Implementable Strategies for Shifting to Direct Usage-Based Charges for Transportation Funding (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

18 3. RELEVANT PROGRAMS, PROPOSALS, AND STUDIES The concept of distance-based user fees has received considerable attention over the past decade, with examples spanning the range from preliminary research and pilot projects to planning and full-scale implementation. At least three factors have motivated a heightened level of interest. First, the challenges associated with declining state and federal fuel tax revenue have grown acute, motivating a willingness on the part of decision makers to consider innovative funding options (note that while transportation funding regimes are different in other countries, many face comparable funding shortfalls). Second, technological advances have enabled a broad range of electronic tolling options offering the potential to develop new forms of road pricing not possible in decades past. Third, as the problems associated with auto-dependency – traffic congestion, sprawling patterns of development, dependence on foreign sources of oil, and the threat of climate change – have become more severe, decision makers and analysts have grappled with a wide range of strategies for reducing auto use and promoting less energy- intensive transportation alternatives. There now appears to be a growing recognition that the means of financing transportation can have significant influence over such outcomes. To elaborate on this last point, if drivers are charged more for traveling during the peak hours, many will choose to shift their travel times or travel by some other mode, thus helping to reduce traffic congestion. In a similar vein, if drivers are charged for the level of harmful pollutants that their vehicles emit, there will be a strong financial incentive to purchase more environmentally- benign vehicles, thus helping to improve air quality. As a final example, many automotive costs that are currently fixed – such as insurance and leasing costs – could be charged on a per-mile basis instead. This would provide additional incentive for reducing automotive travel (or, to be more precise, to limit automotive travel to an economically rational level). Taking stock of planning efforts, pilot tests, and programs implemented in recent years, one can discern three broad categories of distance-based pricing concepts that may offer insights in the technical and administrative design of a VMT-based system for road use charges. The examples within these categories may also yield important insights into a range of technical, political, administrative, and legal obstacles that would need to be overcome in order to implement distance-based road use fees. The three categories are as follows: • General-purpose distance-based road use charges. Examples in this category involve the application of distance-based road-use charges that would apply to all light-duty vehicles (e.g., passenger cars), and potentially apply to trucks as well. • Weight-distance truck tolls. Conceptually similar to the previous category, the key distinctions here are that (a) the charges only apply to heavy trucks, and (b) the per-mile rate varies by some measure of vehicle weight to account for road wear. • Pay-as-you-drive (PAYD) insurance/leasing. Automobile insurance and leasing costs are often fixed, structured as a set price for a fixed period of time (e.g., $1000 per year for insurance). The idea behind PAYD insurance and leasing is to vary these costs on a per-mile basis such that the less one drives, the less one owes.

19 The remainder of this chapter looks at each of these categories in greater detail, listing for each some of the best known examples and discuss policy goals, pricing structure (as a function of policy goals), technical implementation options, enforcement approaches, administrative structures, and public acceptance concerns. Note that Appendix A provides brief summaries for many of the cases referenced in this chapter; for additional discussion, see Sorensen and Taylor (2005, 2006). At the end of the chapter is a summary of several high-level findings that emerge from the review of existing programs, proposals, and studies. More detailed insights regarding specific technical and administrative design issues are deferred to subsequent chapters that focus on those topics. 3.1. General-Purpose Distance-Based Road Use Charges Distance-based systems for levying road use charges that would apply to all vehicles on the road have yet to be implemented, but more limited systems do exist. In New Zealand, for example, a system of distance-based road use charges that applies to passenger vehicles that do not pay fuel taxes (principally diesel-powered vehicles) as well as vehicles weighing in excess of 3.5 tons has already been instituted (Land Transport NZ 2008). The idea of developing more robust and flexible distance-based road-use charging systems, however, has received considerable attention in recent years. Examples. Well-known examples in this category include: • Oregon Department of Transportation Road User Fee Pilot Program (Whitty 2003, Whitty 2007, Whitty 2008) • Puget Sound Regional Council Traffic Choices Study (PSRC 2008) • University of Iowa Mileage-Based Road User Charge study (Forkenbrock and Kuhl 2002, Forkenbrock 2006, Kuhl 2007, Kuhl 2009a) • Georgia Tech “Commute Atlanta” study and trials (Guensler and Ogle 2004) • The Netherlands proposal (Alternative Payment for Mobility Project Team 2008) • The New Zealand road user charge (Land Transport NZ 2008, New Zealand Ministry of Transport undated) Policy Motivations. One of the central motivations underlying this idea is to develop an eventual replacement for motor fuel taxes that would provide more stable and sustainable revenue over time. A second motivation stems from the observation that by varying the per-mile charge according to certain vehicle characteristics (e.g., axle weight or emissions class) and travel characteristics (e.g., time and location of travel), it would be possible to create a set of financial incentives that would support a broad range of policy goals such as reducing traffic congestion, road wear, and harmful emissions (Forkenbrock and Kuhl 2002, Whitty 2003). In the New Zealand program, the goal is simply to capture road use charges for vehicles that do not pay fuel taxes (Land Transport NZ 2008).

20 Pricing Variations. At a minimum, proposals and programs in this category involve levying road use charges on the basis of distance traveled. Depending on the program or proposal, the specific rate could vary with such factors as vehicle weight, vehicle emissions class, vehicle fuel- economy, jurisdiction of travel (enabling different jurisdictions to set their own rates), or specific time and location of travel (to charge more for peak hour travel on congested routes). Technical Implementation Options. The most sophisticated proposals and trials make use of in-vehicle units (also known as on-board units, or OBUs) equipped with global positioning system (GPS) receivers and digital maps to enable charge rates that vary with such factors as jurisdiction, route, and/or time of travel. Because a GPS signal is not always available (e.g., in a tunnel or in an area with many tall buildings), such systems often rely on a redundant means of metering mileage (e.g., a connection to the on-board diagnostic port or the odometer feed) and then use the GPS signal principally to determine the location of the mileage. Several electronic options for communicating billing data have been considered, including the use of “smart cards” (data cards that could be removed from the OBU and then used to upload billing data via the internet or via conveniently located card-reader stations), short-range wireless communications (e.g., communications with an electronic reader device at the fuel pump), and longer-range wireless communications (e.g., cellular communications with a central billing agency). The New Zealand program stands as an outlier with respect to technical complexity; because the sole goal is to account for miles traveled and only a subset of vehicles must pay the charge, the billing system relies on manual inspections (odometers for passenger vehicles, hub-odometers fitted to vehicle tires for trucks). Verification and Enforcement Options. Common to the more sophisticated technical configurations is the idea of ensuring that the OBU is tamper-resistant – i.e., that it can’t (at least not easily) be temporarily disabled to avoid charges. At the more extreme end, one could wire the system such that the vehicle would not be operational unless the OBU is also operational. Such an approach, however, could lead to a range of unintended consequences (e.g., disabling a vehicle if the device breaks down or is accidentally dislodged). The more common approach, therefore, is to design the OBU with a certification seal affixed during the installation process and ensure that the device cannot be disabled without breaking the seal. Enforcement then becomes a matter of verifying that the seal has not been broken through periodic inspections. Beyond the tamper-resistant OBU concept, additional options include manually verifying that the mileage count on the OBU aligns with a vehicle’s odometer reading and setting up road-side or overhead communications devices that electronically query a vehicle’s OBU to ensure that it is functional. In the New Zealand example, the approach is again much simpler. Vehicle owners subject to distance-based road use fees are required to pre-purchase blocks of mileage, and law enforcement officers can stop drivers and verify that the current odometer reading does not exceed the miles that have been purchased. Administrative Options. Two broad administrative options have been considered. The first is to expand or modify existing revenue channels. In the approach developed by Oregon, for example, road use fees would be added, and fuel taxes subtracted, with the purchase of fuel (vehicles not yet equipped with mileage metering devices would continue to pay fuel taxes). Funds would then flow through existing administrative channels, though some additional complexity would be

21 required to reconcile the amount of fuel taxes paid at the point of wholesale with fuel taxes debited at the retail level (Whitty 2003). The second option is to create a new entity that would collect and distribute revenues. This option becomes especially appealing, perhaps necessary, if the program is intended to span multiple states or apply on a nation-wide basis. In principal, the entity could be structured as a public agency, a non-profit organization (perhaps involving a compact among states, similar to IRP and IFTA), or a private firm. Some have argued that allowing the function to be fulfilled by a private or non-profit entity could (a) improve the efficiency of collecting and distributing funds across multiple jurisdictions, and (b) allay to some extent privacy-related concerns regarding government access to detailed information about the driving patterns of individuals. User Acceptance Concerns. Several user acceptance concerns arise with proposals to institute mileage-based user fees. First, with the use of GPS, there is concern that the government would be able to monitor and track the location of individual drivers, a perception often reinforced by poorly-informed press accounts. In fact, existing technical proposals have incorporated strategies to protect private data (for instance, transmitting information about the total bill owed but not data about the time or location of travel – see Forkenbrock and Kuhl 2002), but it appears that further public education and outreach efforts would be needed to convey this fact. Second, environmental advocates have argued that a flat per-mile fee, in comparison to existing fuel taxes, would reduce the incentive to purchase more fuel-efficient vehicles. This, of course, could be addressed by varying the per-mile fee based on vehicle characteristics such as weight or fuel- economy. More generally, it is not clear that a significant portion of the electorate understands the motivations for switching from fuel taxes to a mileage-based system of fees, making the political prospects for such proposals less promising absent considerable outreach efforts. For example, there is little evidence to suggest that the public recognizes any shortcomings of the gas tax as an adequate mechanism for funding transportation. It may be necessary to present to the public generally a better developed analysis of why the gas tax will not work in the future before a shift to an alternative funding system would be acceptable. 3.2. Weight-Distance Truck Tolls Weight-distance truck tolls are not a new concept, and many states have in the past instituted this form of road-use charges. Previous programs were implemented through cumbersome manual means, however, and only four states currently levy weight-distance road use fees. Within the past decade, though, several European nations have successfully implemented weight-distance truck charges through the use of electronic tolling technology, and this has stimulated a renewed interest in the concept. Examples. Well-known examples in this category include: • The Austrian GO program (Schwarz-Herda 2004) • The Swiss Heavy Goods Vehicle Fee (HVF) program (Balmer 2004, Werder 2004) • The German Toll Collect program (Kossak 2003, Rothengatter 2004, Rothengatter and Doll 2002, Ruidisch 2004) • The United Kingdom proposal (Worsley 2004)

22 • The Oregon concept (Whitty and Svadlenak 2009) Policy Motivations. A principal motivation for the application of weight-distance road use charges is to capture and allocate the maintenance costs associated with heavy truck travel (Rothengatter and Doll 2002). Additional motivations include automating (streamlining) the collection of road use charges for trucks (apparent for example with Oregon’s proposal to convert its manually-implemented weight-distance truck toll to an automated program leveraging current electronic tolling technology), ensuring that foreign trucks pay the same road use charges as domestic trucks (Worsley 2004)), stimulating a freight mode shift from trucks to rail (Werder 2004), and providing an incentive for adopting less-polluting trucks (Rothengatter and Doll 2002). Pricing Variations. At minimum, weight-distance truck tolling programs account for the distance of travel as well as some measure of weight (although axle-weight may be the most appropriate measure in this regard, most existing programs employ such surrogates as total laden weight or number of axles). Only vehicles above a certain weight must pay the charges (in Switzerland, for example, the lower limit is 3.5 tons; in Germany, the lower limit is 12 tons). Beyond weight class, these programs may also vary the per-mile charge on the basis of such factors as vehicle emissions class or the type of road on which travel is taking place. Note that some programs only price travel on the main highway network (Schwarz-Herder 2004), while others encompass the entire road network (Werder 2004). The German program initially priced travel just on the highway network, but has subsequently priced travel on some adjacent surface streets to eliminate the incentive for trucks to divert to local roads so as to avoid tolls. Technical Implementation Options. At the simpler end of the spectrum, the Austrian program employs an in-vehicle transponder that communicates with overhead gantries on the highway network to register the charges owed (Schwarz-Herder 2004). This approach works from a cost perspective because only the highway network is subject to tolls; if the program applied to all roads in Austria, it would be prohibitively expensive to install the required infrastructure. The more technically sophisticated implementations rely on GPS. In the Swiss case, the odometer is used to measure mileage, while GPS is used to determine whether or not the travel occurs within Swiss borders (the fees are not differentiated by road type, so greater precision is not required – see Werder 2004). In the German case – by far the most sophisticated – GPS is used to distinguish the specific route of travel, as some roads are tolled and others are not (Rothengatter 2004). The German program also relies on cellular communications for uploading billing data. Note that all of the European examples also have paper-based alternatives that can be used to assess tolls for trucks lacking the necessary in-vehicle equipment (this typically applies to foreign trucks that don’t frequently travel in the country where the charge is applied). Verification and Enforcement Options. Two common approaches have been employed to verify compliance with weight distance truck tolls. The first is to set up gantries with devices that query, via dedicated short range communications (DSRC), passing trucks to verify that their in- vehicle equipment is functioning properly. Non-compliant vehicles can either be pulled over and cited by law enforcement agents, if available, or electronically identified via automated number plate recognition (ANPR) systems (cameras and software capable of reading license plates) and subsequently cited by mail. The second is to ensure that the in-vehicle devices are tamper-

23 resistant – that is, that a driver will not be able to temporarily disable the device without being caught in a subsequent inspection (inspections might be performed during random roadside checks or on a periodic basis, e.g. once per year). Administrative Options. Existing weight-distance truck tolls have been developed with both public and private administration strategies. In the Swiss case, the HVF program is administered by the Swiss Customs Agency, with enforcement support from the Swiss Cantons. The Oregon proposal is another case in which administrative functions would likely be supported by a public agency. In both the Austrian and German cases, in contrast, a private firm (or consortium of firms) performs, for a fee, the operational aspects of program administration. User Acceptance Concerns. Generally speaking, the application of weight-distance truck tolls has resulted in higher road use charges for trucks. To succeed politically, it has therefore been necessary to ensure that the trucking industry also sees some benefits from the program. In the Swiss case, trucks were allowed to carry heavier loads over the Swiss Alps in order to facilitate more efficient goods movement operations (Balmer 2004). In the German case, revenues were dedicated to maintenance and enhancement of the highway network, and rates were structured such that truckers could reduce the charge by up to 50 percent by upgrading to the least polluting vehicles (Rothengatter 2004). With the U.K. proposal, one of the pre-existing concerns for domestic truckers was that their foreign counterparts could purchase fuel in other countries with lower fuel taxes and then conduct business in the United Kingdom, gaining a cost advantage. By developing a weight-distance truck toll that would apply to both domestic and foreign truckers operating in the United Kingdom, the playing field would be leveled (Worsley 2004). It is noteworthy that even though GPS components are used in several of the existing weight-distance truck tolls, privacy issues have not emerged as a significant concern, in part because the trucking agency is already subject to a stricter regulatory regime than passenger vehicles. 3.3. Pay-As-You-Drive (PAYD) Insurance / Leasing The PAYD insurance concept appears to be gaining significant traction, with many companies in the United States and abroad either offering or experimenting with this concept. Note that some examples are more aptly described as mileage-based discounts, in which the rates depend, for example, on whether a vehicle travels between 0 and 2,500 miles in a year, between 2,500 and 5,000 miles in a year, between 5,000 and 7,500 miles in a year, etc. Even in such cases, however, it is still necessary to meter miles of travel. Though PAYD leasing – that is, varying the lease payment on the basis of miles traveled – has also been explored, to our knowledge this related concept has yet to be implemented. Examples. Implemented programs in this category include: • Massachusetts, multiple companies offering mileage-based insurance discounts (Bingham 2009, Boston Consumer’s Checkbook undated) • GMAC, offering mileage-based discounts for OnStar customers in 34 states (OnStar 2007) • MileMeter, offering PAYD insurance in Texas (MileMeter undated)

24 • Progressive Insurance, offering PAYD insurance in nine states (Donohue 2008, Progressive Insurance undated) • Aviva, offering PAYD insurance in Canada (Bettencourt 2005, Insurance-Canada 2005) • CoverBox, offering PAYD insurance in England (CoverBox undated). • Hollard Insurance, offering PAYD insurance in South Africa (Hollard Insurance undated) • Nedbank, offering PAYD insurance in South Africa (Nedbank undated) • Real Insurance, offering PAYD insurance in Australia (Pay As You Drive undated) There are additional companies offering PAYD insurance in Austria (Uniqa), Belgium (Corona Direct), Germany (DVB Winterthur, Swiss Re, and WGV), Israel (Aryeh), Italy (SARA), Japan (Aioi), Netherlands (STOK and Polis Direct), and Spain (MAPFRE), though the research team was unable to gather much information in English about these programs. There are also many examples of deploying in-vehicle equipment to monitor driver risk factors (e.g., speed); if these did not also encompass a pay-as-you-drive feature, they were not included in the review. Policy Motivations. Programs in this vein are privately operated, and the decision to offer PAYD insurance is, at root, a market-based decision. One can infer, then, that insurance firms believe that an appreciable number of drivers may value the opportunity to lower their insurance bill in return for driving fewer miles. It is also the case, however, that PAYD insurance programs can yield important public benefits (e.g., reducing traffic congestion and pollutant emissions by creating a financial incentive to drive less), so this concept has enjoyed support from the public sector as well. In California, for example, legislation was recently enacted to allow insurance firms operating within the state to offer PAYD insurance options. Pricing Variations. PAYD insurance programs, at minimum, vary the cost of insurance on the basis of distance traveled (with PAYD insurance, the rate is structured on a per-mile basis; with mileage-based discounts, distance is measured in more aggregate terms). It is also typically the case that rates will vary based upon demographic characteristics of the insured driver such as age, gender, and residential location. In a number of programs, rates vary by the time of travel as well. For instance, the per-mile rate may be higher when driving during rush hour or late at night than when driving during off-peak daytime hours. In relatively few cases, the rates also vary based on such additional risk related factors as the speed of travel. Technical Implementation Options. All of the programs reviewed in this study appear to rely on one of three options: the odometer, an OBU connected to the on-board diagnostics (OBD-II) port, or an OBU equipped with a GPS receiver. Programs in which the rate varies solely by driver characteristics and miles traveled tend to rely on the odometer, as this is the cheapest option (in terms of equipment costs) to deploy. If the rate also varies with such factors as time of day or speed of travel, it becomes necessary to make use of an OBU equipped with either OBD- II connection or a GPS receiver. As the latter involves more expensive equipment, programs employing GPS-equipped devices often leverage the technology to offer additional value-added services such as pinpointing location for road-side assistance or tracking stolen vehicles. With regard to collecting mileage data, the odometer-based systems rely on periodic (e.g., once per

25 year) checks of the odometer. More sophisticated systems often employ cellular communications to automatically upload mileage data to the insurance provider, facilitating a monthly billing cycle. Verification and Enforcement Options. Odometer-based systems do not generally involve verification and enforcement activities beyond periodic odometer checks. For more technically complex programs involving in-vehicle equipment, efforts are made to ensure that the device cannot be temporarily disabled (e.g., through the use of a certification seal which, if broken, would indicate that the equipment has been tampered with). Administrative Options. All of these programs are implemented and administered privately. In terms of how payments are structured, three main options exist: (1) monthly payment depending on the amount of usage, similar to a utility bill; (2) standard scheduled payments (monthly, quarterly, etc.) with adjustments based on mileage; and (3) pre-payment of a fixed number of miles or for a certain period of time. User Acceptance Concerns. Because enrollment in PAYD insurance programs is optional, there are no major user acceptance concerns. It is noteworthy, however, that many of the programs rely on GPS equipment, suggesting that the privacy-related concerns often associated with such technology may be lessened (at least for some drivers) by the knowledge that more detailed travel information is being monitored by a private firm rather than by the government. Another potential explanation is that even if privacy concerns persist, motorists are willing to set aside those concerns for the chance to save money on their insurance bill. 3.4. Summary of Observations Reviewing the existing program, studies, and proposals in the areas of general-purpose distance- based road use fees, weight-distance truck tolls, and PAYD insurance, several high-level observations emerge: • A broad array of metering mechanisms are feasible. Options range from simple odometer readings to sophisticated in-vehicle equipment featuring GPS to determine the time and location of travel. All of the options have been demonstrated as feasible, either in existing programs or trial tests. • Metering capabilities vary considerably across the options. Simpler metering mechanisms are only capable of metering total miles, while more sophisticated options can determine the time of travel, the jurisdiction in which travel occurs, and even the specific route of travel. • Desired policy goals influence technology choice. Intended policy goals imply a minimal set of metering capabilities (e.g., to levy congestion tolls, it is necessary to meter the time and location of travel). The choice of technology, therefore, will inevitably be based, at least in part, by the policy goals that underlie the program. • There are no “low cost” options that can be easily verified and enforced. The only low cost option observed in this study involves self-reported odometer readings, and this mechanism is difficult to verify or enforce. Other options require either official odometer

26 inspections (entailing high operational costs) or sophisticated in-vehicle equipment (entailing high capital cost). • Concerns over privacy remain a significant barrier to the use of GPS equipment to support general-purpose VMT fees. Existing proposals and trials have taken significant steps to ensure that the privacy of travel data can be protected (Forkenbrock and Kuhl 2002, Whitty 2003), but the perception that GPS will be used to track and monitor travel remains a potent public concern. Beyond education and outreach, factors that may help overcome privacy concerns include (a) providing the opportunity to save money through use of the equipment (e.g., with pay-as-you-drive insurance), and (b) using the GPS technology to provide additional user features (e.g., navigation, real-time route-specific traveler information). • For weight-distance truck tolls, industry concerns center on the distribution of costs and benefits. In many cases, trucking costs would rise with weight-distance truck tolls. To forestall strong stakeholder resistance, existing weight-distance truck tolls have been structured with additional features that benefit the trucking industry – for example, allowing larger truck loads in certain corridors (Balmer 2004), leveling the playing field with foreign competition (Worsley 2004), and dedicating the resulting revenue to highway investments that will benefit truckers (Ruidisch 2004). • Drivers respond to price signals. Existing trials and programs demonstrate that drivers do respond to price signals embedded in the rate structure. Charging more for peak hour travel in busy corridors, for instance, will encourage drivers to shift their travel to other times or routes of travel (PSRC 2008, Whitty 2007), while charging a higher rate for more polluting vehicles will stimulate more rapid adoption of less polluting vehicles (Ruidisch 2004). The implication is that the concept of leveraging a system of VMT fees to achieve other policy goals has merit.