3. Selected Excise Duties in OECD Countries

Although excise may be levied on a broad range of products, excise taxes on alcohol, tobacco and hydrocarbon oils in particular raise significant revenues for governments in all OECD countries. In recent decades, governments have increasingly used these taxes not only to raise revenue but also to influence customer behaviour where consumption of certain products is considered harmful to health or to the environment.

This chapter presents an overview of the key characteristics of excise duties and the evolution in their use by governments in selected areas (Section 3.2). It then looks in some further detail at the excise rates structure for three main categories of products: alcoholic beverages (Section 3.3), tobacco products (Section 3.4) and mineral oil products (Section 3.5). This is followed by a brief description of the impact of differences in excise rates between countries on cross-border trade (Section 3.6) and on their distributional effects (Section 3.7). A more detailed analysis of the impact of excise duties on motor vehicle and aviation fuels is provided in Chapter 4.

Excise, unlike other general consumption taxes (incl. value added taxes - VAT1) is levied only on specific goods. Although many products can be subject to excise, such as chocolate, coffee and orange juice, this chapter focuses on three principal product groups that are subject to excise in all OECD countries: alcoholic beverages, mineral oils and tobacco products.

Before looking at the key characteristics of these three groups and their comparative treatment in different countries, it is useful to recall the following general characteristics of excise duties:

  • Excise duties are generally calculated by reference to the weight, volume, strength, or quantity of the product, combined in some cases with the value, but sometimes on a value basis only.

  • Excise duties normally become payable when the goods enter free circulation. Transfers of ownership of excisable goods can take place within a controlled warehousing environment or between registered operators without creating an excise charge.

  • The excise system is characterised by small numbers of taxpayers that are active in the manufacturing, wholesale stage or importation of the three main product groups.

Unlike VAT, which is collected through a staged collection process by all the stakeholders in the value chain until the final consumer (see Chapter 2), excise duties are normally collected only once, from a registered operator, at the time the goods are released for consumption. In the European Union, the movement of excisable products between Member States takes place under a duty-suspension arrangement until the moment these products are released for free circulation. In the United States, excise duties are levied by the federal government and by many states and local governments. Federal excise taxes are collected by the Internal Revenue Service while states may impose the tax according to their own rules and rates.

The level of revenues raised by excise duties and the economic impact of these taxes depends on their structure. There are two main ways for levying excise duties on excisable products: ad valorem and ad quantum. Under an ad quantum excise (also referred to as a “specific” excise), a fixed amount of tax is levied per unit of the product (e.g. USD 1 per litre), which means that this is a tax on the volume of sales. Under an ad valorem excise, the tax is levied as a proportion of the product price (e.g. 20% of the selling price), and it is thus a tax on the value of sales. In a number of instances (e.g. tobacco taxes as presented in Annex Table 3.A.4) excise duties can be levied on the basis of a mix of ad valorem and ad quantum taxes. Ad quantum taxes requires a precise definition of the nature and characteristics of the tax base (e.g. a litre of unleaded gasoline with 94 RON) while an ad valorem tax is simply based on the price of the excisable good.

Most excisable products naturally present a bundle of different characteristics (volume, weight, strength, octane, alcoholic or carbon content, etc.). Ad quantum taxes remain unaffected by changes in the product characteristics that have not been defined as being relevant for the tax base, whereas ad valorem taxes bear on all the characteristics of the product that are reflected in the price. Depending on the structure of the excise, its impact on production and consumption is different. For example, a specific ad quantum excise on beer (per % absolute alcohol in volume) may encourage brewers to develop varieties of beer, including more luxurious products that could be offered at higher prices while remaining subject to the same level of excise as the cheaper product. On the other hand, ad valorem taxes may discourage costly improvements in product quality or encourage consumers to switch to low-cost products. An ad quantum excise may be easier to administer, because it requires only the determination of the physical quantity of the product taxed.

The tax revenue raised from excise duties and the revenue impact of changes of the design of these taxes are influenced by a wide range of factors. When assessing the impact of change to excise duty regimes, it is important to note that excise is normally part of the VAT base, i.e. VAT is usually levied on the duty-paid value of the excise products. The revenue impact of any changes in an excise tax will thus also be influenced by the level of the VAT rate on a given excisable product. Elasticity of the demand for the taxed products is a key factor in determining the revenue impact of excise duty reforms. The more elastic the demand is, the higher the likelihood that an increase in the price from an increase in excise will lead to a lower demand and thus to lower tax revenue. A price increase from higher excises may induce households to shift to other, more lightly taxed products so that the revenue impact will also depend on the tax rates levied on close substitutes. The impact of these various factors will ultimately depend on the extent in which any tax increase is passed onto consumers in the form of higher after-tax prices. If producers reduce the before-tax price in response to a tax increase, this increase will not be (fully) reflected in the consumer price as producers absorb it (partially or fully) through a reduction in their profit margin. It is also noteworthy in this context that empirical experience suggests that ad quantum tax increases tend to be more than fully passed through to consumers (prices rise by more than the tax increase), whereas ad valorem tax increases tend to be less than fully passed through (Sassi, Belloni and Capobianco, 2013[1]). In general, ad quantum taxes produce a more predictable revenue stream than ad valorem taxes, as revenue does not vary with the price of the product. On the other hand, ad valorem taxes may keep pace with inflation better than ad quantum taxes (although it is possible to adjust ad quantum taxes for inflation).

From a distributional perspective, there may be a case for ad valorem rather than ad quantum taxation. If one assumes that high-income taxpayers purchase more expensive products than low-income consumers, then an ad valorem tax could be assumed to impose a higher tax burden on high-income taxpayers relative to low-income consumers. However, this is not entirely straightforward: the exact distributional impact will depend on consumption patterns, and even with an ad valorem tax, high-income taxpayers may still end up paying less tax relative to their income than low-income households. Addressing redistributive goals is likely to be better achieved through a progressive personal income tax which directly links taxes paid to income (Brys et al., 2016[2])(see also Section 3.4 below).

There may be a case for a combination of ad quantum and ad valorem taxes if the tax is aimed at discouraging consumption of, or maximising revenue from, both high and low value products. Where there are large differences in prices of a product, an ad quantum tax will be less likely to reduce demand for the high value product, and will raise less revenue from it than an ad valorem tax. Additionally, higher income consumers who are more likely to consume high value products may be less responsive than low-income groups to the imposition of a given tax (although ad quantum taxes may reduce the price differentials). Imposing a higher aggregate tax on these expensive products will therefore be needed to affect behaviour. To achieve this, an ad valorem tax can be combined with an ad quantum tax, which is common in tobacco taxation (see Section 3.3 below). Setting the “optimal” balance between ad quantum and ad valorem components of excise will depend of the products concerned, the market structure and the government’s objective, hence there is no optimal balance between the two taxes in absolute (KEEN, 1998[3])

Finally, illicit trade and opportunities for cross border shopping are other factors that might influence the revenue potential and the impact on consumption of excisable products.

All these factors need to be taken into account by governments depending on their policy objectives i.e. to reduce consumption of products considered harmful to health or increase revenue or both.

In the OECD countries, the relative share of excise duties in total tax revenue has seen a long decline between 1975, when they accounted for 10.5% on average, and 2018 when these taxes represented 7.2% of total tax revenue on average (see Annex Table 1.A.5). Although some large differences between countries can be observed, with excise accounting for 2.6% of total tax revenue in New Zealand and 14.9% in Turkey, the weight of excise duties is between 5% and 10% of total tax revenue in the majority of OECD countries (26 out of 37). These taxes account for less than 5% of total taxes in eight OECD countries (Australia, Canada, Israel, Japan, New Zealand, Sweden, Switzerland and the United States). They account for more than 10% of total tax revenue in 5 OECD countries (Estonia, Latvia, Lithuania, Poland and Turkey), down from twelve OECD countries in 2010.

The main characteristics of excise duties and their policy objectives as revenue raisers and tools to influence consumer behaviour are largely shared amongst OECD countries. However, their rates and structure differ significantly and showing the order of magnitude of the total tax burden on specific excisable goods is therefore not straightforward. For example, standard excise rates on beer may be tempered by the application of reduced rates on small breweries. For tobacco products, different duty rates applicable to substitutes (cigarettes and rolling tobacco) may also blur the picture. Similarly, excise duties on road fuels show only a part of automotive taxation policy that also includes road tolls, taxes on registration and use of vehicles, taxes on insurance, etc.

The following sections provide some more detail on the main differences in the structure of excise duties and on their increasing use as an instrument to influence behaviour.

Alcoholic beverages exist in a wide variety across the world and are produced from a wide range of fermented or distilled ingredients (grapes, apples, malt, rice, etc.). The Customs Combined Nomenclature Code (CN) provides a classification of alcoholic beverages with which excise categories are intrinsically linked. The CN includes six main categories of alcoholic beverages: beer made from malt (code 22.03); wine of fresh grapes, including fortified wines (code 22.04); vermouth and other wine of fresh grapes flavoured with plants or aromatic substances (code 22.05); other fermented beverages (for example, cider, perry, mead), mixtures of fermented beverages and mixtures of fermented beverages and non-alcoholic beverages (code 22.06); undenatured ethyl alcohol of an alcoholic strength of 80 % pure alcohol by volume (abv) or higher (code 22.07) and undenatured ethyl alcohol of an alcoholic strength of less than 80 % abv (code 22.08). There are inevitably subdivisions within each of these broad categories but the use of the internationally accepted nomenclature enhances consistency and helps to avoid contradictory definitions in applying rates. Member States of the European Union apply a harmonised structure for excise duties on alcohol and alcoholic beverages (Council Directive 92/83/EEC). Except otherwise mentioned in country notes, Annex Table 3.A.1 and Annex Table 3.A.1 cover products under CN codes 22.03 and 22.04. Annex Table 3.A.3 covers products not included in Annex Table 3.A.1 and Annex Table 3.A.2.

Given the long history of alcohol taxation, several methods and measures have been developed over time for assessing the alcoholic content of a product. The alcohol by volume (abv) is now the standard measure of the level of alcohol contained in an alcoholic beverage. It is defined as the number of litres of pure ethanol present in 100 litres of solution at 20 °C, expressed as a percentage of the total volume. Annex Table 3.A.1 and Annex Table 3.A.2 provide an overview of excise taxation of beer and wine, whereby the alcoholic content is expressed in % abv. In some countries, the excise taxation of beer calculates the alcoholic content in degree Plato (measuring the density of beer wort in terms of percentage of extract by weight). To allow cross-country comparison, Annex Table 3.A.1 shows the estimated amounts of tax per % abv for these countries based on a conversion from the amounts of tax per degree Plato. There is no precise method to convert from degrees Plato and alcohol per volume but for tax purposes it is assumed that 1% abv is equivalent to 2.5 degrees Plato. The tax amounts per degree Plato have thus been multiplied by 2.5 to obtain the rates in degree abv for the relevant countries in Annex Table 3.A.1.

Excise can be applied to alcoholic beverages in two main ways. The duty can be either ad quantum in relation to the alcoholic content of the product or ad valorem calculated according to the value of the product. The two methods are generally combined to include both the volume (based on alcohol content) and value. One exception is Mexico where the rate of tax on alcoholic beverages is calculated exclusively on the value of the product, with a graduated rate for beer that takes into account the alcoholic content of the product.

Annex Table 3.A.1, Annex Table 3.A.2, and Annex Table 3.A.3 in respect of excise duties on beer, wine and other alcoholic beverages illustrate the complexity of the computation of excise duties in many instances. The existence of differing subcategories and specific rates (e.g. for low-alcohol products and for small breweries) and calculations based on both the value and the nature of the product, make it difficult to show an exact estimate of the precise excise tax burden at the consumer level. Nevertheless, Annex Table 3.A.1, Annex Table 3.A.2, and Annex Table 3.A.3 show the large differences of taxation levels between countries. Excise on beer (Annex Table 3.A.1), for instance, varies from less than USD 5 per hectolitre per % abv (Czech Republic, Germany, Luxembourg, Slovak Republic and Turkey) up to more than USD 20 (Finland, Ireland, New Zealand, Sweden and United Kingdom) with the highest levels observed in Finland (USD 41.01) and Israel (USD 66.01). Three quarters of OECD countries (29 out of 36) apply reduced rates to small breweries, with a progressive increase in the tax rate according to their annual production in many cases. Country notes to Table 3.A.1 illustrate the wide diversity of these tax regimes.

Excise rates on still wine (Annex Table 3.A.2) also vary widely across OECD countries from zero (Austria, Czech Republic, Germany, Greece, Hungary, Israel, Italy, Luxemburg, Portugal, Slovak Republic, Slovenia, Spain and Switzerland) to more than USD 4 per litre in Finland and Ireland and more than USD 6 per litre in Norway. In addition, if almost all OECD countries apply the standard VAT rate to alcoholic beverages, Colombia and Luxembourg apply a reduced VAT of respectively 5% and 14% to still wine. Four OECD countries do not apply any ad quantum excise to wine but only ad valorem taxes (Australia, Chile, Korea and Mexico).

For other alcoholic beverages (Annex Table 3.A.3), the excise duty rates also vary across OECD countries, but to a lesser extent than for wine and beer, as there are no zero-rates or reduced rates for small producers. These range from less than USD 15 per litre of absolute alcohol (Hungary, Germany, Czech Republic) up to about USD 90 per litre of absolute alcohol in Norway and USD 126 in Iceland. The only country to apply a reduced VAT rate (5%) to alcoholic beverages is Colombia.

Historically, as for alcohol taxation, the primary motivation for tobacco taxation was the raising of government revenue. Nearly all OECD countries have taxed tobacco products for many decades and even for centuries in some cases. The significant tobacco consumption and the relatively low elasticity of demand for tobacco products (i.e. the less than proportionate response of tobacco product consumption to a moderate price increase) along with the small number of producers made these products a particularly attractive target for excise and other taxation. Overtime, the clear evidence of the negative health consequences of tobacco use have turned tobacco taxation increasingly into a tool reduce tobacco consumption. The World Health Organisation provides economic evidence of the effectiveness of increased tobacco taxation and pricing in reducing tobacco use. It shows that tobacco taxation is highly cost-effective, combining the potential for massive impact with a low implementation cost as returns and economic benefits from tobacco taxation have proven to be several times higher than the cost of this measure (WHO, 2019[4]).

As with alcohol and mineral oils, tobacco products are subdivided into a number of categories i.e. cigarettes, cigars, cigarette rolling tobacco and pipe tobacco. New tobacco products have also emerged in recent years such as Heated Tobacco Products (HTP), and Electronic Nicotine Delivery Systems (ENDS), which may be subject to specific tax rates.

Unlike excises on alcoholic beverages and mineral oils, which are almost exclusively ad quantum, the majority of countries use a combination of ad quantum and ad valorem elements to calculate excise on tobacco products.

Annex Table 3.A.4 and Annex Table 3.A.5 show large differences in tax rates between countries. Difference in tax levels can also exist within a jurisdiction that allows excises to be levied at the sub-national level such as in the United States where, for instance, local excise rates on cigarettes (on the top of the federal tax) range from USD 0.17 in Missouri and USD 0.30 in Virginia to USD 4.50 in the District of Columbia and USD 4.35 in New York per pack of 20 cigarettes (Federation of Tax Administrators, 2020[5])

It should be noted that the individual rates or amounts of tax (ad valorem/ad quantum excise, VAT, duties, etc.) per type of tobacco product as shown in Annex Table 3.A.4 are not sufficient to assess the overall tax burden on those products. Indeed, a high ad valorem tax on a given product group can be balanced with a low ad quantum excise (or vice versa) when excise is levied on the basis of a mix of both ad valorem and ad quantum taxes for this product group. Ad valorem excises can be assessed on a range of different bases (producer price, import price, retail price). The combined effect of the VAT rate with excise duties needs to be assessed etc.

A better understanding of the relative taxation levels may therefore be gained by calculating the total tax burden (TTB) as a share of the total retail selling price (RSP) of a given product to the final consumer. To illustrate this, Annex Table 3.A.5 shows the total tax burden (ad quantum excise + ad valorem excise + VAT) for cigarettes as a share of the retail selling price (RSP) of a pack of 20 cigarettes in OECD countries. This table shows that the total tax burden for a pack of 20 cigarettes varies widely between countries, from 39.47% of the RSP in the United States (national average estimate of federal and local taxes) and 55.49% in Iceland to 82.69% in the Czech Republic and 87.41% in Finland. The tax burden is above 60% of the RSP in all the OECD countries, except Iceland and the United States; and above 75% (as recommended by the WHO) in 21 OECD countries.

Annex Table 3.A.5 also shows that there may be substantial differences in the pre-tax prices, depending on the structure of the market, the geographic location (in particular with respect to cross-border shopping) and the relevant tax structure.

Determining the rate for tobacco taxation depends on the policy objectives of the tax. If the tax is primarily intended to raise revenue, then the tax rate will be determined in light of the revenue targets and the elasticity of the demand for the taxed products. A moderate rate may be sufficient to generate stable revenue without creating significant political economy difficulties. If the tax is intended to have a significant impact on customer behaviour, then a higher tax rate is likely to be required to achieve the desired health outcomes. In the specific case of tobacco, research (Goodchild, Perucic and Nargis, 2016[6])has shown that higher taxes and prices on tobacco reduce both prevalence (i.e. users quitting) and intensity of tobacco use (i.e. users consuming less), in particular for vulnerable populations (young people and low-income households). The monetary burden of higher tobacco taxes also appears to fall more heavily on the wealthiest users, whose tobacco use declines less, while most of the health and economic benefits from reductions in tobacco consumption accrue to the most disadvantaged populations, whose tobacco use declines more when taxes increase (Belinda Loring, 2014[7]). Political economy factors (e.g., industry lobbying, public opposition) may make imposing a higher rate difficult to achieve. Earmarking (part of) the revenue from the taxes for specific health related purposes such as funding health programmes and/or tobacco control activities may increase public support, although it reduces the flexibility in government budgeting. It may also provide a regular source of funding for health care and health promotion programmes that is not subject to annual budgetary review (World Health Organisation, 2016[8]). Concerns about cross-border trade and bootlegging between countries with high price differentials may make it difficult to impose a high tax rate in some countries in the absence of efficient regional co-operation.

Mineral oils are usually sub divided into product categories on the basis of technical specifications. The main product categories for excise duty purposes are unleaded gasoline, diesel oil, and heavy fuel oil. Some OECD countries also tax other energy products such as natural gas, electricity and coal through excises or specific taxes on energy products, such as carbon taxes. This publication only considers VAT and excise taxes, or taxes considered as such or included in the excise amount. These taxes are set ad valorem or ad quantum rather than by energy or CO2 content of the fuel.

In most OECD countries, heating oil is taxed at lower rates than motor fuels. In the European Union, a minimum rate is established by the Energy Tax Directive for heating fuel (EUR 0.021/litre), which is much lower than the one for motor fuel (EUR 0.3/litre of diesel). Only a few member countries (Czech Republic, Hungary and the Netherlands) apply practically the same excise duty rates for heating oil and diesel oil. Ireland, Luxembourg and the UK also apply a reduced VAT rate for heating oil, while none of the OECD countries reported reduced VAT rates for diesel or gasoline for cars.

The large differences in the excise duty rates are reflected in the wide variation in heating oil prices across the OECD, ranging from USD 0.705/litre in Luxembourg and 0.751/litre in Belgium (which have also the lowest excise rates) to USD 1.509/litre in Denmark and USD 1.877/litre in Israel.

Differences in excise rates between countries often result from national traditions, social, environmental and health policy, local production and government financing needs. Such differences are not without impact on the cross-border movement of goods. The development of integrated markets (e.g. the European Union) and elimination of border controls at frontiers have shed light on the sometime considerable difference in excise between neighbouring countries to the extent that market forces are affected. In such circumstances, the effects of cross border shopping can have a significant economic impact on businesses and put pressure on the relevant tax authorities to seek closer approximation of excise duty rates with their neighbours. Differences between certain neighbouring countries may also encourage cross-border “bootlegging” activities (McKee Laura MacLehose Ellen Nolte et al., 2004[9]). Although some would argue that market forces should encourage a move towards convergence of rates, this is contradictory with other policy factors when issues such as health are taken into account in setting the rates.

While excise taxes can reduce the demand for the taxed goods that are considered harmful and thus the money that households spend on them over time, they can increase immediate budgetary pressure for lower income households if introduced without further support measures. It is thus important to understand the likely distributional impact of excise tax policy and to complement excise tax increases with compensatory measures, ideally through the social transfer and benefit system, to reduce adverse distributional consequences where appropriate. Flues and Van Dender (Flues and van Dender, 2017[10]) notably show that just one-third of the additional revenues raised from stronger excise taxes on energy may be sufficient to not only fully compensate lower income households for the tax increase but to even reduce their energy bill through cash transfers.

The distributional impact of excise taxes varies across the products on which they are imposed. A study by the OECD for 20 mainly European OECD countries showed that the combined impact of excise taxes on alcohol, tobacco and transport fuels tends to be regressive both when measured as a percentage of income and expenditure (OECD/KIPF, 2014[11]). These results imply that as households earn more they spend a smaller proportion of their income and total expenditure on these products. Exact burdens can vary depending on tax design. For example, if richer households consume more expensive alcohol and tobacco than poorer households, richer households will face relatively higher tax burdens with an ad valorem tax that taxes the value of the product compared to an ad quantum tax that taxes the quantity consumed.

For the tables in annex, references to the ‘European Union and its Member States’ includes the UK as a Member State for January 2020 and as an addition to the Member States (‘Member States and the UK’) for the period 1 February 2020 until the end of December 2020.

References

[7] Belinda Loring (2014), Tobacco and inequities, World Health Organisation; European Union.

[2] Brys, B. et al. (2016), “Tax Design for Inclusive Economic Growth”, OECD Taxation Working Papers, No. 26, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jlv74ggk0g7-en.

[5] Federation of Tax Administrators (2020), State Excise Tax Rates on Cigarettes, FEDERATION OF TAX ADMINISTRATORS, https://www.taxadmin.org/assets/docs/Research/Rates/cigarette.pdf (accessed on 12 November 2020).

[10] Flues, F. and K. van Dender (2017), “The impact of energy taxes on the affordability of domestic energy”, OECD Taxation Working Papers, No. 30, OECD Publishing, Paris, https://dx.doi.org/10.1787/08705547-en.

[6] Goodchild, M., A. Perucic and N. Nargis (2016), “Modelling the impact of raising tobacco taxes on public health and finance”, Bulletin of the World Health Organization, Vol. 94/4, pp. 250-257, http://dx.doi.org/10.2471/blt.15.164707.

[3] KEEN, M. (1998), “The Balance between Specific and Ad Valorem Taxation”, Fiscal Studies, Vol. 19/1, pp. 1-37, http://dx.doi.org/10.1111/j.1475-5890.1998.tb00274.x.

[9] McKee Laura MacLehose Ellen Nolte, M. et al. (2004), Health policy and European Union enlargement, http://www.openup.co.uk.

[11] OECD/KIPF (2014), The Distributional Effects of Consumption Taxes in OECD Countries, OECD Tax Policy Studies, No. 22, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264224520-en.

[1] Sassi, F., A. Belloni and C. Capobianco (2013), “The Role of Fiscal Policies in Health Promotion”, OECD Health Working Papers, No. 66, OECD Publishing, Paris, https://dx.doi.org/10.1787/5k3twr94kvzx-en.

[4] WHO (2019), WHO report on the global tobacco epidemic 2019: offer help to quit tobacco use, WHO.

[8] World Health Organisation (2016), Earmarked Tobacco Taxes - Experience from nine countries.

Note

← 1. VAT may also be referred to as Goods and Services Tax (GST). For ease of reading, all value added taxes will be referred to as VAT in this chapter.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2020

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.