copy the linklink copied!2. Special feature: Identifying environmentally-related tax revenues in Revenue Statistics
This chapter provides a special study on identifying environmentally related taxes in OECD countries from 1995-2017.
copy the linklink copied!Introduction
Environmentally related taxes are key policy instruments available to policy makers to influence environmental outcomes, in addition to their function in raising government revenues. They have received increasing attention from policymakers and researchers in recent years. Internalising the external environmental costs that remain unpriced by the market is a key element of cost-effective environmental policy (OECD, 2017[1]). There is a broad consensus in the economics literature that taxes allow this goal to be achieved in a manner that is environmentally effective, economically efficient1 and socially inclusive (provided possible negative distributional impacts on vulnerable households are addressed through targeted measures). Many countries also use these taxes in complement to other measures to address pollution externalities.
Data on environmentally related tax revenue (ERTR) are used extensively in OECD country reviews (e.g. Environmental Performance Reviews, Economic Surveys), indicator reports (e.g. Environment at a Glance, Green Growth Indicators) and work on policy integration and structural policies.
This Special Feature builds on the work of the OECD’s Environment Policy Committee (EPOC) to reconcile data on environmentally related taxes in the OECD Policy Instruments for the Environment (PINE) database with the OECD’s Revenue Statistics and with Eurostat’s National Tax Lists. Related work has focused on developing the OECD methodological guidelines on compiling accounts of ERTRs in line with the System of Environmental Economic Accounting, SEEA (OECD, 2019[2]).
Reconciling the data in Revenue Statistics and in the PINE and Eurostat databases will ensure coherence and comparability of data from multiple sources across countries and thus improve the quality of data used for policy analysis.
This chapter discusses differences in reporting and classifications between the key sources of data on ERTRs in OECD countries. It then highlights challenges in identifying ERTRs in Revenue Statistics, and presents information on environmentally related taxes in OECD countries based on the Revenue Statistics database and the PINE database. The chapter concludes by presenting proposals aimed at facilitating the identification of ERTRs in future editions of Revenue Statistics.
copy the linklink copied!Definition of taxes and environmentally-related taxes
The OECD Interpretative Guide (OECD, 2018[3])includes a definition of taxes2 which is consistent with the System of National Accounts (SNA) ( (United Nations, 2009[4])), the System of Environmental Economic Accounting (SEEA) ( (United Nations, 2014[5])) as well as the OECD PINE database (http://oe.cd/pine) and the OECD guidelines for compiling ERTR accounts (OECD, 2019[2]):
In the OECD classification, the term “taxes” is confined to compulsory unrequited payments to general government. (OECD, 2018[3])
This definition also includes revenue generated from permit schemes (e.g. auctioning or selling permits or certificates for greenhouse gas emissions or water effluents), which are classified as “non-recurrent taxes on the use of goods and on permission to use goods or perform activities” in the OECD classification.
Identifying environmentally related taxes in Revenue Statistics
The OECD classification of tax revenues used in Revenue Statistics, set out in the Interpretative Guide at Annex A of Revenue Statistics, classifies taxes according to the economic function of their base (income, property, provision or consumption of goods and services), rather than according to the sector of the economy that they apply to or to the intended purpose of the tax. For example, a tax on fuels is recorded as an excise tax (category 5121) on the basis that it is a tax on a specific good or service, whereas revenues from recurrent motor vehicle taxes are recorded as a recurrent tax on the use of goods (5210). The wide variety of forms taken by environmental taxes means that there is no single, nor even a common set, of classifications that can be used to identify ERTRs in the Revenue Statistics classification. It is necessary to look beyond a classification according to the economic function of tax bases to identify these revenues within the Revenue Statistics framework.
In the economics literature, environmental taxes are motivated by internalising externalities through the alignment of tax rates with marginal external costs, or – more loosely – by using taxes to reduce environmentally harmful behaviour.3 In practice, the identification of environmentally related taxes is made by considering the environmental relevance of tax bases ( (OECD, 2006[6])), i.e. independent of the economic function of the base, and without requiring environmental motives for the application of the taxes, or the alignment of tax rates with marginal external costs. These considerations, in line with existing OECD work on this issue, yield the following definition of environmentally related taxes:
“Environmentally related taxes are taxes whose tax base is a physical unit (or a proxy of it) of something that has a proven, specific, negative impact on the environment.” (United Nations, 2014[5])
Ideally, environmentally related tax bases will comprise physical units that are directly linked to environmental pressures (e.g. emissions, pollutant concentrations). However, for practical or administrative reasons, these tax bases are sometimes not directly taxed. Instead, inputs or outputs of activities closely linked to environmental pressures (e.g. fuel consumption, ownership of cars) are used. In such cases and in line with existing approaches, these tax bases should be used to identify environmentally related taxes. (Annexe 2.A., drawing on Table 1 in (OECD, 2019[2])) provides a list of tax bases to identify environmentally related taxes. The list describes the currently identified tax bases with environmental relevance. There is no straightforward objective criterion that could be used to define what is “environmentally related” since such taxes cut across the standard tax classifications in the OECD Interpretative Guide.4 Therefore, the list of environmentally related tax bases is a practical way to identify relevant taxes. The guidelines distinguish four environmentally related tax base categories: energy, transport, pollution and resources; with two sub-categories covering energy-related greenhouse gas (GHG) emissions (under energy taxes) and non-energy-related GHG emissions (under pollution taxes).
As ERTRs cross-cut the standard tax classification used in Revenue Statistics, the four tax base categories are not directly comparable to the tax categories in the SNA and Revenue Statistics. However, the large majority of ERTRs are likely to be included in the disaggregated tax categories on production and imports in the SNA (e.g. excise taxes, car registration taxes, pollution taxes) and on goods and services in Revenue Statistics (category 5000 of the classification, e.g. excise taxes, recurrent taxes on motor vehicle taxes, and revenues from trading permits).
copy the linklink copied!Existing sources of data on environmentally related tax revenue
The key sources of data on ERTRs include Revenue Statistics, the PINE database and the Eurostat National Tax Lists. They are detailed below.
Revenue Statistics
Revenue Statistics provides harmonised and detailed data on all tax revenues for OECD countries from 1965 to 2018, although data for all countries is not available until 1995 and data for 2018 is still preliminary. More recently, the Revenue Statistics series has expanded to include three regional publications in Africa, Asia and the Pacific, and Latin America and the Caribbean. Data for all countries from 1990 (where available) is included in the Global Revenue Statistics Database. As of 2018, detailed tax revenue statistics are available for over 95 countries.
The wide country coverage and the use of the OECD Interpretative Guide to classify the data in all countries included in the Global Revenue Statistics Database provide a unique source of comparable and reliable tax revenue data for use in tax policy analysis and reform.
Although data on ERTRs does not fall within the common tax categories set out in the OECD Interpretative Guide, there is a high level of breakdown shown in country-specific taxes underneath the common categories set out in the Interpretative Guide. These country-specific taxes allow many environmentally related taxes to be identified in each individual country.
OECD PINE database
The OECD Policy Instruments for the Environment (PINE) database, which was established in 1998, hosts a unique set of detailed information on more than 3500 environmental policy instruments in over 100 countries. PINE currently contains information on six types of market-based policy instruments (taxes, fees and charges, tradable permits, environmentally motivated subsidies, deposit refund schemes and voluntary agreements) tagged into 12 environmental domains: biodiversity, land management, natural resource management, air pollution, water pollution, waste management, land or soil contamination, noise, ozone layer, climate change, transport and energy efficiency. For more information, see http://oe.cd/pine.
The greatest value of the PINE database is that it collects detailed and structured information at the instrument level, usefully complementing other sources of more aggregated statistics. For tax instruments, in addition to annual revenue, other information is included such as the tax base, tax rate and exemptions. Such information is indispensable for an evaluation of the environmental impacts of the tax. The information is collected via a network of country experts, mostly in government agencies. Identification of environmentally related tax bases and their categorisation follows the guidelines described in (OECD, 2019[2]) see Box 2.1.
Annex 2.B provides an example of the correspondence between data reported to the OECD PINE database and Revenue Statistics.
EUROSTAT National Tax Lists
Eurostat publishes revenue data for individual taxes (the National Tax Lists or NTL) imposed by 28 EU-member states as well as Iceland, Norway and Serbia from 1995 onwards.5 Data are collected based on the European System of National and Regional Accounts (ESA 2010).
Information on the economic functions and categories of individual taxes is particularly helpful for this reconciliation exercise. Eurostat NTLs distinguish different categories of taxes: property taxes, recurrent taxes on immovable property, alcohol and tobacco taxes, and environmentally related taxes.6 The recording of these data is fully compatible with the OECD definition of environmentally related taxes and the System of National Accounts.
The compilation of Eurostat tax lists is done with the National Tax Lists (the numbers presented here were extracted on 14 January 2019). The identification of the relevant taxes is done using the four different categories of environmentally related taxes: energy, transport, pollution and resources. In total, 650 tax revenue entries out of the 6700 are categorised as environmentally related.
Annex 2.B provides an example of the correspondence between data reported to the OECD PINE database and Eurostat NTLs.
The System of Environmental Economic Accounting (SEEA) is an internationally agreed statistical framework that integrates economic and environmental data in a systematic, consistent and internationally comparable way to provide a comprehensive and multipurpose view of the interrelationships between the economy and the environment.
Environmentally related tax revenue (ERTR) accounts form part of the SEEA. ERTR accounts can be seamlessly integrated with a large range of conventional economic, environmental and social data, thus facilitating international comparisons and integrated policy analyses. Additionally, ERTR accounts are disaggregated by industries and households and can therefore be used in input-output and other industry-level studies.
The OECD has a long tradition of working on environmentally related taxation and has recently developed practical methodological guidelines for compiling ERTR accounts (OECD, 2019[2]). These guidelines intend to support building global SEEA databases and further advancing the implementation of the SEEA. The OECD guidelines are in line with the SEEA and the System of National Accounts (SNA), and build on existing guidance for applications in countries. They ensure, to the extent possible, coherence with available national and international data sources and manuals, including Revenue Statistics, the OECD Policy Instruments for the Environment (PINE) database and the IMF Government Finance Statistics Manual.
Therefore, the ERTR accounts are generally aligned with Revenue Statistics by sharing a common definition of taxes, applying accrual accounting and recording revenue acquired at different levels of government. Nevertheless, ERTR accounts cut across the standard tax categories set out in the OECD Interpretative Guide, given their focus on environmentally related taxes. Thus, the tax categories in the ERTR accounts (i.e. energy, transport, pollution and resource taxes) are not directly comparable to most tax categories in Revenue Statistics. However, the ERTR are implicitly covered in the more aggregated tax categories on goods and services recorded in Revenue Statistics (e.g. excise taxes, recurrent taxes on motor vehicles).
Work on the OECD guidelines started in 2018 and provided an opportunity to reflect on the conceptual foundations of existing accounting methods and identify aspects that might deserve to be further developed in the future:
greater focus on taxes levied on greenhouse gas (GHG) emissions;
four memo items (i.e. information items not included in the total) on certain land taxes, taxes on oil and natural gas extraction, resource rent taxes, and elevated value-added taxes levied on environmentally related tax bases; and
a common platform between compilers of the industry-level ERTR accounts and the instrument-level information reported in the PINE database.
The guidelines were successfully tested in four pilot countries. The next steps will involve the collection of ERTR accounts from OECD members and interested accession and partner countries, starting in October 2019, with a view to integrating such reporting into the OECD’s regular data collection process under the auspices of the Working Party on Environmental Information.
copy the linklink copied!Identifying environmentally related taxes in Revenue Statistics
A reconciliation between data on environmentally related taxes in Revenue Statistics, the PINE database and the EU National Tax Lists was conducted in 2018, drawing on ENV/EPOC/WPEI(2018)7. The concordances have been validated by countries (WP2 delegates and PINE contacts) and the revised results on ERTRs in Revenue Statistics and the PINE database are summarised below.
Environmentally related tax revenue data reported to Revenue Statistics and PINE
Differences between Revenue Statistics and the PINE database arise primarily due to (i) lack of suitable disaggregation of revenue streams in Revenue Statistics to allow ERTRs to be identified and (ii) generic labels of taxes in Revenue Statistics that do not permit a match to the more detailed description of the environmentally related tax in the PINE database.
In order to reconcile the differences between these sources, the following approach is adopted:
1. Identify environmentally related taxes in Revenue Statistics and, for EU member countries, in National Tax Lists submitted to Eurostat (hereafter Eurostat-NTLs).
2. Compare with the list of taxes in OECD PINE and construct correspondence tables.
3. Compare the level of the revenue reported.
The first step of the exercise is specific to each data source and was done independently. Once the lists of environmentally related taxes were compiled, the comparison with PINE tax lists, the building of the correspondence table and the comparison of revenue levels are done with all data sources simultaneously.
The identification of environmentally related taxes in Revenue Statistics is done using (i) the hierarchical nomenclature of the OECD Interpretative Guide (set out in Annex A of Revenue Statistics) and (ii) a keyword search through the names of the 8250 taxes included in the global Revenue Statistics databases (data was extracted on 13 December 2018). The following keywords were used to pre-screen potential candidates of environmentally related tax:
Energy, CO2, Power, Hydro, Oil, Coal, Gas, Diesel, Fuel, Electric, Petrol, Nuclear, Benzol, Pipeline.
Transport, Motor, Car, Road, Air, Driving, Flight, Park, Driver, Highway, Aviation, Departure, Travel, Exit, Automobile, Boat, Vehicle, Passenger, Traffic, Dock, Rail, Ship, Tonnage, Plate, Tractor, Aero.
Pollution, Hazard, Disposable, CFC, Lead, Insect, Plastic, Herb, Ozone, Fertilizer, Pesticide, Manure, Emission, Tyre, Waste, Sulphur, Nitrogen, PVC, Batter, Packaging, Packing, Rubber, Noise.
Resources, Water, Hunt, Fish, Raw, Marine, Mining, Mine, Wood, Timber, Stump, Forest, Wild, Bauxite, Titanium, Gold, Silver, Metal, Logging, Protected, Fur, Animal.
Ecology, Environment, Green, Natural, Climate, Kyoto.
The objective of the keyword search was to identify potential candidates for inclusion in PINE, and to easily spot new candidates in future releases of Revenue Statistics. Note that these new taxes could be either newly introduced, or a disaggregation of taxes that were previously reported under an aggregate heading.
Several categories of the OECD Interpretative Guide are important for the identification of environmentally related taxes. In particular, the classification of Taxes on goods and services (5000) is the most relevant. It includes all Taxes and duties levied on the production, extraction, sale, transfer, leasing or delivery of goods, and the rendering of services (5100), or in respect of the use of goods or permission to use goods or to perform activities (5200). In addition, some taxes of environmental relevance fall under Other taxes (6000), and few others concern some Property taxes (4000); e.g., forest tax, property tax on boats, payments for the change of use of forest land, forestry levies. Annex 2.B provides an example of the correspondence between data reported to Revenue Statistics and to the OECD PINE database.
Challenges and correspondence found
The main challenge for the reconciliation exercise is the aggregation of reported revenue streams in Revenue Statistics. Some countries report only at the third or fourth level of the OECD Interpretative Guide classification hierarchy without providing country-specific breakdowns. At these levels, the only environmentally related revenue that can be directly identified in the standard classification is found in the fourth-level classifications 5211 and 5212: recurrent taxes in respect of motor vehicles. This aggregation in reporting is partly due to the nature of the OECD classification of taxes which is organised according to the economic characteristics of taxes (e.g. income, payroll, property, goods), whereas the identification of environmentally related taxes is done according to the environmental relevance of the tax base (e.g. freshwater abstraction, pesticide use, energy products, waste management). Ideally, environmentally related information in Revenue Statistics should be available at the individual-tax level, with environmentally related taxes reported separately, i.e. the lowest level of the hierarchy, which is country-specific.
Even where country-specific breakdowns are available, some individual-tax revenue streams are reported as aggregates and grouped with other items. For instance, Revenue Statistics includes country-specific entries such as “Ecological taxes”, “Taxes on pollution/resources”, “Other environmental taxes” and “Other taxes on the environment”, that do not allow the individual taxes to be easily identified. Similar issues arise in the Eurostat NTLs with generic entries labelled as “Pollution taxes”, “Other environmental taxes”, “Taxes and charges on environment and pollution”. In these cases, the revenue level of multiple PINE entries was compared against these aggregated revenue streams, and a tentative breakdown was developed.
In addition, individual-tax revenue streams are sometimes reported under rather generic labels that do not allow the precise nature of the tax-base (e.g. “Environmental tax”, “Environmental charge”, “Environmental incentive fee” or “Ecological tax”) to be reliably identified. In such cases, revenue streams from Revenue Statistics were considered for the reconciliation if they belonged to the OECD Interpretative Guide category 5000, while revenue streams from Eurostat-NTLs were considered only on the basis of their environmental category, independently of the label. The comparison of revenue levels with data in the PINE database allowed for some of the taxes with generic labels to be identified and even some mislabelled data to be detected (e.g., “Excise duties on petrol” in PINE and Revenue Statistics correspond to “Excise duties on gas” in Eurostat NTL for the Netherlands).7 Slightly more descriptive labels of the reported taxes would facilitate the identification of the environmentally related revenue streams.
Another challenge for the reconciliation arises from differences across these data sources on how they report revenue at different levels of government. Revenue Statistics reports revenue collected by federal or central government, plus the combined aggregate of states/regions and the combined aggregate of local government revenues. The policy instruments included in the OECD PINE database are entered at the level of the government that introduces and controls the policy. As such, in the OECD PINE database, most of the revenue information is collected at the national level for nation-wide economic instruments (and some supra-national ones, such as European Union Emission Trading System (EU-ETS)), while sub-national tax revenue data are available for individual states or regions in respect of sub-national policies (e.g. Belgium).
The results of the 2018 reconciliation exercise are:
Construction of a correspondence table, which maps individual tax entries among the three sources: Revenue Statistics, OECD PINE and Eurostat NTLs. The table contains about 700 PINE-Revenue Statistics and PINE-NTL correspondences. The remainder are unmatched entries, of which, about two thirds are sub-national policy instruments (e.g. India, United States) and one third are national instruments. Annex 2.B includes an example of the correspondences.
The reconciliation exercise also allowed the inclusion of the revenue from auctioning of tradable permits, which had been previously excluded from the aggregates for most countries. In particular, this concerns the inclusion of the revenue from auctions of permits under the EU-ETS. It is important to mention that the PINE database captures information on tradable permits separately to taxes; this is different to Revenue Statistics and Eurostat NTLs, where such revenue is reported as a tax item.
Calculation of country-level statistics
The reconciled list of taxes is aggregated at the country-level across distinct environmentally related tax categories (see Annex 2.A for a more complete list of tax bases):
Energy:
Energy products (fossil fuels and electricity) including those used in transportation (petrol and diesel).
CO2 taxes
Transport: One-off import or sales taxes on transport equipment, recurrent taxes on ownership, registration or road use of motor vehicles, and other transport-related taxes. (Note that this definition excludes all taxes on automotive fuels.)
Pollution:
Air pollution: Taxes on e.g. SOx and NOx emissions.
Ozone-depleting substances: Taxes on specific substances, such as chlorofluorocarbons (CFCs), carbon tetrachloride, chlorofluoromethane (HCFCs) and other ozone-depleting substances.
Water pollution: Taxes on discharge of wastewater; pesticides and fertilisers.
Waste management: Taxes on final disposal of solid waste, on packaging (e.g. plastic bags) and other waste-related taxes (e.g. batteries, tyres).
Resources:
Mining and quarrying: Mining royalties, excavation taxes (e.g. sand, gravel).
Freshwater: Taxes on freshwater abstraction and piped water. (Fees and charges related to water supply are not included.)
Hunting and fishing taxes.
The correspondence of taxes and tradable permits to the environmental categories presented above is done according to the tax base of each instrument. Information from PINE on the introduction and discontinuation of instruments is used to impute zero revenues. In particular, the field “year of introduction” is used to impute zero revenue before the introduction of the instrument. The last year marked in the “year of revision” is used for those instruments which have been discontinued, and zeros are imputed for subsequent years.
copy the linklink copied!Trends in environmentally related taxes in OECD countries
Environmentally related tax revenues (ERTRs) in OECD countries form a small but important part of total tax revenues, accounting for 5.1% of total tax revenues on a weighted average8 basis and 6.9% on a simple average basis in 2017, with a range from 2.8% in the United States to 12.5% in Slovenia and Turkey. Similarly, as a share of GDP, ERTRs account for 1.6% of GDP on a weighted average basis and 2.3% on a simple average basis, with their shares in GDP ranging from 0.7% in the United States to 4.5% in Slovenia (Table 2.1).
Since 1995, the earliest year for which Revenue Statistics data is available for all OECD countries, most countries have seen a decline in ERTRs as a share of tax revenues. Because of this, both the weighted and simple average shares of ERTRs to total tax revenue fell significantly (from 7.3% to 6.9% of tax revenues on a simple average basis and from 6.1% to 5.1% on a weighted average basis). In 2017, ERTRs were lower as a share of total taxation than in 1995 in 26 OECD countries, with an average decrease of 1.7 percentage points in these countries. (Figure 2.1, Table 2.1). This fall may be due to a number of factors, discussed further in Box 2.1.
Note: OECD-W refers to the weighted average (as shown in PINE) and OECD-S to the simple average. For Canada, Korea and Israel, data for 2014 are used for subsequent years as data for later years are not available in the PINE database and ERTRs cannot be identified in Revenue Statistics with sufficient precision to be included. For the same reason, data for 2016 are used for 2017 in Australia and the United States.
Source: OECD (2019c) OECD Environment Statistics (database).
The remaining 10 OECD countries had increases in ERTRs over the period as a share of total taxation, with an average increase of 2.9 percentage points in these countries. The largest increases were seen in:
Latvia, which saw an increase from 3.3% of total tax revenues to 12.2%, due to higher energy taxes;
Estonia, from 2.7% to 8.8%, entirely from energy tax revenues increasing; and
Turkey, which saw an increase from 7.2 to 12.5% with roughly equal increases in energy and transport taxes.
Overall, the distribution of OECD countries around these averages has changed slightly. In 2017, the distribution was skewed to the left, with 25 OECD countries having ERTRs of between 4% and 8% of total tax revenues, whereas in 1995, the distribution was more even, with 22 countries having ratios between 6% and 10% (Figure 2.2). This indicates that there has not been a general shift towards ERTRs as a revenue source in OECD countries.
Note: For Canada, Korea and Israel, data for 2014 are used for subsequent years as data for later years are not available in the PINE database and ERTRs cannot be identified in Revenue Statistics with sufficient precision to be included. For the same reason, data for 2016 are used for 2017 in Australia and the United States.
Source: OECD Environment Statistics (database) (OECD, 2019[7]).
As a share of GDP, ERTR in OECD countries have remained relatively steady between 1995 and 2017, with no clear trends emerging. On a simple average basis, ERTRs increased as a share of GDP in the late 1990s to a high point of 2.6% of GDP in 1998 and 1999 (which were was also the high point for the weighted average, at 1.9% of GDP), before decreasing relatively steadily until 2008, picking up in 2009, and slowly increasing since. Half of the OECD countries had ERTRs of between 2.1% and 3.2% of GDP in 1999 and between 1.7% of GDP and 2.9% of GDP in 2017, as shown by the interquartile range in Figure 3. The interquartile range decreased after 1999, before increasing slowly after 2008. The median followed a similar trend to the simple mean across the period, although increased more sharply after the financial crisis to a small peak in 2010, followed by a greater decrease (Figure 2.3).
The tax-to-GDP ratio is the principal indicator used in Revenue Statistics (see Box 1.2 in chapter 1 for further information). Taxes are expressed as a share of GDP both to allow comparability across countries and over time and to provide an indication of the scale of tax revenues against the production of the economy from which they were generated. The indicator is also used for individual tax categories, as unlike the share of total tax revenue, the share of a tax to GDP is not contingent on the levels of other taxes in the economy.
Although the indicator is a useful tool, it has some limitations when applied to ERTRs, as their tax base is not necessarily linked to GDP and may follow quite different trends. Ideally, the tax base for environmentally related taxes is a close proxy for the environmental harm itself such as the emission of CO2 or a pollutant, or the purchase, import or use of a unit of fuel. Over time, GDP and environmentally related tax bases can decouple, for example as the economy becomes less carbon-intensive as a consequence of environmental policy or other developments. Such a decoupling has been seen, for instance, in several countries that have implemented environmentally related taxation, including Sweden and Norway. This will result in a fall in the share of ERTRs to GDP, but is in fact the intended outcome of such a policy. Similarly, a country with a carbon-intensive economy may have higher ERTRs as a share of GDP than a country with a less carbon-intensive economy, even if the tax rate on environmentally related bases is lower.
Other factors are also important in interpreting changes in ERTRs expressed as a share of GDP, which may result from a shift within the base of the environmentally related tax. For example, a structural shift from a high-taxed fuel to a lower-taxed fuel will result in lower ERTRs. Similarly, deployment of new technologies such as solar and wind power, or fuel-efficient heating technology, could result in a decrease in ERTRs.
Given that environmentally related tax rates are typically defined in physical units (e.g., per tonne) and hence are set in nominal terms. Without inflation adjustment, the value of these rates decreases in real terms over time. While some countries, for example Denmark, the Netherlands and Sweden have implemented such adjustments, most OECD countries do not yet apply inflation adjustments.
Finally, it is important to consider other indicators alongside the ERTRs-to-GDP ratio when evaluating a country’s environmental policy. One alternative approach is to compare the level of environmentally related taxes to the environmental harm associated with the tax base. For example, the OECD’s Taxing Energy Use and Effective Carbon Rates publications compare taxes on energy use and carbon prices resulting from emissions trading systems against a benchmark value for the climate damage caused by the use of the energy products. Indicators developed in these publications, such as the carbon pricing gap, provide important complementary information in understanding countries’ environmental tax and carbon pricing efforts.
In 2017, revenues from ERTR in OECD countries ranged from 0.7% of GDP in the United States to 4.5% in Slovenia (Table 2.1). Across the period, Denmark had the highest level of ERTRs as a percentage of GDP from 1995 to 2011 (with the exception of 1998), and Slovenia had the highest level in 1998 and from 2012 to 2017. The lowest level of ERTRs as a share of GDP was seen in Mexico in 1995 and 1996, the United States from 1997 to 2004, Mexico from 2005 to 2014, during which time this share was negative in several years, 9 and the United States from 2015 onwards.
In monetary terms, the greatest amount of tax revenues, in USD adjusted for purchasing power parity, were raised by the United States, at USD 121 229 million in 2016 (2017 data are not available). Italy had the second highest revenues from ERTRs at USD 69 879 million in 2017, and three other countries (Japan, Turkey and the United Kingdom) also had revenues of greater than USD 60 000 million (Table 2.2).
Note: OECD-S refers to the simple average. For Canada, Korea and Israel, data for 2014 are used for subsequent years as data for later years are not available in the PINE database and ERTRs cannot be identified in Revenue Statistics with sufficient precision to be included. For the same reason, data for 2016 are used for 2017 in Australia and the United States.
Source: OECD Environment Statistics (database) (OECD, 2019[7]).
The greatest share of ERTRs in OECD countries in 2017 comes from taxes on energy, predominantly on road fuels, with this share accounting for 71% of total revenues on both a simple and weighted average basis, and over 50% of total ERTRs in all countries (Figure 2.4).
Transport taxes accounted for the second largest share of ERTR in 2017 (both on average, at 26% weighted and 25% unweighted, and in every country except Estonia, France, Lithuania, the Slovak Republic and Slovenia). Pollution and resource taxes play a more minor role: in all but eight OECD countries (Czech Republic, Estonia, France, Iceland, Lithuania, Netherlands, Slovak Republic, Slovenia) they accounted for less than 5% of ERTRs in 2017.
The situation in 1995 was relatively similar: energy taxes then played the biggest role in the composition of ERTRs, at 73% on a weighted average basis and 70% on a simple average basis, although they were under 50% of ERTRs in five countries (Figure 4). Transport taxes accounted for a slightly lower proportion, at 26% (weighted) and 24% (unweighted) on average.
Note: OECD-W refers to the weighted average and OECD-S to the simple average. For Canada, Korea and Israel, data for 2014 are used for subsequent years as data for later years are not available in the PINE database and ERTRs cannot be identified in Revenue Statistics with sufficient precision to be included. For the same reason, data for 2016 are used for 2017 in Australia and the United States.
Source: OECD Environment Statistics (database) (OECD, 2019[7]).
On both a weighted and a simple average basis, there has been a shift towards taxes on energy following the recent financial crisis and away from transport taxes. While revenues from both dropped as a share of GDP in 2008 and 2009, revenues from energy taxes recovered to pre-crisis levels relatively quickly and have increased slowly since. Revenues from transport taxes have not recovered to the same extent and, on a simple average basis, represent a comparatively lower share of total ERTRs, on average, in 2017 than earlier in the period. OECD countries have become more similar, in having a higher share of revenues from energy rather than transport taxes in 2017 compared to 1995.
Finally, taxes on pollution and natural resources represent only a small portion of ERTRs in most countries and on average. This could change in the future if countries start placing a greater emphasis on addressing water pollution, sustainable resource use or conservation of biodiversity.
While comparisons of ERTRs in OECD countries provide a useful starting point for analysing the impact of environmental taxation, comparing only the levels of revenues does not provide the full picture of a country’s environmental policy as it does not provide information on the levels of tax rates or the exemptions applied. Other parts of the OECD PINE database, including information on tax rates and exemptions, allows deeper assessment of the environmental impacts of the taxes. In addition, governments may choose to implement environmental policy using a range of other instruments, including fees and charges, expenditures (both direct and subsidies) and regulation, some of which are also detailed in the PINE database (see http://oe.cd/pine for information on the use of alternative instruments in countries).
Note: For Canada, Korea and Israel, data for 2014 are used for subsequent years as data for later years are not available in the PINE database and ERTRs cannot be identified in Revenue Statistics with sufficient precision to be included. For the same reason, data for 2016 are used for 2017 in Australia and the United States.
Source: OECD Environment Statistics (database) (OECD, 2019[7]).
copy the linklink copied!Conclusion
Reconciling data on environmentally related tax revenue (ERTRs) in Revenue Statistics with the PINE database achieves a greater coherence of data across sources of similar information. This requires identification of environmentally related tax revenue in Revenue Statistics, which can be complicated due to (i) aggregation of multiple revenue streams and (ii) generic labels used in countries’ submissions. Countries are encouraged to report revenue streams from environmentally related taxes separately as a country specific tax in Revenue Statistics (or using a breakdown that follows the ERTR tax categories) and to use labels that allow such taxes to be more clearly identified (e.g. “tax on NOx emissions” instead for “environmental tax”).
The reconciliation of Revenue Statistics with the PINE database has allowed several conclusions to be drawn:
ERTRs in OECD countries in 2017 range from 2.8% of total tax revenues in the United States to 12.5% in Slovenia and Turkey, accounting for 5.1% of total tax revenues on a weighted average basis and 6.9% on a simple average basis. Similarly, as a share of GDP, ERTRs account for 1.6% of GDP on a weighted average basis and 2.3% on a simple average basis, with their shares in GDP ranging from 0.7% in the United States to 4.5% in Slovenia.
Since 1995, there has been no trend that indicates OECD countries are moving toward sourcing a greater share of tax revenues from environmentally related taxes bases, and OECD countries have not become more similar in the levels of ERTRs observed either as a share of total tax revenues or of GDP.
The largest share of ERTRs in OECD countries in 2017 was derived from energy taxes, both on average and in nearly every OECD country, accounting for nearly three-quarters of ERTRs.
On average, the reliance of OECD countries on energy tax revenues (as a share of all ERTRs) increased after the financial crisis, offset by a decrease in revenues from transport taxes.
As part of its work on developing methodological guidelines for compilation of ERTR accounts in line with the SEEA, the OECD is also developing a joint platform to ensure that the instrument-level information in PINE is consistent with the more aggregated statistics reported in ERTR accounts. Similarly, one of the goals of this Special Feature is to achieve consistency between Revenue Statistics on the one hand, and the OECD PINE database and ERTR accounts on the other hand, by encouraging a higher disaggregation of ERTR data in Revenue Statistics.
References
[7] OECD (2019), “Environmental policy: Environmental policy instruments”, OECD Environment Statistics (database), https://dx.doi.org/10.1787/data-00696-en (accessed on 25 November 2019).
[2] OECD (2019), Environmentally related tax revenue accounts: OECD methodological guidelines in line with the SEEA, OECD.
[3] OECD (2018), Revenue Statistics 1965-2017: Interpretative Guide, http://www.oecd.org/tax/tax-policy/oecd-classification-taxes-interpretative-guide.pdf (accessed on 25 November 2019).
[1] OECD (2017), Environmental Fiscal Reform: Progress, prospects and pitfalls, http://www.oecd.org/tax/tax-policy/tax-and-environment.htm. (accessed on 28 November 2019).
[6] OECD (2006), The Political Economy of Environmentally Related Taxes, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264025530-en.
[5] United Nations (2014), System of Environmental-Economic Accounting 2012 Central Framework, https://unstats.un.org/unsd/envaccounting/seeaRev/SEEA_CF_Final_en.pdf (accessed on 25 November 2019).
[4] United Nations (2009), System of National Accounts 2008 European Commission International Monetary Fund, https://unstats.un.org/unsd/nationalaccount/docs/SNA2008.pdf (accessed on 25 November 2019).
Notes
← 1. Split incentives (e.g. owner-tenant), steep marginal damage costs (e.g. exposure to hazardous substances), etc., are some of the well-known exceptions when the literature tends to argue in favour of regulatory instruments (e.g. maximum emission limits, minimum efficiency standards and other technology prescriptions) or some combination of the two (i.e. taxes and regulations).
← 2. There is on-going work that may lead to the revision of the OECD definition of taxes to include also payments to supra-national institutions.
← 3. For a more detailed discussion see OECD (2017).
← 4. For instance, in the SNA, taxes are classified regarding taxes on production and imports, income and wealth, and capital. In the OECD Revenue Statistics, taxes are classified according to income, profits, capital gains, earnings, payroll, number of employees, property, goods and services, and others. Nevertheless, the large majority of ERTR is likely to be part of taxes on production and imports in the SNA and on goods and services in the OECD Revenue Statistics.
← 5. Aggregated statistics are also published for Liechtenstein and Switzerland.
← 6. Called environmental taxes by Eurostat.
← 7. This confusion arose because of the similarity of gas and gasoline (or petrol) in English and is being corrected in the National Tax Lists.
← 8. The weighted average is calculated using the shares of revenues in USD, adjusted for PPP, for each country. See footnote below for an explanation of why these figures will differ slightly from those reported in PINE.
← 9. The OECD Revenue Statistics database and the PINE database treat revenues from Mexico’s fuel price controlling formula differently, which was in place from 2006 to 2014. Revenue Statistics includes only the positive elements of this formula as revenues, whereas the PINE database includes also the implicit subsidy. Data in this Special Feature for Mexico are based on the information provided by delegates for the PINE database and may therefore differ from those reported elsewhere in Revenue Statistics.
← 10. This figure represents the total, rather than the weighted average.
