This paper presents the results of an investigation into the factors contributing to toll lane subscription choice by using data from the MnPASS high-occupancy toll lane system operated by the Minnesota Department of Transportation. The paper estimates a binomial logit model that predicts, on the basis of aggregate characteristics of the surrounding area, the likelihood of a household having a subscription to MnPASS systems. Variables in this model include demographic factors as well as an estimate of the incremental accessibility benefit provided by the MnPASS system. This benefit is estimated with the use of detailed accessibility calculations and represents the degree to which a location’s accessibility to jobs is improved if HOT lanes are available. The model achieves a rho<sup>2</sup> value of .634, and analysis of the results suggests that incremental accessibility benefits play a statistically and practically significant role in determining how likely households are to hold a toll lane subscription.
“2013-08-30: It’s not the average value of time saved (VoTS) that’s relevant to users of toll express lanes running alongside free lanes. It’s the higher time saved values of a small proportion of trips that choose the express lanes that determines how much the minority using express lanes will pay – the VOTS of the top decile or two of the distribution of total trips in the corridor.
Randall Pozdena, managing director and senior economist at ECONorthwest in Portland OR makes this point in comment on our reporting of a University of Minnesota study of motorist behavior on the 394 and 35W toll express lanes in the Minneapolis area by Michael Janson and David Levinson. We’d reported the researchers’ surprise at the very high values of time saved they’d observed in the Minneapolis toll lanes – that drivers in express lanes regularly pay $1 to $2 per minute saved, or $60 to $120/hour.
That compares to full tollroad average driver payments of $15 to $30/hour.
The main thrust of our report was on the Janson-levinson discovery that toll rates can have a definite proxy effect – high rates being seen as a proxy indication of high congestion shifting the demand curve to the right.
But the other newsworthy findings of the study seemed to be the very high VOTS measured and the Janson-Levinson explanation that motorists may have an exaggerated sense of the time they save. They are behaving with imperfect knowledge, and systematically biased perception of what time saving they’re buying.
Pozdena sees it quite differently.
He says people using toll express lanes are acting rationally and while their knowledge is imperfect he sees no systematic misperception.
‘In a toll express lane setting, implicit values of time are very likely to be that high ($1 to $2/minute saved.) The reason is that the ‘conventional’ value of time is actually an average, but in the real world, there is a log-normal distribution around that average that reflects not only the distribution of income, but also the urgency of the trip at the instant. Thus, the users who self-select to pay to join the toll express lane are drawn selectively from the UPPER TAIL of the VoT distribution, and would not be expected to display average values of time.’
He says ECONorthwest has developed a model for toll optimization model that is based on calculating the shape of the distribution of values of time.
‘We take the regional model average VoT and form a distribution around that mean in a key step in the data flow called ‘VoT Transform.’
He adds that at ECONorthwest they ‘routinely find our model predicting that the MARGINAL value of time of those who are skimmed off to the HOT lane is several multiples of the AVERAGE value of time, and the average of those above this cutoff would be even higher. Thus, I don’t think that one can conclude that HOT lane users think they are getting ‘greater time savings’ than they really are. They are just the folks with long-tail values of time.’
Pozdena says the proxy effect is not a problem if the pricing system is working as it should and prices according to density of vehicles. It is conveying good information to motorists about conditions they can expect.”
Obviously there is a long-tail (one cannot have a value of time below zero, but one can have a value of time of $200/h). Further, people using the toll lanes must have a higher value of time than those who do not use the lanes, at that time. The question is the composition of rational and fully-informed travelers vs. misperception (for which, contra Pozdena, we do have evidence) vs. emotionality vs. concern about reliability vs. principal agent problems (someone else’s money) vs. any of the other factors that influence route choice and lane choice. These are all components in the choice of different individuals, and we don’t know enough to accurately ascribe the choice to each component.
I shriek the wail of every scientists: more research is needed.
The Twin Cities freeway system, like many places, is designed so everyone who wants to use a freeway can do so on-demand by showing up and getting on the road (or queueing at a ramp meter waiting for the opportunity). The Twin Cities also has a MnPass system on some of its freeways, which it plans to expand.
The MnPass High Occupancy/Toll (HOT) lanes guarantee free-flow travel times, but have tolls which vary systematically over time-of-day (rising during the peak period), or in modern installations, dynamically. In the case of dynamic tolls (like MnPass), they are intended to ensure the toll is high enough to prevent congestion (ensure the level of traffic is far enough below capacity that bottlenecks are not activated, and if temporary queues are formed they are quickly discharged.) However with the dynamic toll, travelers don’t know what it is until they are about to decide between using the priced lane or not. (It is capped at $8.00, but the price has variability).
Almost everywhere else, we guarantee the price on a road ($0) with variable travel time.
We thus have either uncertainty on price with certainty on time, or certainty on price and uncertainty on time. We should be able to have certainty about both of these, assuming we allow the price to be non-zero.
First-come, first-serve is not the only way to allocate space. We don’t allocate table space that way during prime time at nice restaurants. We don’t generally allocate airline seats that way. We don’t have to allocate roads that way.
We could, for instance, have reservation pricing (which has been proposed for intersections in a real-time way). At its most basic level, for instance, every day the commuter pre-purchases a ticket to use a particular road segment (e.g. between exit 400 and exit 401) during a particular time slot (say between 7:45 and 8:00). The ticket cost is known in advance before departure, like a plane ticket. The road agency would only sell as many tickets as the road would accommodate (without congestion) at that period. The ticket would be validated electronically through some form of Electronic Toll Collection.
But, roads are not like airplanes, the traveler might arrive at 7:44 or 8:01, does the road agency force the car onto the shoulders? More likely they just charge a penalty which increases with deviation from the purchased window. So if the charge were $2, there might be a $0.10/minute surcharge added for each minute early or late the traveler was.
If the system were deployed universally, congestion would be a rarer occurrence (seeing only non-recurring congestion, due to crashes and other incidents, not the daily recurring congestion because of excess demand for the available capacity). Further the system would know whether the traveler or the agency was the caused the earliness or lateness. If there were non-recurring congestion, it might waive the penalty. If on the other hand the traveler left too early or too late to reasonably make their slot under planned for circumstances, the surcharge would stand.
But travelers don’t want to map out their route every day. OK, the agency can probably just sell a ticket allowing travelers to be “on the roads” (as opposed to being on a specific road), and let individual travelers sort out the best path. The losses from not micro-managing spatially are relatively small, compared to the gains from spreading traffic out by time of day. See the literature on Macroscopic Fundamental Diagrams about this.
In this case, the traveler pre-purchases a ticket to use any metropolitan area road between 7:00 and 7:30 am, but the number of passes is limited by system capacity. If purchase is made far enough in advance there is a lower price then if the ticket is bought real time. But no-one wants to plan their schedule that far in advance, or log-in daily spending even 5 minutes to buy tickets for a 20 minute trip.
Here the road agency managing the system can be a little bit more clever. They could sell various types of season passes (just like transit agencies). A traveler might buy an unlimited use pass for a premium, but there would be a limited number sold to residents of each zip code (if spatially fixed, passes would not be transferrable, the electronic transponder would be linked to a license plate) (or any other local geography that makes sense, the key is there is still finite space on the roads, and too many pass-holders from a wealthy suburb won’t save time just because there is excess capacity elsewhere). Or they could buy a more limited use pass at a lower price. And of course, travelers would buy these as recurring subscriptions, billed to a credit or debit card.
People without passes could take their chances with same-day tickets which might be more expensive if traffic is on the edge of congestion, or cheap if traffic is low that day.
If set up properly, these passes replace existing revenue sources for the agency.
How should the agency allocate these passes? Clearly it should not just give them away. But setting a fixed price and selling them does not real allow discovery of demand patterns. Here Dutch or Vickrey auctions might be appropriate. For instance, the passes would be available at a posted price (‘buy-it-now’), but the agency would also accept lower bids. Suppose there were 100 passes, there would be a bidding period, and at the end of the period, the top 100 bids would win and the price would be set at the willingness to pay of the 100th bidder. There are many variations on auctions, which each have advantages for buyers or sellers in terms of maximizing revenue or price discovery or fairness.
“Altshuler bases his position on a couple surveys conducted in metro areas that have adopted HOT lanes in the recent past. One was done in San Diego circa 2001. At that time, about 80 percent of low-income respondents agreed with the concept that people should be able to use an express lane on Interstate 15 for a fee — a greater percentage of agreement than people from high-income brackets (70 percent). Additionally, two thirds of people who didn’t even use the lanes still supported them.
A similar survey was done in 2006 in Minnesota. That work showed a 60 percent approval rate for HOT lanes on Interstate 394. A stronger analysis of this corridor, done by Tyler Patterson and David Levinson [PDF], found that income levels did predict use of the express lane (with higher-income drivers using them more often), but that lower-income drivers could also benefit from the shift of traffic out of the free lanes (as well as always having the express option in a time crunch).
(And a far more recent survey, released in April, showed that two-thirds of people making less than $50,000 a year said they’d use express toll lanes — the same percentage as people making more than that.)”
A key point is that HOT lanes also enable freeway BRT where it might otherwise be unaffordable to construct. The express lanes are uncongested and can be used by buses to maintain speed. An example is the I-35W corridor (Orange Line) south of downtown Minneapolis, which is not complete (Lake Street Station is still missing, e.g.), but has a BRT station at 46th.
Roads exist largely to serve two purposes: movement and access (specifically, access to property). Different types of roads have characteristics of different types of goods based on their functional classification. In other words, there is a correspondence between the functional classification of a road and the type of good it represents. What types of roads are which type of goods?
Limited access highways (freeways) and some arterials with signalized intersections and few access points, could be considered private goods, since it is possible to identify and exclude users with appropriate toll technologies. These roads are also rivalrous since, in the absence of pricing or other measures to limit demand, an additional user can affect the use of the road by others.
It is very clear that users could pay for private goods directly. The most obvious way is a toll, a bit less obvious is a gas tax. Nefarious politicians in a few states have come to realize that a wholesale tax on gas is more hidden than a retail gas tax, even if the incidence is essentially the same. The gas tax collected on users of freeways at the state and federal levels would more than pay for freeways (even if some is diverted to transit and other uses).
Local roads lie on the other end of the spectrum in terms of functional classification, since they exist primarily to provide property access. Local streets can be excludable if access to them is restricted. Access restrictions may take many forms, ranging from the simple posting of signs indicating that access is restricted to residents to actual physical restrictions, such as gates. The latter type of restriction is typically associated with gated communities or other forms of private residential development. Local streets are also generally non-rivalrous in that their low levels of traffic tend to preclude problems with congestion. This combination of characteristics (excludability and non-rivalry) indicates that some local streets may be considered club goods. The oldest such example in the United States is that of Benton Place in St. Louis, Missouri, where adjoining property owners were required to join a private association which was responsible for road maintenance, with assessments being levied on each association member.
These kinds of local roads are essentially club goods. The ideal club might be the homeowners association, the downside is diseconomies of local management of roads and potential interruptions in the local road network (e.g. more cul de sacs and fewer gridded blocks). Division of labor is a good thing, but requires scale. So the club is usually the local municipality rather than the homeowners association. The revenue that is collected for this tends to be a property tax, paid for by everyone. Tolls are impractical for local streets because the cost of collection outweighs the benefits. A local option gas tax may also be impractical because the purchase area of gasoline exceeds the size of the municipality (e.g. our most used gas station might be in St. Paul, even if we live in Minneapolis). Any jurisdiction that tried to raise too much from this fee would see more out-of-municipality purchases, and the gas station (or the landowner of the gas station, if they differ) would bear much of the incidence of the charge.
An additional challenge associated with fuel taxes for roads is that some local roads, when privatized or used as a club, restrict auto traffic. In the Seward neighborhood Milwaukee Avenue is essentially a club good for bikes and pedestrians. In such cases property taxes are much easier and effective for financing the roads. St. Louis, MO, has long featured private local streets in all types of communities.
Local streets are typically provided by local governments with no restrictions on access. In the absence of access restrictions local streets may be both non-rivalrous and non-excludable, leading them to take on more of the character of a public good. Note the term “public good” in this case is defined by the economic characteristics of the good, and not simply by the fact that it is supplied by the public sector.
Some local streets are not excludable because of their functional design connecting places (one of the drawbacks of a connected network is that it is used to connect people who are not local). So design of the network greatly affects how it is best managed and funded.
Between limited-access highways and local streets are a middle level of road, collectors, that link local streets with limited-access highways. These “linking collectors” serve both access and mobility functions, since they may also provide access to some adjacent properties. These roads may be considered “congesting” or common goods.
The characterization of roads in terms of functional classification may also inform decisions about which level of government should be responsible for providing a given road (assuming the decision is made to provide the road publicly). Local units of government seem best suited to providing local streets, since they are closest to the problem. Roads that provide for a higher level of movement, such as limited-access highways, ought to be provided by higher-level jurisdictions, such as states. Of course, there are tradeoffs involved in each of these decisions. Smaller jurisdictions may not be able to fully realize scale economies, while larger jurisdictions may encounter problems with span of control. Between these extremes there is some optimal mix of expenditures between different levels of government that minimizes capital and operating costs.
The larger jurisdiction (the state) could set a higher fuel tax to redistribute back locally bysome formula, but there still remains different preferences in different municipalities for different levels of service, which require some locally different levels of funding. How is that to be collected if not a property tax?
The property tax can be thought of as a charge for having the option to access roads and receive public services such as police and fire. There are other types for local tax that can be used; a variety of land value capture mechanisms, from the land value tax or split rate tax, to transportation utility fees change the basis of collection, but even TUFs are still not strictly proportional to use. If a mileage-based usage fee were already widely deployed, it could be varied by municipality to collect more revenue, but that ideal revenues collection scheme is not worthwhile to implement if the only user were local governments, and would require standardization and enforcement to enact.
Regardless of whether one favors or opposes subsidy for roads, distributional considerations of how the subsidy monies are generated are important. We discussed fuel taxes, tolls and property taxes, but many cities and states turn to sales taxes to pay for roads (and transit. For instance, the Minnesota Legislature is considering an increase in sales taxes dedicated to transit. Sales taxes are a shift away from direct user fees and impose the burden of subsidy broadly. Sales taxes also tend to be regressive in that households who end up paying the largest share of their income in sales taxes receive less than proportionate value in return. So subsidy is an issue of distributional fairness depending on how the money is raised as well as how it is spent. At the very least money used for subsidy should not rely on regressive tax policy.
Rationales for subsidizing roads:
Universality: Everyone uses roads, everyone should pay.
Midwest Energy News quotes succinctly: “There isn’t a person in the United States who doesn’t get some use out of the roads,” says Levinson, who also writes the Transportationist blog. Even people who don’t drive still benefit from things like fire protection, ambulance services, and mail delivery — all of which depend on roads. “I suppose you could be Ted Kaczynski, but even he had to use the U.S. Postal Service to mail his bombs.”
Even if you don’t drive, or bike, you still use roads. We had roads before cars and bicycles, and will continue to have them even when we have flying cars in the future. Their nature will of course vary, but at least some of the costs are communal. We collectively want the option to be accessed by ambulance should the need arise, or fire or other emergency services. We all make use of ground delivery for things like the mail or packages. We all partake of public utilities running along public rights-of-way. Even bus riders use roads. We also like to have things delivered. When we order a pizza we use the roads, but we have substituted our travel to the restaurant for the delivery person’s. When we order things online, FedEx, UPS and other carriers use the roads to deliver our stuff. It is impossible to “opt out” of using the road network.
Democracy: Most people drive, therefore it is a subsidy from everyone to almost everyone else, which is more democratic than a subsidy from everyone to the very few.
Administrative efficiency: Paying for roads out of general funds, or with imperfect gas taxes, is a lot easier to administer than trying to enforce specific payments for specific roads at specific times of day. Tolls are costly.
Cost structure: Uncongested roads are on the left-side of the U-shaped cost curve, and charging for them leads to suboptimal levels of use.
Non-excludability: There is not a good non-governmental finance mechanism for signalized arterials, roads which are rivalrous but not excludable. One could establish a congestion charging zone, but those are very expensive, and have yet to be tried outside a few center cities.
Agglomeration benefits: Accessibility leads to positive spillovers for the urban economy and increase total development.
Network spillovers: A better connected network makes all other roads more valuable. Much like the temporal Mohring effect in transit, there is a similar spatial for roads effect, each additional road reduces travel costs between places, increasing demand on other roads (and raising overall usefulness of the transportation network still more). Like the Mohring effect, this too reaches diminishing returns, but it is important when networks are sparse.
Mutually assured subsidy: Transit is subsidized, therefore roads should be subsidized.
Off-mode effects: More roads reduce congestion on other modes (transit, rail, air). Subsidizing roads can help support car-free roads, since bikes and pedestrians currently don’t directly pay for their infrastructure. However, relying on drivers to pay for roads used primarily by non-drivers may lead to principal-agent problems.
Overconsumption: Subsidies induce overuse. This can lead to congestion.
Negative externalities: Roads, and in particular cars, generate negative externalities which we cannot properly price. At least we shouldn’t subsidize their production.
Off-mode effects: Reducing demand for transit (by subsidizing roads) worsens the strong positive feedback system that drives transit, lowering congestion on transit is seldom the critical problem, encouraging demand is more likely to be important.
Mutually assured subsidy: A key point is that just because we subsidize roads does not mean we should subsidize transit, and vice versa. The economics of the technologies differ significantly. One bad subsidy does not deserve another. Just because transit is subsidized is not a reason to subsidize roads. It may be an argument to remove the subsidies that exist. Mutually assured subsidy is the “fairness” logic of a 4 year old complaining about a sibling.
Obsolescence: Following the Great Gretzky, “Skate to where the puck will be.” Roads will be made obsolete by upcoming technologies, we should cut our losses now.
So what is the net?
In the short run, states should raise their gas tax to replace the general (property tax) revenue from a baseline set by lowest common denominator jurisdiction within their domain with user charges. That is, figure out which jurisdiction spends the least per capita on roads, and raise the gas tax to replace the property tax by at least that amount of money for each jurisdiction. In all cases states should be extremely wary of using sales taxes to pay for roads. (States will also need to cover the declining federal gas tax, but that is separate.)
Over time, states should move toward a vehicle mileage fee varying by weight (for trucks and other heavy vehicles), location and time of day to replace the motor fuel tax. This should be phased in with EVs (and Hybrids) which don’t pay (much) motor fuel taxes, and trucks which would be charged for weight and distance, going first. Off-peak discounts would encourage peak-spreading. Rates would vary by area to account for different costs of running networks.
Road networks should be organized and operated like public utilities, managing to generate revenue from users to pay for cost of operations. Restrictions on usage should be allowed in this model, where auto and truck traffic can be limited to specific times of day or excluded altogether. Road design that allows access for emergency services can be regulated.
“In 2007 Minnesota legislature approved a $5,000,000 project in order to demonstrate technologies which will allow for the future replacement of the gas tax with a fuel-neutral mileage charge. The Minnesota Department of Transportation (MnDOT) organized a study to examine the implementation and operation of a mileage based user fee program (MBUF), which might allow for the supplementation or replacement of traditional gas taxes. The primary objectives of the study were to: assess the feasibility of using consumer devices for implementing Connected Vehicle and MBUF applications. These applications included localized in-vehicle signing for improving safety, especially for rural areas, and the demonstration of the proposed Connected Vehicle approach for providing location-specific traveler information and collecting vehicle probe data. The study consisted of 500 voluntary participants, equipped with an in-vehicle system comprised of entirely commercially available components, primarily a smartphone using an application capable of tracking participant vehicle trips. Successfully meeting its primary objectives, the system was capable of assigning variable mileage fees determined by user location or time of day, as well as presenting in-vehicle safety notifications which had measureable effect on the participants driving habits. MnDOT contracted Science Applications International Corporation (SAIC) to perform research for the project and an evaluation of its findings. This document is the final report from SAIC, providing a summary of the study, its findings and an evaluation of the project as a whole.”
The rates tested were:
1. Outside Minnesota miles – $0.00 per mile;
2. Inside Minnesota miles – $0.01 per mile;
3. Twin Cities (“Metro Zone”) – Peak miles – $0.03 per mile; and
4. “Non-Technology” miles – $0.03 per mile.
It looks like there were some technology problems in the experiment (having worked with GPS and in-vehicle devices for research, I believe this is still emerging technology with imperfect reliability, insufficient for mainstream application):
As mentioned previously, trip data was only available for 57 percent of trips generated by the system. Of the 43 percent of trips where trip data was not available, 69 percent of the trip data loss was due to a vehicle detection failure. Trip data was only recorded if the system could both detect the device was in the correct vehicle and a valid GPS signal was found. Therefore, the remaining 31 percent of the trip data loss can likely be attributed to poor GPS signal during trips. Although the log messages associated with GPS availability cannot be extrapolated to measure the number of trips or miles impacted, the loss of trip data resulting from vehicle detection failures or lack of GPS signal during trips clearly identifies GPS availability as a significant system issue. The deployment team’s report provides additional insight into the accuracy of the system as it relates to GPS connectivity and accuracy. Intermittent GPS signal was reported as a contributing factor to lower device miles compared to odometer miles collected.
Just as a random statistic, which probably doesn’t mean a lot, the report includes the word “success*” 27 times and “fail*” 33 times.
I believe MBUF or an equivalent will come eventually, but here and now the gas tax (or wholesale fuel tax, if we want to hide it), properly indexed to inflation and fuel economy, is where we need to be focusing for the revenue required to operate road systems in the US.
“Two favorites in my mind come from transport scholar David Levinson, who suggests road fees for general road pricing (and peak road fees for road pricing aimed at heavy congestion), and urban planner Laurence Lui, who recommends road fares. What’s nice about road fare is that it parallels mass transit, has an intuitive purpose, and offers flexibility. You can alter it to suit a specific situation — peak road fare, midtown road fare, etc. — without obscuring the basic meaning.”
The term “fare” definitely has a public transit connotation, its definition is: “The price of conveyance or passage in a bus, train, airplane, or other vehicle.”
The etymology is more general though:
Old English fær “journey, road, passage, expedition,” strong neuter of faran “to journey” (see fare (v.)); merged with faru “journey, expedition, companions, baggage,” strong fem. of faran. Original sense is obsolete, except in compounds (wayfarer, sea-faring, etc.) Meaning “food provided” is c.1200; that of “conveyance” appears in Scottish early 15c. and led to sense of “payment for passage” (1510s).
late 13c., from Old French fieu, fief “fief, possession, holding, domain; feudal duties, payment,” from Medieval Latin feodum “land or other property whose use is granted in return for service,” widely said to be from Frankish *fehu-od “payment-estate,” or a similar Germanic compound, in which the first element is cognate with Old English feoh “money, movable property, cattle” (also German Vieh “cattle,” Gothic faihu “money, fortune”), from PIE *peku- “cattle” (cf. Sanskrit pasu, Lithuanian pekus “cattle;” Latin pecu “cattle,” pecunia “money, property”); second element similar to Old English ead “wealth.”
OED rejects this, and suggests a simple adaptation of Germanic fehu, leaving the Medieval Latin -d- unexplained. Sense of “payment for services” first recorded late 14c. Fee-simple is “absolute ownership,” as opposed to fee-tail “entailed ownership,” inheritance limited to some particular class of heirs (second element from Old French taillir “to cut, to limit”).
I could go either way, but I think “fee” is better established and more likely to be adopted.