Abstract
As travel costs fall with new capacity, the quantity of travel increases. This concept—induced travel—has profound implications but remains unevenly embraced in practice. Do instructors teach it in transportation engineering classrooms? What explains their pedagogical decisions? Interviews with university instructors revealed remarkable variation. Whereas some featured induced travel as a key takeaway, others omitted the idea entirely. Instructors also varied in their willingness to critique standard engineering practices; some were largely uncritical while others sought to “counteract conventional wisdom.” In justifying their choices, instructors offered a range of overlapping concerns. Those who “believed” in induced travel but did not teach it often lacked expertise in the area and were uneasy teaching “soft” concepts. Because teaching was seen as a lower priority than conducting research, instructors had little motivation to overcome those challenges. Instructors also advanced pragmatic concerns about the need to prepare students for the Fundamentals of Engineering exam and their careers. Instructors who were more skeptical of induced travel wondered whether seemingly new travel was instead shifted or previously suppressed. Some of these instructors argued that even if new travel was indeed induced, engineers still had a responsibility to accommodate it. Finally, the contested language of induced travel can lead parties to talk past each other. “Believers” and “skeptics” sometimes have more in common than initially thought. However, there are still profound disagreements—about induced travel, standard engineering practices, and indeed the very purpose of engineering. In these debates it will be essential to operate from a shared vocabulary.
Widening highways doesn’t fix traffic. So why do we keep doing it? -January 6, 2023 The New York Times (
1
) Induced demand is one of the most irrational theories I’ve ever heard. Correlation is not causation. If the transport system exceeds public travel needs, there will be very little traffic. I support anything that improves traffic, as this negatively affects almost everyone. -Tweet from Elon Musk, December 28, 2019 (
2
)
These quotations exemplify an ongoing debate between “believers” and “skeptics” about induced travel. (Although a believer–skeptic framework masks considerable nuance, it helps introduce the debate. We embrace nuance later.) The foundational economic theory is deceptively simple: Any increase in highway capacity (supply) reduces the generalized cost of travel, especially on congested highways … When any good (in this case travel) is reduced in cost, the quantity demanded of that good increases (
3
).
And yet the believers and skeptics cannot seem to agree on much of anything, with both archetypal camps asking “leading questions in the opposite way” ( 4 , p. 2). The believers ask, “isn’t it futile to keep trying to build our way out of congestion? After all, it will be just as congested as before.” Meanwhile, the skeptics wonder how much of this so-called induced travel was previously suppressed by congested conditions and is now able to be expressed? Moreover, the skeptics inquire, isn’t it our job to get people from A to B quickly and conveniently?
Believers and skeptics hold diverging views about what induced travel implies about the central purpose and practice of transportation engineering. If we take induced travel seriously, the believers argue, the logical foundations of 20th century transportation policy become unstable ( 3 , 5 ). We could no longer justify infrastructure investments with anticipated travel time savings. Moreover, transportation investments would be less about relieving congestion and more about deciding how to grow, a much thornier issue without a straightforward engineering solution. The skeptics tend to take a slightly different view. They contend that the foundations of the predict-and-provide paradigm remain strong and that we should continue to do our best to forecast anticipated travel demand and provide a transportation system that reasonably accommodates that demand ( 4 , 6 , 7 ).
The “believer–skeptic” dichotomy can lead to black and white debates about transportation investments ( 4 ). On the one hand, some “believers” think the existence of induced travel should automatically put an end to any and all roadway widening. On the other, some “skeptics” refuse to admit that any travel is ever induced because doing so would be an admission of defeat. Yet the truth lies between these extremes. We need more conversations about whether, when, and where to expand roadways.
Moreover, given the reality that nearly all transportation investments induce at least some new travel, we need a more careful accounting of the true costs and benefits of our proposed investments. Induced travel should be systematically incorporated into our environmental assessments and benefit–cost analysis so that claims about congestion relief, travel time savings, and emissions reductions are accurate ( 8 , 9 ).
The process of incorporating induced travel into practice is complicated by incompatible definitions and imprecise language. The terms induced demand, latent demand, and generated traffic mean different things to different people and are occasionally used interchangeably (see Gorham for the lack of precision in induced-travel conversations [ 4 ]). Some suggest that the term induced travel should be reserved for truly new travel, not diverted travel. But distinguishing between the two is not cut and dry ( 4 ).
It is no surprise that the idea is often misunderstood; wider roads seem like an intuitive fix for gridlock. Indeed, a 2020 survey found that 66% of the U.S. public and—more troublingly—fully 30% of domestic transportation engineering students–mistakenly believed that widening would relieve congestion in the long term ( 10 ). Induced travel may be even less well understood in some international contexts where the public has less experience with the frustrating aftermath of widening projects. This collective misunderstanding gives political cover to costly ongoing expansions in the United States (for example in Houston, TX, Los Angeles, CA, and Jersey City, NJ [ 1 ]) and abroad ( 11 ). A better, more widely shared understanding of induced travel and its implications for practice would improve outcomes by enabling clearer, more accurate conversations about the costs and benefits of projects.
The purpose of this article is to explore the contours of the induced-travel debate within transportation engineering classrooms. Our hope is that this exploratory effort will serve as a conversation starter for engineering instruction and debates about induced travel. We ask three questions:
To what extent do instructors discuss induced travel or related terms in class?
To what extent do instructors take a critical view of engineering practice?
Among those who omit induced travel and/or rarely critique engineering practices, what explains their pedagogical decisions?
In answering these questions, we draw on a series of in-depth interviews with engineering instructors. The richness enabled by interviews allows us to move beyond a skeptic–believer binary and to deeply explore instructors’ concerns and motivations for their pedagogical decisions. The qualitative approach used here complements previous analyses of transportation engineering pedagogy based on surveys of instructors ( 12 – 14 ) and students ( 10 , 15 ).
Background: Defining Induced Travel
Before detailing the methodology and results, we begin by defining our terms. We use “induced travel” to refer to any travel stemming from improved travel conditions (typically a reduction in travel time or travel time variability). Although travel can be induced on any mode, we are concerned here primarily with induced vehicle travel, which is measured as an increase in vehicle miles traveled (VMT). Induced travel has several direct and indirect components that proceed over the short, medium, and longer term. The direct responses are most familiar. In the immediate aftermath of a capacity expansion, people may change their route, mode, or time of a trip (i.e., Downs’ Triple Convergence [ 16 ]). Of these diversions, only a modal shift is likely to induce much truly new travel. People also respond to lower travel costs by selecting more distant destinations, which induces new travel ( 4 , 6 ). Some additional travel may be induced if people further alter their trip-making patterns by, for example, unlinking their trip chains into several trips or visiting some destinations more frequently. In the longer term, businesses and households respond to new capacity by locating in more outlying areas. Capacity improvements can also indirectly induce new travel. For example, a household who relocates to the fringe in response to lower commuting costs may also find it necessary to drive farther to more distant supermarkets or schools.
Dozens of articles and several literature reviews have been written on the topic of induced travel, outlining its conceptual underpinnings, documenting its extent, and exploring its implications for practice (3–9, 17, 18). Our aim here is not to duplicate previous efforts or adjudicate thorny, ongoing empirical debates. Instead, we focus squarely on whether and how induced travel appears in engineering classrooms and the rationale that instructors give for its inclusion or omission.
Methods
We conducted in-depth qualitative interviews with engineering instructors in the United States during the fall term of 2021. The 16 interviewees represent 8.8% of the transportation engineering instructors at accredited programs in the United States (n = 180). A strength of the interview methodology is that we can capture rich detail about instructors’ pedagogical decisions and their views on engineering practice ( 4 ). A weakness is that we cannot comment on the representativeness of the views described below. A systematic survey of transportation instructors could offer some representative information but would offer considerably less texture and insight into what goes on in educators’ minds. Further, surveys can only capture what the survey authors already know to ask; interviews help capture “unknown unknowns” ( 4 ).
Our aim was to characterize the diversity of views about transportation engineering instruction. To that end, we specifically recruited individuals who we expected to have diverging perspectives. We were careful to speak with instructors of various experience levels, from post-docs and recently hired assistant professors to full and emeritus professors. Our interviewees came from a range of institutions (private and public), program sizes (large and small), and student instructional level (undergraduate and graduate). Most taught in-person, but a few taught online courses (particularly during the COVID-19 pandemic). All instructors taught within an engineering program, but some held joint appointments in other programs (planning or public policy). Most instructors in our sample were academics, but some had considerable practical experience (i.e., worked in practice during or before their time as an instructor). Finally, our sample includes twice as many men as women, a fact that largely reflects prevailing gender differences in employment, particularly among senior faculty. To conceal the identity of our interviewees, we report just the rank, school type, and gender of our interviewees and redact identifying information (e.g., city or state, course titles).
We recruited the sample by identifying highly ranked engineering programs with required transportation courses and then emailing the instructor listed online (or in some cases the program administrator). Occasionally, we interviewed two instructors at the same university. In two cases, we specifically contacted instructors in our broader professional networks who were known to employ more progressive teaching practices. Of the email solicitations to 74 unique individuals, 13 replied that their research did not fit our topic and 45 did not respond. The 16 successfully completed interviews indicate a 22% response rate, which is in line with qualitative research norms and expectations.
Interview Guide
The average semistructured interview lasted for 47 min (range: 25 to 63 min). Interviewees began by sharing basic information about their courses (title, student type, pedagogical style, etc.) and the broader context of the program (number of students, sequencing of courses, etc.). Instructors then described the key ideas that they hoped students would take away from the course and reflected on how they prioritized topics given their limited time with the students. Finally, we asked instructors to specifically reflect on traffic congestion and induced travel. For those who did not teach induced travel, we asked them to reflect on the reasons for this omission. Finally, interviewees reflected on their students’ experience applying induced travel and other ideas in internships and other professional settings.
Analysis and Expectations
We transcribed all interviews and reviewed them using inductive coding ( 19 ), a well-established qualitative method in which the researcher starts by carefully reviewing one observation (in this case, a single interview) and comes to recognize broader patterns and themes as they systematically incorporate additional observations. The approach is highly iterative; the researcher repeatedly reads and rereads the texts, refining the themes over time. Inductive coding is best understood in contrast to deductive coding in which the researcher starts with a preexisting theoretical framework and uses data to support or refute it. Inductive coding is preferred when conducting exploratory research.
From the outset, we expected engineering curriculums to be somewhat tailored to help students successfully pass the Fundamentals of Engineering exam or FE, which students take at or near the completion of their undergraduate degree. The exam requirements are outlined in the “Body of Knowledge,” a publication of the American Society of Civil Engineers ( 20 , 21 ). Before undertaking the interviews, we carefully reviewed the Body of Knowledge and found that induced travel and related terms were not explicitly mentioned, nor were they listed in the associated FE reference materials or practices tests. Given the importance of these materials for student success and accreditation, we suspect that this omission may lead some instructors to focus their limited class time on other topics.
We also expected some engineering instructors to omit induced travel owing to two related concerns about credibility ( 22 ). We suspected that some engineering educators may be skeptical about the evidence base for induced travel, particularly if they are only familiar with early studies, which suffered from methodological shortcomings ( 3 ). Some engineering instructors may also be skeptical of the “relative knowledge” of planners, further deteriorating credibility ( 22 ). In this view, engineers may be seen as scientific, precise, and pragmatic whereas planners can be unscientific, fuzzy, and idealistic. If some instructors hold this view and see induced travel as a planning concept, they may find the concept easier to disregard.
Results
Do Instructors Discuss Induced Demand?
Instructors in our sample were evenly split on whether they taught induced travel (though we caution that this is a result of our recruitment strategy and may not reflect the distribution among all instructors). Some took great pains to explain the concept, with several describing induced travel as one of the key takeaways from their course. One professor specifically warns, Don’t expect to decongest a roadway by adding capacity. It hasn’t happened in the recorded history of mankind and we won’t be the first ones to do it (#1, Full, Public, M).
Another explicitly uses the term and emphasizes behavioral adjustments: I do very specifically talk about induced demand. So we talk about the fact that you increase capacity on a roadway and that is going to pull more people onto that roadway who might be taking a different mode, choosing not to travel and we talk through all of the travel behavior aspects of that (#2, Associate, Public, F).
This professor is careful to take a multimodal perspective, explaining, But then in parallel, I also tell them that when we build a new transit line, it’s not likely going to reduce congestion either. And I think cycling is similar. Anything we look at, people are always going to be attracted to that empty capacity on the best serving roadway (#2, Associate, Public, F).
By contrast, several other instructors do not teach the concept. One explained, We don’t really get into the whole concept of “if you build more, that’s just going to add more demand, and you’re still going to have the same problems as you do already.” We don’t spend a ton of time talking about anything like that, to be honest (#3, Assistant, Public, M).
A different instructor noted that they have never taught the concept in any of their engineering classes: We never talk about “Okay, now we’ve added three more lanes, how many more vehicles are going to be attracted to this roadway as a result of adding more lanes to the roadway?” I don’t think I’ve ever talked about induced demand at all in the classes that I’ve had (#4, Full, Public, F).
Between these extremes, some instructors cover induced travel, but often in an informal or ad hoc manner: I kind of mention in passing that if you build more highway capacity then new demand will come along to fill up that capacity. We might talk about some instances where that happened, but it’s not terribly well-organized (#5, Full, Private, M).
One instructor is careful to supplement formal problem sets with brief lectures on induced travel. However, they are quick to acknowledge the limits of this approach when the problem sets—the central feature of the course—tell an opposite story: They have a homework problem where they look at the level of service on the road and then calculate the new level of service after they’ve added a lane. The result is “look they’ve improved the level of service.” And I am pretty careful to say things like “There’s also human behavior. Adding an additional lane will just make more people want to use the road, and so we don’t know that this is actually going to be an effective, long-term solution.” But the homework problem is still set up for them to find the number of lanes that will match the design level of service (#6, Associate, Private, M).
Other instructors supplement their problem sets with discussions or in-class activities where students were asked questions like, “Are there other ways to manage this demand?” and “What are the unintended consequences of accommodating this demand?” (#2, Associate, Public, F).
Finally, some instructors “want folks to have at least a kind of qualitative sense of how the process pans out in the transportation network,” but explicitly avoid using terms like induced travel, induced demand, or latent demand (#7, Full, Public, M). When asked, “do you ever use those phrases explicitly in class?” this instructor responded, No, mainly because, at least in my experience, those are not well-defined terms. They seem to mean different things to different people. It is more of an advocacy-related term (#7, Full, Public, M).
How Much Do Instructors Critique Engineering Practice?
Instructors also exhibited considerable diversity in the extent to which their explanation of induced travel (and the class as a whole) critiqued established engineering practices. Whereas the extent of critique falls along a continuum, for simplicity we describe three categories: 1) those who were largely uncritical of engineering practices, 2) those who offered modest critiques, and 3) those who advanced major critiques.
Largely Uncritical
Instructors in the first group tend to be largely uncritical of engineering practices. Many of these instructors stress that traffic congestion is harmful and merits action. One put engineers’ goal succinctly: “Avoiding delay: This is what we are trying to do” (#4, Full, Public, F). In their first lecture, this instructor emphasized the “costs of congestion and shared photos of people in other countries in gridlock.” When asked the key idea from their course, another instructor explained, “What happens when capacity isn’t enough? That would be number one” (#5, Full, Private, M).
Many instructors used standard engineering problem sets to reinforce the need to accommodate capacity. For example, for a class project the instructor has students “deliberately add a lot more cars to the network so that it’s going to need improvements” (#14, Full, MW, Public, M [emphasis added]). Similarly, another instructor explained that “a standard question in my homework examples and test questions is, ‘If we make this change, how does it improve the level of service?’” (#4, Full, Public, F). Notably, these instructors do not supplement these problems with a lecture, discussion, or assignment that describes induced travel. Instead, instructors in this group tended to focus squarely on mastering problem sets and understanding the foundational logic of relevant equations and engineering guidance. For example, one instructor, who spends “75% of the course on the four-step model” explained that their goal was for students to understand “how those numbers were achieved, and what went into them.” Notably absent, however, was any critique of the methods: “I don’t even get to the point of critiquing LOS [level of service] in this class. I literally just introduce it and that’s it” (#8, Assistant, Public, F).
Modest Critique
Some instructors offered modest critiques of engineering practice, most often by describing important practical limits to our ability to accommodate all potential demand for travel. For instance, one instructor emphasized the great difficulties in expanding capacity: Adding a new lane in the software is so easy because you click one lane and convert to two lanes. But adding a new lane on the existing network? It’s terrible because so much construction is involved and so much cost as well. It’s ugly (#14, Full, MW, Public, M).
This instructor, like some others, hoped that students would carefully consider financial tradeoffs, urging them to identify “the lowest cost improvements.” In a class project, this instructor required students to estimate the costs of their road widening recommendations. The instructor underscored “that those are real dollars that somebody has to spend. Don’t come back and say, ‘I’m adding 100 cars here and I want you to spend a million dollars’” (#14, Full, MW, Public, M).
Another instructor wanted students to have “at least a kind of qualitative sense” of how people respond to new capacity (#7, Full, Public, M). “People will respond to what you do,” they tell students “and you should take that into account when making your planning and design decisions.” These behavioral adjustments affect forecasts in predicable ways: If you predicted that widening the road would reduce travel time because the same X number of people would be using it, you have a good chance of being disappointed (#7, Full, Public, M).
Given these and other capacity-expansion challenges, instructors in this group agreed that it is sometimes “okay” or “necessary” to have some congestion “in the peak of the peak” (#14, Full, MW, Public, M). Beyond these exceptions, however, instructors in this group continued to emphasize the “need” for traditional capacity expansion.
Major Critique
A third group of instructors introduced induced travel as part of a broader critique of standard engineering practices. They taught students “critical thinking skills about systems” (#9, Assistant, Public, F) and used limited class time to critically examine several engineering concepts including LOS, building for the peak of the peak, and linear growth models. More broadly, these instructors often embraced a very different interpretation of the purpose of engineering. Whereas other professors emphasized the need to prevent congestion, these instructors suggested that engineers should ensure that people have travel choices. Some sought to reduce driving and increase the usefulness, convenience, and safety of other travel modes. For example, My view is that the way transit and active transport get better is by making cars worse. People don’t shift to transit if cars are cheap and free and easy to travel with. So my preference is to de-emphasize cars (#6, Associate, Private, M).
Multiple instructors expressed a need to correct conventional wisdom. For instance, one instructor clarified that a central aim of their course is to “try to erase preconceived notions about the Highway Capacity Manual and things like that” (#2, Associate, Public, F). Another hoped that students would be more critical of the way transportation projects are selected and justified: [The public] has been told for years that we have to make one public investment or another to make their lives better. And whether it’s widen the freeway or build the train to clean our air and decongest our roads, the messaging typically is: “We’re going to solve your problem, despite the fact that the approach we’re taking is dead wrong and we know it’s wrong” (#1, Full, W, Public, M).
When exploring reasons to avoid road widening, instructors in this group advanced broader and deeper critiques than previous groups. In addition to exorbitant costs and the paucity of available land, instructors in this group explicitly considered other road users: If we over-design it so that we’re really only focused on the peak of the peak, as a byproduct of that, we’re not making space available for all the other uses. We’re doing a disservice to the full usage of that roadway (#2, Associate, Public, F).
What Reasons Do Instructors Give to Justify Their Coverage of Induced Travel?
What explains instructors’ pedagogical decisions? Instructors raised seven interrelated concerns. The first three were raised most frequently by instructors who “believed” in induced travel and were interested in teaching it but struggled to do so. The next two concerns were raised by instructors who worried about the realities of preparing students for the FE and the job market. The final two concerns were raised more by instructors who were somewhat skeptical of the extent of induced travel and its implications for practice.
Concern #1: It is Outside My Area of Expertise
Some instructors admitted to being poorly prepared to teach induced travel because it was outside their area of expertise. A common refrain was, “I don’t spend a lot of time thinking about this stuff. I don’t consider myself an expert on it” (#8, Assistant, Public, F). Instructors noted that it was commonplace to teach beyond their expertise, particularly in small departments where “we’re trying to cover a lot of things and we’re just spread thin” (#7, Full, Public, M).
When teaching beyond their area of expertise, instructors largely drew on their time as a student. One used a confessional tone in sharing, “I use my graduate student notes from 30 years ago, so don’t judge me” (#4, Full, Public, F). Another humorously reflected that they had last worked closely on these topics as a PhD student, which was, “Oh boy, decades ago” (#7, Full, Public, M). Because the scholarly literature on induced travel was less developed when instructors were students, many of them do not have much familiarity with the topic: I feel like this makes me sound very simple, but I don’t remember having this discussion as an undergraduate or graduate student (#4, Full, Public, F).
Instructors also rely on resources from textbooks and other sources to supplement their knowledge. But there was a sense that induced travel was largely missing from standard engineering textbooks. As one instructor explained, “If it was in my textbook, I 100% would have taught it” (#3, Assistant, Public, M).
Concern #2: I’m Uneasy Teaching Softer Concepts
Although many engineering concepts are mathematical and technical, others are “more of an art than a science” (#2, Associate, Public, F). Some instructors were uneasy teaching these softer skills. In part, this preference for a quantitative focus comes from the students, who are thought to have, A real strong desire for problems. “I track down a formula and the instructor gives me the numbers to plug into it and I do the algebra and the calculus and I get the answer and everything is just great” (#7, Full, Public, M).
Discomfort also originates from professors themselves. One explained, “I don’t always feel as comfortable teaching those [because there is] no right answer” (#3, Assistant, Public, M). Another clarified that these concepts “can’t be defined and engineers hate things that can’t be defined.” They went on to explain that, I don’t want to have to guess whether something’s right or not. If it’s math, it’s math and it has an answer. I think a lot of people in our field are like that (#10, Assistant, Public, F).
Compounding the struggle was some instructors’ limited knowledge of the empirics of induced travel. As one instructor explained, There are no methodologies for capturing [induced travel] in the Highway Capacity Manual, or any other material that I know of, where they say, “You just added a new lane to this highway. What happens?” (#4, Full, Public, F).
This instructor was frustrated that the Highway Capacity Manual’s coverage of induced travel lacked empiricism: The discussion of induced demand, as far as I can recollect, has a lot of narrative. It doesn’t translate into how to use that information if you are an engineer or planner and you’re trying to decide if this is a good methodology to add a lane or add three more lanes on this roadway. If you can’t quantify it, then it’s more than likely not being discussed in any detail (#5, Full, Public, F [emphasis added]).
This instructor was not alone in their unfamiliarity with research quantifying the amount of induced travel. In describing why they do not teach it, another instructor explained, There’s no induced demand equation that you can use to calculate what it’s going to be for this roadway. Maybe there should be (#3, Assistant, Public, M).
Yet another instructor wondered aloud, If there was an induced demand equation, maybe it would pop up more. I don’t know. Do they have those? I haven’t seen them (#10, Assistant, Public, F).
Similarly, another instructor explained, I want to show the students an analytical model or a simulation model that captures latent demand or induced demand. Show me where it is and I’ll teach it tomorrow (#11, Full, Public, M).
These quotes suggest that some instructors might be willing and able to teach induced travel if they were familiar with quantitative resources that are more consistent with engineering’s emphasis on empiricism and problem solving.
Concern #3: Teaching is a Low Priority
Instructors raised concerns about their ability to teach induced travel given their time constraints: I had to get this class ready in a semester. I came in, and they asked me to teach this class and my research is not specifically focused on this stuff. If I have to teach this, I’m going to go take a textbook and use that as my skeleton for the class and then slowly add stuff on (#3, Assistant, Public, M).
Even with a plan to “add stuff on” over time, time pressures made it difficult for instructors to update course materials. This instructor admitted, Once the stuff is made, it’s hard to motivate yourself to add a lot more and find new innovative stuff. I definitely thought I was going to overhaul a lot more than I’ve ended up doing. Especially once the semester starts, you’re just like “on the train” (#3, Assistant, Public, M).
Yet another made “piecemeal changes” to a course they developed 10 years ago, but admitted the course “needs revamping in a really bad way” (#4, Full, Public, F). The collective failure to keep curricular materials up to date is a problem because, as one instructor explained, “if you’re teaching the same thing for 20 years, you’re probably missing an awful lot” (#10, Assistant, Public, F).
As several interviewees noted, instructors who fail to update their curriculum are responding rationally to incentives around promotion and tenure: Teaching is lowest on the totem pole to be honest in terms of what they expect. I need to go get a certain amount of funding so I can get tenure (#3, Assistant, Public, M). The bottom line is a professor should not be wasting his time learning [other topics]. He should be doing research because that’s what they’re judging him on (#14, Full, MW, Public, M).
With so much pressure to produce research, “if you can teach the same thing next year, you’re probably going to” (#10, Assistant, Public, F).
Concern #4: I Need to Cover the FE
There is only a limited amount of time in a semester and instructors must make decisions about what to include and what to exclude. Whereas all instructors face these tradeoffs, those who omitted induced travel mentioned overcrowded curriculums more frequently and appeared to share a sense that they had an obligation to teach other topics instead. They said things like, “there’s just so much that you have to teach” (#10, Assistant, Public, F [emphasis added]), and “we have a limited number of class periods and a lot we need to cover” (#7, Full, Public, M).
Reference to time pressures were especially prominent for instructors who focused on preparing students for the FE exam. As one instructor explained, As a program, like almost all other civil engineering programs, we want to have almost all our students take the FE before or immediately after graduating. And since this is the only transportation class they get, I spend probably 80% to 90% of my time teaching to the FE (#8, Assistant, Public, F).
This emphasis on the FE was shared by many (but not all) of our respondents. The perceived pressure to prepare students was palpable. As one instructor explained, When you have licensure, you really have to ensure your students know the standard stuff (#10, Assistant, Public, F).
The demands of the exam profoundly shape these instructor’s teaching practices: We pull right out of the FE reference book. I literally teach to those exact nine or ten pages. I cover every single equation (#8, Assistant, Public, F).
Another instructor echoed this approach: Basically, we try to make sure that every question they could be presented with on the fundamental of engineering exam, they’ve had a homework question similar to that and on their tests (#6, Associate, Private, M).
Because the FE does not include questions about induced travel or critiques of engineering practices, those topics got short shrift in engineering classrooms. One instructor put this decision plainly, noting that, Induced demand is not really much of a part of the exam, so that’s not something I need to worry about (#3, Assistant, Public, M).
Concern #5: I Must Consider Job Training
In making curricular decisions, instructors must balance critiques with the realities of job training. Although there was a sense that some firms encourage their junior employees to question standard engineering approaches and go after new kinds of projects, most firms appear to take a more traditional approach. Interns who question standard practices might get the message, “Don’t worry about it, little kid. Just do what I’m telling you to do” (#2, Associate, Public, F). As one instructor explained, junior employees, Are being taught by more senior traffic engineers who learned these concepts in much more standard, old-school ways. It’s been drilled into their head so much that if the road looks like this, you add capacity (#2, Associate, Public, F).
This instructor went on to explain: They were taught, “You do what the manual says. You don’t question these things.” And so I think it’s almost like they’ve been trained in a way that it’s hard to reverse (#2, Associate, Public, F).
Economic pressures compound established professional norms. The need to attract and retain paying clients can make it difficult for consulting firms to directly challenge existing practices for fear of losing contracts.
More common than not, folks are afraid to say anything. There’s one thing to learn about [induced travel], and it’s another to actually figure out how to influence a client or influence your superiors if you work for the DOT. Especially if you’re a new person. You have no say (#10, Assistant, Public, F).
Given this context, instructors are wary of engaging in too much critique: I can critique the DOT practice guides, but I don’t think it’s appropriate for me to say the practice guides are wrong and you should ignore them (#6, Associate, Private, M).
Another explained, I guess I worry about eroding their confidence. The stuff that we teach, it’s been around for a while. It works reasonably well most of the time, if you’re careful. I don’t want to go shooting that down (#7, Full, Public, M).
Concern #6: Is It Really New or Just Shifted?
Some instructors do not teach induced travel because, in their eyes, what looks like new travel is actually existing demand that has simply shifted to a new location.
I keep telling them that, “Is it induced demand or is it shifted demand? Is it really new traffic that you created that wasn’t there or did you attract that from other crowded roads? (#14, Full, MW, Public, M).
Yet another instructor drew on their experience from practice to distinguish between induced travel and what they call “pure overcapacity situations”: People argue that if you have another travel lane you are going to induce more demand. As a practitioner, I would argue that no, that demand already existed. It’s just that they had nowhere else to go (#12, Adjunct, Private, M).
By emphasizing modest shifts in route selection, these instructors tended to deemphasize other short-term behavioral responses that may increase driving (i.e., longer trips to more distant destinations, less trip chaining, and more frequent trips) and thus probably underestimate the potential of capacity expansions to increase VMT.
Concern #7: Our Role is to Accommodate Demand
Some instructors did not necessarily question the logic of induced travel but did question its implications for practice. In their eyes, even if new capacity really does generate new travel, it is not a bad thing. It is the job of engineers to help people get from A to B. These instructors viewed their role within the traditional predict-and-provide paradigm and endorsed the need to accommodate existing and anticipated demand.
Accommodating more travel is good because travel is a derived demand. Every trip that is taking place benefits the person who chose to take it, otherwise it wouldn’t be taking place. More capacity and more trip making is not a bad thing (#1, Full, W, Public, M).
Another explained that capacity expansion is still worthwhile, even in the presence of induced travel: Say you add capacity and about 90% of it is eaten up right away. That is the worst case. But with 10% more capacity than you had before, that suggests lower cost, not higher costs. That means you have more people traveling at a lower cost and in a cost–benefit analysis, that’s a plus (#14, Full, MW, Public, M).
In making these arguments, some instructors explicitly highlight the economic motivations that practicing engineers face. As one explained, for many practitioners, Any position that leads to more transportation projects is the right position. If building gantries for congestion pricing means we get to build more gantries, congestion pricing is a good thing. If adding capacity means that we get to build the capacity, then adding capacity is a good thing (#1, Full, W, Public, M).
Another echoed these sentiments when they said, There’s a certain amount of self-serving nature to that because the more problems that causes, the more work there is going to be for civil engineers down the road (#12, Adjunct, Private, M).
Political pressure compounds these economic incentives: If you are the State DOT Administrator, your purpose in doing the simulation work is a means to an end because you need to get reelected or your boss needs to get reelected (#11, Full, Public, M).
Ultimately, these instructors disagreed with the suggestion that induced travel should put a halt to roadway widening. As one instructor explained, I see a lot of planners ask, “Why build more capacity? It’ll just fill up again.” Well, why build more schools? Why build more hospitals? They fill up immediately? Because you’re providing a service. That is the answer to that question (#14, Full, MW, Public, M).
Beyond Believers and Skeptics: An Illustration
We close with a final example, which we quote at length because it illustrates the usefulness of interviews to fully capture knotty, layered perspectives that could be easily overlooked. It initially appears that this instructor is wholly dismissive of induced travel: A lot of people think that if you provide more supply that somehow generates demand. That’s not how economics works. It’s just patently false. I want to get away from this idea that when you put capacity out there and people that weren’t traveling before start traveling just because it’s cheaper (#14, Full, MW, Public, M).
The instructor compares travel to an everyday good to illustrate the lack of behavioral response: If they dropped the price of a can of corn at the store, I’m not going to run out and buy more corn (#14).
Many “believers” would take issue with this explanation of the economics of travel (and most goods); after all, drivers do indeed adjust their behavior as travel costs fall. Some “believers” may tune out, assuming that there is no room for consensus with this instructor. Others may engage in adversarial listening, looking for opportunities to point out errors in judgment and win the argument ( 23 ).
Yet a careful, open-minded listener will find that this instructor’s seemingly absolutist statements belie a more complicated view. Elsewhere in the interview, they readily admit that drivers respond to new capacity: Now that doesn’t mean that there’s not certain things that occur. For example, if there’s a lot more capacity, and that capacity was just kind of excess and thrown out there, people might start traveling a little further or by different modes (#14 [emphasis added]).
While this instructor readily acknowledges that new capacity “can induce VMT,” they are adamant that “it’s not induced demand.” This seemingly irreconcilable distinction is possible because the instructor employs a narrow definition of the term: My contention is that the term induced demand should be reserved for added trips that wouldn’t have been there had the change not been made (#14 [emphasis added]).
In sum, this instructor acknowledges a variety of behavioral responses to new capacity—including longer trips—and agrees that faster, wider roadways are likely to induce VMT. This is in line with most of the induced travel “believers.” Yet this considerable consensus might be easily overlooked because they do not share the same definition of induced travel. This is not to suggest that the believers and skeptics agree on everything—clearly, they do not. Rather we highlight this example to illustrate that in discussing this topic, it is essential to move beyond a believer–skeptic dichotomy.
Discussion
Our interviews revealed remarkable variation in engineering instruction. Some professors featured induced travel as a central theme, whereas others omitted the idea entirely. Instructors also varied in their willingness to critique engineering practices. Some were largely uncritical, some advanced practical critiques of the limits to roadbuilding, while others engaged in extensive critique and worked to counteract conventional wisdom.
After sharing our findings at the 2024 Annual Conference of the Transportation Research Board, we were struck by diverging reactions to the results. Many audience members were confident that induced travel is already taught in all engineering programs. Some even went so far as to suggest that the concept is no longer controversial and that it is well-understood by students and the public, though surveys suggest otherwise ( 10 , 24 , 25 ). Meanwhile, other conference goers peppered us with questions that revealed a very different perspective. “Wait,” they asked after hearing our results, “what exactly is induced travel?” Others took a more adversarial and condescending tone, asking “Why should instructors teach a myth?” It is clear from these interactions that many induced travel “believers” and “skeptics” operate in echo chambers where they hear mostly from like-minded peers and rarely encounter opposing views on this topic ( 26 ). People suffer from both the curse of knowledge ( 27 ) and the false consensus effect ( 28 ); they assume that others know what they know and that they share their views. But this couldn’t be further from the truth. As this work revealed, induced travel remains poorly understood (in some circles) and hotly contested.
The differences in classroom instruction uncovered in this work are likely to have a profound influence on engineering practice domestically and internationally. A student trained primarily to complete problem sets about capacity expansion is likely to make very different choices on the job than a student who is familiar with induced travel and well-versed in critiques of engineering practice. Differences in training are likely to become more important as students gain experience and potentially begin making upstream decisions about how to evaluate and prioritize projects. Although higher-level decisions about transportation are often couched in technical, apolitical language, they are fundamentally about values and priorities (see the values implicit in the use of LOS [ 29 ], in multiple domains of transportation practice [ 30 ] and governance [ 31 ], and in policy controversies more broadly [ 32 ]). We suspect that a practitioner’s values and preferences are shaped, at least in part, by their training at university (as they are in economics education [ 33 ]).
What explains the divergent teaching practices uncovered here? In justifying their pedagogical decisions, instructors offered a range of overlapping concerns. Those who “believed” in induced travel but did not teach it often lacked expertise in the area and were uneasy teaching “soft” concepts. Because teaching was not seen as a priority for their promotion and tenure, these instructors lacked the motivation to overcome these challenges. The next pair of concerns centered on the pragmatic realities of preparing students for the FE and their future careers. Instructors who voiced these concerns also tended to “believe” in induced travel but felt that there were other things they had to cover instead and worried that an overly critical curriculum would be inconsistent with professional norms and employer needs. The final pair of concerns were voiced by those more skeptical of induced travel. They suggested that what is seen as induced travel is really shifted from other locations or is relieving overcapacity situations. More importantly, even if new travel was indeed induced, some instructors argued that accommodating that new travel was the role of engineers.
We heard the strongest reluctance to embrace induced travel from some of the more senior scholars (full professors). At the same time, the instructors most enthusiastic about teaching induced travel and critiquing standard engineering practices tended to be relatively newer to engineering instruction (assistant or associate professors). Together, these results suggest that a generational transition may be under way. However, we caution that this transition is far from guaranteed. In fact, several of the most junior interviewees were the least likely to cover induced travel, principally because the topic was outside their area of expertise and they were too pressed for time to develop comprehensive instructional materials on the topic.
How to Increase Coverage in the Classroom
Increasing coverage of induced travel in transportation engineering classrooms would improve outcomes by enabling more accurate assessments of projects and clearer conversations with planners, policy makers, and the public. We offer four suggestions for increasing coverage of induced travel.
Recommendation #1: Share Resources for Teaching
First, instructors can make use of a large and growing set of resources available to help them teach induced travel. We direct instructors to the Induced Travel Calculator (https://travelcalculator.ncst.ucdavis.edu/) and its supporting documentation. The empiricism of the calculator serves as a useful bridge between the equations preferred by engineers and the softer concepts many of them shy away from. We have also made available our own teaching materials, which we developed and refined over years in our classrooms (https://bit.ly/TeachingInducedTravel). In sharing these materials, we hope that instructors will use and adapt them for their own purposes. Beyond these resources, it would be helpful to widely publicize induced travel resources through the American Society for Engineering Education’s monthly newsletter or through workshops offering Continuing Education credits from the Institute of Transportation Engineers.
Further developments are needed. Ideally, teaching materials would incorporate real-world data and open-ended questions by which students can grapple with whether to widen specific roadways. Engaging in these debates will cement students’ understandings of the concepts and bring them into important policy discussions.
Recommendation #2: Value and Reward University Instruction
Whereas open-source materials will be helpful, more foundational changes to university instruction are probably needed. Given the realities of university teaching described above, the process of incorporating induced travel is likely to be slow. But tuition-paying students deserve competent up-to-date instruction. At a minimum, instructors should have ample time to prepare new courses and, whenever possible, should have stability in their course offerings so they can refine materials over time. Instructors should be encouraged to regularly update course resources and periodically conduct major overhauls. Instructors should be given financial and strategic support to accomplish these tasks. A cultural shift may also be needed. The intellectual rewards for excellent teaching are undervalued relative to research and professors face considerable pressures to prioritize publishing and grants ( 34 ). One approach to balance these competing needs could be to hire full-time teaching professors with subject area expertise and training in pedagogy. However, there is a risk that teaching faculty may be treated as inferior to traditional research faculty ( 35 ).
Recommendation #3: Revise the Body of Knowledge
A third approach is to add induced travel to the Body of Knowledge, the document that determines the content of the FE exam and guides accreditation ( 20 , 21 ). Many of our interviewees shared our assessment—based on a careful review of the materials—that induced travel is missing from the Body of Knowledge and the FE. Moreover, because these materials are so central to instructors’ curricular decisions, adding induced travel is likely to increase student exposure to the idea. We recognize that changes will be controversial and that adding new requirements will be difficult given competing demands and limited space in the curriculum ( 12 , 36 ). Despite similar challenges, recent revisions to the Green Book and the Manual on Uniform Traffic Control Devices are auspicious signs that adapting to new needs is possible, even in foundational documents governing transportation practice ( 37 , 38 ).
Recommendation #4: Change the Policy Context
Finally, engineering instruction is likely to change if the policy context demands it. For example, California Senate Bill 743 became law in 2020, mandating changes to traffic impact analysis and requiring transportation models to incorporate induced travel ( 39 , 40 ). According to one interviewee, most practitioners in the state were reluctant to embrace these new ideas and methods until they were forced to do so by state policy (#13, Postdoc, Public, M). An instructor elsewhere echoed these sentiments, arguing that changes to engineering practice would require “a permanent change to state policy.” Without it, “typical traffic engineering firms who work on state projects are going to produce the same old standard engineering outputs” (#1, Full, Public, M).
Given the perceived need for an external push, we were heartened by a recent call from U.S. DOT for an “Induced Demand Assessment Framework” ( 41 ). The aim of the project is threefold: 1) “to assist DOTs to better understand induced demand, especially at the project level”, 2) to consider how induced travel affects “priorities, policies, funding decisions,” and 3) to outline “DOT options to respond to induced demand directly or indirectly.” We suspect that this project and others like it will further increase instructors’ familiarity with induced travel and encourage more of them to teach it in their classrooms.
Caveats and Future Research
Because this work is exploratory, it raises several additional questions. For instance, our interviewees rely heavily on textbooks, but felt that induced travel was not well covered in existing materials. Further research should assess whether this is indeed the case. Second, by focusing on instruction in U.S. classrooms, we leave open the question of whether international students are being exposed to induced travel and critiques of standard engineering practices.
A survey of transportation engineering instructors could determine the prevalence of the views expressed here. A survey could also clarify the extent to which pedagogical decisions varied by teaching experience, university setting (public versus private and rural/suburban versus urban), program size, class size, student type (undergraduate versus graduate), and other potentially relevant characteristics.
This research illustrates the challenge of preparing students for a profession in transition. Do students who embrace critiques thrive in the job market or do they struggle to find their footing? Relatedly, this research poses a challenge for those interested in breaking the reinforcing cycle between government agencies and the private consulting firms they hire. As our interviewees explained, when agencies request standard practices, consultants have little incentive to press for innovative approaches or wholly new strategies. A question remains: how to stop or amend this cycle, especially when both groups have economic and political incentives to maintain the status quo?
In closing, we note that conversations about induced travel are just one of many important debates as engineering practice—and instruction—evolves to respond to 21st century challenges. Although relieving congestion and increasing mobility were our priorities for decades, a host of other goals—like protecting the environment, reducing emissions, promoting justice, and improving livability—are now ascendant. How should transport professionals prioritize these aims, especially when new and old goals conflict? How might standard engineering practices be refined (or replaced)? Fundamentally, what is the appropriate role of an engineer? What kind of broader curricular metamorphosis is warranted to more holistically prepare our future planners and engineers? In these debates it will be essential to listen carefully and use precise language.
Footnotes
Author Contributions
The authors confirm contribution to the paper as follows: study conception and design: K. Ralph; data collection: E. White, K. Ralph; analysis and interpretation of results: K. Ralph; draft manuscript preparation: K. Ralph. All authors reviewed the results and approved the final version of the manuscript.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
