dwelling will “generate” eight daily vehicle trips somewhere on the network.
Unfortunately, the trip generation inputs used in most traffic models are static and are not adjusted to reflect levels of congestion or other factors. In fact, many traffic models do not even use local data for trip generation, relying instead on national tables published by the Institute of Transportation Engineers or other sources. So although we know that there is latent demand for travel and that people will adjust how much they travel in response to congestion and other variables, we do not take that into account in designing our traffic models and thus we do not take this into account in developing transportation improvement programs.
Changes in location. The final category of induced demand is the impact of the transportation system on land development patterns and, then, the impact of land development patterns on traffic and travel. A number of specific factors are at work here. At the most obvious level, when a commuter rail line or freeway is extended into the countryside outside a city there will be increased demand for residential development near the stations or interchanges. Of course, other more subtle forces are also at work. For example, if a state Department of Transportation has rigorous rules and procedures governing access management (the addition of new driveways and intersections), that will shape how commercial development responds to highway corridor investments. There might be less of a tendency for commercial strip development to occur, which might in turn encourage more of a village pattern for commercial businesses, which in turn will affect where traffic goes.
However, the major factors driving induced demand will be the influence of new or expanded facilities on where people choose to live, where employers locate new office complexes or industrial facilities, and where commercial enterprises choose to locate stores and businesses. These in turn will produce new traffic at new locations that will consume at least part of the new capacity provided by the new or expanded facilities.
Transportation planning in the United States generally takes account of the upper set of relationships but not the lower. We can readily predict how the transportation system will respond to land use patterns because we control the outcome: we decide publicly what corridors to build, where to provide interchanges and so forth. It is more difficult to forecast how development trends will respond to transportation investments. The response of land development patterns to transportation investment is the sum of many decisions made by many land owners and developers. These decisions are influenced not only by transportation spending, but by economic conditions, land use characteristics (topography, etc.), and other infrastructure programs (especially sanitary sewer line placement), as well as by a wide range of poorly understood factors--culture, development inertia, mortgage lending trends and so forth.
Policy Implications of Induced Travel Demand
At a time when relationships between vehicular travel trends and such issues as climate change, energy dependency, and air quality are increasingly important, the implications of induced travel demand have taken on greater importance. Most highway investment programs are presented to the public in terms of “congestion relief” or traffic alleviation. What if the actual impact of building new or expanded highways is to increase traffic and not to alleviate congestion?
The policy implications of induced demand are tied in part to the objectives of transportation investment in the first place. If the purpose of a particular project or program of projects is to increase land development or redevelopment in a particular area, then induced traffic would presumably be expected and accepted as an outcome. However, if a project or program is promoted based on congestion alleviation, that particular benefit may not be realized and the project or program may be judged to have failed.
Some researchers have taken pains to make it clear that even when the impact of a transportation capacity investment is to produce unanticipated development and induced traffic, someone may still benefit. For a short period of time, traffic may flow better. Maybe a landowner is able to sell land for development and use the proceeds to send family members to college. Perhaps housing contractors will have more demand to build new houses and that will create jobs. So the policy implications of induced travel demand are complicated not only by technical challenges associated with measuring and forecasting but also by different perspectives on how desirable or undesirable the outcomes of induced traffic are.
Generally, policy makers can agree on at least one principle: local, regional, and state governments should take induced travel into account in planning and should make transportation decisions in the light of factual information about the secondary effects of transportation investments.
Examples and Resources
Blueprint Planning Processes
One of the more interesting and productive approaches to rationalizing transportation planning in light of induced demand considerations is the Blueprint Planning Process implemented by the State of California. Initiated in 2005, the Blueprint Planning process was designed to enable public officials and other participants to more realistically evaluate future land use patterns and their potential impacts on the region’s transportation system, housing supply, jobs-housing proximity and balance, environment, and natural resources. This integrative approach to land use and transportation planning is also known more generally as “scenario planning.”
California initiated a grant program to transportation planning regions of the state that was designed to create a planning process that would better inform regional and local decision-making through proactive engagement of all segments of the population to foster consensus on a vision and preferred land use pattern. Most importantly, the process was designed specifically to address potential land use responses to different transportation investment programs and the implications of resulting land development patterns for the efficacy of transportation investments. The process was also designed to take into account the regional nature of these issues by allowing and funding collaboration between neighboring regional transportation planning agencies. Some examples of Blueprint Plans produced through the California process can be found at these websites:
The Southern California Association of Government’s “Compass Blueprint” process.
The Sacramento regional approach to Blueprint Planning.
A useful overview of “scenario planning” can be found in a paper written by Keith Bartholomew.
The following are websites and research reports that are available for more in-depth discussions of induced travel demand.
Cervero, R. 2001. Road Expansion, Urban Growth, and Induced Travel: A Path Analysis, University of California-Berkeley.
Fulton, L., Meszler, D., Noland, R. and Thomas, J. 2000. A Statistical Analysis of Induced Travel Effects in the U.S. Mid-Atlantic Region, Journal of Transportation and Statistics.
Goodwin, P. 1996. “Empirical Evidence on Induced Traffic: A Review and Synthesis,” Transportation, Vol. 23, pp 35-54.
Hansen, M. and Huang, Y. 1997. “Road Supply and Traffic in Urban Areas: A Panel Study.” Transportation Research, Vol. 31A, pp. 205-218.
Heanue, K. 1997. “Highway Capacity and Induced Travel: Issues, Evidence and Implications.” Transportation Research Circular, Vol. 418, pp. 33-45.
Johnston, R. and Ceerla, R. 1996. “The Effects of New High-Occupancy Vehicle Lanes on Travel and Emissions.” Transportation Research, Vol. 30A, No. 1, 35-50.
Litman, T. 2009. “Generated Traffic and Induced Travel: Implications for Transport Planning.” Victoria Transport Policy Institute.
Noland, R. 1999. “Relationships Between Highway Capacity and Induced Vehicle Travel.” Washington, D.C. Paper presented at the 38th Annual Meeting of the Western Regional Science Association.
Noland, R. and Cowart, W. 2000. “Analysis of Metropolitan Highway Capacity and the Growth in Vehicle Miles of Travel.” Washington, D.C. Paper presented at the 79th Annual Meeting of the Transportation Research Board.
Transtech Management, Inc. and Hagler Bailly. 2000. “Assessing the Issue of Induced Travel: A Briefing on Evidence & Implications from the Literature.” Washington Metropolitan Council of Governments.