Figure 5.5. Revenue from taxation of new cars based on energy use per vehicle km and powertrain type, 2017

Notes: EU G4 includes France, Germany, Italy and United Kingdom. Average fuel consumption per powertrain is based on fuel economy values from the IEA Mobility Model and the Argonne National Laboratory GREET model, linked to the 2017 Worldwide Harmonised Light Vehicle Test Procedure (WLTP) values and market shares of vehicle size segments of the IEA-GFEI database on new LDV registrations with a real-driving correction of 20% (ANL, 2018; IHS Markit, 2018) (IEA, 2019b). PHEVs are assumed to be in all-electric mode for 60% of their vehicle kilometres and one-third using gasoline. Hybrids are assumed to use gasoline. Electricity taxes for India and United States are based on the weighted average of residential electricity consumption, sales tax and local taxes per state. Values for Japan, EU G4 and the United States (except US electricity tax) are based on the IEA World Energy Prices 2018 database (IEA, 2018e). Tax rates for China and India are estimates based on Bankbazaar (2018) and Avalara (2019).

Sources: IEA assessment based on IEA (2018e), EIA (2019), Tax Foundation (2017), Bankbazaar (2018), OECD (2018), Government of India (2016), Avalara (2019) and IHS Markit (2018) for the IEA-GFEI database.

There are large differences in tax revenue per vehicle kilometre between powertrains and among countries.

Solutions

Near-term options

Ensuring that the policies defining road use, vehicle and energy taxes in transportation are ready to adapt to the changes in the vehicle and fuel markets is the minimum near-term solution needed to handle the challenge posed by the technology transition towards electric mobility.

Key examples of features that would enable this adaptation capacity include:

•Adjustments of the emissions thresholds (or the emissions profile) that define the extent to which vehicle registration taxes are subject to differentiated fees (fee bates or bonus/malus), as well as those that define the charges applied in annual circulation taxes, in order to ensure that these schemes can maintain their overall characteristics (i.e. keep

providing financial incentives for the adoption of vehicles offering good environmental performance) while adjusting the total amount of revenue they generate.60

60 These changes should be implemented in a way that ensures consistency of the policy signals for OEMs, given their long-term planning horizon needs for investment in low-carbon technologies.

PAGE | 192

IEA. All rights reserved.

61 In the Norway bus lane example, it is worth mentioning that the municipality of Oslo restricted the bus lane access on two specific corridors during rush hours to only electric cars with two or more persons on-board. This was due to congestion during rush hour in these corridors, a phenomenon that increased with the growth of EV sales (IEA, 2018g).

PAGE | 193

IEA. All rights reserved.

Table 5.5. Taxes and charges relative to revenue stability, management of external costs and ease of implementation

aThe one-off character of a registration or a purchase tax also renders revenue dependent on fleet turnover and the business cycle.

bDistance-based charges are a promising tool to manage a transition to more automated vehicles and shared services.

cFuel quality influences pollutant emissions, but vehicle technologies are the main determinant of vehicle performance in this respect.

eDifferences in vehicle efficiency make fuel taxes uneven if converted once they are expressed in “per km driven” terms. This could be corrected setting taxation levels in a way that accounts for differences in vehicle efficiency. In the case of electricity, this would be challenging since it would require different levels of taxation for various end-uses and separate metering.

fFuel taxation is well suited to apply carbon taxes, also for EVs and fuel cell electric vehicles, especially if carbon taxes are embedded in electricity (or hydrogen) prices and are uniformly applied across all sectors. If fuel taxation aims to collect infrastructure costs, it would face implementation challenges for EVs because of a need for different levels of taxation across end-uses and separate metering.

gFor example, in Oregon’s experimental distance-based charging programme and the German truck tolling system, driving related data are destroyed as soon as drivers pay their road user charge.

Source: OECD (2019c) with IEA elaboration

PAGE | 194

IEA. All rights reserved.

Indications in Table 5.5 suggest that gradually increasing taxes on carbon-intensive fuels, combined with the use of distance-based charges to recover infrastructure costs and to reflect the costs of pollution and congestion (which requires the variation of distance-based charges depending on the extent of the pollution and congestion levels) can support the long-term transition to zero-emissions mobility while maintaining revenue from transport taxes.62 Distance-based charges are also well suited to manage the impacts of disruptive technological change in road transport, including those related to electrification, automation and shared mobility services. To date, however, only a few places have considered or implemented distance-based charges (Box 5.6). Alternative approaches would need to consider a broader set of instruments, not only limited to the transport system.

Box 5.6. Distance-based charges in use today

Distance-based charges in use today consist primarily of tolls (typically applied to highways and bridges). In a limited number of cases, mostly urban areas, distance-based charges are used primarily to manage travel demand (by adapting the charges for different times of the day) and mitigating pollution (by increasing the cost of travelling for vehicles leading to higher rates of pollutant emission per passenger-km).

For intercity travel, the highway tolls are the main instrument to collect distance-based charges. The main alternative to distance-based pricing for highway driving has been the use of a fee-based system, unrelated to distance driven. An example is the “vignette” in Switzerland. In the European Union, distance-based charges for trucks apply within the framework of the Eurovignette Directive (1999/62/EC) and subsequent amendments (2006/38/EC and 2011/76/EU). This aims to recover costs associated with the construction, operation and maintenance of road network (user pays principle), complementing it with a light version of the polluter pays principle allowing toll rates to vary (to a limited extent) in order to account for environmental (air pollution and noise) or traffic management objectives (OECD, 2019c).

To date, distance-based charges are not widely used in cities because of implementation challenges, but there is growing interest in the concept. To date, four cities have implemented road charging schemes (London, Milan, Singapore and Stockholm). Of these, Singapore has a zonal differentiation (bringing them close to a distance-based charge approach), while London, Milan and Stockholm adopted a cordon-based approach, having no (or weaker) links with the amount of kilometres driven. New York decided in 2019 to introduce charges from 2021 to enter Manhattan’s most congested neighbourhoods (Hu, 2019).

There are no examples of nationwide applications of distance-based charges for light-duty

62 For example, Jenn (2018a) indicates that fuel taxes will become increasingly outdated as a mean to fund transportation infrastructure, and the OECD (2019c) points to the adoption of a gradual increase in fuel or carbon taxes to cover the external costs closely related with fossil fuel use in vehicles and the phasing-in of distance-based charges for cars to reflect external costs closely related to driving. OECD (2019c) also warns about the need to ensure that the revenue potential from distance-based charges is adequate for the challenges posed by the technology transition, mentioning for example that in the European Union, this would need a revision of the maximum values allowed in the legislative framework regulating distance-based charges to reflect the full driving-related external costs.

PAGE | 195

IEA. All rights reserved.