(Unedited) Podcast Transcript 515: Highway Robbery
January 8, 2025
This week on the Talking Headways podcast we’re joined by Ben Ross and Joe Cortright to discuss their article in Dissent Magazine discussing how modeling is being used to expand highways around the country. We chat about their critiques of highway modeling, politics, and some potential solutions to the problem.
You can listen to this episode at Streetsblog USA or find it on our hosting site.
Below is a full unedited AI generated transcript:
[00:01:14] Jeff Wood: Ben Ross and Joe Cortright, welcome to the Talking Headways podcast.
Joe Cortright: Great to be here.
Ben Ross: Great to be with you. Yeah.
Jeff Wood: Good to have you all here. And I’m really glad Ben to see you after all these years, we had Joe on a couple of times on the show. So it’s nice to see you again too, Joe.
[00:01:31] Before we get started, can you tell us a little bit about yourselves and we’ll start with Ben and then go with Joe, because Joe has been on two episodes already, as I mentioned, episode 52, I think it was like nine years ago. And then. 87 of the Monday show, which is more recent where you’re talking about the Rose quarter expansion. [00:01:45] So Ben, could you tell us a little bit about yourself? [00:01:48] Ben Ross: I’ve been a political activist my whole life and always an urban oriented person, and in the 1990s, I got involved in trying to get the purple line built in the Maryland suburbs. Of Washington, D. C. and this was from my point of view, the ideal political issue for me, it was basically 1 of these 99 percent versus the 1 percent issues, because the purple line was held up for years, because it uses an old railroad right away that runs through the middle of 1 of the most expensive country club golf courses in the Washington area country club. [00:02:30] Heavily populated by lobbyists and so we had what i think is one of the epic nimby contests and we won impressive and. As part of that i got more and more into land use also wrote a book called dead end. That I know you’ve seen, but more relevant to this really is my professional background. I have a degree in physics and was a modeler, a groundwater modeler, but the groundwater models are mathematically very similar to the traffic models. [00:03:08] Joe Cortright: Yeah, I’m Joe Cortright. I’m an economist based in Portland, Oregon, but last decade, I’ve run a little think tank called city observatory, which focuses on what are the factors that seem to be making the most difference to urban success in the 21st century and do a lot of work on a range of urban issues from inequality, segregation, housing and economic development, but then also to urban form and transportation. [00:03:36] Thank you Because we know from looking across the country that the way cities are laid out for transportation and the kinds of investments that they make, or don’t make in transportation have a profound impact on not just environmental issues, but also their economy, their livability and the fairness of communities. [00:03:57] And in addition to that professional interest for the last, oh, more than 10 years now, I’ve been involved in a number of the freeway fights in Portland, where we’re. Pushing back against plans to spend literally tens of billions of dollars on a handful of freeway expansion projects that, if history is any guide, we’ll only make all of our transportation and urban problems worse. [00:04:21] And it’s a concern about the ways in which a lot of technical information has been wielded to try and justify. These massive freeway projects that’s led me to take a close look at this work. And then, yeah, I’ve crossed paths with Ben, who, from the other side of the country, and from a different disciplinary perspective has reached, I think, many of the same concerns that I have about the work that’s being done here. [00:04:50] And that’s what led us to collaborate on this article for dissent. [00:04:54] Jeff Wood: It’s interesting, one person in Oregon and one person in Maryland. How did you all start talking about this? [00:05:00] Ben Ross: I got very deeply into this through the plan to add toll lanes to the beltway around Washington, D. C. which was my main activity for several years was fighting that and the more we got into it, the more fraudulent it seemed. [00:05:16] And we’ll probably talk about that some more after we went through this well publicized battle. I wanted to get more information about this. So I worked with transportation for America, and they had an intern for the summer who helped with this. To develop this survey and show it going on more widely. [00:05:39] And meanwhile, Joe had started publishing about the bridge over the Columbia River with the same thing had been going on. So I called him up and I had met him at meetings before. And I called him up and we agreed to collaborate on it. I’ve been writing for dissent for many years. So I checked with them and they were interested in the article. [00:06:01] Joe Cortright: Yeah, I’ve known of the work that Ben has done in the D. C. area for a number of years, both with the purple line and then the beltway project and read with great interest. His critiques of the traffic modeling. There and knew that it raised many of the same concerns that arose here and in our discussions. [00:06:19] And I think in our discussions with other folks who are familiar with this field, we came to the realization that there are some practices in this field that are being used to take what is ostensibly an objective and scientific process and twist it in particular ways to support the conclusions of project sponsors. [00:06:41] And so I think what we try to do is spend some time digging into the practices and then the institutional factors that seem to be driving those practices. [00:06:51] Jeff Wood: So I want to be honest here. I had a hard time writing questions for this because I don’t know much about modeling and it’s hard for me to get my head around thinking about that. [00:06:59] I took a trans CAD class in college and I hate TAZs with a passion. I found out anyways in that class. Class and I became a GIS person, so I did a lot of mapping and data analysis, but it was mostly for census blocks and census tracks and things like that. And so one of the things that I found was that like traffic modeling, isn’t really my jam. [00:07:17] And so I haven’t really talked about it a lot on the show. And I know a lot of smart people that do modeling, but one of the things I think that might be actually the problem is that it’s hard for people to get their heads around the models and how they work and, what goes into them, even in somebody like me who, went to planning school, even took a class in this as still like four step modeling, all this stuff, it’s like really complicated. [00:07:38] Like, why is it so hard to get into and understand, even if you’re interested in the subject? [00:07:43] Ben Ross: Yeah. Maybe I should talk about how I evolved my involvement with this. My background with groundwater models, I have written groundwater models and I’ve critiqued a lot of groundwater models. And so I had general feelings about modeling and 1 of them is that. [00:08:05] They are overly complex and there’s actually a mathematical theory called control theory that has proved mathematically that you can have too many variables in a model. When you add more variables, it makes the model less reliable. Now, that’s only proven mathematically for simple cases, but it’s clearly true to from more complicated cases. [00:08:31] And in the case of a graph of a traffic model, the number of potentially adjustable inputs is far greater than the amount of data they have. So you can adjust the inputs for anything you want and get any answer you want. Not any answer you want, but lots of different answers. I came to this thinking that based on all my previous experience, they must have fixed the inputs. [00:09:04] You change the inputs and to get the outputs you want, because there are so many inputs you can change. And when I got into this, I was really surprised and it had to hit me over the head from looking deep into the numbers that they hadn’t done that. They just changed the numbers that they got. [00:09:24] And that’s. To me, coming from a different field, that’s just totally fraudulent. No one would ever do that. It was really a surprise to me. And I think to Joe, too, that they weren’t fudging the models in the usual ways by making assumptions that they wanted, but actually changing the numbers. [00:09:44] Joe Cortright: I think we’re maybe jumping a bit ahead in the story. [00:09:47] Ben Ross: Yeah, [00:09:47] Joe Cortright: and I don’t want to dispute what Ben has said, which is actually true. I think there are 3 things here. 1 is. We should address how fundamental the models are to rationalizing these big investments and analyzing their environmental impacts. This is the technocratic foundation of. [00:10:06] All of these big highway projects and the rationale for their construction. This is not a trivial technical exercise. This is supposedly the scientific case for these multi billion dollar investments. 2nd, thing is, there is a set of models and has been alluded to. They’re extremely complex as, they’re opaque to ordinary human beings. [00:10:29] And in part, that’s by design. Because the people who run the models don’t want people looking at what’s going on. But despite their complexity, that doesn’t mean that they’re accurate or realistic. We know that the standard four step model, which has been around for 40, almost 50 years now, Is actually a really poor representation of the way the world works and that’s 1 of the 1st tests of any model is does it give you actionable information in business terms? [00:10:58] We would say, do you mark your model to market? Does the model produce results that are observed in the real world? And there are a lot of biases there. And then finally, as. Ben just alluded to, we found that, notwithstanding the fact that they had these opaque models and that they were based on questionable assumptions and couldn’t produce accurate results that in the pinch, what a lot of these agencies do is simply just put their thumb on the scale and change the outputs. [00:11:27] To get exactly what they want, so they don’t even they don’t even go through the charade really of relying on these twisted assumptions. They simply alter the outputs to get what they want. And the term of art there is something called post processing where we didn’t like what the model produced and, it doesn’t fit with the way we think things should look. [00:11:47] So we’ll change the numbers. [00:11:50] Ben Ross: Yeah, I would just correct 1 thing. I don’t think they’ve given up the charade. They hide the fact that they change the numbers and it really took a lot of digging and some public information act requests to confirm that they just changed the [00:12:07] Joe Cortright: numbers. Yeah, that’s absolutely true. [00:12:09] You have to be extremely well versed in this, and I spend a lot of time working on modeling issues. I’m the chair of the Oregon Governor’s Council of Economic Advisors. We look over the state economist’s shoulders as they construct a macroeconomic model for the state. So I’m very well versed in those technical issues. [00:12:25] Ben does as well. And then you have to file repeated public records requests in order to get access to just Really fundamental information about what’s going on in these models in order to critique them, which means from the standpoint of ordinary citizens, this stuff is impossible to understand or question or evaluate because it is hidden behind these multiple both technical and legal walls that keep people from understanding and questioning what’s going on. [00:12:58] Jeff Wood: I want to go back again to 30, 000 feet as well. And I think that’s important to think about like why we’re doing these models for regional transportation planning and what in the federal system is set up such that you have to do that model. So why is it that all of these, state DOTs, for example, are putting together, these traffic models that are telling them to build more roads? [00:13:18] What is it in the law or in, policy that is pushing them in that direction? [00:13:23] Ben Ross: A lot of this comes from National Environmental Policy Act, the EIS requirement. [00:13:28] Jeff Wood: Environmental impact statement. Yeah. [00:13:30] Ben Ross: Yeah. Environmental impact statement, to be able to know what the impact of the traffic is, you have to know how much traffic there’s going to be. [00:13:37] So it’s very vital from that point of view. And also. Not really part of NEPA, but in practice is what’s the benefit of this thing going to be? How much faster will traffic move? So they have to politically have some prediction of what it will do. [00:13:59] Joe Cortright: Yeah, there’s a kind of manifest destiny quality to traffic forecasting traffic forecasts tend to. [00:14:07] Inevitably predict that traffic will grow without limit, and therefore we need to widen roads to accommodate it. And we know that’s not true because we know that the empty in the U. S. went down. The vehicle miles travel went down for the better part of a decade after 2000 for when gas prices rose. We also know as a practical matter when we don’t build capacity, the traffic doesn’t grow. [00:14:31] The absence of capacity on a system leads people to change their behavior and they don’t drive as much. Those basic facts are not reflected in these models, because in part, these State Highway Departments like models that show that traffic will grow inexorably no matter what happens, because that provides A basis for arguing that we need to widen these roadways to accommodate more traffic. [00:15:01] Ben Ross: I think in fairness to the modelers. What I’ve learned from this whole exercise, I think is that this is a really tough problem. And it’s not obvious to me that anyone really has a better way of predicting this stuff. And I’m inclined to the view, and I know Joe agrees that a much simpler model would actually be more accurate, but the much simpler model would be just to say that nothing you do will ever change the level of congestion on any congested highway and that would be, although obviously not very far from exactly what will happen. [00:15:45] That was really. Probably the best forecast you can make right now, but you don’t need a computer to do it. So a lot of people would be out of work. [00:15:56] Jeff Wood: Yeah, I think that’s an important point too, is the idea of the politics that get into this as well, right? Because it’s not just the, in deference to modelers, because there’s a lot of TRB papers on the accuracy of models and things like that. [00:16:07] But there’s also the politics behind this and what I’ll call the highway industrial complex. There’s a lot of money behind highway building. There’s a lot of politicians who hear from constituents, right? Yeah. You get this feedback because there’s a lot of auto dependency out there that is requires driving. [00:16:21] There’s even here in San Francisco, you try to take away a couple of parking spaces and people go nuts. And so I think there’s some of that too, where the politics above everything and even in places like Texas, where we’ve learned recently, where the governor is pushing for. Put his thumb on the scale for complete streets and other things. [00:16:36] And yeah, San Antonio specifically, but also expanding the highway in Austin, expanding the highway in Houston, the I three 45 connection in Dallas, all that stuff that Megan Kimball talked about in her book, the governor is put his thumb on the scale. And so there’s a higher level of politics that’s getting into this too, above the modelers heads that, pushes in that direction as well. [00:16:53] And I’m wondering if you all can speak to that a little bit. [00:16:56] Ben Ross: We certainly had that in Maryland where, I wrote this letter pointing out that they must have changed the numbers by hand in their outputs and it made a big brouhaha because it was holding up the approval of the EIS and Governor Hogan. [00:17:14] Then Governor Hogan wrote a, to President Biden, they then shuffled it around and got a non committal technical analysis and said okay, we approve it. No, it’s very political. [00:17:28] Joe Cortright: It is political, but this very complex, very technocratic. Process serves as a smoke screen for those politics, right? [00:17:39] People say, oh, it’s not, not me as a politician saying this. We have this projection. That’s really precise down to. We know, 40 years from now within 50 cars, how many cars will drive on a particular stretch of roadway? And again, that’s opaque and unassailable to most folks. Helps legitimize the politics and if the modelers were more realistic, more accurate and more open, then presumably some of the politics of this could be changed. [00:18:13] But essentially, the modelers are giving cover to that highway industrial complex. And unsurprisingly, traffic firms in this country make money by giving highway departments, the studies that prove what they want and justify what they want. So there’s a strong selection effect here. If there were somebody who produced more realistic forecasts, they wouldn’t get much business from state highway departments. [00:18:39] And I’ll just give you an example from. From the I 5 bridge project here in Portland, we had modeling done more than 15 years ago, 18 years ago, they did modeling and their modeling predicted how much traffic there would be on the I 5 bridges in the no build scenario. If we did nothing. That model said that today there would be about 160, 000, 170, 000 vehicles. [00:19:05] On the I 5 bridges. In reality, it’s about 135, 000. There’s actually been no growth on the I 5 bridges. So we actually know, we’ve run the experiment. We know that their model is wrong, but the people who did that model never looked back, never looked to see whether their model made accurate forecasts. [00:19:24] They were never accountable for that. And in fact, the interstate bridge, Replacement project 18 years later hired exactly the same person to do the modeling for them, even though his previous estimates were wildly wrong. And the reason they did that is he gave them the numbers that they wanted that create this false picture of. [00:19:46] Future increases in traffic that can be used to justify the project. So there’s not only is there no accountability, but essentially there were rewards for producing fictitious information that helps justify these. [00:20:00] Jeff Wood: We’ve all seen those charts online, looking at the models and how basically we know now that VMT is decoupled from economic growth and you see the flat line, or at least a little increase, 3 percent or something like that line along there, and then you see all of the models from the past go straight up. [00:20:15] In a line hockey sticks for 1990, 1995, 2000, 2005. And you’d think at some point we’d start to get to a point where you look at those hockey sticks and you’re like maybe we’re not doing something right. Because it keeps on going this direction, even if we were adding more capacity. [00:20:29] Ben Ross: Yeah. I think there is a growing degree of skepticism. But there’s a tremendous amount of, I would say, institutional momentum and a political interest group power that has to be overcome. [00:20:43] Joe Cortright: And one of the concepts that I think have gotten a lot of attention and a lot of understanding is the notion of induced travel or induced demand that when we widen highways, we stimulate additional travel more trips, longer trips. [00:20:56] And we also stimulate. Or disperse sprawling patterns that create additional travel and car dependence, which I think is what Ben was alluding to when he was talking about how we know that it doesn’t matter if you widen the roads, because people’s behavioral response wipes that out. That’s a real thing, and it’s very well established in the scientific literature. [00:21:16] Now it’s referred to as the fundamental law of traffic congestion that essentially additional unpriced roadways in urban settings are going to be saturated by traffic. The moment they open and. These models essentially deny the existence of induced travel, and they do that in a very devious specific way, which is. [00:21:39] They assume that traffic will increase dramatically no matter what you do. Essentially, they have a baseline forecast or a no build forecast that says, even if we do nothing, traffic will grow exponentially, even though, and this is true in the case of interstate 5 bridge project in Portland, the structure is actually a physical capacity. [00:21:59] It cannot handle more traffic, but the models show that even though it handles about 4800 vehicles per hour at the peak hour. But somehow in the future, that’s going to rise to 7, 000 vehicles per hour in the peak hour. There’s no physical way that facility can accommodate that, but by pretending that traffic will increase anyhow they’ve essentially erased the idea of induced demand because they have a high level of travel, whether you build the project or not. [00:22:29] And that’s just plainly unrealistic and unscientific. But nothing has been done to correct that. And just by the way, that’s the basis of the legal challenge that will likely be filed to the I 5 bridge project is that essentially they’ve simply ignored the science of what we know about how traffic works and made projections that are simply unrealistic, but we’ll have to see whether a court can be made to recognize that’s a fundamental flaw in this analysis. [00:22:57] Ben Ross: Yeah, let me add to that. One of the things that I learned doing the research for this, or I wouldn’t say learned, but came to think, is that I’m not sure that you can really add induced demand to this type of model. I don’t think anybody’s figured out a way to do it. You would need a completely different way of Constructing the model and a lot of people talk about you left the induced demand out of your model, but, there’s part of induced demand is the trips that people currently would like to take, but don’t take because traffic is so bad. [00:23:35] How do you know how many of those trips are out there? There’s no way to measure that. [00:23:39] Joe Cortright: Yeah, [00:23:39] Ben Ross: the only thing you can check your model against is present conditions, but present conditions involve a lot of. Congestion and trips that aren’t taken. It’s a concern of mine that people are reacting to this problem induced demand by saying, go add induced demand to your model, but I’m skeptical that there’s a way to do that. [00:24:02] With any accuracy, [00:24:04] Jeff Wood: could you reverse engineer like you mentioned the takedown of the Embarcadero freeway here in San Francisco. Could you like reverse engineer like the trips that disappeared because of the highways that were taken down? I guess they’re not enough of them to do that, but I’m just throwing stuff up there. [00:24:17] Yeah. [00:24:17] Ben Ross: That’s why I say, I think I can’t prove this, but I think the best prediction you can make right now is that no matter what you do, the level of congestion will be the same. I think that’s better than anything you can do with a complex model. [00:24:35] Joe Cortright: I think that’s right. And just 2 things. 1, that’s what the modelers do. [00:24:39] They go. Oh, it’d be really too hard to include induced travel. So we’ll keep this incredibly unrealistic model because it would be difficult to do that. But I think to Ben’s point, and to your point, we know from examples of traffic evaporation that when. There are sudden constraints on traffic capacity, whether it’s a construction project, like the Sepulveda pass in L. [00:25:01] A, or a collapse of a bridge, like in a Minneapolis or temporary closures, like the ones that we saw, in Baltimore in Atlanta and in other cities that traffic adjusts almost instantaneously. To changes in capacity and what that should tell us from a very fundamental standpoint is traffic will respond to capacity and that’s true in both directions. [00:25:25] If you build more capacity, traffic expands. If you restrict capacity, traffic declines or stays the same and has been saying, you don’t need a complicated model to validate that very fundamental notion of how traffic works. [00:25:39] Jeff Wood: Another issue that’s come up over the last few years that I’ve seen articles about is just like the self certification that happens for highway agencies. [00:25:46] There’s Texas and California know how to have that happen in their NEPA processes. So there’s like kind of the wolf guarding the hen house kind of situation going on in a lot of these as well. [00:25:55] Ben Ross: I haven’t seen FHWA do any real oversight, to be honest. Now, understand the same models are used for FTA, for rail transit projects [00:26:08] Jeff Wood: to predict the ridership. [00:26:09] Ben Ross: And [00:26:10] Jeff Wood: those are frustrating me as well, if I’m honest. [00:26:13] Ben Ross: But. FTA heavily oversees them completely different from FHWA because they’re deciding who gets the money, but their oversight is really aimed at making sure that it’s a level playing field rather than making sure it’s accurate. I think there are some technical reasons why these models will perform better for a brand new rail line, or for that matter, a brand new roadway than they will for widening an existing roadway. [00:26:46] I also think that the FTA models for new rail lines would probably be just as well. And I suspect would give you pretty much the same numbers. If you just assume that nothing you do will change the amount of traffic on the highways. [00:27:02] Joe Cortright: And it would be a lot less expensive. Yeah, I think Ben’s absolutely right. [00:27:06] And there’s clearly a conflict of interest when you let Texas pick its own models. But frankly, and I’ve looked in detail at the I 5 project, the FHWA review raised a few issues, but then at the end of the day, didn’t hold anybody’s feet to the fire. There’s no accountability here. So it’s air, paper, tiger. [00:27:26] They could, in theory, impose some restrictions, but in reality, they don’t. They did the same in Houston too. Yeah, [00:27:33] Ben Ross: I should add that in Texas, what we’ve seen also in Florida is that when you go deep. Into the report, the page 3 dash 72 of appendix D, you find out that they never actually bothered to run the model. [00:27:50] They just came up with some numbers. We’ve seen that for some roads in Miami that we looked at. That’s what they did for I 35 in Austin. [00:28:02] Jeff Wood: This goes to another thing. And going back to the transit and highway discussion, it’s just the idea of planning for transportation in regions, right? Thinking about what you should do in a regional context. [00:28:11] And I think. One of the things that I, think that we should be thinking more about, and we’ve had folks on the show talk about this is that scenario planning ideas where you can use models, but then you look at the different scenarios that you want for a future and think about all of the different modes that are in the hopper in terms of like how you want your region. [00:28:28] Look, you had, envision Utah. You had envisioned central Texas, you had all of these different, processes where you looked at the models and they gave you an output, but you had different scenarios where people were like, I want this final output. I want this, world to live in. [00:28:42] And I’m wondering what you all think about how we should be thinking about transportation planning from that regional perspective. So we don’t continue to get these highway projects that are shoved down people’s throats. [00:28:55] Ben Ross: I’m not sure that the models are a help to that. The traffic models may be impediment for that because they’re always going to tell you, you need highway capacity. [00:29:04] Joe Cortright: I think there is some value to thinking about scenario planning and starting with some sort of fundamental analysis from what we know about cross sectional relationships in the country. So one of the things I’ve worked on is we’ve looked at variations in VMT per capita across metropolitan areas, which we know is strongly correlated with housing markets, with land use patterns, and to some extent with transportation investments. [00:29:25] But, we can show for Portland, for example, that the 20 percent less. Then the average resident of a typical large metropolitan area in the United States that’s metro wide and that saves Portland households on the order of a billion dollars a year in transportation expenditures, which then is money that’s available to be spent in the local economy. [00:29:46] So that’s really significant. Arguing from a scenario planning perspective, I would say. Things that you do that discourage the growth of VMT are going to produce significant economic gains for your region, and that should be a guiding principle. I think the problem with the modeling is it’s being employed so narrowly in the context of rationalizing a particular project that it never informs those larger strategic decisions about where we invest our money. [00:30:18] And that’s unfortunate, and I’ll just I’ll give you 1 more example for Portland. We have a regional transportation plan required by the federal government for funding that also is subject to state laws. And we have a climate policy in the state of Oregon that says we’re going to reduce vehicle miles traveled in the Portland area by about 12 percent over the next in the aggregate over the next 2025 years. [00:30:43] The trouble is. The traffic forecasts that were used for the IBR project, which are separate from that regional transportation plan contemplate a 25 percent increase in driving over that same period of time. So there’s a complete disconnect between what our strategy says. We’re going to do for climate reasons and what the modeling says. [00:31:04] That’s used to rationalize this highway expansion is being presented with and 1 is saying we’re going to decrease driving by 10%, which essentially means we don’t need any additional highway capacity, but the highway departments in league with our regional government, unfortunately, have said, oh, no, we’re going to plan for 25 percent more driving when it comes to actually sizing roadways. [00:31:25] And again, I think that’s another legal, just flat out ignored state climate law in here in Oregon and moving ahead with this project. It’s all concealed deep within the appendices as Ben said of these highway plans. They didn’t acknowledge that you have to sift through it to find out that there’s a complete disconnect between what our plan says, and what the highway planning modeling is doing. [00:31:52] Jeff Wood: It also points to something you guys mentioned in the article, which is the tolling aspect of things and how we could actually make some headways in reducing congestion and making sure that the VMT doesn’t increase departments are scared of it and people are scared of it. For some reason, [00:32:07] Joe Cortright: I think there’s very powerful evidence that the reason you have this induced travel is because all highway capacity does is lower the price of driving. [00:32:15] The time cost of driving, and it’s very straightforward economic principle that when you lower the cost of something or lower the price of something, people use more of it. And the reason we have traffic congestion in the United States is because we don’t have pricing to reflect back to people. The cost of the very high cost of providing highway capacity at peak hours. [00:32:37] Now, I’m. Optimistic that in what about 2 months now, when we finally turn on congestion pricing in New York City yeah, fingers crossed we’ll get a manifest demonstration in the United States of how pricing can play a role and actually produce benefits. I think that the unfortunate thing about pricing is people just look at. [00:32:56] I’m getting told for something, and I’m not getting any benefit, but I think in reality, if it’s done correctly, pricing can actually produce significant benefits to folks, both for the people who pay the toll will get faster travel times. But more importantly, I think it helps all the other modes of transportation work better and chiefly buses. [00:33:15] And we know that buses are especially true in New York City. The buses are. Unattractive because they’re stuck in traffic. So if we can ease the flow of traffic, then buses become not only much more attractive, but much more efficient as well. They carry more passengers further and faster. So here’s hoping that we get a positive result from congestion pricing in New York shortly. [00:33:37] Ben Ross: Yeah, I agree with that, but I also am concerned that unfortunately so much of the U. S. doesn’t really have an alternative to driving and I don’t think congestion pricing will work for people unless there’s a good transit alternative and unfortunately, there are limited number of places in the U. S. [00:34:00] where that good transit alternative [00:34:02] Joe Cortright: exists. I have to disagree somewhat with that on this issue, because I think a lot of people do have an alternative. It’s to travel at a different time. And that’s another flaw in these models. The models, essentially, they have a level of demand, a number of trips, and they say between 8 a. [00:34:19] m. and 9 a. m. and between 4 p. m. and 5 p. m. this many 100, 000 people will travel on this route, and that’s an artifact of the way the model is constructed. In reality, A significant portion of those trips are flexible. People could take them earlier. They could take them later. And what peak hour pricing in particular does is provide people with incentives to change their travel time, even if they don’t change their travel mode. [00:34:47] And that can be a really significant way of reducing traffic congestion. And I think if we wait until we get enough transit to make transit an alternative for every trip, you essentially never do pricing. But there are a lot of benefits to be gained from pricing in terms of signaling to people how other destinations and other times would be better and that then that would produce significant benefits for the people who are traveling. [00:35:14] Jeff Wood: Considering 19 percent of, at least when I last looked at the data, which was pre pandemic, 19 percent of trips are for work, but most of those trips take place at that time period. And so if you spread them out a little bit more, they might be actually a little bit more easy to provide those in some way. [00:35:29] It makes me think too, as we were seeing what’s happening with the pandemic, after the pandemic, in terms of. Work trips and things like that. And the spreading of all other trips throughout the day, it makes me wonder like what the, the new paradigm might be and how the models aren’t really going to address that either. [00:35:44] Ben Ross: Yeah. The models basically only do rush hour and then what they do is they just apply some multiplier. To get the rest of the day, and I think that the modelers are a long way from catching up to the changes that have happened since pandemic. [00:36:04] Joe Cortright: That’s a great point because and we’ve looked at that with the, with regard to the interstate bridge project here in Portland, they essentially just blow off anything that happened during the pandemic. [00:36:14] It’s industry standard practice to ignore. What happened in the shift to work from home? They haven’t even tried to adjust their models to show what’s going on there. And that should speak volumes to anybody who is serious about sort of the science of this. This is probably the most dramatic, both short term and long term change that we’ve seen to Travel behavior, occurred during and then subsequent to the pandemic and we’re seeing it, in the collapse of commercial office values in cities throughout the country that, this market has fundamentally changed and yet these models are pretending that. [00:36:53] Nothing happened in terms of work from home, and I wrote a piece for city observatory that they’re planning for a world that doesn’t exist. That is something that should call into question all of these projects. [00:37:07] Ben Ross: No, I just want to jump on that phrase industry standard practice, because what we basically found in our paper is that fraud is an industry standard practice. [00:37:20] Joe Cortright: Yeah, and again, they have post processing, which there’s actually a kind of a legitimate basis for applying a model when you have to, if you have a regional travel demand model, and it has a very high level of aggregation and you want to tease out results for a particular roadway that’s not maybe modeled directly in the model, or for a particular time that’s not modeled. [00:37:43] You may have daily volumes and you may know for a whole quarter and so post processing. Is designed to be well, the model didn’t predict for this stretch of roadway, but can we infer from that model? What the hourly volume on this particular road would be that’s legitimate. But the way it gets used in the business in the trade, the industry standard practice is. [00:38:06] We just go in and we change the values that are there and we’ll call that post processing. And in the case of the I 5 bridge project, they raised some traffic volumes on one road, and then they lowered them on that same road at different times with no explanation. It’s not post processing. It’s just, we didn’t like the results of the model. [00:38:26] So we, we changed and put in different numbers. And that really is fraud as Ben says. What’s been the response to the [00:38:33] Jeff Wood: article so far [00:38:35] Ben Ross: among friendly people have been very impressed I have not heard any feedback from the establishment. They’ve just been ignoring it as far as [00:38:44] Joe Cortright: I know. And that’s the PR strategy here is ignore it and it will go away. [00:38:51] The traffic modelers simply refuse to engage. In any discussion of critiques of the modeling and the modeling process. And again, this is hiding behind this veneer of scientific legitimacy that the modelers are, have to have these strange and wonderful and inexplicable ways of coming up with really precise numbers. [00:39:13] That tells us what will happen in the future, but they’re not going to engage at all on whether those are accurate or reliable or defensible. So we get no response whatsoever. And you would think, given the severity of the criticism we made that they would have some basis for telling us we’re wrong. [00:39:31] But maybe Ben can correct me, but I haven’t gotten anybody right back to us and say, Fort Wright and Ross are wrong. And here’s why. Nobody has said anything like that. [00:39:41] Jeff Wood: Nobody. What are some of the solutions then? Is it in front of the policy perspective? Do you have to go into NEPA? What’s the, what’s the prescription here for fixing some of these issues? [00:39:52] Ben Ross: I think there are some real NEPA issues here. But it’s tough to untangle the scientific from the legal. I think, on a practical level, what you need to do is step back from these giant models and really rely more on your common sense. As to what a highway expansion will do and rely on the experience that it won’t solve congestion. [00:40:22] I think it’s really something that has to be handled at a policy level, you have to reach the scientific conclusion, which there’s plenty of scientific guidance for that widening highways doesn’t solve congestion and then that doesn’t call for a new fancy and different fancy model. [00:40:41] It calls for stopping building new highways. No new highway capacity on major highways. Now, obviously, as someone who lives in the suburbs, there’s an urgent need for, localized grid streets. That’s a different issue. But New Jersey made that decision a few years ago and unmade it, but I think. [00:41:02] There really has to be a political decision to stop adding [00:41:07] Joe Cortright: highway capacity. Yeah, I think, as Ben says, there is a NEPA issue. NEPA at its core is basically a requirement that you bring best available science to the question of what are the impacts of these projects. And that’s probably 1 of the most egregious. [00:41:25] Parts of this is that the process being perverted by highway agencies who are misrepresenting what the real impacts of these proposals are. So 1 of the things we’re hoping is that courts will hold them accountable. We would hope also that the US Department of transportation who is basically funding. [00:41:42] Both the projects, but then also the planning for this, they’re paying for the state departments of transportation to hire all these modelers to generate all this information. And there’s essentially no quality control over this. There’s no, no check on the veracity of this. And as we’ve shown, you essentially rehired the people incorrectly predicted what was going on last time to do the same thing again. [00:42:06] That shouldn’t be subject to, State Highway Departments picking their favorite planner to come up with an exaggerated forecast. There should be some national standards that say how this is done. Unfortunately, I think, in effect, FHWA is complicit in this. They know what’s going on. [00:42:23] They’re not stupid. But they simply roll over and let states do this. Because they’re about building highways, too. So there has to be some, I think, external direction to, to more accurately insist on more accurate analysis than what have been provided. But ultimately, I think it’s, in the case of Portland, we’re talking about 2 projects, the rose quarter interstate bridge that are more than a 1Billion dollars a mile more than a 1Billion dollars a mile for roadway. [00:42:51] And the fiscal implications of that are. Astonishing it’s, it’s bankrupting our state Department of transportation and taking away all the resources for all the other oxygen for transportation in the room. And so if anybody is concerned about transportation, and if we have money to maintain the system, much less make investments in smart alternatives, that’s getting consumed by these projects that are being rationalized by these thoroughly wrong traffic models. [00:43:21] Ben Ross: I want to just throw in a caveat here that a lot of the people who are actually writing the models are doing it in good faith. Typically, the model is written by your regional transportation, metropolitan planning organization. And then the actual runs on the highways are done by consultants working for the state agencies, or by the state agencies themselves. [00:43:52] The people who are writing the models. Very often are trying to do a good, honest job. I think in some cases, they’re trying to do the impossible, but they’re doing it honestly. And the really bad stuff comes in with the applications for individual highways. Which is typically done by a different group of people. [00:44:19] Jeff Wood: I appreciate that point. I have a number of friends who are actually are modelers and they, we’ll go to TRB and they audit a lot of papers and there’s a lot of literature about them and making sure that they’re, trying to do the best they can. And a lot of the modelers that I know are going to do models that are local and more related to transit and, active transportation and those types of things, or city models, for example. [00:44:40] But. I do think that there’s something, there’s some perverse incentives at that higher level for the highways. And that’s, a big frustration. I know that Joe, you’re mentioning the FHWA and, during the Biden administration, Stephanie Pollack sent out that memo saying that we should do fix it first and we should focus on environmental impacts and things like that. [00:44:57] And some States, completely freaked out and they even had Mitch McConnell and some other senators on them. And it was really, unfair to Stephanie and in a memo that wasn’t even enforceable. It was just like, these are our policy. Prescriptions and so that’s what you’re dealing with in certain states and it’s not all states, but it’s certain states that are really pushing back hard against any change and they want to just build their highways. [00:45:19] I should note that. A couple of days ago, I posted an article from inside climate news about the zombie highway in Birmingham, Alabama, and the way that got put into the system, richard Shelby said this is on the Appalachian highway network, this line. And so when the Appalachian highway network gets funding through the IIJA, then that highway, that. [00:45:40] Probably didn’t need to be on that system gets funded. And so it continues and continues, even though the state’s this isn’t part of our plan. What do we, why do we continue to do this? But it has money. So let’s build it. Let’s build 4. 5 billion highway on the outskirts of Birmingham to, as a SOP, probably to developers, in my opinion. [00:45:56] So those types of things are the things that keep it going. Even if there are folks that are working really hard to get the models right. And it’s frustrating to see people push back and then get, Overwhelmed by we just talked about tsunami of pushback. [00:46:10] Joe Cortright: There’s one other thing I’ll mention too, which is we have state highway departments who are producing these wildly exaggerated traffic forecasts and oftentimes they’re doing it for facilities that are going to be told or have tolls as a portion of the project. [00:46:23] And the really interesting thing to me is private financial markets go. No, we don’t buy your modeling at all. And in fact, there’s a term of art in the business investment grade analysis. So if you’re going to sell bonds to pay for part of your project, folks on wall street will not let you come in with your silly forecast. [00:46:45] They will make you hire somebody else who will be much more realistic. About traffic levels, and we had investment grade analysis prepared for the predecessor of the interstate bridge project and where our state highway department said, Oh, 185, 000 vehicles a day are going to go across that bridge, even after you told it roughly 180, 000. [00:47:03] The people who did the investment grade analysis said, no, you charge a toll on that project and fewer than 120, 000 people will go across it, which is the difference between needing to expand the bridge and not needing to expand the bridge at all. So the interesting thing is if you get people who do the modeling who aren’t. [00:47:23] Being selected by state highway departments to rationalize these giant projects, you get a very different answer. And in fact, using essentially the same tools, but just plugging in slightly different assumptions. And I think that’s 1 of the things that I would walk away from this. I tell people is this isn’t so much an Oracle. [00:47:44] As it’s an Excel spreadsheet, right? People think, oh, the model produces this specific answer. No, the model is like an Excel spreadsheet. And if you change some formulas, or change some of the values that feed into those formulas, you get totally different results. And the real question that nobody ever seems to dig into is. [00:48:02] Are those models and the assumptions that go into them and the values that are plugged into those models. Correct? Because you can use those same models. And as Ben said, in some cases, you can get any result you want, but clearly, when different people who have money on the line are asked to do the modeling, you get a very different result. [00:48:22] And that speaks volumes about. The veracity and the reliability of the forecasting that state highway departments are doing. [00:48:31] Ben Ross: I once helped write a groundwater model on an Excel spreadsheet, but I have to add to that, though, this overoptimism in the Wall Street stuff too, even though it’s nowhere near overoptimistic as the state highway administrations, all kinds of people have lost money building toll roads and investing in toll roads. [00:48:55] Yes. Because there’s a lot of salesmanship goes on Wall Street too. [00:49:00] Joe Cortright: They’re just not as egregiously bad as the state ones, but you’re right they on average tend to overestimate future traffic levels. [00:49:08] Jeff Wood: The articles in Dissent Magazine, folks can check it out. I’m going to have it in the show notes as well as linked in the article where this is posted. [00:49:15] Where can folks find you if you wish to be found? [00:49:17] Joe Cortright: You can find me at City Observatory, cityobservatory. org and over the last couple of months, we published a series of articles that dig into more detail in the way that the interstate bridge project has falsified or exaggerated traffic projections to try and sell this 7. [00:49:33] 5 billion dollar bridge project between Portland, Oregon, Vancouver, Washington. You can find me [00:49:39] Ben Ross: on social media on Mastodon at Ben Ross transit at Mastodon dot [00:49:45] Jeff Wood: social. Awesome. Ben and Joe, thanks so much for joining us. We really appreciate your time. [00:49:50] Joe Cortright: Thanks Jeff. Keep up the great work. [00:49:52] Ben Ross: Yeah. Keep it up.
