Patrick DeCorla-Souza, Tolling and Pricing Program Manager, FHWA
Lee Munnich, Humphrey Institute, University of Minnesota
Kenneth Buckeye, Minnesota Department of Transportation
John Doan, SRF Consulting
Office of Innovative Program Delivery
Federal Highway Administration
Eighth Part of a Webinar Series on Overcoming the Challenges of Congestion Pricing.
Thank you, good afternoon or good morning to those of you on the West and welcome to the overcoming the challenges of congestion pricing webinar series. My name is Jennifer Symoun, and I'll be moderating today's webinar, which will focus on economics of congestion pricing and impacts on business. Please be advised today's seminar is being recorded.
Before I go any further, I do want to remind you that are calling into the teleconference for the audio that you do need to mute your computer speakers or else you'll be hearing audio over your computer as well.
Today we'll have three presenters - Zabe Bent of the San Francisco Transportation Authority, Ewa Tomaszewska of HDR, and Jose Holguin-Veras.
Today's webinar will last 90 minutes. We'll take questions following each presentation and then take questions at the end if time allows. If during the presentations you think of a question, please type it into the chat area, send it to everyone, and indicate which presenter your question is for. The presenters will not be answering the questions during their presentations but will take about five minutes after each presentation to answer the questions that have been typed into the chat box. If we're unable to get through all of the questions in the time allotted, we'll attempt to get written responses from the presenters and send them out with the follow-up information.
The PowerPoints used today are available for download from the file download box in the lower right corner of your screen, and again, the session is being recorded, and the recorded presentations and transcripts will be posted to the pricing website within the next week or so and I'll send out a notice once they are available. We'll now go ahead and get started.
Our first presenter will be Zabe Bent of the San Francisco Transportation Authority. Zabe, I'll bring up your presentation and then you can get started.
Great. Good morning for those of you on the West Coast and good afternoon again. Really excited to hear some feedback from people. To give you a little bit of background on our study, we conducted a feasibility study of congestion pricing in San Francisco, and it's focused on an approach to managing congestion in San Francisco. I'll spend just a few minutes talking about the program itself and then I'll get into the sort of details of the economic analysis that we did.
So the first thing we wanted to do was establish a case for why we should study congestion pricing in San Francisco, and it does focus of course on congestion itself. We are the Congestion Management Agency for San Francisco and we also do the long range planning for the city as the Transportation Authority. We also wanted to incorporate some of the other factors that we're seeing on the ground, so we looked at not just the technical details but also where is congestion worse in San Francisco, where does it impact not only motorists but also transit performance. There is a lot of concern and advocacy to look at pedestrian impacts and such, so we wanted to look at not just the auto additions on the street but the whole picture. In addition to that, we did include impacts on our economy. We sacrificed over $2 billion to congestion in 2005 and we expect that to increase in the time frame of the project, as well as very long range planning and environmental impacts. We also have a few documents in the city that require us to evaluate congestion pricing, or in some cases actually call for implementation of congestion pricing.
So just to sort of focus in on what the actual scenario is: as I mentioned, we wanted to focus on when congestion is worse in the city and so we looked at a peak period. Is everyone hearing me okay?
There's a little bit of an echo. I don't know if anybody has their computer speakers on as well as the presenters, but there's a slight echo in the audio - but keep on going.
Okay. So we did focus on the peak periods. We heard from a lot of people that we don't have congestion like New York or London, where you might see 10 hours out of the day where there's sort of bumper to bumper traffic, but we do have very peaked traffic. So, we focused on three or four hours in the morning peak and three or four hours in the evening peak, and not in the middle of the day or in the off-peak and not on the weekends either. We also looked at discounts, and this was both to respond to public feedback as well as to respond to some of the concerns that we heard from the business community and from low income residents about some particular groups that needed some protection of some kind. And then, finally, we looked at a couple of sort of programmatic constraints or designs, concerns, and what is a $6 daily cap that helps you address concerns about parents of school aged children as well as businesses that are transportation-based or transportation-related. Businesses that may have to make multiple deliveries throughout the day essentially said we don't want to see a fee of $21 over the course of the day just because we have to make multiple trips in and out of the zone; we want to have some sort of protection so that we can predict what the fee is going to be, so we established a cap, and then a rebate on bridge tolls for existing bridge toll payers. We have two bridges that approach San Francisco, from two corridors but not our third, and so people felt that there should be an acknowledgment of geographic equity, and then a fleet program for businesses to help minimize the administrative burden, so instead of having to reconcile their invoices every day, they could look at them over the course of a month or a quarter.
So we've modeled about - I want to say we looked at about 100 different scenarios when we were looking at the number of model runs that we included for the study, but we looked at about a couple dozen discrete scenarios geographically. We had a lot of different permutations, but we really ended up with just a couple dozen. At the end of the day though, we wanted to look at the best performing scenario, and that needed to be a question of not only the performance of the scenario - is it reducing traffic and congestion and delay - but also what are the impacts on transit, what are the impacts on the sort of shift to transit. Then, finally, is it financially feasible? Do we know that we can actually come up with a program that can support itself in terms of paying for the program as well as paying for some of the transportation investments that need to accompany the program? So, we looked at very, very small zone and then a city-wide zone and ended up somewhere in the middle, and this represents about a fifth of the city. The underlying map is a map of the trends and auto speeds that are operating below eight miles and ten miles per hour, respectively, in the peak period, and as you can see, we didn't want to say "well, let's draw a line around all of the congestion," but we wanted to find a way to focus on where the congestion is worst. This represents, as I said, about a fifth of the city in terms of land area, but in terms of development that we expect - both within the time frame that we could deliver a congestion pricing program as well as sort of looking again at the longer term - this represents an area that will accommodate about half of the city's job growth and a significant portion of the city's housing growth. So, this is an area where we wanted to make sure we could keep people moving in the context of a pretty mature network where we don't really have the space to sort of build the lanes or build our way out of the problem.
So, we did also include - as I mentioned - a package of investments and different programs. We wanted to make sure that not only could that package accommodate all of the different growth to the different modes (people are going to eventually decide to drive at a different time of day or take transit or walk or bike to their destinations in this zone or outside of this zone if they're going the opposite direction), we wanted to make sure that we could improve transit significantly enough so that the number of people on those transit services were not sort of seeing an undue burden, and that the loads were not too excessive on those different lines and services. And so, we devised a program of up-front improvements as well as ongoing improvements and then fed that back into our financial analysis to make sure that the net revenue that we expected the program could generate was actually sufficient to accommodate all of those improvements. And you see here the way we split the up-front costs and the ongoing costs and the types of programs. We do expect somewhere between $60-80 million a year from the different types of scenarios, and I'll explain those in just a moment.
So this is a comparison of the scenarios that I mentioned. These are the three best performers for a couple of different reasons, and these are just a couple of the - or just a few rather, of the metrics that we evaluated. So the first scenario is the Northeast corridor, and that's the one that I mentioned previously. That's the best performer. It would be a $3 fee in both directions in the a.m. and the p.m. peak hours. And we heard from a lot of people that this seems like it could be a good sort of permanent scheme, but we wanted to evaluate for something that would be a sort of more measured approach, for something that would be a first step so that we could pilot and demonstrate the concept and improve the concept, and then move to a more permanent system. and so we looked for two additional scenarios that addressed public concern or specifically focused on a key area of the city that is the most congested. First is the Northeast corridor, and this is a p.m. outbound-only charge, and it would be a $6 fee, and this was in response to a lot of businesses that said: "well, what about a fee to get out of the city rather than to get into the city, so that people can still come in to go to the symphony or have dinner with friends or what have you?" And then the third scenario that you're seeing here is the Southern gateway, and it evaluates a $3 fee, again in the a.m. and p.m. in both directions, but instead of, [as] in the Northeast corridor, demonstrating the geographic area that is most congested, focused only on the corridor that is most congested. For those of you who are familiar with the Bay Area, this would essentially be pricing trips in and out of San Francisco, to and from Silicon Valley or the South main peninsula.
So we looked at a range of different metrics and wanted to understand how the different metrics would impact the priced area as well as the city as a whole, and one of the things that I would say is the first sort of reading criteria, or evaluation criteria, on the economic side that we looked at was the change in daily person trip. So, a lot of people were concerned that a congestion pricing fee would suppress trip-making overall, and what we found is that we just don't expect that to happen, and part of it is because it's a peak-period-only fee, and so we expect people to shift the time of day that they might travel as well as to shift to other modes. And the other aspect is that we expect to be able to use that net revenue to again reinvest in improvements in the transit system and walking and biking facilities. So that was our first cut of looking at the different changes in the scenarios. If we found a scenario that had significant trip suppression, we would call that scenario infeasible.
So a lot of people are really curious about how we looked at the economic analysis as well as trip changes in general. One of the first things we did after looking at the overall change in the volume of trips: we looked at how trip patterns might change. We are very blessed with an activity-based model, and so what we can do is just aggregate some data and begin to sort of follow individuals in our model. We looked - we call these simulated versions of individuals or residents of the Bay Area, and we wanted to understand what types of decisions might people make, and sort of how their changes might impact them. So here are just four examples. We wanted to pick people in different parts of the region and also taking different types of modes or making different types of travel decisions, so we have here someone who's driving alone and then picking transit, and also carpooling. And to be able to demonstrate not only to ourselves, but to the public, the types of changes that people might make - and this is based on not only our sort of general travel demand model, but also based on market research where we asked 700 drivers to and from the downtown area what types of decisions might they make if they had a $.50 fee and a more flexible work schedule, or a $5 fee and a more frequent transit available to them. And what we found is that people would be able to take advantage of different options. A person driving alone from within San Francisco could potentially save 15 minutes a day, and so even though their costs may change slightly, they're saving time in some ways, and in some cases they might actually save on money. So this was helpful just to make sure that our model is behaving as we might expect, and also to be able to demonstrate to people how their different trips might change or their trip patterns might change.
So then, we also aggregated all of this information to look at the overall user benefits and sort of economic and social benefits. So first looking at the congestion charges, the travel time savings and the vehicle operating cost savings. So people might be traveling faster or traveling less because they have more direct route, they have a faster route, what have you. Obviously, the travel time savings - we found that about half of a trip is currently lost to congestion-related delay today, so any travel time improvements due to reduced congestion would obviously accrue to all of the different travelers into and out of the congestion zone. Then we began to sort of take a very conservative look at the safety, health and environmental benefits (and we based this on work that we've done with other partner agencies, as well as work that we've seen in London and Stockholm for more active transportation), so walking and biking, a reduction in pedestrian incidents and bicycle incidents, as well as air quality improvements and such, and estimated the costs here. And then for the total benefits, obviously the annual cost and benefits and the overall annual social benefits, which you can see is at the bottom line, is somewhat robust. And so we felt fairly confident that the overall changes would be pretty significant on the positive side. We didn't stop there, though.
We wanted to then look at how those travel time benefits accrued to drivers and transit riders, and then to San Francisco travelers and regional travelers. And part of it was understanding what are the benefits to motorists, making sure that motorists are actually seeing a benefit for the fee that they're paying, and then also, looking at the sort of system-wide benefits, both within the zone and also the ripple effects that might accrue outside of the zone. And you can see here again positive net benefits on that side.
So the next thing that people wanted us to look at - and sort of the next couple of slides really focus on addressing some of the concerns that people had as well as some of the misperceptions that people might have about how people get around in San Francisco. The first thing is looking at business impacts, specifically retail, and sort of recreational trip-making in the downtown. A lot of people said "we want to have lots of drivers because lots of drivers means congestion, and congestion means economic activity," and we said well, let's figure out how many drivers are coming to the downtown, how they're spending when they come downtown, particularly in the sort of later hours of the off peak - sort of between three and six in the afternoon and evening. We wanted to understand how are they coming, how much are they funding, and how frequently are they coming. And it was shocking to a lot of business owners to find that even though drivers do come and spend a lot of money loading up their car, they don't come as frequently as transit riders or pedestrians do, and so over the course of an average month, they're getting about the same, or if not more, from transit riders than pedestrians. That, coupled with the improvements that we could make to transit, net benefit - or net revenue, rather - from the program and also to walking in and biking facilities could increase the foot traffic, and then over time begin to increase the sort of, or maybe even some of, the retail impacts that businesses perceive. And that we did as sort of an adjunct to the study just to understand how people are traveling and spending.
The next question that we got on sort of economic impacts was impacts to lower income travelers. And specifically, people were concerned that lower income travelers were going to be feeling even more of the burden from this particular program than a typical driver, and what we found first of all is that most of our peak period travelers are actually not low income drivers, and less than 5% of peak period travelers are low income drivers. And sort of having an understanding of how many people might be affected was the first thing, and the second thing was that people wanted to see how we are incorporating this into the discount package, so a discount of 50% for low income travelers was included in what we modeled. And then also, what we heard from people was "well, that's all good and well, you know, a discount for low income drivers sounds great, but what would be even better and preferred would be a means-based fare assistance." So, instead of having a discount for drivers, having a discount for low income transit riders, to sort of encourage them to make transit decisions and sort of reduce the fare of transit. And less so within the City of San Francisco, because we already have a sort of lifeline pass, but as you get to some of the more regional transit providers, those costs can be somewhat significant, particularly if you're traveling for a far distance.
So another concern that we heard from businesses and from people on sort of the economic impact of a program like this was: people in the peak are - all of them have to be traveling the peak. They're already making decisions to move to a different time of day if they can, and everyone traveling in the peak is going to work or school, and we found that simply isn't the case. There are quite a bit of discretionary trips after it trips taken during the peak, as you can see from this chart. And obviously, during the peak it's a very high number compared to the sort of daily average, but just about 28% of daily travel is actually work-related in San Francisco. In the peak, that's obviously more than 50%, but that still leaves a healthy portion of trips where people could begin to make different decisions about the time of day they travel, and that's pretty much pointed out by the information that we found from our market research.
I'm seeing a lot of questions and I promise I will come back to them. I only have a couple more slides to share and we'll come back for some questions.
So: distribution of trips in the downtown area. A lot of people said "we already have pricing on the North Bay and the East Bay with the Golden Gate Bridge. We think it's all those people from all of those other places that are coming in and bringing their cars and causing all of the congestion. We should have a fee for the southern gateway. We should focus on those regional trips and not charge San Francisco." What we found, however, is that people from San Francisco are about 50% or more of the trip-making. When you take into account even trips within the Northeast corridor, what we found is that they're well over 50%--almost 70%--of total trip-making or auto trip-making in the downtown area. And this was shocking to a lot of people but it was helpful to be able to point out that having a regional program would do two things: one, it wouldn't have as much of an impact on congestion reduction; and two, because it's a regional program, it would be charging those regional travelers, and we would have to return that money to folks in that corridor and not spend any of the money within San Francisco. So having the sort of improvement for the corridors within the city as well as the regional corridors didn't really appeal to people, and understanding that it had to be a local and a regional solution because of both the benefits and impact was key.
Another concern that people said is that a lot of people don't have transit trips available to them. We're already a very transit-oriented city, and if they could take transit they would. We found that was actually not the case: about 80% of travelers felt they have a transit option, but our transit is highest in the peak at about 41 or 42%, so what we found is that only about half as many people as could be taking transit are. And so again, one of the goals was to figure out what types of improvements do we need to make in our network so we can get to a greater transit share for those people who could make a different decision, and what types of improvements we need to make those options more attractive.
So in our last round of outreach for the study, we conducted several different types of online and in-person sort of surveys. We had webinars, we had electronic town halls, and we also had a voter response during our public workshops. And so we compiled all of that information and ask participants to tell us their opinions on implementing a congestion pricing project in the next three to five years (so not immediately, but continuing to work towards something in the next three to five years). So, the first three options indicate some willingness to implement congestion pricing in some way, shape, or form, whether it's a permanent program. Obviously, a lot of support for a pilot or some modification of the existing scenarios that we presented. In all of our different discussions with people, there is a pretty healthy portion of the population who would like to see something else, even anything else, just anything that doesn't include pricing essentially, but that was relatively small compared to the number of people who were willing to consider at least a demonstration project.
So finally, our status. We had board action in December of this past year, accepting the report and the concept of congestion pricing is technically feasible. And looking forward to more information and further rounds of analysis through our environmental analysis, focusing on the economic evaluation-- wanting to see more information there, and then also wanting to see development of an expenditure plan for investments. A lot of people are very concerned that funds from congestion pricing might be diverted to other uses rather than focus on transportation - and specifically transportation for improvements in this area - and that an implementation plan for those improvements, so they can see they're happening at the same time that the program might be rolled out. And then the final concern people had moving forward is looking at a parking-based alternative. We do have parking pricing here in San Francisco already, and people feel strongly that it's something we already have control over. We already know what we can do here, so let's advance that program and see if there is a scenario that could be as robust as the corridor-based approach.
With that, I'll take any questions or come back to questions.
Okay, thank you. We're going to take about five minutes for questions, and like I said before, if we don't get through all of them, then we will have time at the end, or I'll get written responses. So I'll just start up at the top. The sound was cutting in and out. Some asked if you could repeat the information that goes with Slide 10, and that was about the spending travel patterns and the 3-6:00 p.m. time frame.
Uh-huh. So what we did was a survey of about 1,400 travelers to the downtown area in the holiday period as well as in the sort of typical period, and we asked people between I believe it was 3:00 and 6:00, or might have been actually 2:00 and 6:00. We wanted to get people who were sort of on the tail end of the off peak period, or the sort of shopping hours, as well as some during the peak to understand what their spending habits might be, and what we found is that even though drivers may come and load up their car and spend a lot of money on an individual trip, they don't travel and spend in the area quite as frequently as transit riders or pedestrians do, and so the spending habits actually on average are actually greater for transit riders and pedestrians than for drivers typically.
Okay. The next question is: wouldn't a discount for higher spenders affect accessibility for all criteria for transportation?
Not sure I understand what a discount for higher spenders is. Basically, I'm not sure I understand the question. It wouldn't be a discount for higher spenders. It could be a discount for low income residents or low income drivers, which could then be changed into a sort of discount for transit passes for low income travelers, but it wouldn't be based on high spenders. It would be based on your income. What we found is that low income drivers are actually a very small population, even in this area, particularly during the peak. A lot of our low income travelers in the Bay Area are already on transit. So we can certainly come back to that one, but--
It sounds like actually he clarified. He said that was just due to the audio cutting out but it was the other way around like you mentioned.
Next question: was the analysis of business impacts from San Francisco limited to the retail and recreational sectors? What about construction and other sectors? Also, was there any disaggregation for small and large businesses? And the person further clarifies and says they asked because broadly neutral impacts in London masked actual impacts on specific types of businesses when disaggregated.
So we didn't go into a lot of detail on this, but we did want to understand what sectors were more or less impacted positive or negatively, and then also what sort of size business was more or less impacted. And we can say pretty broadly that we wouldn't expect the type of impact on businesses by size to vary that greatly because the economic analysis focused on the sector as a whole, and San Francisco is something like 80% small businesses, so the overall analysis should apply, but that's one thing we would want to evaluate in more detail in the next phase of analysis - the environmental analysis. In terms of sectors, we didn't focus solely on the recreation and retail sector. It is a sector during the sort of first stab at some of the economic analysis that was flagged as a potential impact area, but then we looked back at some of the other analyses and sort of off-model analyses and the feedback that we got from different business types (and specifically businesses in this area), as well as the increase in foot traffic, we found that it probably wouldn't be quite as impactful to this particular sector. And so that's the reason we wanted to drill down on the sectors - because there was some instances that were raised in London and Stockholm - as well as looking at the sort of travel time impact.
Okay, there's three more questions I see. I'm going to try to get through them quickly here. How will fees be assessed from drivers? Through cameras, transponders or some other means?
It would be a combination of both. Essentially, we have an existing system in the Bay Area called Fast Track. It's very similar to Easy Pass or Sun Pass or - I'm not even sure where you're from, whoever is asking, but very similar to transponder systems that you'd have in other parts of the country. What we would do is link peoples' plate numbers to their Fast Track account, and between transponders at some of the sort of more highway-oriented entry points, we could have some sort of things there, but for the most part, people felt strongly that we needed to have a more London-based approach with the camera system because it can have a smaller footprint in the urban fabric. And we do have a pretty robust sort of neighborhood, even in the downtown area or surrounding the downtown area. So essentially as you drive through, your plate number would be detected.
Okay, doesn't 60% in favor of pricing also mean that 40% don't agree to it, so for every ten people you have four disappointed people. What could be done with them? I think that's more of an opinion-based question, but interested your thoughts.
Are the other slides in here? So I think the big question that we had - I'm sorry, I'm looking for a particular - there we go - a particular slide. The biggest issue that we found here is that people needed more information, and the 40% was really 20-some odd percent, I forget exactly, who were unsure and wanted more information and wanted to see the sort of next phase of analysis, and then 16% who simply wanted another choice. So the big concerns from the 16% were often the income impacts, which we have shown are not as great as people believe, and also from the business impacts, which again, we don't feel are as great as people believe. And so one aspect is simply designing the system and sort of refining it to understand how to address those particular concerns. The other is simply more information. A lot of people don't know that something like this would work in San Francisco. A lot of people have misperceptions or misunderstandings about how the program might be designed, so one of the things we asked in the slide that you're seeing here, right before we started all of the different feedback sessions, we asked people, do you or don't you support, or would you or wouldn't you support congestion pricing. And we asked the same question at the end, and it was about an hour-long session, and found that support and opposition flipped almost exactly, and it was interesting to see that people just needed to understand more about the program. And we'll continue to work on those concerns that people have in the next phase of analysis, but we do think that more information over time is the key there.
Okay, one last question and then we'll move on. On Slide 4, it showed yellow-depicted auto speeds less than 10 and highway speed. Does highway speed mean speed limit or something else?
It means actual speed. So this is based on our congestion management observations, which are actually observed speeds, and they're below 8, 10, and 30. Not at 8, 10, and 30, so in some cases they're much, much lower than that.
Alright, thank you, Zabe. We're now going to move on to the next presenter. If anybody does think of additional questions for Zabe, please feel free to type them in. Zabe, you're welcome to type in answers, or we can send you them to get answers later on.
We're now going to move on to our next presenter, Ewa Tomaszewska of HDR. Ewa if you give me a second, I'll bring up your presentation and you can get started.
Thank you, Jennifer, and good morning and good afternoon to everyone. I will start my presentation with a brief explanation of our study background and purpose, just to give a general context. Then I will provide an overview of approach, implementation, and sample of study results.
In 2006, New York City was assessing hypothetical scenarios of congestion charges formation and their effect on traffic patterns, average speed, and congestion across the New York City region. Certain business groups were expressing concerns over above plans and the costs to individual businesses that would or potentially could result under congestion pricing. Just as Zabe mentioned, for example, businesses were concerned that congestion pricing would reduce the number of trips by car, and that this would negatively affect the revenues. And at the same time, there was a sense of understanding that the current situation with current heavy congestion is not sustainable in the long run, and that congestion causes other costs in the economy. And so to provide a broader perspective on congestion, HDR was hired by the partnership for New York City to assess the economic costs and impacts of congestion in the region.
Our study focused on the effects of excess congestion: that is, congestion above the economically efficient level. This approach recognizes that some level of congestion is actually beneficial to the economy and that maintaining free-flow at all times would underutilize growth capacity, if not really waste it. Our point was that for a given road capacity, there was an economically efficient level of traffic congestion, and traffic above this level is excess because reduced speeds increase travel time and create various costs across the economy.
We define the efficient level of congestion as the traffic volume that would result if people would take into account in their travel decisions, by themselves, all the externalities that they create or delays they cause to all of the other travelers on the road. The actual level of congestion is above this level because people do not behave in this way. They took into account the cost of travel and they found the difference between the actual traffic and the efficient traffic is the excess congestion. Technically, the efficient level of congestion is determined by the point where the travel demand curve intersects the social marginal costs of driving curve rather than the average private costs of driving, and the graph on the next slide illustrates this relationship.
This graph is a typical graphical illustration, which many of you may have seen in textbooks or other scholarly discussions. It shows the actual traffic and optimal traffic that would result if all costs of externalities were taken into account in each of the decisions, and the relationship in this graph serves as the basis for the development of the economic cost model. In other words, we estimated the distance to start to zero, the excess traffic congestion and the reduction in optimized speeds, and the speed as compared to the speeds corresponding to the start.
In the specific approach, our argument was that it was the excess travel and speed due to excess congestion that generates various costs throughout the economy. Our study assessed three categories, or three manifestations, of congestion costs, and that is: time lost to travel for commuting and other general purposes; time lost in work travel or travel for business purposes; and other economic costs.
Regarding time lost to travel, excess travel time for commuting and other personal travel compared to actual travel times can be converted into monetary values using value of time assumptions, and this represents loss to travelers. They have less time for leisure and personal pursuits, as well as the lower average speeds increase vehicle operating costs due to reduced fuel economy.
Regarding time lost in work travel, similar methodology can be used. Excess travel time for work purposes can be converted into monetary values, using again value of time assumptions, and this represents loss to employer, loss in productive work time, or lost productivity.
Regarding other economic costs, which are discussed in more detail on the next slide, our study evaluated two broad categories of these costs: labor demand impacts due to high commuting costs, and industry level effects on revenue operating costs and employment due to lower speeds.
Regarding labor demand impact, it is argued that high congestion leads to an increase in labor costs and reduction in demand for labor. Now it should be pointed out that direct evidence is relatively limited, but it does underscore the importance of the issue - the issue of congestion - and it indicates that employers feel the need to compensate, at least partially, the employees for higher commuting costs or longer commutes in an effort to continue to attract or retain suitable employees. But this may increase some of the above levels that an employer can actually afford, and as a result, this reduces our demand for labor.
Regarding industry-level effects on revenues, operating costs, and employment, our study argues that lower average speeds increase total private cost of travel and affect peoples' decisions for travel, such as trips for shopping or entertainment. They make travel longer, more unreliable, including commercial deliveries of merchandise and business service trips, and this effect negatively affects business performance.
Again, direct evidence is relatively limited, but it is routinely recognized in transportation models that time is an important factor in trip-making decisions; that people do take this into account. Also, it is recognized that efficient logistics management is critical to overall economic efficiency of many businesses, and that these costs are highly sensitive to the magnitude as well as variability of transit times.
So our study argues that first, higher costs of travel reduces the number of some trips, and thus business revenues of businesses that rely on these trips; second, long and unreliable commercial delivery times inhibit cost-saving strategies in inventory and logistics management and does increase operating costs; third, long travel times reduce productivity of business services that rely on travel and efficient access to client locations. These types of costs are particularly relevant for industries such as retail, trade, restaurants, arts and entertainment, construction, manufacturing, taxicabs, services, and repairs, and these would be industries which our study evaluated.
In terms of implementation, based on the arguments discussed and shown on the previous slide, logic models would be from bottom-up, so to say, models which show the underlying cause and effect relationships that show how various variables combine to determine outputs, and in terms of loss in revenue, increasing operating costs, loss in employment, and so on. Then, the spreadsheet based economic model was developed and populated with data and input assumptions for theoretical variables such as value of time, cost of travel, average vehicle speed, elasticity of demand, traffic volume, elasticity of logistic costs, etc., and many other variables.
Major sources of data our study included: there was a previous study by another consultant on the traffic in the New York City region and potential impact of congestion charges. We also used data from population and economic census, publications from New York metropolitan transportation council, as well as related economics literature and previous studies of impacts of improving traffic congestion/traffic conditions.
All the costs and impacts were estimated by New York City sub-area or county and summed across, provide total, although it should be mentioned that due to data limitations, some impacts in that specifically applies to industry impact were estimated for one or two areas and then extended or pro-rated to the entire New York City region.
And then the next few slides provide a sample and key results of the study that we reported. This slide provides the summary of the impacts and the overall results, and they show that we estimated total travel costs - it's nearly $5 billion - and car-commuting costs of $2 billion; car for travel for businesses at about $600 million; industry logistics costs increasing - those costs were estimated at about $1.99 billion. Vehicle operating costs were estimated at the range of between $200 million and $2 billion (the range is so broad because the data that we were using was not necessarily the most recent and was not necessarily based on the most recent usage or usage rate and fewer efficiency, so there was some uncertainty - that would raise uncertainty around those estimates); and the regional product was estimated at about $4 billion; and loss in employment at about 51,000 [jobs].
The next few slides show more specific results for the costs borne by the industry in terms of revenue operating costs and employment. Not all of the facts were estimated for all industries, and that is the reason why there were some blanks or N/A for some of the industries. As this table shows, in total, for the entire New York City region, reduction in industry revenues due to excess congestion was estimated at about $4.6 billion, increasing operating costs at about $1.9 billion, and reduction in employment at about 22,000 [jobs].
Now to put this in perspective, we compared to these costs to total - to industry size - basically total industry revenue, and as this table shows, the industry which appears to be most-affected is business repair and maintenance industry, and it accounts for about 3.6% of the current revenues. And for other industries this effect is much smaller, and it is less than about 1.5%. Manufacturing and construction are the second and third industry, with the largest cost at about 1.3%.
The next slide shows the regional distribution of congestion based on the employment loss, so out of the total of 22,000 jobs lost across the region, as the graph shows, nearly 40%, or exactly speaking 39.1%, would be incurred in the central business district of Manhattan. The second largest effected area is New Jersey, followed by Connecticut, Queens and Brooklyn. And this concludes my presentation.
Well thank you, Ewa. Going to give a second - looks like someone might be typing in a question, so give a second for the question to pop up because right now I don't see any other questions typed in. So let's just wait a minute or so and see if anything comes in. Looks like a few people might be typing some questions in. We will give it a minute. On Slide 13 are those increases in operating costs and decreases in revenue?
Okay, let's see what else comes up here. I think there's another person typing in a question. Okay, if there was a net benefit to drivers overall, wouldn't more people drive?
Well, I'm not sure, what is the specific reference to people. What we are looking at was the cost in terms of excessive time spent in commuting or in traveling for business trips - for example, for meetings or to visit client location - so with excessive congestion, those travel times are longer, so there was a higher cost in terms of the time lost. If congestion would be smaller, then those costs would be smaller as well.
And I guess maybe the question, maybe that relates actually to the next question that if more people drive it could cause more congestion. So the next question is that somebody asked: How might you convince the public there is an efficient level of congestion? And the person says that this might be beyond your scope but maybe just offer any thoughts on that.
Yes, that is a very good question and how that could be achieved. We really didn't go into the policy implications of the results of the study. One implication is really going - the one implication is the one that is going back to Zabe's presentation, and that is congestion charges. Congestion charges would be the instrument which would convince people to travel less.
And actually, Zabe and Jose I'll open that question up to you if either any of you have any thoughts on that?
About how to convince the public?
Yes, if there is an efficient level of congestion.
We haven't actually determined or haven't actually presented there is an efficient level of congestion. Our main focus has been simply explaining to people the goal is not to eliminate congestion altogether, but to reduce it so that we're reducing the congestion impacts, and that is both delay as well as operating cost as well as economic activity and greenhouse gas emissions. And I think the question will differ depending on who you ask. Our environmental partners have said well, there is a sweet spot in terms of the speed that a vehicle operates, so if you can get your sort of congestion or your traffic levels to a point where you're flowing at that rate, then that's your sort of efficient level of congestion. But the big question is who are the different groups that we're speaking to because each group has a different perception of what their sort of efficient level is, and we've tried to tailor the concerns about the congestion impacts to those different groups as we speak with them.
I have nothing to add to that.
Okay, let me just wait here. It sounds like there may be one more question coming in, so let me just wait on that one. Does Herbert Mohring's University of Minnesota work state the efficient level of congestion is when the marginal cost equals the marginal benefit?
I think this is correct. So that's basically what our approach encompasses.
Can I add something to that?
Sure, please do.
I remember this from my days in economic classes ages ago, but even today I think what we found is to get to the economic definition of efficient level of congestion, you would have to charge well more than you could politically and typically hear, and so that's one of the reasons that our goal has not been to get to that sort of economic definition of efficiency. It's more been about: how do we get people going to where they need to go more efficiently than they are today, and sort of how do we manage the broad range of benefits and impacts on the system.
Okay. Well I think then at this point we'll move on to our final presenter. Thank you, Ewa, and also thank you to Zabe for jumping in on some of those questions. We're now going to move on to our final presenter, Jose Holguin-Veras. So, Jose, if you give me one second I'll bring your presentation up and you can get started.
Fantastic. It's really a great pleasure to have this opportunity to tell you a bit about the impact of, in this case, freight road pricing on businesses, and basically what I'm going to do is go over the empirical evidence on this. From the get-go, let me tell you the empirical evidence on the impact of pricing, freight pricing in this case, is very limited. There are a number of toll roads that have implemented a differential pricing on trucking, but the realities of this study, these cases have not always been analyzed. As of now, there are basically three major cases, London, New York City, in both cases, and the Ohio Turnpike, and there was also an ex-post analysis, I mean, done after-the-fact.
Basically, they analyzed the Ohio turnpike and in essence, what they did was to analyze the experience it had over a number of years and the experience of the Ohio turnpike. In essence, in the 1990s, they increased - the Ohio turnpike increased tolls in order to finance a number of construction projects they had, and what happened was immediately after - basically trucks began shifting routes, and then totals at the end were basically lower to attract the toll traffic back, and usually in this case a subsidy from the State.
In essence, what my theory and Swan did was to somehow relate the truck VMT at the Ohio turnpike to the truck VMT at the US level, at the national level, and to leave that to a number of independent variables, including tolls and travel speeds.
This is basically a model that I got in the essence, as you can see, and it has a very high level of statistical accuracy and it's really good.
Now, what are the key findings of this? Well, the first thing is that some of the trucks were able to shift route. Now, if they shift route, whether it changes from the schedule that they imply, they are basically - in this case - they are trying to avoid the toll. That means if they try to avoid the toll it's because they simply couldn't pass the toll costs to customers. Now, the reality is that we know very little about the carriers that remain used in the turnpike. We don't know how many of them pass the total costs, we don't know how many of them simply had to absorb the toll cost, and we know it's reacted in any other ways differently. That means we don't know much about this.
Now, moving now to the urban case. In the urban case, as I indicated, there have been two major cases: London, and in this case, New York City, and I'm going to tell you the key findings from New York City now. In essence, in 2001 the Port Authority of New York and New Jersey implemented a time of day pricing initiative. They simply raised tolls for trucks traveling the peak hour. Now, what I'm going to present here is basically the result of the significant study with it that collected behavioral data, complemented that with in-depth interviews with industry representatives and also with a focus group, with carriers, receivers, and different participants in the Supply Chain. Basically, here is the toll schedule: as you can see, there was a significant increase in the tolls that were in the peak hour from $4 to $6, and depending on the use of Easy Pass, which is the electronic toll collection system that went from $3.60 to $5 per axle, which was a significant number.
This is a figure showing the Port Authority facilities that basically you can see concentrate in the three crossings that separates New York City from New York. Basically, what it means: internal New York City is not impacted by the tolls.
Well, we collected data about this and as you can see here, everything that was interviewed it was a random sample of about 200 carriers, and basically this is typical in conditions in which the truckers have to deliver to multiple customers, and now the implication is that any change is a change behavior. If they want the change time of travel in response to pricing, they have more customers to coordinate with, because the truckers cannot simply change unilaterally time of travel, because they have customers that expect their supplies. That is the first implication of this.
Now, the second implication: we asked them, why do you travel at the time you do? And as you can see here, I mean, external constraints in this case where the drivers reported customer requirements, and those that have to deliver during the normal hours, and they represent a big chunk of the cases, between 65-73% of the total. We also asked about tolls that the tolls typically, it's not the primary factor in their decision. Why? Because they are basing it on the time-sensitive industry in which they have to travel, whatever the customer demanded.
In the final slide, it basically represents the flexibility windows they have. We asked the carriers how early or late they could arrive and still make the customer constraints. As you can see, private carriers with those who are delivering for their own company, and they have a wider flexibility window, from 79 minutes to 55 minutes, and it's much more than the one for the current carriers. For-hire carriers have a better type of window, and it's about 50 minutes in total. That basically imposes tight constraints on the drivers that they have to meet, because otherwise, I mean, they lose this.
We also asked the carriers if there was an opening in the process - after the fact. We asked them how they reacted to pricing, to the time of the pricing initiative. Their responses were quoted in this pyramid that basically has at the very top, you have changes in facility use and you have cost transfers, and then productivity increases. The point in between - the midpoint of this represents combinations of the strategies, while the center represents the carriers that did all these things: facility use, transferring costs to customers and productivity increases.
Now, organizing the results in this fashion basically sorts the impacts into three different bands. At the lower left corner, the carriers do only productivity increases in response to tolls, and they're the ones absorbing the impact of tolls. In the event in between, both carriers and receivers are impacted, and at your end represented the top right part, only the receiver is impacted, and basically, that will tell you about who has the power. Basically, here is the response we got. As you can see at the very top of the figure, basically here, no carrier simply changed facility use. That was basically, clearly indicate that this is not a preferred attendant. As you can see here, the bulk of them basically implemented carriers in productivity to cope with the toll increase, and implementing in productivity knows by heart that this is something that you do basically when you have no over attendance, and the essence is basically the toll increases were basically felt by the carriers.
Now, here, in summary, basically what you see is that 20% of the carriers interviewed changed behavior, but not in the way that we had anticipated, that we're expected, because people expected them to simply reduce the amount of travel. Why couldn't they change behavior? In the second bullet, 70% of the sample said customer requirements. They simply could not change time of travel just like that. In only 9% of the cases, say one out of 11 trucks, they were able to pass the toll costs, and these cost increases were relatively small, about 15% of the increase, when the toll increases they got were about 50%. That means they couldn't pass the toll cost to their customers, and that indicates an area of weakness over the delivery industry.
The reasons for not changing behavior: in this slide, you have the breakdown, and as you can see, the main reason by far was basically the customer that cannot change schedule due to the customer requirements.
Now, when we took a look at the breakdown of those carriers that were able to pass the toll costs to the customers, we found out that basically only segments with some degree of market power that were able to pass the toll cost, and this is completely consistent with economic theory that basically indicates that only if you have market power, you are able to pass costs like this is a fixed cost to the operations. And what it means is that the rest [that] are in very competitive industry segments simply couldn't do much to pass the toll to the customers.
Now, here, as you can look and see, we have two differences, two different environments. In the inter-city case, the truckers were able to pass, to change route (at least some of them), and the difference is that in the inter-city case, when facing a toll the truckers could simply change route, as a better route they could use, and by changing route they may be able to still meet the customer requirements. And if that's the case, they are going to do it. Now, in the case in particular where you have a toll, basically the only intent the truckers have is to change time of travel, but changing time of travel is going to impact the way in which the receivers will gather goods, and they are going to oppose that because in essence, you get quality condition, and they prefer to have deliveries whenever they have the stuff present. That means in the regular hours and the urban case, the time of the pricing for freight simply doesn't work much, doesn't work very well. Impacts on businesses: so what all of this indicates is in essence we have two key agents involved. We have carriers/receivers and we have the shippers, but in essence the carriers are basically the Supply Chain, and this is so because after many years of deregulation, there is a significant amount of oversupply, and oversupply leads to the regulation of the market power of the trucking industry. Now which of the agents are mostly impacted depends on this element of market power, and whoever has more clout is going to impose its will on it, and it also depends on the container, but it is basically the number of cases. The first case is the carriers have alternatives (and in this case, alternative routes). The respective behavior is going to depend on the size: if those are small they are going to pay the toll and get going, and now if tolls are large, they may seek alternative routes. And now that says they are going to produce an impact on alternative routes that might be local and might lead to community complaints.
Now in terms of who paid the tolls, those carriers with market powers most likely are going to pass the toll cost to the others. Those without simply have no alternative than absorbing the tolls.
In cases where - like in the case in which - because of the way in which the toll system, basically they are assigned to prevent an alternative. In essence this boils down to a straightforward base on market power. Carriers with market power will try to pass the toll costs to the customers. In the case of New York City, that's basically about 11%. Those without simply will swallow the tolls, will simply absorb the tolls because they simply have no choice.
Basically, empirical evidence on the impact of freight road pricing on businesses is that they seem to indicate that the bulk of the impact falls on the carriers, and other agents/shippers/receivers, who share the cost to the carriers seem to be impacted much less. And now, other implications. Well, the first one is no major impacts are going to reach the end consumers, because the carriers are basically taking the brunt of the cost, and then obviously, the carriers profitability suffers. The behavior change basically, the implication is that the constraints imposed by the customers simply do not provide enough room for the carriers to change behavior and in essence, what we need to do is to take this element carefully into account in order to make sure, to the extent we could, that the impacts of this are basically distributed among all participants. And with that I would like to end. Thanks a lot.
All right! Thank you, Jose. It looks like we have some questions coming in, so we'll just give it a minute or so to see what's coming up here. Looks like we have a few questions being typed in. Still not coming up yet. We'll wait a second more.
The lack of information on intercity toll impacts is troubling. Is NCFRP, which is the national cooperative freight research program, or anyone else funding further research on this that you're aware of?
My understanding is that there is an NCFRP project dealing with that. I don't know anything else though.
If anybody wants to type anything in, if anybody is aware of that please feel free to. Let's see if anything else comes up here. I think we have someone from VDOT who might be typing in a question so we'll wait a second.
It would appear that tolls would be much like fuel costs and that they are delivery costs that would get passed on top of the receiver since all carriers would be subject to paying tolls. Is this not the case?
No, it is not. It is not. Tolls, let's say fuel costs, depend on the unit of output. The unit of output in this case might be the kilometer and things like that, and that's why they're able to pass - let's say if you have increasing fuel tax, for instance, then we would be able to pass that to the customers. The toll is not. The toll is a fixed cost because you could insert additional customers in the tour without having to pay additional toll costs. There is only one exception to this, which is basically when a new customer, when an additional customer implies an additional trip. In that case, if that were the case, the toll is variable. Another way to look at this is to take a look at the way in which the contracts are signed. Most contracts have freight determined by the number of miles, which is kind of a measure of variable costs, and that is the case.
Okay. Could a road pricing policy be designed with longer operating hours for receivers as a possible mitigation measure for the impacts on carriers?
Could you repeat the question?
Could road pricing policy be designed with longer operating hours for receivers as a possible mitigation measure for impacts on carriers?
Well, whenever you deal with the receivers - for the receivers, extending into the off hours implies additional costs and basically, that is what provides, that's why there's a look into it. Now, people have suggested in the past instead of charging the truckers, charging the receivers, as they want to generate the demand. From the point of view, that is ideal, but might lead to great political opposition. What we are doing in a different project in New York is to basically provide incentive for the receivers to switch to the off hours and by doing so, they're going to pull the truckers. But in essence without - the fundamental thing is that without addressing the receiver concerns about moving into the off hours, the carriers do not have the clout to push them to do so.
Okay. Let's see. Did you conduct any research related to how to spend the pricing revenues?
No, we did not.
Okay. But aren't tolls treated as hidden costs in the pricing of services, maybe not specifically enumerated in contracts?
The reality is that in competitive markets like the delivery industry, the carriers cannot record the fixed costs: not only the tolls but also even the price of the truck, and that's why truckers go broke constantly. Basically, the delivery industry is in a constant state of turmoil because truckers go broke and they simply cannot recoup the cost of maintaining the truck, and when the truck goes broke, they go out of business, and another hopeful individual will make a replacement, and the process keeps going like that. So basically, it's basically a major source of concern for the industry, because this is something that - this turmoil produces a lot of problems and in essence, if you talk to the truckers - if you talk to the drivers, I mean - they will tell you that they "eat the trucks." In the trucking industry, this is how they refer to the fact that they cannot recoup the fixed costs.
I see the e-mail about when truckers go broke, there are less truckers. That is not the case. Let me tell you why.
You have to go over the industry structure. Let's say if you take a look, for instance, at the industry in which we have a relatively small set of companies: when one company goes broke, the others have the ability to jack up the prices, because the number of competitors goes down. In the trucking industry, in essence - what you have in practice is infinity supply, and for every trucker that goes broke, there is one individual waiting there basically to step into the market. And that is reality, and they go broke, and somebody else comes to take their place, and the cycle keeps repeating. And that is the main difference between the trucking industry and the industry like airlines, in which the number of companies is much less. In here, there are no barriers to enter the trucking industry, and that is what allows the cycle to keep repeating itself.
Okay, thank you. Well, we are actually about out of time today. We've had three great presentations and really good questions as well, so I want to thank all three presenters and thank everybody in attendance for their questions. I'm actually going to bring back up the slide that has some information about obtaining further information to the congestion pricing website. So then, thank you to everybody. The next webinar will be on October 27th and will be about integrating transit with congestion pricing and increasing congestion pricing acceptance. This webinar is now open for registration, and I do encourage everybody to visit the congestion pricing website and register, and in a minute the slide should be coming back up on the screen that has that web address on it. So with that we're going to go ahead and close out for today. The recording and the transcript and the presentations will be posted online within the next few days, probably about a week or so, and I'll send out an e-mail to everybody who registered once it is available. So thank you, everybody, and enjoy the rest of your day!