Policy Research Working Paper 10647 Vehicle and Fuel Taxation for Transport Demand Management Learnings from the Literature through a Development Lens He He Chaeyoung Kim Infrastructure Chief Economist Office December 2023 Policy Research Working Paper 10647 Abstract Correctly pricing private vehicles and their use is paramount evasion of taxes; how the effectiveness of taxes as policy to building sustainable, safe, and equitable transportation instruments also depends on the availability of alternatives systems. However, determining the “right” price—the com- to driving; and what the emergence of electric vehicles bination of taxes on vehicle purchase, ownership, and use means for optimal taxation. Importantly, the paper con- —is a complex problem. Although a rich literature exists on siders how these lessons, mostly derived from high-income the subject, it is built on evidence from developed countries. countries with mature automobile markets, apply to devel- This paper synthesizes the lessons learned from the literature, oping contexts. In addition to the policy discussion, the theoretical and empirical, on vehicle and fuel taxation for paper conducts two exercises compiling empirical evidence. managing private vehicle demand. In particular, the paper It compiles and compares estimates of the externality costs examines the efficiency and distributional impacts of pur- associated with private vehicle use, including congestion, chase, ownership, and use taxes. The literature is unequivocal local air pollution, greenhouse gas emissions, injuries, and that taxing use dominates taxing purchase or ownership on noise. Similarly, it compiles and compares demand response efficiency grounds. Nonetheless, the latter instruments can elasticities to vehicle purchase, ownership, and use taxes. still have important roles to play, for example, addressing Both serve as useful references for researchers, development specific market failures, for equity and political acceptability practitioners, and policy makers. considerations, or for ease of enforcement. The paper also discusses the practical challenges of saliency, gaming, and This paper is a product of the Infrastructure Chief Economist Office. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at hhe2@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Vehicle and Fuel Taxation for Transport Demand Management: Learnings from the Literature through a Development Lens◊ He He, Chaeyoung Kim Keywords: taxation, externality, international development, motorization JEL: O18, R41, R48 ◊ The authors thank Jevgenijs Steinbuks and Stéphane Straub for their insightful comments. Opinions expressed in this study are of the authors alone and do not necessarily reflect the views of The World Bank, its Board of Directors, or its member states. 1. Introduction Correctly pricing private vehicles and their use is paramount to building sustainable, safe, and equitable transportation systems. Although cars and motorcycles provide great mobility, their use is disproportionately responsible for road injuries and deaths, congestion, transport-related greenhouse gas (GHG) emissions, and local air pollution. With the accelerated motorization in many low- and middle-income countries, managing vehicle demand is an increasingly pressing challenge. The urgency is underlined by the World Bank’s Country Climate and Development Reports (CCDRs), which have consistently found transport sector emissions to be among the most rapidly growing (World Bank, 2023). Sound management of transportation demand is crucial for achieving the decarbonization objectives of the Paris Climate Accords and contributes to several Sustainable Development Goals (SDGs), including building inclusive, safe, and sustainable cities and communities, reducing inequality, and taking climate action. Price instruments on private vehicles and their use – including taxes, subsidies, fees, and rebates – are important tools in the transport policy toolbox. Using these instruments to change the price of vehicles that households face enables policy makers to incentivize more sustainable behavior. However, in many developing and developed countries alike they are underutilized or inappropriately designed for this purpose. Determining the “right” price, i.e., combination of instruments on vehicle purchase, ownership, and use, is a complex problem. There are many, and often competing, objectives involved in determining what to tax and by how much. In addition to being useful levers for building sustainable transport systems, vehicle and fuel taxes serve as important sources of government revenue. In many cases, portions of these revenues are earmarked for transportation infrastructure investment and maintenance. Hence, any changes will have wider implications on fiscal budgets. On the consumer side, there is often a broad aversion towards increasing taxes. Making matters worse, taxation of vehicles and fuels will generally burden some people more than others, and choosing winners and losers inevitably creates contention. Consequently, the determination of vehicle and fuel taxes is intensely political, and concerns about emissions and economic efficiency often end up falling by the proverbial wayside. Furthermore, the optimal tax scheme for a given country today is different from that of next year or that of a neighboring country. The optimal tax depends on consumer preferences, which continue to evolve with spatial development and technological innovation. Crucially, the proliferation of electric vehicles (EVs) can fundamentally change the calculus of fuel and vehicle 2 use taxes. It is also important to recognize that vehicle and fuel taxation should not be viewed in isolation. Their effectiveness for managing transport demand depends on numerous contextual factors, such as income levels, built environment, and the availability and quality of mobility alternatives, particularly public transportation. The importance of vehicle and fuel tax policies has not escaped the attention of the academic literature. On the contrary, the climate crisis has shone a spotlight on emissions in the transport sector and brought out a wealth of studies in recent decades. However, the empirical evidence supporting this work comes overwhelmingly from developed economies with mature transport systems and automotive markets. Transportation systems in developing contexts differ in a number of meaningful ways. It is typically characterized by a combination of: low but very rapidly growing motorization; ubiquity of two- and three-wheelers and older, more polluting vehicles; poor mobility alternatives including public transportation services and infrastructure for non-motorized travel; and distorted markets and lacking enforcement. How well the lessons from high-income countries apply to developing contexts remains an open question. The purpose of this paper is to serve as a guide for researchers, policy makers, and development practitioners to the recent literature. In particular, it aims to provide a conceptual understanding of different price instruments, including their advantages and limitations vis-à-vis economic efficiency, distributional impacts, and enforceability. This will enable the reader to engage in policy discussions with an understanding rooted in first principles rather than merely standard practice. Additionally, the paper collates a series of relevant parameters, including external costs of vehicle use and price elasticities of vehicle and fuel demand, which the reader can use as reference for both back-of-the-envelope calculations and further reading. We organize the remainder of the paper as follows: Section 2 motivates the use of price instruments by discussing and quantifying the external costs of private vehicle use; Section 3 reviews the central themes in the vehicle and fuel taxation literature as they relate to transport demand management; Section 4 collates relevant elasticities; Section 5 discusses what lessons can be learned and applied to development practice, focusing particularly on contextual differences and potential pitfalls; finally, Section 6 concludes. 3 2. The External Costs of Private Vehicle Use In deciding whether or not to purchase a vehicle, a household considers the utility the vehicle provides relative to the costs of the vehicle over its lifecycle, including both the initial outlay, maintenance costs, and cost of day-to-day use. Similarly, at the trip-level, a car owner decides whether or not to drive by comparing the travel time, costs, and comfort of driving relative to those of other modes. Notably, these decisions take into account the costs borne by the consumer. However, driving a private vehicle imposes substantial costs on the wider society. These so-called externalities take several different forms, including crashes leading to injuries and deaths, congestion, GHG emissions, and local air and noise pollution. Numerous studies have quantified the various external costs of private vehicle use. One such study determined the average external cost of each kilometer traveled by road is $0.24 (Sovacool et al., 2021). In other words, every kilometer driven costs society 24 cents in addition to the costs already paid by the driver, such as value of time lost in traffic, price of fuel, and vehicle wear and tear. At an estimated 36.5 trillion kilometers traveled per year globally, this translates into an annual external cost imposed on society at large of $8.8 trillion, corresponding to 8.7% of annual global GDP. Table 1 summarizes the externality costs of private vehicle use, measured in dollars per 1,000 vehicle-kilometers. The numbers in the table represent simple averages across estimates from the literature. Please refer to Appendix A for the full list of 89 estimates with notes on geographic context, spatiotemporal scope, fuel types, and other study details. It should be noted that the averages in Table 1 only include estimates reported by vehicle kilometers or units directly convertible to that. Estimates in other units, e.g., dollars per metric ton of CO2, can be found in Appendix A but are excluded from the Table 1 summary calculations. 4 Table 1: Externality costs of passenger transport by private vehicle [$/1000 vehicle-km] Air GHG Scope Congestion Crashes Noise Total pollution emissions All 244.65 17.10 9.74 26.87 9.13 307.49 Spatial1 Urban 457.81 31.56 17.12 51.38 17.87 575.75 Rural 69.79 8.13 13.51 19.26 0.32 111.01 Temporal1 Peak 275.45 21.40 15.04 32.85 5.11 349.84 Off-peak 53.42 12.07 16.26 29.57 13.76 125.09 1 Estimates that do not match the specified scope have been excluded from the spatial and temporal breakdowns. Several patterns emerge comparing these data. Notably, congestion costs make up the bulk of the total externality costs associated with vehicle use at 80% of total mean costs. On the other hand, GHG emissions, which has received increasing attention recent years, contribute only 3% of the external costs. This value may be on the low end, since several studies in the catalogue of literature we reviewed are somewhat dated and the impacts of climate change have become clearer in recent years. Nonetheless, the highest cost estimate for GHG emissions in our catalogue ($44.44 / 1,000 vehicle-kilometers) still only represents a relatively small fraction of the total external cost of private vehicle use. That said, the point here is not to downplay the importance of decarbonization. On the contrary, climate change is one of the major challenges of this generation. Rather, the point is that managing demand for private vehicles should be a priority for numerous reasons, and it contributes benefits along multiple dimensions that are important to justify policies and investments. The potential imbalance of attention has been noted in the literature previously (Lindsey & Santos, 2020). Breaking down the estimates by their spatial and temporal scope reveals some of the underlying variation. As expected, externality costs are higher in urban compared to rural areas because the external costs naturally apply to more people in dense areas. For the same reason, externality costs are more severe during peak hours than during the off-peak. Noise pollution is a notable exception to this trend. This is likely a result of higher sensitivity to noise at night. 5 Figure 1 shows that there is considerable variation in the externality estimates. Note that the vertical axis for congestion costs is 10x the other externalities. Some of this variation can likely be attributed to the differences in spatial and temporal scope across studies as discussed before. However, it undoubtedly also reflects uncertainty and differences in estimation methods and assumptions. Finally, it is important to point out that very few of the estimates come from developing contexts. Rather, most studies concentrate on high-income countries, particularly countries in Europe and the United States. Figure 1: Distribution of externality costs of passenger transport by private vehicle That these costs exist is not the problem from an economic efficiency point of view – they are part of the cost of mobility. Rather the economic issue lies in who pays these costs. Since drivers, i.e., the consumers of travel, do not pay the full cost of driving, they tend of overconsume compared to the socially optimal level. In other words, if drivers of private vehicles had to pay the full cost of driving, they would drive less as a consequence. Figure 2 illustrates this point and the deadweight loss that arises from uninternalized social costs. 6 Figure 2: Overconsumption and deadweight loss arising from uninternalized social costs. The textbook solution for restoring economic efficiency is a Pigouvian tax. That is, a tax set equal to the marginal cost, in turn forcing the driver to account for, or internalize, the externality in their decision making. In theory, this restores economic efficiency and maximizes social welfare outcomes (Pigou, 1920; Knight, 1924). However, in practice, translating Pigouvian principle into policy is far from straightforward. Considering the wide distribution of externality cost estimates, there is rarely consensus around a single optimal tax rate (Parry, 2014). Additionally, political acceptance and enforcement capacity often prevent the implementation of such a tax anyway. We elaborate on these challenges in the subsequent sections. 3. Price Instruments for Transportation Demand Management 3.1. Overview of Instruments From excise taxes to environmental levies, from scrappage subsidies to cordon charges, nearly every aspect of private vehicle ownership and use has been the subject of a price instrument. To make sense of the myriad of instruments in a systematic manner, we consider three categories of instruments. Namely, these are: instruments on use, instruments on ownership, and instruments on purchase. As their names imply, each category of instruments targets a particular aspect of behavior related to private vehicles. The instruments within each category can be vastly different 7 in terms of political justifications, fiscal implications, and enforcement challenges. However, they are tied together by conceptually similar expected consumer responses. This subsection provides an overview of commonly used price instruments in each of the three categories – summarized in Table 2. The remainder of the section takes a deeper dive into the key considerations in designing an appropriate taxation scheme, namely efficiency, distributional impacts, gaming and enforcement, interactions with public transportation, and the emergence of EVs. Table 2: Overview of common vehicle and fuel price instruments for transport demand management Instrument Type Examples Vehicle excise tax Import duty Purchase Environmental levy or feebate scheme Vehicle attribute standard Scrappage credit Registration fee Ownership Licensing fees Certificates of Entitlement Fuel tax Use Distance-based tax Congestion charge 3.1.1. Instruments on Purchase Purchase taxes are the simplest and most ubiquitous form of price instrument on vehicles. In many cases, these taxes are motivated by priorities, past and present, other than transport demand management. First, vehicles are subject to the general consumption value-added tax (VAT). However, governments have numerous reasons to tax vehicles beyond the baseline level. Historically, for example in the UK, road taxes, i.e., tax revenue earmarked for building and maintaining road infrastructure, were charged upfront on vehicle purchase. In countries with a domestic automotive manufacturing sector, governments will commonly impose import taxes on vehicles to promote the competitiveness of that sector. Finally, in many developing countries, where private vehicles remain unaffordable to most, governments tax vehicle purchases as a 8 luxury goods tax to raise additional revenue without angering broad swarths of the population. The primary advantage of purchase taxes, and the reason for their ubiquity, is their relative ease of collection. Being a one-time payment, whether at import, transaction, or registration, purchase taxes are far more straightforward for governments to collect and enforce. Purchase taxes can also be designed to encourage substitution towards more sustainable vehicles. Where the policy motivation is decarbonization of the transport sector, governments often opt for a revenue-neutral approach to make the policy more politically palatable. Such approaches can take the form of a combined fees and rebates scheme – so-called feebates. For example, consumers can be incentivized to purchase vehicles with better fuel economy by subsidizing high fuel economy vehicles and conversely charging a tax on low fuel economy ones. Similarly, vehicle age and engine size are commonly used as proxy attributes for how polluting a vehicle is. Although not directly price instruments, standards can be formulated to produce equivalent outcomes. There is a symmetry between price instruments and quantity instruments. Intuitively, this can be understood as intervening along either of the two axes in the supply-demand curve in Figure 2 to achieve the optimal outcome. The most well-known example comes from the environmental regulation literature with the equivalence between a carbon tax (price instrument) and a cap-and-trade program (quantity instrument). For example, rather than a feebate scheme, the US has made use of the Corporate Average Fuel Economy (CAFE) standards to incentivize manufacturers to improve vehicle fuel economy since 1975. The CAFE standards require each auto manufacturer to ensure that the average fuel economy of all the vehicles they sell meets a certain target, lest they pay a fine. In turn, manufacturers adjust the prices of vehicles in their portfolio so as to reach this target – effectively achieving the same outcome as a feebate scheme. This symmetry between price and quantity instruments means that the intuitions, advantages, and limitations of purchase taxes broadly apply to standards, and vice versa. See Gillingham (2013) for a comprehensive examination of the symmetry between fuel economy standards and feebates. Finally, scrappage programs – despite their name – can also be considered instruments on purchase. A typical scrappage program awards car owners credit towards buying a new and more fuel-efficient car when they scrap an old one with poor fuel economy. The credit effectively works as a purchase subsidy. However, linking it to the disposal of an old vehicle enables scrappage programs to affect the existing stock of vehicles that are otherwise out of reach for 9 typical price instruments on purchase. In addition to improving fuel economy, scrappage programs provide economic stimulus targeted at the auto industry. This led to their rise in popularity following the 2008 global recession in many countries with mature automotive markets, including the US, Japan, and much of Europe. 3.1.2. Instruments on Ownership Instruments on ownership, such as registration fees or licensing fees, serve as property taxes on private vehicles. They are typically collected annually and often increase with engine size. A positive property tax discourages ownership across the board, while a taxation structure that penalizes more polluting vehicles pushes consumers towards greener choices. Compared to typical taxes on purchase, ownership taxes are more effective for driving changes in the vehicle population, since they affect both new purchases and existing stock. However, they require somewhat greater institutional capacity to enforce. Consequently, they are not as ubiquitous, but common in high-income countries and to a lesser extent in middle-income countries. In Singapore, a Certificate of Entitlement (COE) is required to own a car. Only a limited number of COEs, set by the government, are available at any given time. Singaporeans looking to own a car must first acquire a COE through an official auction. Once acquired, the certificate entitles the holder to own a car for 10 years, after which the COE expires and must be renewed. Additionally, should the COE holder wish to deregister their car before the 10-year term, they will be refunded the prorated cost of their COE according to the time remaining on it. With the continued growth of the Singaporean economy and the cap on quantity, COE prices recently reached US$100,000 (Tjoe, 2023). Interestingly, the COE program functions as an instrument on both purchase and ownership. The COEs’ 10-year term must be paid for upfront. Additionally, the refund value is prorated according to the price originally paid for the COE rather than the live COE auction price, making subsequent changes to COE quantities an ineffective policy instrument for reaching existing COEs. However, at the same time, the prorated refunds and 10- year renewal requirements provide car owners incentives to deregister their vehicle. 3.1.3. Instruments on Use Fuel taxation is the most prominent price instrument on private vehicle use. From a transport demand management point-of-view, the purpose of fuel taxes is to put a price on the externalities of vehicle use. However, for many countries this purpose is superseded by fuel taxes’ 10 considerable impact on government budgets. Similar to purchase taxes, this prominence is owed to the ease of collection, specifically at import (Timilsina & Dulal, 2008). Fuel taxes have also historically been, and in many places continue to be, earmarked for road funds. However, with the advent of alternative fuel vehicles, particularly EVs, traditional fuel taxes are at risk of becoming ineffectual, both for raising government revenue and for internalizing the externalities of private vehicles that remain beyond the transition to a greener fuel source. To address the eventual decline in fuel consumption and associated tax revenue, many economists have touted the advantages of distance-based taxes (Heine et al., 2012; Langer et al., 2017). As their name suggests, distance-based taxes, also colloquially known as VMT (for vehicle-miles traveled) or VKT (for vehicle-kilometers traveled) taxes, charge drivers a fee based on the distance their vehicles travel. This provides policy makers an alternative, and arguably more direct, instrument to charge for the use of private vehicles. The fee can in theory be differentiated by relevant vehicle attributes, when and where the vehicle was driven, and even characteristics of the vehicle owner, such as income. The flexibility enables the design of a more efficient price instrument. However, granular targeting of these taxes requires similarly granular data on both vehicle use and the user. Not only does this put higher demands on tracking technology, but the collection of such data, especially location data, naturally raises privacy concerns among the public. As a consequence of both the technological and political acceptance barriers, distance-based taxes have yet to become widespread. Numerous pilots have been conducted. Most notably, vehicle owners in Oregon state in the US have since 2015 been able to sign up to pay distance-based taxes in exchange for fuel tax credits and reduced vehicle registration fees. Participants install a third-party telematics device in their vehicle that logs mileage, fuel consumption, and optionally, GPS location. Several demand management measures from the transport engineering practice are also price instruments on use. Cordon charges, road tolls, and parking fees all put a price on the externalities of vehicle use. However, unlike fuel and distance-based taxes, these congestion- related instruments are inherently spatial. For example, the toll on a congested road should be high to reflect the external cost of delays imposed on the many fellow travelers. Similarly, where space is at a premium, parking fees should be expensive to reflect the opportunity cost of occupying the space with a parked vehicle. Fuel taxes cannot directly capture these spatial 11 considerations. Conversely, congestion-related instruments cannot directly capture the emissions-related externalities. Hence, the instruments serve as complements. 3.2. Efficiency Restoring efficiency is the primary economic argument for taxing private vehicles and their use. Absent government intervention, equilibrium car use yields excessive miles driven and sub- optimal social welfare due to the negative externalities. These externalities arise from the use of private vehicles, as opposed to their purchase or ownership. Hence, instruments on use align most closely with the Pigouvian tax principle and are broadly the most appropriate for restoring efficiency. Despite its conceptual simplicity, implementing an optimal Pigouvian tax is exceptionally complex in practice. As discussed in Section 2, there are numerous different externalities associated with vehicle use. Unfortunately, no single existing tax instrument can by itself capture all these externalities. For example, fuel taxes can accurately internalize the damages caused by GHG emissions. However, local air pollution impacts depend not only on how much a vehicle is used but importantly also on where it is used. That is, driving a polluting car in a densely populated area multiplies the number of impacted people and by extension the total external cost. Conversely, congestion-related instruments, such as cordon charges, can more accurately capture spatiotemporal variations but are inherently crude vis-à-vis the actual use (Santos et al., 2010a). Furthermore, estimates of the magnitudes of each externality span wide ranges as documented in Table 1. In other words, determining the optimal taxation structure requires, first, selecting a combination of instruments with adequate granularity and dimensionality to match that of the different externalities, e.g., spatial, temporal, emissions, etc.; and then, setting the tax rates such that they most accurately approximate the external social costs of private vehicle use. Complicating matters further, a Pigouvian tax is strictly speaking optimal when considered in isolation. However, in practice, any new price instrument will exist within an existing system of taxes on substitutes and complements. If these are suboptimal and unalterable, the optimal new price instrument will not generally be a Pigouvian tax (Sandmo, 1975). In the transport sector context, this is particularly relevant when fuel taxes are set below optimal levels but cannot be raised, e.g., due to political opposition. Despite these complications – not to mention equity, enforceability, and other practical considerations – the broad consensus remains that taxing the 12 use of private vehicles is the preferred price instrument for restoring economic efficiency (Anderson & Sallee, 2016). Additionally, even though the Pigouvian ideal is rarely achievable in practice, it still serves as useful intuition and guiding principle. The direct effect of a fuel tax is straightforward. It encourages drivers to consume less fuel, whether by driving less or by using more fuel-efficient vehicles. However, a fuel tax also has several indirect effects on both drivers and vehicle manufacturers that are slightly less obvious. The first indirect benefit relates to road injuries and fatalities. In case of a vehicle collision, being hit by a heavier vehicle increases both the injury severity and fatality risk, while being the driver of a heavy vehicle is safer for the occupant. In other words, driving a heavier vehicle further externalizes the social costs of road injuries and fatalities. Over the past half century, the average vehicle weight has grown continuously in the U.S. (U.S. EPA, 2022). Some authors hypothesize that, all else equal, this favors the purchase of increasingly heavy vehicles, and has likely contributed to creating an “arms race" effect in vehicle weight. A simple distance-based tax can price in the social cost of road injuries and fatalities at average incident rates, i.e., the average cost of road incidents per kilometer driven. However, such a tax fails to account for vehicle weight. Conveniently, a vehicle’s fuel consumption scales positively with its weight. Hence, a fuel tax can simultaneously internalize mileage and weight externalities. This can be another argument for raising fuel taxes (Anderson & Auffhammer, 2014). A second indirect benefit of fuel taxes relates to their saliency and the signal they send to consumers. Although changes to fuel taxes and the tax-exclusive components of fuel prices affect consumer wallets the same, empirical evidence suggests that consumers respond more strongly to the former (Li et al., 2012). Specifically, a fuel tax increase yields larger reductions in fuel consumption and larger improvements to fuel economy of sold vehicles than a commensurate increase to the tax- exclusive components of fuel price. The difference can possibly be attributed to fuel tax changes being more salient, e.g., due to news coverage, than changes to the tax-exclusive components of the price. Additionally, for deciding fuel economy when purchasing a vehicle, consumers might see fuel taxes as a reflection of a longer-term policy direction, whereas changes to the tax- exclusive components are more likely interpreted as short-term price fluctuations. In other words, the strength of fuel taxes as a policy instrument is amplified by its saliency and signaling effects. Consequently, evaluating the impact of a fuel tax policy using the total price elasticities could actually underestimate the behavioral response but overestimate the tax revenue. Last but 13 not least, fuel taxes also affect vehicle manufacturers by encouraging them to offer more fuel- efficient options in the long term (Klier et al., 2020). When fuel taxes increase, making fuel more expensive, demand for more fuel-efficient vehicles similarly grows. Consumers’ increased willingness-to-pay makes it worthwhile for automakers to invest in innovation to improve fuel economy. Furthermore, if automakers can sell a larger quantity of fuel-efficient vehicles, those vehicles benefit more from scale economies in both innovation and manufacturing. That is, automakers can spread the fixed costs of innovation and manufacturing equipment for fuel- efficient vehicles over a larger number of sales, thus lowering the average cost. Simply by increasing the upfront cost of cars, purchase taxes reduce the demand for private vehicles, and in turn reduce the negative externalities associated with their use. Additionally, cleverly designed tax structures, such as feebate programs, can incentivize substitution towards more sustainable vehicles. However, all evidence points to purchase taxes being less capable of restoring economic efficiency. In particular, the separation between the taxed activity – vehicle purchase – and the externality-creating activity – vehicle use – makes purchase taxes a crude instrument. For example, the welfare costs of reducing gasoline consumption via the CAFE standards in the US were estimated to be three times those of an equivalent direct gasoline tax (Austin and Dinan, 2005). Furthermore, price instruments on vehicle purchase do not have any direct effect on the existing stock of vehicles (Sallee, 2010). Rather, existing vehicle owners will only be reached once they choose to replace their vehicle. The aforementioned separation between the taxed and externality-creating activities creates misaligned incentives, which can lead to unintended outcomes. Although a tax on purchase does not have a direct effect on existing stock, it can in fact have a harmful effect indirectly. An upfront tax makes vehicle replacement more costly, which extends the life of old, generally more polluting, vehicles (Gruenspecht, 1982). This so-called “Gruenspecht Effect” – named after the author who first discussed the issue – is estimated to have offset 13-16% of the fuel savings of the CAFE standards in the US (Jacobsen & van Benthem, 2015). Several different strategies can be taken to alleviate the Gruenspecht effect. Instituting a revenue-neutral feebate scheme rather than a strictly positive tax, such that replacement does not necessarily become more expensive, can go some way to lessening the impact. However, those most financially constrained, and consequently most likely to delay replacing an old vehicle due to purchase taxes, are likely the 14 same people in the market for the cheaper, usually less fuel efficient, vehicles, which are penalized by typical feebate schemes. Hence, a revenue neutral feebate scheme based on fuel economy may still prevent them from replacing their old inefficient vehicle. Alternatively, combining purchase taxes with an ownership tax scheme that penalizes ownership of older vehicles would also counteract the Gruenspecht effect (Jacobsen et al., 2022). Fuel economy is commonly the target for feebate schemes, vehicle standards, and sometimes, indirectly, ownership fees. Over time, such policies push car buyers and manufacturers towards more efficient vehicles. However, as a byproduct, this lowers the marginal use cost of vehicles. In other words, driving becomes cheaper, which in turn encourages more driving. Additional vehicle-kilometers traveled induced by this so-called “rebound effect” negated 20%-25% of the fuel savings from fuel economy improvements between 1966-2001 in the US (Small & Van Dender, 2007). Another study puts the emissions reductions lost due to increased driving at 30%- 80% (Frondel et al., 2017). The extra vehicle-kilometers traveled does not just cancel out some of the fuel savings, but they also generate their own additional non-emissions-related externalities, which broadly dominate the emissions-related ones as we saw in Table 1. Despite these drawbacks, arguments can still be made for purchase taxes to have a role to play in managing transport demand. Beyond being simple to administer, a single upfront payment can be more salient and thus effective for changing behavior. In particular, this is useful if consumers are myopic in their vehicle purchasing decision. That is, failing to fully account for the future fuel savings of driving a more efficient vehicle would lead car buyers to purchase a car with a lower fuel economy than what is rationally optimal. In such cases, introducing an upfront price instrument that depends on vehicle fuel economy can correct for the undervaluation of future savings (Sallee, 2010). Numerous studies in the literature have attempted to quantify the severity of consumer myopia in the context of vehicle purchases. However, the empirical evidence remains inconclusive (Greene, 2010; Busse et al., 2013; Allcott & Wozny, 2014; Sallee et al., 2016; Grigolon et al., 2018; Gillingham et al., 2021; Leard et al., 2023). 3.3. Distributional Impacts Recognizing that taxation of vehicles and fuels creates winners and losers is fundamental to effective policy making. Until now, we have made a case for the use of price instruments to improve social welfare by restoring economic efficiency – growing the pie, so to speak. 15 However, who gets the bigger slice? And do taxes and subsidies change that? In theory, the net welfare gain can be redistributed via lump sum transfers to compensate those directly harmed by the corrective policies to prevent anyone from being worse off overall. Unfortunately, such Pareto improvements are not feasible in practice (Sallee, 2019). Identification of winners and losers is imprecise and redistribution between them is costly (Parry & Williams, 2010). This is exacerbated by the diffuse nature of the externalities associated with private vehicles and the heterogeneity of use patterns. Hence, the costs of targeting transfers with adequate precision can quickly offset the welfare gains created by the price instruments. In other words, even an ideal Pigouvian tax creates winners and losers. Effective policy making requires consideration for the distributional impacts of price instruments and their resulting political acceptability. In our review of price instruments for restoring efficiency, theory, such as Pigouvian taxation, served as guiding principles. Unfortunately, it is not generally obvious a priori if a given formulation of a price instrument is progressive or regressive. Instead, more often than not, we are left with the rather unsatisfying answer, “it depends”. Take for example the US, where extensive research has been conducted on the distributional impacts of various instruments. As Table 3 shows, income is a strong predictor of vehicle-related consumption. Higher income households own more, newer, and more expensive vehicles. They also use their vehicles more, driving longer total distances, and consequently spending more on fuel. However, private vehicle ownership is ubiquitous in the US. Even households in the lowest income quintile own, on average, 0.9 vehicles. Despite their lower consumption, vehicle and fuel expenditure make up a larger percent of income for low-income households. 16 Table 3: Average household-level vehicle-related consumption use by income quintile in the US, 1980-2014 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 (lowest 20%) (highest 20%) Income, before tax 13,607 27,212 44,478 69,444 132,768 Vehicles 0.9 1.3 1.7 2.0 2.4 Vehicle age 10.2 9.4 8.4 7.7 6.8 Price per vehicle 8981 9927 11,750 14,115 18,930 VMT 7274 11,042 15,180 19,218 24,344 Fuel expenditure 1008 1545 2125 2673 3290 Vehicle price / income 0.66 0.37 0.26 0.20 0.14 Fuel expenditure / income 0.07 0.06 0.05 0.04 0.03 Adapted from Greene & Welch (2018) Regressivity is a common argument levied against fuel taxation – but reality is not that simple. Poorer households spend a larger share of their income on fuel. Hence, the direct distributional impact of raising fuel taxes is regressive. However, poorer households are generally also more price responsive. They are more willing to reduce vehicle-kilometers traveled and switch to alternative modes. Their larger price elasticity mitigates the regressivity of fuel taxes if adequate alternatives are provided, e.g., public transportation, non-motorized transport infrastructure, appropriate land use planning, etc. (West, 2004). Additionally, the money paid in fuel taxes do not disappear. They enter fiscal budgets, and economists increasingly recognize that how they are recycled determines the distributional impact of fuel taxes (Davis and Knittel, 2019). Due to the larger absolute expenditure of high-income households, a flat recycling scheme has been found to be progressive and welfare-enhancing for the bottom 40 percent (Bento et al., 2009). Finally, a switch to distance-based taxes allows more precise targeting while still taxing vehicle use. For example, a relatively simple urban-rural differentiated VMT tax is both more efficient and more progressive in the US context (Langer et al., 2017). Although less apparently so, instruments on purchase are broadly regressive in the US context. The CAFE standards in the US have contributed to the improvement of fuel economy of vehicles across all market segments. In turn, this has lowered the per mile fuel consumption and total fuel expenditures. However, to satisfy the standards, manufacturers have had to develop more efficient vehicles. Concretely, a one mile-per-gallon tightening of the CAFE standards is estimated to have cost manufacturers $9-$27 per vehicle (Anderson & Sallee, 2011). The costs of 17 this innovation inevitably pass through to consumers, who ultimately fund the technological development. High-income households more frequently purchase new vehicles that are also more expensive, thus incurring a larger share of these costs. This has sometimes been used as an argument for the progressivity of instruments on purchase. Indeed, considering only the impact on the market for new cars, such the CAFE standards have been found to be progressive. However, if the analysis accounts for the propagation of price impacts to the used car market, the distributional impact becomes regressive (Jacobsen, 2013; Davis & Knittel, 2019). Additionally, instruments on purchase based on fuel economy subsidize high fuel economy vehicles, either implicitly through standards or explicitly through feebate schemes. Efficient vehicles are generally more expensive and more often purchased by high-income households. Conversely, inefficient vehicles are generally cheaper and more often purchased by low-income households. These consumption patterns tend to further exacerbate the regressivity of price instruments on purchase based around fuel economy (West, 2004). International evidence underlines the context-dependent nature of distributional impacts. Although possibly more nuanced, the findings above may be specific to the US. Examining the international evidence underlines the dependency on local income, spatial development, and car use patterns. A study of fuel taxation across seven countries in Europe, namely France, Germany, United Kingdom, Italy, Serbia, Spain, and Sweden, found that their distributional impacts were approximately proportional, i.e., neither progressive nor regressive (Sterner, 2012). The largest outlier is Serbia, for whom fuel taxation is relatively more progressive. Notably, Serbia is the lowest income country in the sample. An evaluation of fuel taxes in Brazil also found the distributional impacts to be progressive (Proque et al., 2022). These findings align with the intuition that taxation of private vehicles and their use is progressive in lower income countries, where cars remain more of a luxury good. However, as countries become richer, cars become more ubiquitous, and taxation of private vehicles becomes increasingly regressive (Berri et al., 2014). The distributional impacts also depend on which and how externalities are accounted for. In Sweden, a study found that people in rural areas are far more adversely affected by motor taxes than in urban areas. These distributional differences were not counterbalanced by various recycling schemes (Eliasson et al, 2018). In France, a feebate scheme introduced in 2008 encouraged adoption of diesel over gasoline vehicles. Diesel engines emit fewer GHG emissions but more local air pollutants. In other words, the policy changed the distribution of externalities 18 across space. Specifically, a study found that while the policy had a net positive impact on welfare, local air pollution increases most in areas where high-income household live (Durrmeyer, 2022). Hence, the feebate scheme had a progressive distributional impact, albeit in a somewhat roundabout way. 3.4. Poor Saliency, Gaming, and Evasion Even well-designed policies can sometimes yield smaller than expected behavioral responses, for example producing emissions reductions below predicted levels. Numerous practical considerations can underlie such discrepancies. In the context of vehicle and fuel taxation, the literature primarily concerns itself with three challenges. Namely, these are: poor saliency of instruments to consumers, manufacturers gaming regulations, and illicit evasion of taxes. The saliency of a price instrument increases its behavioral response. Conversely, consumer behavior changes less when taxes are not salient. For example, posting tax-inclusive prices in grocery stores have been shown to reduce demand by 8 percent (Chetty et al., 2009); and, directly relevant to transport, electronic road toll collection has resulted in equilibrium tolls being 20 to 40 percent higher than were they collected manually (Finkelstein, 2009). A study of price instruments in France, Germany, and Sweden aimed at reducing vehicle GHG emissions adds further supporting evidence (Klier & Linn, 2015). The instruments in all three countries were tied to vehicle emissions rates. However, while France introduced an instrument on purchase with discrete tax brackets, Germany and Sweden both introduced ownership taxes that scaled linearly with vehicle emissions rates. The authors found the approach in France to be considerably more effective. They attributed the differences to the increased saliency of both paying upfront, as opposed to periodically, and the discrete brackets, as opposed to the more nebulous linear function. Manufacturers will take advantage of loopholes to maximize profits. Improving vehicle efficiency and technology is costly for manufacturers. Hence, they will take advantage of incomplete regulation to avoid doing so to the extent legally possible. In particular, this gaming of the regulatory system typically involves taking advantage of discrete “edges” in regulation, whereby minor changes that do little to ameliorate the externalities targeted by the regulation but yield preferential tax treatment. Common gaming strategies can take the form of, for example, vehicle design tweaks to move vehicles to lower tax bracket, relabeling of vehicle categories to 19 take advantage of a more lenient regulatory classification, or temporal shifts in sales to benefit from temporary tax rebates (Sallee, 2010). Depending on instrument design, the scale and impact of gamification can be quite considerable. The 2015 EU-wide vehicle emissions standards yielded a 14% decrease to sales-weighted emissions by official calculations. However, of these, two-thirds were estimated to be attributable gaming (Reynaert, 2021). Thus, while discretized regulation may be advantageous for simplicity and saliency, such design also creates opportunities for gaming. Recognizing this trade-off is important for designing effective price instruments. The larger the tax, the larger the incentive to evade. This follows from common economic intuition and is borne out in practice. In Finland, importing a vehicle requires paying an import tax that is a set percent of the value of the vehicle. For used cars, the value is discounted by the vehicle’ mileage. In other words, a vehicle with high mileage is liable for lower import taxes, all else equal. However, the Finnish customs authorities, who are responsible for collecting the import tax, do not regularly verify the reported mileage, creating an opportunity for tax evasion. Comparing the reported mileage with odometer readings during subsequent inspections reveals that approximately 10% of imported used cars over-reported their mileage (Harju et al., 2020). In 2012, import taxes were reformed to also depend on carbon emissions rates. The reform saw evasion become more prevalent for cars affected by the tax hike but remain unchanged for low- emissions vehicles. This supports the intuition that higher taxes result in a greater incentive to evade. Recognizing the enforcement gap, in 2013, the Finish customs authorities acquired access to odometer readings from downstream inspections. An information campaign randomized controlled trial (RCT) showed that awareness of the more stringent enforcement was effective at reducing tax evasion. 3.5. Interactions with Public Transportation Sustainable mobility requires a multi-pronged approach that involves public transportation and non-motorized travel in addition to private vehicle demand management. Price instruments on private vehicles and their use push travelers towards more sustainable and efficient choices. However, pushing can be far more effective if complemented by a pull. That is, if there are viable alternatives to private vehicles, such as high-quality public transportation or built environment designed for non-motorized travel, drivers are much more likely to change their 20 behavior in response to vehicle and fuel taxes. Absent these alternatives, taxing private vehicles is simply pushing against a figurative wall. In economics terms, driving is a derived demand. Consumers make trips to access work, school, shops, leisure destinations, etc., and private vehicles are the means to that end. Inelasticity in a derived demand arises when the demand for the final “good” is inelastic and there are no viable substitutes. Although telecommuting and e- commerce have begun to change the former in certain sectors and contexts, leaving the house remains broadly necessary for the vast majority in the foreseeable future. So long as that is the case, investing in public transport will remain an essential component of any sustainable mobility strategy. This interdependency between private vehicle use and public transportation is also evident empirically. For example, a study of the impact of a fuel tax reform across metropolitan areas in Spain found that the reform more effectively reduced carbon emissions and energy use from transportation in cities with greater public transport penetration (Danesin & Linares, 2018). Several studies in the US examine the elasticity of transit demand with respect to fuel prices (Mattson, 2008; Lane, 2010). This cross-elasticity reflects the extent to which public transportation serves as a substitute to driving in different contexts. Transit demand is relatively inelastic in small cities, where lacking service quality and coverage likely prevent more significant substitution. By comparison, transit demand is more elastic to fuel price changes in mid-size cities with population in the 100,000-500,000 range. Interestingly, the evidence is mixed for large cities. The suggestion of an S-shaped elasticity curve with respect to city size could be indicative of ridership saturation given existing capacity constraints and urban form in certain large metropolitan areas. Conversely, investments in public transportation have been shown to help reduce road congestion and improve air quality. However, the effectiveness of such investments similarly and unsurprisingly differs considerably by context (Beaudoin et al., 2015). Several studies have examined the viability of using revenues from taxation of private vehicles, specifically fuel taxes, to cross-subsidize public transportation. This approach is particularly attractive because it improves welfare while having a clear mechanism for maintaining revenue neutrality for government budgets (Barros & Prieto-Rodriguez, 2008). Despite the conceptual appeal, the redistributive nature of such tax reforms inevitably makes them politically 21 contentious. Thus, it is crucial to consider carefully the winners and losers in designing the tax and subsidy schemes (Victor-Gallardo et al., 2022). Additionally, it is also important to understand the impacts on transit operations. At first glance, one might think that increasing transit ridership increases its revenue. However, the vast majority of transit systems operate with government subsidies and would generate considerable deficits through operations alone. In Germany, fuel taxes were found to exacerbate the operational deficit because it primarily shifted demand for commuting trips, whereas shopping and leisure trips remained relatively inelastic (Storchmann, 2001). In other words, the transit system became even more crowded during peak hours, when the system was already at capacity, forcing additional investment. Meanwhile, it did little to improve efficiency during off-peak hours, when the system is relatively empty and operating with large deficits. 3.6. Emergence of Electric Vehicles The proliferation of EVs poses challenges to the conventional wisdom. In China, more than a third of new vehicles sold are EVs (Li et al., 2022; Li & Goh, 2023). Although few countries, if any, can match China’s fervor in policy making to promote EV adoption, EVs’ ascendancy appears a matter of time in most contexts. Their rise in popularity has drastic consequences for taxation of vehicles and fuels. Business-as-usual tax schemes that rely heavily of fuel taxation will inevitably leave a revenue gap as traditional fuels are increasingly replaced by electricity (Jenn et al. 2015). For example, in the US, each EV results in more than $300 of foregone fuel tax revenue per year (Davis & Sallee, 2020). Even at less than 1% of the fleet of registered vehicles, this creates a fiscal gap of $250 million annually. In places where fuel taxes are earmarked for transportation infrastructure maintenance and capital investment, such revenue gaps will have direct adverse impacts on mobility and accessibility. Furthermore, until EV technologies mature and the price of EVs becomes competitive with internal combustion engine vehicles (ICEV), EVs will mostly be affordable to high-income households. Consequently, the lack of a fuel tax equivalent for EVs will have equity implications and change the current redistribution calculus. The technological innovation calls for policy makers to evolve their tax structures accordingly. However, optimal taxation of EVs is not straightforward. Compared to ICEVs, EVs produce fewer externalities, particularly in terms of local air pollution and GHG emissions – provided 22 that the electricity fueling the vehicles were produced in a cleaner manner. However, other externalities, such as congestion and road safety, remain largely unchanged. Hence, although electrification has rightly been brandished for its role in decarbonizing the transport sector, EVs are merely the “lesser evil” considering the full suite of external costs of private vehicle use. The first-best tax on EVs follows the Pigouvian principle of internalizing the external costs. This can be implemented through a distance-based tax – ideally one that is spatiotemporally disaggregate – but such schemes remain in their infancy, as discussed in previous section. As a more direct substitute for fuel taxes, several states in the US have introduced legislation to tax electricity at public EV charging stations (Jaros & Hoffer, 2023). However, with the availability of private home chargers, it is unclear if such policies are effective or merely an inconvenience. With all that said, the first-best tax is not relevant in most cases anyway. In practice, few countries tax the use of ICEVs adequately with fuel taxes shown to be below optimal rates in the US (Parry & Small, 2005) and many developing countries (Granger et al., 2021). If these existing distortions cannot be ironed out, the optimal tax on EVs is below the first-best level – and potentially even a subsidy – so as to encourage substitution away from the larger distortion. For example, in the US the optimal distance-based tax can be positive or negative depending on the spatiotemporal context (Davis & Sallee, 2020). On the other hand, in Germany, where fuel taxes are relatively higher, EV use should unequivocally be taxed (Hirte & Tscharaktschiew, 2013). Numerous countries have made use of subsidies or tax credits to encourage consumers to buy EVs. For example, as early as 2008 the US introduced a federal tax credit program that subsidized EVs purchases up to $7,500 (IEA, 2017). Not to be outdone, in 2010, China announced a ¥60,000 subsidy (corresponding to around $9,000 at the time) for each EV sold paid to the manufacturer (Motavalli, 2010). While instruments on purchase are not effective at internalizing the external costs of vehicle use, they serve a different purpose in the early stages of EV adoption. In their infancy, new technologies are expensive for numerous reasons, including, chiefly, the costs of research and development. However, other car manufacturers, drivers, and society at large subsequently benefit from the technological innovation. In other words, early- stage EV adoption has positive externalities, making purchase subsidies an appropriate instrument. Importantly, as technologies mature, the subsidies will have served their purpose and should be phased out. Hence, many programs similar to the aforementioned ones have built-in sunset clauses, ending the subsidies after a target number of vehicles has been sold. It should also 23 be noted that these subsidies can quickly become expensive. Early adopters tend to be individuals who are enthusiastic about new technologies and likely would have made the switch to EVs regardless of subsidies. In economics terms, their demand tends to be relatively inelastic with respect to price. Each $1,000 subsidy between 2010-15 has been estimated to increase EV uptake by 2.6% (Jenn et al., 2018) and 7.5% (Wee et al., 2018) in the US. In other words, each additional EV sale cost the federal government $38,462 and $13,333 by the two estimates, respectively. Beyond merely the cost, it is once again important to keep in mind the equity considerations. For example, a study of a suite of policies to encourage EV adoption in Colombia showed that is the interventions largely benefit high-income households, making them regressive (Callejas et al., 2022). Alternatively, policy makers can also opt to encourage EV adoption indirectly by subsidizing charging infrastructure. This has been shown to be an effective strategy (Li et al, 2022). For example, results of a survey conducted in China in 2017 showed consumers valued having a charging station every 5 km higher than a 50% discount off charging fees (Ma et al., 2019). Additionally, increasing the ubiquity of charging infrastructure lowers the demand on range. With batteries making up a considerable portion of the weight of EVs, lowering battery capacity can save non-negligible weight. All else equal, this lowers the severity of injuries in the event of road collisions, damage to road infrastructure, and importantly, the energy required to move the vehicle. Finally, there is a positive feedback effect between charging infrastructure investment and consumer EV adoption (Li et al., 2017). That is, more charging stations encourage more EVs, and vice versa. This feedback makes subsidizing the two sides of the market complementary. Governments can kickstart the feedback loop by upgrading the fleet of public vehicles away from traditional fuels (Corts, 2010). 4. Elasticities Quantifying the response to price instruments is fundamental for designing taxation schemes for demand management. In this section, we present a compilation of response elasticities to price instruments found in the literature. Table 4 shows inverse-variance weighted average elasticities of new vehicles, fuel economy, fuel demand, and traffic volume with respect to purchase, ownership, and use taxes. Appendix B provides a description of the summarization methods and the complete list of elasticities and their geographic and temporal scopes. 24 Table 4: Weighted average elasticities Weighted average elasticities4 Instrument type New vehicles1 Fuel economy Fuel demand2 Traffic volume3 -0.42 -0.16 0.00 Purchase - (0.080) (0.046) (0.120) -0.09 -0.13 -0.09 Ownership - (0.008) (0.033) (0.035) 0.12 -0.11 -0.04 Use4 - (0.190) (0.005) (0.002) Standard errors in (parentheses). 1 Includes studies examining sales and new registrations. 2 Includes studies examining fuel demand, gasoline demand, and CO2 emissions. 3 Includes studies examining traffic volume and vehicle miles travelled. 4 Fuel taxes only; congestion-related taxes not directly comparable. Examining the weighted average elasticities, the empirical evidence broadly supports the hypotheses from theory with all elasticities appearing with the expected sign. However, the instruments on purchase and ownership appear to have larger elasticities than anticipated in several cases, particularly for fuel demand. It should be noted that the averages for instruments on purchase and ownership are supported by relatively few studies. Table 5 shows the number of studies used to calculate each average. While there are numerous studies examining use taxes, specifically fuel taxes, far fewer studies examine the effectiveness of purchase and ownership taxes. Notably, some cells do not have corresponding elasticities estimated. This is understandable for the elasticity of fuel economy with respect to purchase and use taxes, since these do not inherently affect fuel economy. Instead, numerous studies examine various feebate schemes, but there is not a standardized elasticity for comparison. For the elasticity of new vehicle sales with respect to use taxes, some authors estimate the impact of total spending on fuel, but this represents a combination of fuel economy and price. It is also worth pointing out, once again, that there are only a few estimates originating from developing countries with the vast majority of studies conducted in mature vehicle markets. 25 Table 5: Number of studies used for the calculation of averages Total number of studies Instrument type New vehicles Fuel economy Fuel demand Traffic volume Purchase 1 0 2 1 Ownership 11 0 2 2 Use 0 1 56 11 Finally, the average elasticities in Table 4 obscure the temporal dimension. Numerous studies distinguish between short-run and long-run elasticities. In the long-run, households can make greater behavioral adjustments, such as vehicle replacement and residential relocation. Hence, as expected, responses were generally found to be greater in the long-run. However, the definition of short-run and long-run were not always consistent across studies. Typically, one year is considered short-run, while more than five years is considered long-run. Unfortunately, the inconsistent definitions make summaries and comparisons difficult. At the same time, a robust understanding of the impact of various tax instruments over time is crucial to effective policy making considering the rapidly changing landscape of the transportation sector. 5. Lessons for Development Practice As evidenced by the first four sections of this paper, the empirical work on vehicle and fuel taxation has overwhelmingly focused on high-income countries. This focus is a natural consequence of the prevalence of private vehicles in high-income countries and the stronger institutional capacity to enforce a variety of vehicle and fuel related price instruments. Several studies have recognized this gap and called for research efforts focusing on developing contexts (Anderson & Sallee, 2016; Jacobsen et al., 2022). However, more than likely, the empirical evidence specific to developing countries will only gradually become available as their vehicle markets mature – in turn changing the question from “how can we shape development?” to “how can we undo its damage?” In light of the differences in challenges faced, policy makers and development practitioners should not merely adopt existing practices from mature automotive markets and assume direct applicability. Rather, it is all the more important that they acquire an understanding of tax policies built on first principles. In this section, we examine the topic through a development lens and highlight the key areas and patterns where developing contexts 26 tend to differ from the developed ones, where much of the existing evidence originates. We discuss the policy implications of these differences and possible additional considerations that development practitioners should keep in mind. At the same time, it is important to emphasize that this discussion inevitably paints with a broad stroke, drawing attention to some of the common challenges. However, the exact circumstances in each country, city, and community are the result of their unique combination of people, culture, history, climate, geography, etc. Households in lower income countries naturally also have smaller budgets available to address their mobility needs. This has important consequences for vehicle ownership and use patterns as well as policy making targeting these behaviors. For vehicles to be affordable, the automobile markets in many low- and middle-income countries rely on importing used vehicles from high- income countries. In particular, the vehicle fleet in Sub-Saharan Africa is largely made up of used vehicles imported from Europe, Japan, and the US (Baskin et al., 2020). At the same time, vehicle owners are more willing to contend with ageing vehicles, slowing vehicle replacement rates. Thus, the average age of vehicles in developing countries tends to be considerably higher than in developed countries. Furthermore, two- and three-wheelers are extremely popular in many low- and middle-income countries due to their relative affordability. However, these vehicles tend to be involved in road injuries and deaths at a much higher rate and emit more particulate matter than cars. The combination of older cars and widespread two- and three- wheeler use have contributed to unhealthy air quality in many cities across the developing world. For policy makers, the relatively small mobility budgets can make households more responsive to price instruments (West, 2005). Due the disparity in purchasing power, monetary budgets are more often a binding constraint for travel decisions in developing countries, whereas time considerations tend to weight more heavily for such decisions in developed economies. Consequently, policy makers might observe larger response elasticities than those found in Section 4, which were largely based on empirical evidence from high-income countries. Importantly, this applies to both intended and unintended policy outcomes. That is, while well- designed price instruments may well be more effective, poorly designed ones will accentuate the incentive misalignments. For example, a graduated purchase tax based on fuel economy will likely exacerbate the Gruenspecht and Rebound effects due to the higher price-sensitivity. Specifically, if mobility budgets are a binding constraint, the Gruenspecht effect will be more pronounced since making vehicle purchase more expensive prevents, or at least further delays, 27 replacement of old vehicles. Similarly for the Rebound effect, reducing the marginal cost of driving through fuel economy improvements leads to more induced demand if monetary rather than time constraints were binding. The volatility of price instruments in low- and middle- income countries underscores the importance of carefully considering the incentives and implications of policy designs in such contexts. There is a remarkably robust correlation between motorization rates and GDP per capita. In other words, private vehicles are generally less common in lower income countries and often act more as luxury goods. Naturally, this has implications for the distributional impacts of tax policies. Broadly, taxation of private vehicles and their use is more progressive in low-income contexts, as discussed in Section 3.3. However, fuel taxation, in particular, can be a precarious instrument. Low-income households that do not own a private vehicle may still be reliant on motorized transport, and by extension fuel, for their mobility needs, e.g., paratransit and rideshares. Raising fuel taxes increases the costs of these services. Furthermore, petroleum products, in their various forms, are used for a variety of industrial and residential purposes, such as manufacturing and cooking. Hence, indiscriminately increasing taxes on these products may inadvertently raise prices on a broad range of consumer goods. Although high-income households will undoubtedly be affected more in absolute terms, the progressivity of fuel taxes is much less clear relative to income. That said, the collection of taxes only makes up half the distributional equation. Ultimately, equity impacts will depend as much on how the government redistributes the collected tax revenue. Recycling the revenue to benefit low-income household can help ensure progressive outcomes. However, in developing countries, the prevalence of informality and incomplete data can make policy targeting more challenging and costly, consequently, diminishing the efficiency of redistribution (Coady et al., 2004). Although taxation of vehicles and fuels is arguably far from optimal even in mature auto markets, these distortions are far more pronounced in emerging markets. For example, commonplace in Sub-Saharan Africa, private vehicle costs are very heavily front-loaded (Granger et al., 2021). Import duties, vehicle excise taxes, and registration fees add up to exceed the value of the vehicles themselves in many places. Meanwhile, ownership taxes and registration fees are rare, and the cost of vehicle use is often undermined by fuel subsidies. Such distortions encourage import of cheaper vehicles to reduce the upfront taxes. This leads to import 28 of older vehicles, which in turn are more polluting. Once the vehicles have been acquired, their use is inadequately priced. This leads to excessive consumption, i.e., distances driven beyond the social optimum. To combat ageing and polluting vehicle fleets, many countries, particularly in Latin America, have banned the import of used vehicles (Baskin et al., 2020). By itself, such regulation is a crude instrument. Absent complementary policies, it can create an extreme case of the Gruenspecht effect, whereby large segments of the population continue to drive dilapidated vehicles because replacements have become unaffordable. As consequence of poorly designed tax structures, the transport sector in many developing countries imposes undue external costs on society at large through GHG emissions, air pollution, productivity lost in congestion, and injuries and deaths in crashes. Several factors contribute to these inefficient tax structures, including political acceptability and institutional capacity for enforcement. Fuel taxation has a much more volatile impact on fiscal budgets in low- and middle-income countries making them even more politically contentious. Fuel prices are to a large extent determined by the price of crude oil in the global marketplace. Consequently, the same nominal fuel price change, e.g., from market fluctuations, has an outsized real impact in developing economies due to the disparity in purchasing power. To counteract price spikes, many developing country governments institute – what are intended to be – temporary subsidies to absorb the shocks. However, these subsidies can wreak havoc on fiscal budgets. Globally, energy subsidies accounted for 0.7% of GDP in 2013 (Inchauste & Victor, 2017). However, the fiscal impact can be far more significant in developing countries. For example, federal fuel taxes in the US raised $38 billion dollars in 2019 (U.S. Federal Highway Administration, 2020). This made up a not insignificant 1.1% of the total federal government revenue that year (U.S. Congressional Budget Office, 2023). By contrast, in 2022 Senegal’s fuel subsidies amounted to 15.5% of government revenue, leaving a considerable fiscal gap (Konokhova, 2023; World Bank, 2023). For comparison, Senegal’s total gasoline consumption is approximately 0.07% that of the United States’. Unfortunately, energy subsidies are also notoriously difficult to eliminate from a political perspective. Often, reforms to reduce or eliminate them lead to protests, which in fragile contexts can easily contribute to political instability and the toppling governments, e.g., the 2008 Mauritanian coup d’état (Granger et al., 2021), or the 2010 Kyrgyz revolution (McCulloch, 2023). 29 Institutional capacity for enforcement of tax polities tends to be weaker in developing contexts. Collection of distance-based taxes, congestion charges, and annual registration fees requires considerable enforcement capacity and costly technology. Both of which tend to be in short supply in developing countries. Consequently, import taxes on vehicles and fuel, which are relatively straightforward to enforce, are more often than not the instruments of choice. Especially in countries with large informal employment sectors, these upstream taxes on consumption play an outsized role, since income taxes are not collected as consistently as in developed economies. However, given the frequent politicization of fuel taxes, governments are left often with a very small set of feasible instruments in their toolbox. This in turn leads to overloading a single instrument, namely the upfront import taxes on vehicles. Not only does this create distortions, it also encourages evasion of that particular tax instrument. For example, exorbitant vehicle import taxes in Ethiopia have spurred several illegal practices, such as falsification of imported vehicle documents and smuggling from neighboring countries (Metekia, 2022). Keeping in mind the correlation between tax rates, incentive to evade, and enforcement costs is essential to effective policy making in practice. Offering alternatives is crucial to managing demand for private vehicle ownership as incomes grow, but these alternatives remain poor in most cases (Santos et al., 2010b). The aforementioned robust correlation between per capita GDP and motorization also implies that, as incomes rise, many developing countries will experience a rapid growth in motorization rates. This underlines the challenge of managing private vehicle demand in developing contexts. Changing this trend requires high-quality alternatives to private vehicles. However, public transportation systems are in many cases underdeveloped, e.g., with poor coverage and comprising informal operators. They are often overpriced, overcrowded, and unsafe – particularly for women. Dedicated infrastructure for non-motorized transportation is also often lacking. Without sidewalks, pedestrians may be relegated to simply walk along the side of the road, making trips unsafe and stressful. Cyclists can be a rare sight with mixed traffic and poor road conditions often deemed too chaotic and dangerous. The spatial development patterns can also contribute to poor accessibility for households without a private vehicle, which are disproportionately low-income. The prevalence of informal settlements in outlying areas with inadequate infrastructure join and exacerbate all these issues. Such settlements tend to be inefficiently served by mass transit. At the same time, access to employment opportunities, 30 especially the formal job market, may be severely hindered without a use of motorized transportation. Absent a well-functioning public transportation system and built environment conducive to non-motorized travel, those who do not own a private vehicle will be severely disadvantaged. In turn, this reinforces the need to acquire a means of private motorized transport as soon as it becomes financially possible. Although not yet widespread, EVs present important advantages specific to developing countries. Considering the state of the transportation sector in most developing countries, completely decoupling motorization from economic growth is not realistic – at least not in the foreseeable future. Rather, private vehicle ownership remains aspirational, particularly in developing contexts (Moody, 2019). In light of this, EVs offer a potential second-best solution. Specifically, EVs present an opportunity to leapfrog ICEV technology. Insofar that motorization rates are low currently, developing countries can avoid entrenchment in traditional fuel infrastructure and avoid the slow and costly process of transitioning. Substituting ICEVs for EVs can also help alleviate the fiscal impact of fuel subsidies by reducing demand for traditional fuels. Furthermore, in the developing context, EVs, including electric two- and three-wheelers, have the advantage that they contribute much less to local air pollution – a problem severely afflicting many cities in low- and middle-income countries (Briceno-Garmendia et al., 2022). Undoubtedly, leapfrogging will require proactive policy making and investments from governments. For example, Egypt and Bhutan have both banned the import of used ICEVs but allow import of used EVs. Even more stringently, several Chinese cities promote EV adoption by capping the number of ICEVs through license plate auctions and lotteries but making EVs exempt. Such aggressive instruments do naturally require corresponding enforcement capacity to implement successfully. 6. Conclusion 6.1. Summary of Lessons Learned The literature teaches several important lessons. In policy design, the Pigouvian principle, i.e., taxing the externality-creating activity to internalize social costs, serves as a useful guide for restoring economic efficiency. This is an argument for taxing vehicle use, e.g., through distance- based taxes, fuel taxes, and congestion charges. On the other hand, price instruments on vehicle purchase can have unintended negative consequences due to a misalignment of incentives, e.g., 31 the Rebound and Gruenspecht effects. Despite this relatively simple intuition, it is clear that there is no one-size-fits-all solution in practice. Distortions in the design of taxations schemes exist for numerous reasons, be it: lacking technical expertise, equity considerations, political concerns, inadequate enforcement capacity, etc. Understanding the source of the distortions is imperative to successful diagnosis and policy making. Unfortunately, the root cause is rarely just a lack of technical expertise, and other sources of distortions generally call for second-best thinking. This is particularly relevant in developing countries, where greater volatility and rapid growth makes understanding the local challenges all the more important. In general, households in low- and middle-income countries have smaller budgets to address mobility needs. Consequently, private vehicle ownership remains aspirational for many, meaning that vehicles tend to act as luxury goods and vehicle taxation is broadly more progressive than in mature markets. Additionally, relatively more affordable two- and three-wheelers and older, more polluting, vehicles are ubiquitous, making local air pollution and vehicle replacement higher priorities. At the same time, existing tax structures tend to be more distorted, often due to limited technology availability and weaker institutional capacity for enforcement. For example, in many countries in Sub-Saharan Africa, taxation is heavily front-loaded with exorbitant taxes on purchase but inadequate and often subsidized fuel taxation. At the same time, raising fuel taxes is often not an option due to political constraints and the technology needed to implement more sophisticated use taxes might not be available. Under such conditions, partly shifting the upfront taxes to ownership taxes is likely an advisable second-best approach, despite not being optimal absent other distortions. Importantly, this enables tax scheme designs that incentivize the retirement or replacement of dilapidated vehicles, which contribute disproportionately to the severe air pollution in many developing country cities. Alternatively, subsidizing private vehicle substitutes, e.g., public transportation or EVs, beyond first-best levels can also be a viable strategy for reducing externality costs in light of existing distortions. 6.2. Areas for Future Research Previous studies had already noted the absence of academic work rooted in low- and middle- income countries. Apart from literature from the Chinese context, which in many ways has its own unique characteristics, this gap has largely persisted. Not only is there a dearth of empirical evidence, but there is also rather limited thinking about alternative solutions that account for the 32 pre-existing distortions specific to developing countries. Instead, treatments are largely constricted to the same set of tools that have been used in mature markets. Additionally, much of the relevant existing literature originates from the energy and environmental economics fields rather than transportation. While these fields are more well- versed in tax policy, they generally consider fuel consumption and vehicle use in isolation. In other words, private vehicle use is implicitly seen as an end in itself rather than a derived demand. Consequently, potential substitution patterns and spatial considerations are generally not considered. On the other hand, while the transport literature tends to be better about recognizing these facets, it often overlooks the finer nuances and incentive structures of different taxation schemes. Undoubtedly, both emphases are important – and should be thought about in unison. Finally, vehicles represent long-term financial commitments by households. Hence, the full effect of tax policies will not necessarily be evident immediately but gradually take form as people replace old vehicles and purchase new ones. Some studies distinguish between the short run and long run outcomes of tax policies. While this goes part of the way to examine the temporal dimension, it still implicitly considers steady state conditions, which may not exist in practice, and abstracts away the gradual transition between them. Even so, what is deemed short run versus long run often seems ad hoc. A more systematic treatment and understanding of the dynamics of the responses to vehicle and fuel taxation is important – especially given the rapidly changing landscape of mobility technology. 33 References Alberini, A., & Bareit, M. (2019). The effect of registration taxes on new car sales and emissions: Evidence from Switzerland. Resource and Energy Economics, 56, 96–112. https://doi.org/10.1016/j.reseneeco.2017.03.005 Alberini, A., & Horvath, M. (2021). All car taxes are not created equal: Evidence from Germany. Energy Economics, 100, 105329. https://doi.org/10.1016/j.eneco.2021.105329 Allcott, H., & Wozny, N. (2014). Gasoline Prices, Fuel Economy, and the Energy Paradox. The Review of Economics and Statistics, 96(5), 779–795. Anderson, M. L., & Auffhammer, M. (2014). Pounds That Kill: The External Costs of Vehicle Weight. The Review of Economic Studies, 81(2 (287)), 535–571. Anderson, S. T., & Sallee, J. M. (2011). Using Loopholes to Reveal the Marginal Cost of Regulation: The Case of Fuel-Economy Standards. American Economic Review, 101(4), 1375–1409. https://doi.org/10.1257/aer.101.4.1375 Anderson, S. T., & Sallee, J. M. (2016). Designing Policies to Make Cars Greener: A Review of the Literature. NBER Working Paper Series. https://doi.org/10.3386/w22242 Austin, D., & Dinan, T. (2005). Clearing the air: The costs and consequences of higher CAFE standards and increased gasoline taxes. Journal of Environmental Economics and Management, 50(3), 562–582. https://doi.org/10.1016/j.jeem.2005.05.001 Bansal, P., & Dua, R. (2022). Fuel consumption elasticities, rebound effect and feebate effectiveness in the Indian and Chinese new car markets. Energy Economics, 113, 106192. https://doi.org/10.1016/j.eneco.2022.106192 Barros, C. P., & Prieto-Rodriguez, J. (2008). A revenue-neutral tax reform to increase demand for public transport services. Transportation Research Part A: Policy and Practice, 42(4), 659–672. https://doi.org/10.1016/j.tra.2008.01.027 Baskin, A., de Jong, R., Dumitrescu, E., Akumu, J., Stannah, V. R., Mwangi, A., Diabate, F., Quirama, L. F., & Maina, G. (2020). A Global Overview of Used Light Duty Vehicles: Flow, Scale and Regulation. United Nation Environment Programme. 34 Beaudoin, J., Farzin, Y. H., & Lin Lawell, C.-Y. C. (2015). Public transit investment and sustainable transportation: A review of studies of transit’s impact on traffic congestion and air quality. Research in Transportation Economics, 52, 15–22. https://doi.org/10.1016/j.retrec.2015.10.004 Bento, A. M., Goulder, L. H., Jacobsen, M. R., & Von Haefen, R. H. (2009). Distributional and Efficiency Impacts of Increased US Gasoline Taxes. American Economic Review, 99(3), 667–699. https://doi.org/10.1257/aer.99.3.667 Berri, A., Vincent Lyk-Jensen, S., Mulalic, I., & Zachariadis, T. (2014). Household transport consumption inequalities and redistributive effects of taxes: A repeated cross-sectional evaluation for France, Denmark and Cyprus. Transport Policy, 36, 206–216. https://doi.org/10.1016/j.tranpol.2014.08.007 Bickel, P., Friedrich, R., Link, H., Stewart, L., & Nash, C. (2006). Introducing Environmental Externalities into Transport Pricing: Measurement and Implications. Transport Reviews, 26(4), 389–415. https://doi.org/10.1080/01441640600602039 Bonilla, D., & Foxon, T. (2009). Demand for New Car Fuel Economy in the UK, 1970–2005. Journal of Transport Economics and Policy, 43. Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Introduction to Meta‐Analysis. Wiley. https://doi.org/10.1002/9780470743386 Briceno-Garmendia, C., Qiao, W., & Foster, V. (2022). The Economics of Electric Vehicles for Passenger Transportation. World Bank. Brons, M., Nijkamp, P., Pels, E., & Rietveld, P. (2006). A Meta-analysis of the Price Elasticity of Gasoline Demand. A System of Equations Approach. Busse, M. R., Knittel, C. R., & Zettelmeyer, F. (2013). Are Consumers Myopic? Evidence from New and Used Car Purchases. American Economic Review, 103(1), 220–256. https://doi.org/10.1257/aer.103.1.220 Callejas, J., Linn, J., & Steinbuks, J. (2022). Welfare and Environmental Benefits of Electric Vehicle Tax Policies in Developing Countries: Evidence from Colombia. World Bank Policy Research Working Paper. https://doi.org/10.1596/1813-9450-10001 35 Cerruti, D., Alberini, A., & Linn, J. (2019). Charging Drivers by the Pound: How Does the UK Vehicle Tax System Affect CO2 Emissions? Environmental and Resource Economics, 74(1), 99–129. https://doi.org/10.1007/s10640-018-00310-x Chaudhury, P. D. (2006). Competition between Intercity Rail and Road Transport in India: External and Social Costs. Journal of the Asia Pacific Economy, 11(3), 308–330. https://doi.org/10.1080/13547860600764294 Chetty, R., Looney, A., & Kroft, K. (2009). Salience and Taxation: Theory and Evidence. American Economic Review, 99(4), 1145–1177. https://doi.org/10.1257/aer.99.4.1145 Ciccone, A., & Soldani, E. (2021). Stick or Carrot? Asymmetric Responses to Vehicle Registration Taxes in Norway. Environmental and Resource Economics, 80(1), 59–94. https://doi.org/10.1007/s10640-021-00578-6 Coady, D., Grosh, M., & Hoddinott, J. (2004). Targeting of Transfers in Developing Countries: Review of Lessons and Experience. The World Bank. https://doi.org/10.1596/0-8213-5769-7 Coglianese, Davis, Kilian, & Stock. (2015). Anticipation, Tax Avoidance, and the Price Elasticity of Gasoline Demand. https://www.nber.org/system/files/working_papers/w20980/w20980.pdf Corts, K. S. (2010). Building out alternative fuel retail infrastructure: Government fleet spillovers in E85. Journal of Environmental Economics and Management, 59(3), 219–234. https://doi.org/10.1016/j.jeem.2009.09.001 Dahl, & Sterner. (1991). A Survey of Econometric Gasoline Demand Elasticities. Danesin, A., & Linares, P. (2015). An Estimation of Fuel Demand Elasticities for Spain. Journal of Transport Economics and Policy, 49. Danesin, A., & Linares, P. (2018). The relevance of the local context for assessing the welfare effect of transport decarbonization policies. A study for 5 Spanish metropolitan areas. Energy Policy, 118, 41–57. https://doi.org/10.1016/j.enpol.2017.12.019 36 Danielis, R., & Chiabai, A. (1998). Estimating the cost of air pollution from road transport in Italy. Transportation Research Part D: Transport and Environment, 3(4), 249–258. https://doi.org/10.1016/S1361-9209(98)00004-2 Davis, L. W., & Kilian, L. (2011). Estimating the effect of a gasoline tax on carbon emissions. Journal of Applied Econometrics, 26(7), 1187–1214. https://doi.org/10.1002/jae.1156 Davis, L. W., & Knittel, C. R. (2019). Are Fuel Economy Standards Regressive? Journal of the Association of Environmental and Resource Economists, 6(S1), S37–S63. https://doi.org/10.1086/701187 Davis, L. W., & Sallee, J. M. (2020). Should Electric Vehicle Drivers Pay a Mileage Tax? Environmental and Energy Policy and the Economy, 1, 65–94. https://doi.org/10.1086/706793 Delucchi, M. A., & McCubbin, D. R. (2010). External Costs of Transport in the U.S. Retrieved October 2, 2023, from https://escholarship.org/content/qt13n8v8gq/qt13n8v8gq.pdf Dunkerley, F., Rohr, C., & Daly, A. (2014). Road traffic demand elasticities: A rapid evidence assessment. RAND Corporation. https://doi.org/10.7249/RR888 Durrmeyer, I. (2022). Winners and Losers: The Distributional Effects of the French Feebate on the Automobile Market. The Economic Journal, 132(644), 1414–1448. https://doi.org/10.1093/ej/ueab084 Econometrics, C., & Garden, C. (2013). The effectiveness of CO2-based ‘feebate’systems in the European passenger vehicle market context. Eliasson, J., Pyddoke, R., & Swärdh, J.-E. (2018). Distributional effects of taxes on car fuel, use, ownership and purchases. Economics of Transportation, 15, 1–15. https://doi.org/10.1016/j.ecotra.2018.03.001 Eltony. (1993). Transport gasoline demand in Canada. https://jtep.org/wp- content/uploads/2021/02/Volume_XXV11_No_2_193-208.pdf 37 Eskeland, G. S., & Yan, S. (2021). The Norwegian CO2-differentiated motor vehicle registration tax: An extended Cost-Benefit Analysis (OECD Environment Working Papers 178; OECD Environment Working Papers, Vol. 178). https://doi.org/10.1787/ee108c96-en Espey, M. (1998). Gasoline demand revisited: An international meta-analysis of elasticities. Energy Economics, 20(3), 273–295. https://doi.org/10.1016/S0140-9883(97)00013-3 European Commission. Directorate General for Mobility and Transport., CE Delft., INFRAS., Ricardo., & TRT. (2019). Handbook on the external costs of transport: Version 2019. Publications Office. https://data.europa.eu/doi/10.2832/27212 Finkelstein, A. (2009). E-Z Tax: Tax Salience and Tax Rates. The Quarterly Journal of Economics, 124(3), 969–1010. https://doi.org/10.1162/qjec.2009.124.3.969 Frondel, M., Flores, F. M., & Vance, C. (2017). Heterogeneous Rebound Effects in Individual Mobility: Evidence from German Households. Journal of Transport Economics and Policy, 51. Gillingham, K. (2013). The Economics of Fuel Economy Standards versus Feebates. NEPI Working Paper. Gillingham, K. (2014). Identifying the elasticity of driving: Evidence from a gasoline price shock in California. Regional Science and Urban Economics, 47, 13–24. https://doi.org/10.1016/j.regsciurbeco.2013.08.004 Gillingham, K. T., Houde, S., & Van Benthem, A. A. (2021). Consumer Myopia in Vehicle Purchases: Evidence from a Natural Experiment. American Economic Journal: Economic Policy, 13(3), 207–238. https://doi.org/10.1257/pol.20200322 Goetzke, F., & Vance, C. (2019). An Increasing Gasoline Price Elasticity in the United States? SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3498379 Goodwin. (1992). A Review of New Demand Elasticities with Special Reference to Short and Long Run Effects of Price Changes. https://jtep.org/wp- content/uploads/2021/02/Volume_XXV1_No_2_155-169.pdf 38 Gössling, S., Choi, A., Dekker, K., & Metzler, D. (2019). The Social Cost of Automobility, Cycling and Walking in the European Union. Ecological Economics, 158, 65–74. https://doi.org/10.1016/j.ecolecon.2018.12.016 Graham, D. J., & Glaister, S. (2002). The Demand for Automobile Fuel: A Survey of Elasticities. Journal of Transport Economics and Policy, 36(1), 1–25. Granger, H., Nair, V., Parekh, H., Phillips, D., Prinz, D., Seid, E., & Warwick, R. (2021). “Green” motor taxation: Issues and policy options in sub-Saharan Africa. IFS. https://doi.org/10.1920/re.ifs.2021.0197 Greene, D. L. (2010). How consumers value fuel economy: A literature review. United States Environmental Protection Agency. Greene, D. L., & Welch, J. G. (2018). Impacts of fuel economy improvements on the distribution of income in the U.S. Energy Policy, 122, 528–541. https://doi.org/10.1016/j.enpol.2018.08.005 Grigolon, L., Reynaert, M., & Verboven, F. (2018). Consumer Valuation of Fuel Costs and Tax Policy: Evidence from the European Car Market. American Economic Journal: Economic Policy, 10(3), 193–225. https://doi.org/10.1257/pol.20160078 Gruenspecht, H. K. (1982). Differentiated Regulation: The Case of Auto Emissions Standards. The American Economic Review, 72(2), 328–331. Hanly, M., Dargay, J., & Goodwin, P. (2002). Review of Income and Price Elasticities in the Demand for Road Traffic (Contract number PPAD 9/65/93). 2002. Harju, J., Kosonen, T., & Slemrod, J. (2020). Missing miles: Evasion responses to car taxes. Journal of Public Economics, 181, 104108. https://doi.org/10.1016/j.jpubeco.2019.104108 Heine, D., Norregaard, J., & Parry, I. W. H. (2012). Environmental Tax Reform: Principles from Theory and Practice to Date. IMF Working Paper. Hirota, K., & Poot, J. (2003). Do policy incentives affect the environmental impact of private car use? Evidence from a sample of large cities. 39 Hirte, G., & Tscharaktschiew, S. (2013). The optimal subsidy on electric vehicles in German metropolitan areas: A spatial general equilibrium analysis. Energy Economics, 40, 515–528. https://doi.org/10.1016/j.eneco.2013.08.001 Hughes, Knittel, & Sperling. (2006). Evidence of a Shift in the Short-run Price Elasticity of Gasoline Demand. https://www.nber.org/system/files/working_papers/w12530/w12530.pdf Hymel, K. M., Small, K. A., & Dender, K. V. (2010). Induced demand and rebound effects in road transport. Transportation Research Part B: Methodological, 44(10), 1220–1241. https://doi.org/10.1016/j.trb.2010.02.007 IEA. (2017, November 5). Energy Improvement and Extension Act 2008—Tax Incentives. IEA. https://www.iea.org/policies/1817-energy-improvement-and-extension-act-2008-tax- incentives Inchauste, G., & Victor, D. G. (2017). The Political Economy of Energy Subsidy Reform. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1007-7 Jacobsen, M. R. (2013). Evaluating US Fuel Economy Standards in a Model with Producer and Household Heterogeneity. American Economic Journal: Economic Policy, 5(2), 148–187. https://doi.org/10.1257/pol.5.2.148 Jacobsen, M. R., Sallee, J. M., Shapiro, J. S., & van Benthem, A. A. (2022). Regulating untaxable externalities: Are vehicle air pollution standards effective and efficient? NBER Working Paper Series, 121. https://doi.org/10.3386/w30702 Jacobsen, M. R., & van Benthem, A. A. (2015). Vehicle Scrappage and Gasoline Policy. American Economic Review, 105(3), 1312–1338. https://doi.org/10.1257/aer.20130935 Jaros, B., & Hoffer, A. (2023, September 20). How Are Electric Vehicles Taxed in Your State? Tax Foundation. https://taxfoundation.org/data/all/state/electric-vehicles-ev-taxes-state/ Jenn, A., Azevedo, I. L., & Fischbeck, P. (2015). How will we fund our roads? A case of decreasing revenue from electric vehicles. Transportation Research Part A: Policy and Practice, 74, 136–147. https://doi.org/10.1016/j.tra.2015.02.004 40 Jenn, A., Springel, K., & Gopal, A. R. (2018). Effectiveness of electric vehicle incentives in the United States. Energy Policy, 119, 349–356. https://doi.org/10.1016/j.enpol.2018.04.065 Jochem, P., Doll, C., & Fichtner, W. (2016). External costs of electric vehicles. Transportation Research Part D: Transport and Environment, 42, 60–76. https://doi.org/10.1016/j.trd.2015.09.022 Johansson, O., & Schipper, L. (1997). Measuring the Long-Run Fuel Demand of Cars: Separate Estimations of Vehicle Stock, Mean Fuel Intensity, and Mean Annual Driving Distance. Journal of Transport Economics and Policy, 31(3), 277–292. Koyama, S. (2004) Nippon ni okeru jidosha kotsu souon no gaibu hiyo no suikei [Estimation of external costs of noise by vehicular transport in Japan], Shodai Ronshu, 3(4), pp. 121–157 [in Japanese]. Koyama, S. & Kishimoto, M. (2001) Nippon ni okeru jidosha kotsu no gaibu hiyo no gaisan [Approximate figure of external costs of vehicular transport in Japan], Unyu Seisaku Kenkyu [Transportation Policy Research], 14(2), pp. 19–30 [in Japanese]. Klier, T., & Linn, J. (2010). The Price of Gasoline and New Vehicle Fuel Economy: Evidence from Monthly Sales Data. American Economic Journal: Economic Policy, 2(3), 134–153. https://doi.org/10.1257/pol.2.3.134 Klier, T., & Linn, J. (2015). Using Taxes to Reduce Carbon Dioxide Emissions Rates of New Passenger Vehicles: Evidence from France, Germany, and Sweden. American Economic Journal: Economic Policy, 7(1), 212–242. https://doi.org/10.1257/pol.20120256 Klier, T., Linn, J., & Zhou, Y. C. (2020). The effects of fuel prices and vehicle sales on fuel‐ saving technology adoption in passenger vehicles. Journal of Economics & Management Strategy, 29(3), 543–578. https://doi.org/10.1111/jems.12384 Knight, F. H. (1924). Some Fallacies in the Interpretation of Social Cost. The Quarterly Journal of Economics, 38(4), 582–606. https://doi.org/10.2307/1884592 Knittel, C., & Tanaka, S. (2019). Driving Behavior and the Price of Gasoline: Evidence from Fueling-Level Micro Data (w26488; p. w26488). National Bureau of Economic Research. https://doi.org/10.3386/w26488 41 Konokhova, M. (2023). Senegal Raises Fuel Prices to Alleviate Budget Strains. Sputnik International. https://sputniknews.com/20230109/senegal-raises-fuel-prices-to-alleviate- budget-strains-1106159086.html Lane, B. W. (2010). The relationship between recent gasoline price fluctuations and transit ridership in major US cities. Journal of Transport Geography, 18(2), 214–225. https://doi.org/10.1016/j.jtrangeo.2009.04.002 Langer, A., Maheshri, V., & Winston, C. (2017). From gallons to miles: A disaggregate analysis of automobile travel and externality taxes. Journal of Public Economics, 152, 34–46. https://doi.org/10.1016/j.jpubeco.2017.05.003 Lawley, C., & Thivierge, V. (2016). Refining the Evidence: British Columbia’s Carbon Tax and Household Gasoline Consumption. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2874417 Leard, B., Linn, J., & Zhou, Y. C. (2023). How Much Do Consumers Value Fuel Economy and Performance? Evidence from Technology Adoption. The Review of Economics and Statistics, 105(1), 158–174. https://doi.org/10.1162/rest_a_01045 Li, Q., & Goh, B. (2023, September 8). China car sales return to growth in August, Tesla nearly doubles EV share. Reuters. https://www.reuters.com/business/autos-transportation/china-car- sales-return-growth-aug-tesla-nearly-doubles-ev-share-2023-09-08/ Li, S., Linn, J., & Muehlegger, E. (2012). Gasoline Taxes and Consumer Behavior. NBER Working Paper Series. Li, S., Tong, L., Xing, J., & Zhou, Y. (2017). The Market for Electric Vehicles: Indirect Network Effects and Policy Design. Journal of the Association of Environmental and Resource Economists, 4(1), 89–133. https://doi.org/10.1086/689702 Li, S., Wang, B., Yang, M., & Zhang, F. (2021). The Global Diffusion of Electric Vehicles: Lessons from the First Decade. World Bank Policy Research Working Paper. https://doi.org/10.1596/1813-9450-9882 42 Li, S., Zhu, X., Ma, Y., Zhang, F., & Zhou, H. (2022). The Role of Government in the Market for Electric Vehicles: Evidence from China. Journal of Policy Analysis and Management, 41(2), 450–485. https://doi.org/10.1002/pam.22362 Lindsey, R., & Santos, G. (2020). Addressing transportation and environmental externalities with economics: Are policy makers listening? Research in Transportation Economics, 82, 100872. https://doi.org/10.1016/j.retrec.2020.100872 Lipow. (2008). Price-Elasticity of Energy Demand A Bibliography. Littell, J. H., Corcoran, J., & Pillai, V. (2008). Systematic Reviews and Meta-Analysis. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195326543.001.0001 Ma, S.-C., Xu, J.-H., & Fan, Y. (2019). Willingness to pay and preferences for alternative incentives to EV purchase subsidies: An empirical study in China. Energy Economics, 81, 197–215. https://doi.org/10.1016/j.eneco.2019.03.012 Mattson, J. (2008). The Effects of Gasoline Prices on Bus Ridership for Different Types of Transit Systems. Journal of the Transportation Research Forum, 47(3). https://doi.org/10.5399/osu/jtrf.47.3.2101 Mayeres, I., Ochelen, S., & Proost, S. (1996). The marginal external costs of urban transport. Transportation Research Part D: Transport and Environment, 1(2), 111–130. https://doi.org/10.1016/S1361-9209(96)00006-5 McCulloch, N. (2023). Ending Fossil Fuel Subsidies: The politics of saving the planet. Practical Action Publishing. https://doi.org/10.3362/9781788532044 Metekia, T. S. (2022, July 25). Costly taxes drive Ethiopia’s illicit vehicle trade. ISS Africa. https://issafrica.org/iss-today/costly-taxes-drive-ethiopias-illicit-vehicle-trade Moody, J. C. (2019). Measuring Car Pride and its Implications for Car Ownership and Use across Individuals, Cities, and Countries. MIT. Motavalli, J. (2010, June 2). China to Subsidize Electric Cars and Hybrids. The New York Times. https://archive.nytimes.com/wheels.blogs.nytimes.com/2010/06/02/china-to-start-pilot- program-providing-subsidies-for-electric-cars-and-hybrids/ 43 Nash, C. (2003). UNIfication of accounts and marginal costs for Transport Efficiency. 2003. Odeck, J., & Johansen, K. (2016). Elasticities of fuel and traffic demand and the direct rebound effects: An econometric estimation in the case of Norway. Transportation Research Part A: Policy and Practice, 83, 1–13. https://doi.org/10.1016/j.tra.2015.10.003 OECD. (2001). Environmentally Related Taxes in OECD Countries: Issues and Strategies. OECD. https://doi.org/10.1787/9789264193659-en Parry, I. W. H. (2014). Designing Fiscal Policy to Address the External Costs of Energy. CESIFO Working Paper. Parry, I. W. H., & Small, K. A. (2005). Does Britain or the United States Have the Right Gasoline Tax? The American Economic Review, 95(4), 1276–1289. Parry, I. W. H., Walls, M., & Harrington, W. (2007). Automobile Externalities and Policies. Parry, I. W. H., & Williams, R. C. (2010). What are the costs of meeting distributional objectives for climate policy? NBER Working Paper Series. Peirson, J., Skinner, I., & Vickerman, R. (1995). Estimating the External Costs of UK Passenger Transport: The First Step towards an Efficient Transport Market. Environment and Planning A, 27, 1977–1993. https://doi.org/10.1068/a271977 Pigou, A. C. (1920). The Economics of Welfare. Palgrave Macmillan London. https://link.springer.com/book/9780230249318 Proque, A. L., Betarelli Junior, A. A., & Perobelli, F. S. (2022). Fuel tax, cross subsidy and transport: Assessing the effects on income and consumption distribution in Brazil. Research in Transportation Economics, 95, 101204. https://doi.org/10.1016/j.retrec.2022.101204 Reynaert, M. (2021). Abatement Strategies and the Cost of Environmental Regulation: Emission Standards on the European Car Market. The Review of Economic Studies, 88(1), 454–488. https://doi.org/10.1093/restud/rdaa058 Rizzi, L. I., & De La Maza, C. (2017). The external costs of private versus public road transport in the Metropolitan Area of Santiago, Chile. Transportation Research Part A: Policy and Practice, 98, 123–140. https://doi.org/10.1016/j.tra.2017.02.002 44 Sallee, J. M. (2010). The Taxation of Fuel Economy. NBER Working Paper Series. Sallee, J. M. (2019). Pigou Creates Losers: On the Implausibility of Achieving Pareto Improvements from Efficiency-Enhancing Policies. NBER Working Paper Series. https://doi.org/10.3386/w25831 Sallee, J. M., West, S. E., & Fan, W. (2016). Do consumers recognize the value of fuel economy? Evidence from used car prices and gasoline price fluctuations. Journal of Public Economics, 135, 61–73. https://doi.org/10.1016/j.jpubeco.2016.01.003 Sandmo, A. (1975). Optimal Taxation in the Presence of Externalities. The Swedish Journal of Economics, 77(1), 86–98. https://doi.org/10.2307/3439329 Santos, G., Behrendt, H., Maconi, L., Shirvani, T., & Teytelboym, A. (2010a). Part I: Externalities and economic policies in road transport. Research in Transportation Economics, 28(1), 2–45. https://doi.org/10.1016/j.retrec.2009.11.002 Santos, G., Behrendt, H., & Teytelboym, A. (2010b). Part II: Policy instruments for sustainable road transport. Research in Transportation Economics, 28(1), 46–91. https://doi.org/10.1016/j.retrec.2010.03.002 Schmidt, F. L., & Hunter, J. E. (2015). Methods of Meta-Analysis: Correcting Error and Bias in Research Findings. SAGE Publications, Ltd. https://doi.org/10.4135/9781483398105 Sen, A. K., Tiwari, G., & Upadhyay, V. (2010). Estimating marginal external costs of transport in Delhi. Transport Policy, 17(1), 27–37. https://doi.org/10.1016/j.tranpol.2009.09.003 Sheng, M., & Sharp, B. (2019). Aggregate road passenger travel demand in New Zealand: A seemingly unrelated regression approach. Transportation Research Part A: Policy and Practice, 124, 55–68. https://doi.org/10.1016/j.tra.2019.03.005 Small, K. A., & van Dender, K. (2007). Fuel Efficiency and Motor Vehicle Travel: The Declining Rebound Effect. The Energy Journal, 28(1), 25–51. Sovacool, B. K., Kim, J., & Yang, M. (2021). The hidden costs of energy and mobility: A global meta-analysis and research synthesis of electricity and transport externalities. Energy Research & Social Science, 72, 101885. https://doi.org/10.1016/j.erss.2020.101885 45 Sterner, T. (2012). Distributional effects of taxing transport fuel. Energy Policy, 41, 75–83. https://doi.org/10.1016/j.enpol.2010.03.012 Sterner, T., Dahl, C., & Franzén, M. (1992). Gasoline Tax Policy, Carbon Emissions and the Global Environment. Journal of Transport Economics and Policy, 26(2), 109–119. Storchmann, K.-H. (2001). The impact of fuel taxes on public transport—An empirical assessment for Germany. Transport Policy, 8, 19–28. Timilsina, G. R., & Dulal, H. B. (2008). Fiscal Policy Instruments for Reducing Congestion And Atmospheric Emissions In The Transport Sector: A Review. World Bank Policy Research Working Paper. https://doi.org/10.1596/1813-9450-4652 Tjoe, L. N. (2023). COE prices hit new highs; Open category premium soars to $137,000. The Straits Times. https://www.straitstimes.com/singapore/coe-prices-climb-to-new-highs-open- category-premium-surges-to-137000 U.S. Congressional Budget Office. (2023). Revenues. 2023 Historical Budget Data. https://www.cbo.gov/system/files/2023-02/51134-2023-02-Historical-Budget-Data.xlsx U.S. EPA. (2022). The 2022 EPA Automotive Trends Report. https://www.epa.gov/system/files/documents/2022-12/420r22029.pdf U.S. Federal Highway Administration. (2020). Status of The Federal Highway Trust Fund 2019. Highway Statistics Series. https://www.fhwa.dot.gov/policyinformation/statistics/2019/fe10.cfm van Essen, H., Schroten, A., Otten, M., Sutter, D., Schreyer, C., Zandonella, R., Maibach, M., & Doll, C. (2011). External costs of transport in Europe. van Reeven, P. (2011). On the Effectiveness of Pricing Urban Car Usage. Journal of Transport Economics and Policy, 45. Vance, C., & Mehlin, M. (2009). Tax Policy and CO2 Emissions—An Econometric Analysis of the German Automobile Market. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.1374818 46 Victor-Gallardo, L., Zúñiga, M. R., Quirós-Tortós, J., Jaramillo, M., & Vogt-Schilb, A. (2022). Policy options to mitigate the fiscal impact of road transport decarbonization: The case of Costa Rica. IDB Working Paper Series, 27. Wee, S., Coffman, M., & La Croix, S. (2018). Do electric vehicle incentives matter? Evidence from the 50 U.S. states. Research Policy, 47(9), 1601–1610. https://doi.org/10.1016/j.respol.2018.05.003 West, S. E. (2004). Distributional effects of alternative vehicle pollution control policies. Journal of Public Economics, 88(3–4), 735–757. https://doi.org/10.1016/S0047-2727(02)00186-X West, S. E. (2005). Equity Implications of Vehicle Emissions Taxes. Journal of Transport Economics and Policy, 39. World Bank. (2023). Country Climate and Development Reports (CCDRs). World Bank. https://www.worldbank.org/en/publication/country-climate-development-reports World Bank. (2023). Senegal Macro Poverty Outlook. https://thedocs.worldbank.org/en/doc/bae48ff2fefc5a869546775b3f010735- 0500062021/related/mpo-sen.pdf Yan, S., & Eskeland, G. S. (2018). Greening the vehicle fleet: Norway’s CO2-Differentiated registration tax. Journal of Environmental Economics and Management, 91, 247–262. https://doi.org/10.1016/j.jeem.2018.08.018 Zhang, T., & Burke, P. J. (2020). The effect of fuel prices on traffic flows: Evidence from New South Wales. Transportation Research Part A: Policy and Practice, 141, 502–522. https://doi.org/10.1016/j.tra.2020.09.025 47 Appendix A: External Costs of Private Vehicle Use – Methods and Individual Study Estimates Table A.1 presents the full table of external costs of private vehicle use. To calculate the averages presented in the main text, all costs were converted to “dollars per 1000 vehicle kilometers” to ensure standardized and comparable unit. Only those costs that could be converted were included in the averages. For costs presented in non-US currencies, we adopted the currency rates at the time of the respective papers’ publication year, unless the paper explicitly states preferred conversion rates. Table A.1: External costs of private vehicle use – full table Air GHG Vehicle Country Congestion Accident Noise Spatial Temporal Notes Source pollution emissions Fuel 0.014 0.0064 0.0263 0.0231 0.0056 Nash, C. Austria National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) 0.0190 0.0070 0.0100 0.0080 Nash, C. Belgium - National All day Unspecified €/vkm €/vkm €/vkm €/vkm (2003) Belgium 0.036 0.269 0.079 0.002 Mayeres et - Urban Peak Gasoline (Brussels) ECU/vkm1 ECU/vkm1 ECU/vkm1 ECU/vkm1 al. (1996) Belgium 0.056 0.269 0.079 0.002 Mayeres et - Urban Peak Diesel (Brussels) ECU/vkm1 ECU/vkm1 ECU/vkm1 ECU/vkm1 al. (1996) Belgium 0.035 0.002 0.110 0.007 Mayeres et - Urban Off-peak Gasoline (Brussels) ECU/vkm1 ECU/vkm1 ECU/vkm1 ECU/vkm1 al. (1996) Rizzi & De Chile 0.1 44.9 2.0 0.4 - Metropolitan Peak Gasoline La Maza (Santiago) ¢/km ¢/km ¢/km ¢/km (2017) Rizzi & De Chile 0.1 44.9 2.0 0.4 ¢ - Metropolitan Peak Diesel La Maza (Santiago) ¢/km ¢/km ¢/km /km (2017) Rizzi & De Chile 0.1 9.4 2.0 0.4 - Metropolitan Off-peak Gasoline La Maza (Santiago) ¢/km ¢/km ¢/km ¢/km (2017) Rizzi & De Chile 0.1 9.4 2.0 0.4 - Metropolitan Off-peak Diesel La Maza (Santiago) ¢/km ¢/km ¢/km ¢/km (2017) 48 Air GHG Vehicle Country Congestion Accident Noise Spatial Temporal Notes Source pollution emissions Fuel 0.0116 0.0062 0.0095 0.0159 Nash, C. Denmark - National All day Unspecified €/vkm €/vkm €/vkm €/vkm (2003) pkm 0.007 0.011 0.00 0.002 0.007 European indicates Gössling, S. EU All day Unspecified €/pkm €/pkm €/pkm €/pkm €/pkm Average passenger (2019) kilometer. 27.3 van Essen et EU273 - - - - Metropolitan All day Unspecified €/1,000vkm al. (2011) 13.9 5.03 van Essen et EU273 - - - Urban All day Unspecified Other urban €/1,000vkm €ct/vkm al. (2011) 0.33-1.00 Large urban van Essen et EU273 - - - - Urban All day Unspecified €/vkm > 2million al. (2011) Small and 0.11-0.44 medium van Essen et EU273 - - - - Urban All day Unspecified €/vkm urban < al. (2011) 2million 9.0 €/1000 van Essen et EU273 - - - - Urban Peak/Day Unspecified vkm al. (2011) 21.9 €/1000 Off- van Essen et EU273 - - - - Urban Unspecified vkm peak/Day al. (2011) 16.5 €/1000 van Essen et EU273 - - - - Urban Peak/Night Unspecified vkm al. (2011) 39.9 Off- van Essen et EU273 - - - - Urban Unspecified €/1000 vkm peak/Night al. (2011) 1.98 van Essen et EU273 - - - - Suburban All day Unspecified €ct/vkm al. (2011) 7.3 0.00-0.22 van Essen et EU273 - - - Rural All day Unspecified Non-urban €/1,000vkm €/1000 vkm al. (2011) 0.2 van Essen et EU273 - - - - Rural Peak/Day Unspecified €/1000 vkm al. (2011) 49 Air GHG Vehicle Country Congestion Accident Noise Spatial Temporal Notes Source pollution emissions Fuel 0.1 Off- van Essen et EU273 - - - - Rural Unspecified €/1000 vkm peak/Day al. (2011) 0.1 van Essen et EU273 - - - - Rural Peak/Night Unspecified €/1000 vkm al. (2011) 0.4 Off- van Essen et EU273 - - - - Rural Unspecified €/1000 vkm peak/Night al. (2011) 9.6 5.2-30.2 1.45 van Essen et EU273 - - National All day Unspecified €/1,000vkm €/1000 vkm €ct/vkm al. (2011) European 0.14 1.00 41.2 1.41 0.5 EU28 Metropolitan Peak/Day Unspecified Commission euro ¢/pkm euro ¢/pkm euro ¢/pkm euro ¢/pkm euro ¢/pkm (2019) European 0.08 0.86 18.2 0.25 0.04 EU28 Rural Peak/Day Unspecified Commission euro ¢/pkm euro ¢/pkm euro ¢/pkm euro ¢/pkm euro ¢/pkm (2019) 0.0095 0.0048 0.0050 0.0036 Nash, C. Finland - National All day Unspecified €/vkm €/vkm €/vkm €/vkm (2003) 0.0220 0.0050 0.0330 0.0030 0.0080 Nash, C. France National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) 5.36 Jochem et al. Germany - - - - Metropolitan All day ICEV4 €/1000vkm (2016) 0.79 55.1 €/1000v Jochem et al. Germany - - - Urban All day ICEV4 Large urban €/vkm km (2016) 0.28 55.1 €/1000v Medium Jochem et al. Germany - - - Urban All day ICEV4 €/vkm km urban (2016) 21.16 10.5 Jochem et al. Germany - - - Urban Peak/Day ICEV4 €/1000vkm €/1000vkm (2016) 17.63 25.5 Off- Jochem et al. Germany - - - Urban ICEV4 €/1000vkm €/1000vkm peak/Day (2016) 50 Air GHG Vehicle Country Congestion Accident Noise Spatial Temporal Notes Source pollution emissions Fuel 17.63 46.5 Off- Jochem et al. Germany - - - Urban ICEV4 €/1000vkm €/1000vkm peak/Night (2016) 3.89 Jochem et al. Germany - - - - Suburban All day ICEV4 Other urban €/1000vkm (2016) 0.6 Jochem et al. Germany - - - - Suburban Peak/Day ICEV4 €/1000vkm (2016) 1.6 Off- Jochem et al. Germany - - - - Suburban ICEV4 €/1000vkm peak/Day (2016) 3.0 Off- Jochem et al. Germany - - - - Suburban ICEV4 €/1000vkm peak/Night (2016) 3.16 14.11 0.08 21.7 €/1000v Jochem et al. Germany - Rural All day ICEV4 €/1000vkm €/1000vkm €/vkm km (2016) 0.1 Jochem et al. Germany - - - - Rural Peak/Day ICEV4 €/1000vkm (2016) 0.2 Off- Jochem et al. Germany - - - - Rural ICEV4 €/1000vkm peak/Day (2016) 0.5 Off- Jochem et al. Germany - - - - Rural ICEV4 €/1000vkm peak/Night (2016) 7.2 9.8 0.79 55.1 Jochem et al. Germany - Urban All day EV Large urban €/1000vkm €/1000vkm €/vkm €/1000vkm (2016) 7.2 9.8 55.1 €/1000v Medium Jochem et al. Germany 0.28 €/vkm - Urban All day EV €/1000vkm €/1000vkm km urban (2016) 10.5 Jochem et al. Germany - - - - Urban Peak/Day EV €/1000vkm (2016) 23.5 Off- Jochem et al. Germany - - - - Urban EV €/1000vkm peak/Night (2016) 51 Air GHG Vehicle Country Congestion Accident Noise Spatial Temporal Notes Source pollution emissions Fuel 9.9 14.8 0.08 21.7 Jochem et al. Germany - Rural All day EV €/1000vkm €/1000vkm €/vkm €/1000vkm (2016) 0.1 Jochem et al. Germany - - - - Rural Peak/Day EV €/1000vkm (2016) 0.4 Off- Jochem et al. Germany - - - - Rural EV €/1000vkm peak/Night (2016) 0.0134 0.0061 0.0277 0.0232 0.0100 Nash, C. Germany National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) 0.18 0.12 13.22 0.78 0.01 p indicates Bickel et al. Great Britain National Peak Unspecified p/vkm p/vkm p/vkm p/vkm p/vkm pence. (2006) 0.18 0.12 7.01 0.8 0.01 Bickel et al. Great Britain National Off-Peak Unspecified p/vkm p/vkm p/vkm p/vkm p/vkm (2006) 0.57 0.11 53.75 0.01 0.04 Central Bickel et al. Great Britain Urban All day Unspecified p/vkm p/vkm p/vkm p/vkm p/vkm London (2006) 0.42 0.11 20.10 0.01 0.03 Inner Bickel et al. Great Britain Metropolitan All day Unspecified p/vkm p/vkm p/vkm p/vkm p/vkm London (2006) 0.31 0.10 31.09 0.01 0.02 Outer Bickel et al. Great Britain Suburban All day Unspecified p/vkm p/vkm p/vkm p/vkm p/vkm London (2006) 0.11 0.13 4.01 0.01 0.00 Bickel et al. Great Britain Rural All day Unspecified p/vkm p/vkm p/vkm p/vkm p/vkm (2006) 0.0060 0.0020 0.0310 0.0200 0.0020 Nash, C. Greece National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) 0.0610 0.0100 0.0410 0.0090 Nash, C. Hungary - National All day Unspecified €/vkm €/vkm €/vkm €/vkm (2003) 1.90 Electric p indicates Chaudhury, India - - - - National All day p/pkm traction indian paise. P.D. (2006) 52 Air GHG Vehicle Country Congestion Accident Noise Spatial Temporal Notes Source pollution emissions Fuel 2.75 p indicates Chaudhury, India - - - - National All day Diesel p/pkm indian paise. P.D. (2006) India 0.28 4.91 0.067 0.05 Sen et al. - Urban Peak Gasoline (Delhi) Rs/vkm2 Rs/vkm2 Rs/vkm2 Rs/vkm2 (2010) India 0.27 0.32 0.067 0.13 Sen et al. - Urban Off-peak Gasoline (Delhi) Rs/vkm2 Rs/vkm2 Rs/vkm2 Rs/vkm2 (2010) India 1.674 4.913 0.067 0.05 Sen et al. - Urban Peak Diesel (Delhi) Rs/vkm2 Rs/vkm2 Rs/vkm2 Rs/vkm2 (2010) India 1.030 0.316 0.067 0.13 Sen et al. - Urban Off-peak Diesel (Delhi) Rs/vkm2 Rs/vkm2 Rs/vkm2 Rs/vkm2 (2010) 0.0082 0.0043 0.0105 0.0063 0.0092 Nash, C. Ireland National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) Average of Danielis & 18.9 Gasoline, Italy - - - - Urban All day Chiabai ¢/mile Diesel and (1998) LPG 0.0150 0.0050 0.0080 0.0060 Nash, C. Italy - National All day Unspecified €/vkm €/vkm €/vkm €/vkm (2003) Koyama & 1.8 7.3 7.1 Japan - 3.6 ¥/vkm National All day Unspecified Kishimoto ¥/vkm ¥/vkm ¥/vkm (2001) Koyama Japan - - - - 1.10 ¥/vkm National All day Unspecified (2004) 0.0200 0.0120 0.0180 0.0110 Nash, C. Luxembourg - National All day Unspecified €/vkm €/vkm €/vkm €/vkm (2003) 0.0126 0.0058 0.0263 0.0121 0.0026 Nash, C. Netherlands National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) 0.0070 0.0070 0.0020 0.0070 0.0030 Nash, C. Portugal National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) 53 Air GHG Vehicle Country Congestion Accident Noise Spatial Temporal Notes Source pollution emissions Fuel 0.0109 0.0078 0.0174 0.0121 0.0156 Nash, C. Spain National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) 0.0070 0.0060 0.0140 0.0020 Nash, C. Sweden - National All day Unspecified €/vkm €/vkm €/vkm €/vkm (2003) 0.0096 0.0037 0.0106 0.0168 0.0094 Nash, C. Switzerland National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) Parry & 1.9 25 /metric 7.0 2.4 UK - National All day Unspecified Small ¢/mile ton of carbon ¢/mile ¢/mile (2005) 0.0113 0.0052 0.0422 0.0043 0.0126 Nash, C. UK National All day Unspecified €/vkm €/vkm €/vkm €/vkm €/vkm (2003) UK 0.50 0.03 15.08 1.50 0.39 p indicates Peirson et al. Urban Peak Unspecified (London) p/km p/km p/km p/km p/km pence. (1995) UK 0.36 0.02 1.65 1.50 0.39 p indicates Peirson et al. Urban Off-peak Unspecified (London) p/km p/km p/km p/km p/km pence. (1995) Parry & 1.9 25 $/metric 3.5 3.0 USA - National All day Unspecified Small ¢/mile ton of carbon ¢/mile ¢/mile (2005) 2.0 0.3 5.0 3.0 Parry et al. USA - National All day Unspecified ¢/mile ¢/mile ¢/mile ¢/mile (2007) pmt indicates Delucchi & 0.09-6.7 0.06-4.8 0.88-7.5 1.4-14.4 0.0-3.5 passenger USA National All day Unspecified McCubbin ¢/pmt ¢/pmt ¢/pmt ¢/pmt ¢/pmt mile of (2010) travel. 1 The European Currency Unit (ECU) was a unit of account used by the European Community before the adoption of the euro. The ECU was introduced in 1979 and replaced by the euro in 1999. 2 1 USD = Rs. 45.00 (Indian Rupees). 3 EU 27 with the exemption of Malta and Cyprus but including Norway and Switzerland. 54 Appendix B: Elasticities – Methods and Individual Study Estimates The average elasticities were calculated using an inverse-variance weighted method, commonly used for meta-analyses (Littell et al., 2008; ̂ and standard error Borenstein et al., 2009; Hunter & Schmidt, 2015). Specifically, the inverse-variance weighted average elasticity � are given by: ∑ ̂= ∑ ∑ �= ∑ where the weights are: 1 = 2 and and are the elasticity estimates and associated standard errors, respectively, taken from individual studies. Note that we did not control for publication biases in this exercise, which may result in an upwards bias in effect magnitudes. Unfortunately, we did not have an adequate number of studies for the majority of elasticities to assess or tease out the publication bias. The following Table B.1, Table B.2, and Table B.3 present elasticity estimates from individual studies and existing meta-analyses with respect to purchase, ownership, and use taxes, respectively. Note that the estimates from meta-analyses were excluded in our weighted average elasticities presented in Section 4 to avoid double-counting estimates referenced by multiple meta-analyses. 55 Table B.1: Elasticities with respect to purchase taxes Country Dependent variable Elasticity Time Period Notes Source Europe, North America 0.002 Hirota & Poot Traffic volume 1990 and 2002 and Asia1 (0.12) (2003) Europe, North America -0.01 Hirota & Poot Fuel demand 1990 and 2002 and Asia1 (0.12) (2003) -0.417 Klier & Linn France New vehicles 2007 and 20082 Short-run (0.08) (2010) Large cities in Europe, -0.191 Hirota & Poot USA, Canada and the Fuel demand 1990 and 2000 (0.05) (2003) Asia-Pacific region 1 It includes 68 large cities in Europe, North America and Asia. 2 France began taxing and subsidizing purchase in 2008 according to the vehicle's emissions rate, where the level of the tax or subsidy changed in discrete steps. Table B.2: Elasticities with respect to ownership taxes Country Dependent variable Elasticity Time Period Notes Source Europe, North America 0.037 Primarily 1990, some for Hirota & Poot Traffic volume and Asia2 (0.05) 2002 (2003) Europe, North America -0.004 Primarily 1990, some for Hirota & Poot Fuel demand and Asia2 (0.06) 2002 (2003) -0.057 Alberini & Horvath Germany New vehicles Jan 2011 - Mar 2019 (All) (0.01) (2021) -0.095 Alberini & Horvath Germany New vehicles Sep 2017 - Mar 20193 (0.02) (2021) -0.124 Alberini & Horvath Germany New vehicles Jan 2013 - Dec 20144 (0.03) (2021) -0.316 Alberini & Horvath Germany New vehicles Jan 2011 - Dec 20124 (0.04) (2021) 56 Country Dependent variable Elasticity Time Period Notes Source -0.301 Klier & Linn Germany5 New vehicles 2009 Short-run (0.08) (2010) -1.23 Vance & Mehlin Germany New vehicles 1995 - 2005 Small car segment (0.11) (2009) Large cities in Europe, -0.187 Hirota & Poot USA, Canada and the Fuel demand 1990 and 2000 (0.04) (2003) Asia-Pacific region Large cities in Europe, -0.216 Hirota & Poot USA, Canada and the Traffic volume 1990 and 2000 (0.05) (2003) Asia-Pacific region -0.296 Econometrics & Garden Netherlands New vehicles 2005 -2012 Long-run (0.08) (2013) Ciccone & Soldani Norway New vehicles -1.37 2005 - 2011 (2021) Yan & Eskeland Norway Fuel demand -0.06 2006 - 2014 (2018) Eskeland & Yan Norway Fuel demand -0.80 2006 - 2018 (2021) -0.244 2005, 2006 and first Klier & Linn Sweden5 New vehicles Short-run (0.11) quarter of 2007 (2010) -0.084 Alberini & Bareit Switzerland New vehicles 2005 - 2011 (0.04) (2019) -0.296 Cerruti et al. UK New vehicles Jan 2005 - Oct 2010 Short-run (0.07) (2019) -2.36 Econometrics & Garden UK New vehicles 2005 -2012 Long-run (1.06) (2013) 1 12 OECD countries include USA, UK, Japan, Australia, Germany, France, Italy, The Netherlands, Sweden, Denmark, Norway, and Finland. 2 It includes 68 large cities in Europe, North America, and Asia. 3 During this time, Germany adopted new vehicle gas testing procedure, Worldwide Harmonized Light Vehicles Test Procedure. 4 During this time, Germany has reduced the threshold for car emission penalties. 57 5 Germany and Sweden introduced linear CO2-based circulation taxes, while France began taxing and subsidizing purchase in 2008 according to the vehicle's emissions rate, where the level of the tax or subsidy changed in discrete steps. Table B.3: Elasticities with respect to use taxes Country Dependent variable Elasticity Time Period Notes Source 0.22 Brons et al. Meta-analysis Fuel economy n/a (0.02) (2006) -0.1 Brons et al. Meta-analysis Traffic volume n/a (0.03) (2006) -0.32 Brons et al. Meta-analysis Traffic volume n/a (0.13) (2006) -0.53 Brons et al. Meta-analysis Fuel demand n/a (0.04) (2006) Short-run. Dahl & Sterner Meta-analysis Fuel demand -0.26 n/a In Long-run: -0.6 to -0.8. (1991) Espey Meta-analysis Fuel demand -0.23 1929 - 1993 Short-run, median (1998) Espey Meta-analysis Fuel demand -0.43 1929 - 1993 Long-run, median (1998) Graham & Glaister Meta-analysis Traffic volume -0.15 n/a Short-run (2002) Graham & Glaister Meta-analysis Fuel demand -0.30 n/a Short-run (2002) Graham & Glaister Meta-analysis Traffic volume -0.30 n/a Long-run (2002) Graham & Glaister Meta-analysis Fuel demand -0.70 n/a Long-run (2002) 58 Country Dependent variable Elasticity Time Period Notes Source -0.25 Hanly et al. Meta-analysis Fuel demand 1929 - 1991 Short-run (0.02) (2002) -0.64 Hanly et al. Meta-analysis Fuel demand 1929 - 1991 Long-run (0.06) (2002) Short-run, OECD Meta-analysis Fuel demand -0.26 n/a average (0 to -1.36) (2001) Long-run, OECD Meta-analysis Fuel demand -0.58 n/a average (0 to -2.72) (2001) Meta-analysis (Rapid Dunkerley et al. Evidence Assessment Fuel demand -0.1 to -0.5 1990 or later (2014) review) Average of previous -0.16 Goodwin Traffic volume n/a Short-run literature results (0.04) (1992) Average of previous -0.27 Goodwin Fuel demand n/a Short-run literature results (0.03) (1992) Average of previous -0.33 Goodwin Traffic volume n/a Long-run literature results (0.06) (1992) Average of previous -0.71 Goodwin Fuel demand n/a Long-run literature results (0.06) (1992) Johansson & Schipper 12 OECD countries1 Traffic volume -0.30 1973 - 1992 Long-run (1997) Johansson & Schipper 12 OECD countries1 Fuel demand -0.70 1973 - 1992 Long-run (1997) -0.07 van Reeven 15 European countries2 Traffic volume 1990 - 2006 Short-run (0.01) (2011) -0.15 van Reeven 15 European countries2 Traffic volume 1990 - 2006 Long-run (0.03) (2011) 59 Country Dependent variable Elasticity Time Period Notes Source Europe, North America -0.631 Hirota & Poot Traffic volume 1990 and 2002 and Asia3 (0.08) (2003) Europe, North America -0.561 Hirota & Poot Fuel demand 1990 and 2002 and Asia3 (0.13) (2003) -0.05 Sterner et al. Australia Fuel demand 1960 - 1985 Short-run (0.02) (1992) -0.18 Sterner et al. Australia Fuel demand 1960 - 1985 Long-run (0.07) (1992) Australia -0.04 Zhang & Burke Traffic volume 2010 - 2017 Short-run, hourly data (New South Wales) (0.002) (2020) Australia -0.103 Short-run, Off-peak, Zhang & Burke Traffic volume 2010 - 2017 (New South Wales) (0.018) weekday (2020) Australia -0.074 Short-run, Off-peak, Zhang & Burke Traffic volume 2010 - 2017 (New South Wales) (0.020) weekend (2020) Australia 0.064 Zhang & Burke Traffic volume 2010 - 2017 Short-run, Peak, weekday (New South Wales) (0.023) (2020) -0.25 Sterner et al. Austria Fuel demand 1960 - 1985 Short-run (0.11) (1992) -0.59 Sterner et al. Austria Fuel demand 1960 - 1985 Long-run (0.26) (1992) -0.36 Sterner et al. Belgium Fuel demand 1960 - 1985 Short-run (0.05) (1992) -0.71 Sterner et al. Belgium Fuel demand 1960 - 1985 Long-run (0.09) (1992) -0.22 Gillingham USA (California) Traffic volume 2001 - 2003 Medium-run (2yr) (0.03) (2014) 60 Country Dependent variable Elasticity Time Period Notes Source Eltony Canada Fuel demand -0.311 to 0.313 1969 - 1988 Short-run (1993) Eltony Canada Fuel demand -0.975 to -1.059 1969 - 1988 Long-run (1993) -0.25 Sterner et al. Canada Fuel demand 1960 - 1985 Short-run (0.06) (1992) -1.07 Sterner et al. Canada Fuel demand 1960 - 1985 Long-run (0.24) (1992) Canada -0.01 Lawley & Thivierge Fuel demand 2001 -2012 (British Columbia) (0.01) (2016) -0.28 Bansal & Dua China Traffic volume 2016 - 2017 (0.05) (2022) -0.37 Sterner et al. Denmark Fuel demand 1960 - 1985 Short-run (0.06) (1992) -0.61 Sterner et al. Denmark Fuel demand 1960 - 1985 Long-run (0.10) (1992) -0.34 Sterner et al. Finland Fuel demand 1960 - 1985 Short-run (0.15) (1992) -1.1 Sterner et al. Finland Fuel demand 1960 - 1985 Long-run (0.47) (1992) -0.36 Sterner et al. France Fuel demand 1960 - 1985 Short-run (0.08) (1992) -0.7 Sterner et al. France Fuel demand 1960 - 1985 Long-run (0.15) (1992) -0.05 Sterner et al. Germany Fuel demand 1960 - 1985 Short-run (0.07) (1992) 61 Country Dependent variable Elasticity Time Period Notes Source -0.56 Sterner et al. Germany Fuel demand 1960 - 1985 Long-run (0.82) (1992) -0.23 Sterner et al. Greece Fuel demand 1960 - 1985 Short-run (0.11) (1992) -1.12 Sterner et al. Greece Fuel demand 1960 - 1985 Long-run (0.52) (1992) -0.18 Bansal & Dua India Traffic volume 2016 - 2017 (0.06) (2022) -0.21 Sterner et al. Ireland Fuel demand 1960 - 1985 Short-run (0.04) (1992) -1.62 Sterner et al. Ireland Fuel demand 1960 - 1985 Long-run (0.33) (1992) -0.37 Sterner et al. Italy Fuel demand 1960 - 1985 Short-run (0.13) (1992) -1.16 Sterner et al. Italy Fuel demand 1960 - 1985 Long-run (0.40) (1992) -0.37 Knittel & Tanaka Japan Fuel demand 2005 - 2014 Short-run (0.03) (2019) -0.302 Knittel & Tanaka Japan Traffic volume 2005 - 2014 Short-run (0.04) (2019) -0.15 Sterner et al. Japan Fuel demand 1960 - 1985 Short-run (0.03) (1992) -0.76 Sterner et al. Japan Fuel demand 1960 - 1985 Long-run (0.17) (1992) Large cities in Europe, -0.195 Hirota & Poot USA, Canada and the Fuel demand 1990 and 2000 (0.08) (2003) Asia-Pacific region 62 Country Dependent variable Elasticity Time Period Notes Source Large cities in Europe, -0.042 Hirota & Poot USA, Canada and the Traffic volume 1990 and 2000 (0.08) (2003) Asia-Pacific region -0.57 Sterner et al. Netherlands Fuel demand 1960 - 1985 Short-run (0.11) (1992) -2.29 Sterner et al. Netherlands Fuel demand 1960 - 1985 Long-run (0.46) (1992) -0.08 Sheng & Sharp New Zealand Traffic volume 2011 - 2015 Petrol (0.015) (2019) -0.011 Sheng & Sharp New Zealand Traffic volume 2011 - 2015 Diesel (0.018) (2019) North America and Second half of the Lipow Fuel demand -0.17 Short-run Europe twentieth century (2008) North America and Second half of the Lipow Fuel demand -0.40 Long-run Europe twentieth century (2008) -0.43 Sterner et al. Norway Fuel demand 1960 - 1985 Short-run (0.13) (1992) -0.9 Sterner et al. Norway Fuel demand 1960 - 1985 Long-run (0.28) (1992) -0.26 Odeck & Johansen Norway Fuel demand 1980 - 2011 Short-run (0.078) (2016) Odeck & Johansen Norway Fuel demand -0.36 1980 - 2011 Long-run (2016) -0.11 Odeck & Johansen Norway Traffic volume 1980 - 2011 Short-run (0.047) (2016) Odeck & Johansen Norway Traffic volume -0.24 1980 - 2011 Long-run (2016) 63 Country Dependent variable Elasticity Time Period Notes Source -0.13 Sterner et al. Portugal Fuel demand 1960 - 1985 Short-run (0.07) (1992) -0.67 Sterner et al. Portugal Fuel demand 1960 - 1985 Long-run (0.34) (1992) -0.14 Sterner et al. Spain Fuel demand 1960 - 1985 Short-run (0.17) (1992) -0.3 Sterner et al. Spain Fuel demand 1960 - 1985 Long-run (0.37) (1992) -0.23 Danesin & Linares Spain Fuel demand 2000 - 2007 Short-run, Diesel (0.02) (2015) -0.25 Danesin & Linares Spain Fuel demand 2000 - 2007 Short-run, Gasoline (0.04) (2015) -0.28 Danesin & Linares Spain Fuel demand 2000 - 2007 Short-run, Total fuel types (0.03) (2015) -1.67 Danesin & Linares Spain Fuel demand 2000 - 2007 Short-run, Diesel (0.63) (2015) -0.82 Danesin & Linares Spain Fuel demand 2000 - 2007 Short-run, Gasoline (0.32) (2015) -2.49 Danesin & Linares Spain Fuel demand 2000 - 2007 Short-run, Total fuel types (1.66) (2015) -0.3 Sterner et al. Sweden Fuel demand 1960 - 1985 Short-run (0.09) (1992) -0.37 Sterner et al. Sweden Fuel demand 1960 - 1985 Long-run (0.11) (1992) 0.09 Sterner et al. Switzerland Fuel demand 1960 - 1985 Long-run (0.28) (1992) 64 Country Dependent variable Elasticity Time Period Notes Source 0.05 Sterner et al. Switzerland Fuel demand 1960 - 1985 Short-run (0.16) (1992) -0.31 Sterner et al. Türkiye Fuel demand 1960 - 1985 Short-run (0.06) (1992) -0.61 Sterner et al. Türkiye Fuel demand 1960 - 1985 Long-run (0.11) (1992) -0.11 Sterner et al. UK Fuel demand 1960 - 1985 Short-run (0.07) (1992) -0.45 Sterner et al. UK Fuel demand 1960 - 1985 Long-run (0.27) (1992) Bonilla & Foxon UK Fuel economy -0.264 1970 - 2004 Long-run (2009) Austin & Dinan USA Fuel economy 0.22 2001 (2005) Austin & Dinan USA Traffic volume -0.20 2001 (2005) Austin & Dinan USA Fuel demand -0.39 2001 Long-run (2005) Bento et al. USA Fuel demand -0.35 2001 Short-run (2009) -0.37 Coglianese et al. USA Fuel demand 1989 - 2008 Short-run (0.24) (2015) -0.46 Davis & Kilian USA Fuel demand 1989 - 2008 Short-run (0.23) (2011) -0.05 Goetzke & Vance USA Traffic volume 2009 (0.02) (2019) 65 Country Dependent variable Elasticity Time Period Notes Source -0.042 Hughes et al. USA Fuel demand 2001 - 2006 Short-run (0.01) (2006) -0.335 Hughes et al. USA Fuel demand 1975 - 1980 Short-run (0.02) (2006) -0.055 Hymel et al. USA Fuel demand 1966 - 2004 Short-run (0.01) (2010) -0.285 Hymel et al. USA Fuel demand 1966 - 2004 Long-run (0.04) (2010) 0.12 Klier & Linn USA Fuel economy 1978 - 2007 Short-run (0.19) (2010) Long-run, average OECD USA Fuel demand -0.53 n/a (-0.02 to -1.59) (2001) -0.43 Small & Dender USA Fuel demand 1966 - 2001 Long-run (0.04) (2007) -0.18 Sterner et al. USA Fuel demand 1960 - 1985 Short-run (0.03) (1992) -1.00 Sterner et al. USA Fuel demand 1960 - 1985 Long-run (0.15) (1992) 1 12 OECD countries include USA, UK, Japan, Australia, Germany, France, Italy, The Netherlands, Sweden, Denmark, Norway, and Finland. 2 15 European countries include Austria, Belgium, Switzerland, Germany, Denmark, Spain, Finland, France, Ireland, Italy, Luxembourg, the Netherlands, Norway, Sweden, and the UK. 3 It includes 68 large cities in Europe, North America, and Asia. 4 This study used the European measure of fuel economy in liters per 100 km for elasticity. Hence, a reduction in the numerical value represents an improvement in the fuel economy (=inverse of the US measure of fuel economy). Lower number means better fuel economy. 66