Environmental Taxes in Europe

Abstract

This paper provides an overview and a discussion of environmental taxes in Europe. On the whole, most European countries have fairly high levels of environmental taxation - at least compared to the US. This appears broadly speaking to be true of both tax levels and tax revenues. It is partly related to a greater reliance on taxes as an instrument of environmental policy and partly due to a greater acceptance of taxes and maybe a larger public sector overall. It may also be due to a more ambitious goal when it comes to reductions in fossil energy use, particularly for transportation. There is still considerable variation within Europe when it comes to the level of taxation, the use of the revenues collected and other issues.

Full text

Public Finance and Management,
3(1), 2003 pp. 117-142
Environmental Taxes in Europe
Thomas Sterner
and
Gunnar Köhlin
Department of Economics
Göteborg University, Sweden
Abstract
This paper provides an overview and a discussion of environmental
taxes in Europe. On the whole, most European countries have fairly high
levels of environmental taxation – at least compared to the US. This
appears broadly speaking to be true of both tax levels and tax revenues. It is
partly related to a greater reliance on taxes as an instrument of
environmental policy and partly due to a greater acceptance of taxes and
maybe a larger public sector overall. It may also be due to a more
ambitious goal when it comes to reductions in fossil energy use,
particularly for transportation. There is still considerable variation within
Europe when it comes to the level of taxation, the use of the revenues
collected and other issues.
Introduction1
In this article we discuss environmental taxes in Europe. Revenues
from these taxes average around 3% of GDP or 8% of government revenue
(OECD, 2002) and are widely accepted – not only in Germany and the
Scandinavian countries where green parties are strong – but also in many
Southern and Eastern European countries with less pronounced green
movements.
The stylized fact is that the USA prefers tradable permits2 and
particularly freely allocated, grandfathered, permits while Europe prefers
1 Special thanks to Nils -Axel Bråthen at the OECD for useful and enlightening discussions
concerning these data.
2 One striking feature of permits is the dominance of the US experience: outside the US the
only important applications are in fisheries.

118
taxation. It is tempting to see this as a reflection of a stronger belief in
individual private rights and particularly the “prior appropriation” doctrine
(the right of the first user) in the US. The corresponding credo in Europe is
rather in favor of the benign and paternalistic state as a representative of
“societal” interests. There is probably a good deal of path dependency in the
learning and lobbying processes that characterize the development of policy
over time. Originally, environmentally related taxes and fees in some
European countries were intended mainly to finance water supply, sewage
treatment, road construction etc. However, gradually policy makers and
others found that they also led to energy and resource savings that provided
motivation for continued increases of these taxes. As the taxes grew
gradually, society adapted to them and some groups may even have realized
that they benefit from them, such as public employees in general, and in the
case of fuel taxes the producers of energy efficient cars and those who work
for public transportation etc.
A common trait for these industrialized countries is that basic
regulation and licensing forms the backbone of environmental protection.
The market-based and other instruments are not construed as alternatives so
much as complements to the traditional instruments.
A number of formerly planned economies of Eastern Europe
implemented a form of environmental taxes already under the old regime.
These did not really operate as Pigouvian taxes since the firms had “soft
budget constraints”: They did not earn their revenues on a market but were
allocated ministerial funds in relation to “needs”. Additional “taxes” would
thus be paired with additional allocation of “funds” to pay them so there
was no incentive effect at all! The taxes did however raise revenue for funds
that were used for some abatement and restoration. They also raised
environmental awareness and they facilitated the introduction of real tax
instruments as these economies embarked on their transition towards market
economies.
As shown by various authors on fiscal federalism, there are strong
welfare arguments in favor of ‘subsidiarity’ (which is the jargon used within
the European Union for finding an appropriate level of decentralization for
each particular political topic), see for instance Oates (1998). Allowing
local legislations to adapt policies to local conditions can lead to
considerable gains as is well known in countries that are political

119
federations (including the US and Germany). This assumes however that
these legislations are large enough to internalize most of the pollution and
that they have the appropriate instruments in order to implement an optimal
policy. In the case of transboundary pollution or the management of global
commons, international policies or at least coordination of policies is clearly
called for.
Within the European Union the need for common environmental
regulation, as a response to cross-boundary pollution, has been one
important justification for greater harmonization between countries.
Historically the EU has also been a progressive force in countries with less
developed environmental regulation. This argument has also been important
in order to persuade skeptics in countries about to join the EU. There have
even been (green) voices proposing that the development and
implementation of environmental policies should be a major theme for the
EU. In recent policy papers the Environmental Directorate of the European
Commission has given particular emphasis to environmental taxation
(Schlegelmilch, 1998).
Naturally, an important force behind environmental taxes is political
pressure and in many European countries green parties have succeeded in
establishing themselves in both parliament and in some cases government.
There has been quite a popular pressure in favor of environmental tax
reform and several countries have officially carried out such reforms in a
more or less ambitious way. Both taxes and regulations have to be enforced
to be meaningful. When it comes to taxes, experience is that various forms
of tax relief crop up, particularly in the energy intensive industries where
the incentive effect is most needed. Public opinion, even among green
parties tends to want to avoid bankruptcy and thus the affected industries
are usually quite successful when lobbying for exceptions. This is discussed
particularly for the case of Swedish energy taxes below. The difficulty in
enforcing and levying traditional taxes has led to a number of innovations
such as revenue recycling or revenue neutral instruments that will also be
discussed below (see for instance the NOx charge in Sweden).
The menu of policy instruments
There is no universally accepted taxonomy of environmental policy
instruments, but table 1 shows one that may be useful in this context. The

120
main point to retain is perhaps that the choice of instruments is not just a
choice between “taxes or regulations”. In most countries a great variety of
instruments are used for different applications within the general area of
environmental management. In many cases the application itself restricts
the use of instruments in a natural way. The cleaning of public areas is in
almost all countries provided as a “public good”. Still, information
campaigns and fines for littering, as in Singapore, could be complementary
instruments. Many industrial processes are so complicated that they can
hardly be regulated by taxation but there may still be a choice between
information disclosure, labeling, liability and licensing. In many countries
there are layers of regulation and policy instruments, which imply that in
comparing instrument choice between countries, the main difference may
rather be one of degree or of the emphasis put on one type of instrument
such as taxes as compared to other instruments in the mix of policies
applied. This should be kept in mind when we take a closer look at the use
of environmental taxes in various countries.
Table 1. An overview of environmental policy instruments.
Policy Instruments
Information/
Persuasion Creating rights Regulation/legal Price-based instruments
• Public
participation • Property
rights • Standards
Permits • Subsidies (Subsidy Reduction)
• Information
disclosure • Tradable
permits • Bans, Zoning • Environmental charges/taxes
• Voluntary
agreements • Tradable
quotas • Public goods • User charges
• Labeling
schemes • Offset
systems • Liability • Deposit-refund systems
Scientific analysis of policy instruments is quite abundant. The
broad synthetic literature includes many books and articles on the selection
of policy instruments of which we will briefly mention just a few. Natural
starting points are the popular books by Baumol and Oates (1988) and
Tietenberg (1990). Xepapadeas (1997) focuses on the underlying economic
theory while others such as Dijkstra (1999) focus more on the political
economy of environmental policy instruments including a good deal of
game theory. Other central works on the selection of instruments include
Bohm and Russell (1997) Russell and Powell (1996) Stavins (2002), Nordic
Council of Ministers (1999) and US EPA (2001). A textbook covering

121
theory and applications in both developing3 and industrialized countries is
Sterner (2002). A recent survey of European environmental taxes is found
in Ekins (1999).
There are thus many possible ways to structure a review of
environmental taxes. In the following we present a model that distinguishes
between countries and firms to lay the foundation for a comparison of
environmental taxes. We then start the analysis with a comparison of
environmental tax revenues including a discussion of the limitations of such
an approach. This is followed by comparisons of tax rates for gasoline,
carbon, sulphur, nitrogen, lead, water effluent and waste.
Taxes as indicators of environmental policy
It is far from straightforward to use environmental taxes as
indicators of environmental policy. To see this, consider a group of
countries, i, and firms, j, with vectors of production q. Cost functions for
environmental damage are d as in (1) where emissions e depend on
production q, and abatement a as in e=e(q, a) and damage is also affected
by ecosystem attributes (including population density, assimilative capacity
etc) N. Abatement costs are reflected in production costs, c, as in (2): (for a
more rigorous treatment see Xepapadeas 1997).
di = di (Ni, Σeij (qij, aij)) (1)
cij = cij (qij, aij) (2)
Consider social planners in each country who maximize welfare Wi,
a function of net profits (P is a vector of prices) minus environmental
damage of emissions, both of which depend on output q, and abatement a.
Note that (1) assumes emissions are “perfectly mixed” so that aggregate
damage depends on total emissions in that country, (Ei = Σj eij),
Wi = Σj{Pqij – cij (qij, aij)} - d{Ni , Σj eij (qij, aij )}. (3)
The necessary first order conditions for each country4 are:
3 There is a specialised literature on environmental policy making in industrialising
countries, see for instance Panayotou (1998) or Wheeler et al (2000).

122
P = c’q + d’ e’q (4)
c’a = - d’ e’a . (5)
Suppose that some countries regulate some sources of pollution by
firm specific emission targets ê. The firm then maximizes profit subject to
the constraint e = ê and the Lagrangean for this situation is:
L = Pq – c(q, a) + λ(ê - e(q, a)). (6)
The corresponding optimality conditions are:
P = c’q + λ e’q (7)
c’a = - λ e’a (8)
ê = e . (9)
The λ are shadow prices of pollution. With perfect information and
an optimal (welfare maximizing) allocation of permits ê, the λ will all be
equal to the country specific marginal level of damage d’ in equations 4-5
ensuring optimal abatement and allocative efficiency. The marginal damage
d’i will be a function of Ni and thus vary from country to country depending
on such factors as population density, assimilative capacity of the
environment and other factors such as the flow of air and water and so forth
see (1). Also the costs of abatement will vary with the number of industries,
their size, age, cost structures and so forth.
Thus the optimal policy instrument, in this case an emission
allowance êi will not generally be the same for the individual countries i.
Since abatement costs vary we will, generally, not even have the same
target level of abatement between firms within the same country.
Notice that the optimal price should internalize damage costs as in
(4). Assume some countries use a vector of taxes T as their (only) policy
instrument. Consider again a representative firm in a representative country
that seeks to maximize its profit after abatement costs and taxes. It solves
Max Pq – c(q, a ) – T e(q, a ), (10)
4 Omitting subindices for countries and firms when not needed.

123
where T is the Pigouvian tax. The necessary and sufficient first order
conditions for a firm with positive output and abatement costs are (dropping
indices):
P = c’q + T e’q (11)
c’a = - T e’a. (12)
Again a comparison with the conditions for optimality 4-5 shows that the
conditions are identical as long as the Pigouvian tax is correctly set at the
level of marginal damages (T = d’). For the reasons stated above, there is no
reason to believe that the optimal tax rates for individual pollutants should
be the same across countries. The target emission or abatement rates are not
the same since damage and abatement costs vary and similarly the
necessary taxes. In addition to the factors already mentioned there are more
aspects that are worthy of mention. First of all some of the damage may be
local and some regional or global. The willingness to internalize the non-
domestic parts may be seen as moves in a complicated international game
and it is far from clear that all countries would play the same strategy in
such games. Secondly market power and other disturbances that affect the
optimal tax rates, including related environmental regulation, may vary
from one country to another. The extent to which countries rely on taxes
versus regulation or other instruments also varies.
In the European Union context one would expect differentiation of T
according to N for local impacts; harmonization of T for pollution with
regional impact and varying levels of differentiation of T for global
pollutants given national and EU policies. Thus Ti cannot be used as a
definitive indicator for environmental ambition since abatement and
damage curves vary, particularly for local pollutants and because the size of
the tax must be judged together with information on other policy
instruments such as regulations. Tax revenues for a country i, Sj Tijeij may
therefore also be hard to interpret as indicators since tax revenues may be
high either because T is high and emissions moderate or because T is quite
low and thus emissions are high. Similarly, low tax revenues may depend
on either an ambitious environmental tax program or its opposite.
We need to recognize that emission levels take a long time to adapt
so that looking at data for one year might be misleading. The highest tax
revenues, for instance, often occur at the moment in time when a policy
switches from being “lax” to being “ambitious” since this gives (for a short

124
period) a combination of high tax rates and high emission levels. Looking at
several years of data on both production, tax and tax revenues does at least
partly alleviate these concerns. It is also important to look at the
composition of the environmental tax revenues: Generally a dominating
share comes from the energy and transport sectors and can (partly) be
interpreted as targeting global pollution. Since tax elasticities are fairly
low5, high revenues from these taxes will be a reasonable reflection of an
ambitious (environmental) tax policy.
An additional factor that complicates comparisons across countries
is that different countries use different policy instruments to deal with
different environmental issues. Some use regulations of some sort, others
use taxes or charges, tradable permits, voluntary agreements, liability,
labeling or other information disclosure strategies. It is not at all uncommon
for a country to combine several instruments. Although it is not necessarily
easy to judge the combined effect of a tax and a regulation it is clear that
when a small tax is combined with strong regulations the tax is non-binding
and such a tax is mainly fiscal and can not easily be compared with a
classical Pigouvian tax. This means that the optimal level of say abatement
requirements and charges would be different from the corresponding
optimal levels if either instrument were used in isolation.
An international comparison of environmental tax revenues
A practical difficulty in comparing environmental tax revenues is
deciding what to include. A striking example is the USA where the IRS has
an official category “Environmental taxes” which covered only 130 million
USD in 2001 mainly from ozone depleting substances. By this definition
environmental taxes (IRS 2002) have fallen dramatically since 1995 (down
from about 1.5 billion USD) and now only constitute a minute .0007% of
US GDP. However total US excise taxes in 2000 were 54 billion USD of
which around 30 billion USD were taxes on gasoline, diesel and other fuels.
Other large items were alcohol, tobacco, air traffic and telephone. Some
would argue that tobacco taxes are supposed to internalize factors such as
the health damages of secondary smoking and that this is a typical inter-
5 If price elasticities are around –0.7 and the tax share in prices is 50% then the tax
elasticity will only be around –0.35. For sulphur emissions however, elasticities are high
and this relationship would not hold.

125
personal, environmental externality. In the OECD statistics that we use here
(OECD 2002), tobacco taxes are not included but most fuel and
transportation-related taxes are – and in fact they form, by far, the largest
share of what is called environmental taxes here.
Using these statistics, we can compare these energy and
environmentally related taxes as a share of GDP. Figure 1 shows that
Europe does in fact stand out as having consistently higher levels of
environmental tax revenue than the USA. In 2000, the US had a share of
below 1% compared to an average of just under 3% for the European part of
the OECD.
Figure 1. Environmental revenues as percent of GDP in Europe and
some other OECD countries
0,00
0,50
1,00
1,50
2,00
2,50
3,00
3,50
4,00
4,50
5,00
Denmark
Netherlands
Turkey
Finland
United Kindom
Portugal
Iceland
Norway
Luxembourg
Hungary
Czech Republic
Sweden
Ireland
Austria
Italy
Greece
Germany
France
Slovak Republic
Switzerland
Belgium
Spain
Poland
Korea
Australia
Japan
Mexico
New Zealand
Canada
United States
Source: OECD, 2002.
Most European countries are in fact clustered around 3%, although
the range is from 2% for Spain and Poland to almost 5% for Denmark,
while the non-European OECD are on the whole lower (with the exception
of Korea). It is noteworthy that there is no particularly clear pattern within
Europe: One cannot say anything in particular about such groups of
countries as the non-EU countries (Turkey, Iceland, Hungary, Czech
Republic, Slovakia, Switzerland and Poland) or the Southern European

126
countries such as Greece, Italy, Spain, Portugal and Turkey. Some
observers might believe that these countries would on the whole have a less
ambitious environmental profile than for instance the Netherlands, Germany
and Scandinavia that are often considered as being influenced by “green
politics”, but this does not come out in any clear way in the ranking of
countries here (as we would indeed not expect from our earlier discussion).
Following our earlier discussion, we hesitate to comment on environmental
tax revenues for a single year, since a high value might be transitory as a
new tax will lead to erosion of the tax base due to abatement – which may
well be a beneficial and intended outcome. The OECD data covers
environmental revenues for 7 years 1994-2000. We have looked at the
variance in environmental tax revenues and time trends in the data and
found that environmental tax revenues are in fact rather stable in most
countries: The average standard deviation for all the European countries is
0.26 and most of this variation is due to some smaller countries, the average
standard deviation for the major EU economies of France, Germany, UK,
and Italy is only 0.07. This shows that the environmental revenues have on
the whole been very stable and that the figures that we are discussing from
2000 are indeed representative for each respective country. Over the whole
period, the European average has also been virtually constant at 2.8%. Few
countries have any large and clear trend in environmental tax revenues
although a few countries have some variation – most notably Hungary
which is by far the most erratic, its environmental tax revenues rise from 4
to 5 and then fall back under 3% of GDP.
As mentioned, interpretation must be cautious: low tax revenues
may be due to a low tax level (as gasoline taxes in the USA) or low
emissions (e.g. consumption of gasoline per capita as in Japan).
Furthermore the countries import and export goods, services and emissions
in complex ways. Vehicle and fuel taxes are an important part of these tax
data and some countries have a much larger fraction of transit traffic than
others. Some countries such as Luxembourg have even been notorious for
attracting motorists from neighboring countries by having low fuel taxes.
These motorists thus go to Luxembourg to fuel their cars and thus
Luxembourg combines very low tax rates with very high tax revenues.
An aspect that is worth mention is the fact that total tax revenues
varies quite strongly as a share of GDP. In the US, Canada and Australia the

127
0,00
0,50
1,00
1,50
2,00
2,50
3,00
3,50
4,00
4,50
5,00
0 10 20 30 40 50
Total tax share in GDP
Environm tax share in GDP
US
DK
NZ
BEL
TURK
tax share is around 20%. There are a few European countries with taxes that
are in the same range but most of them are actually in the range of 30-40%
irrespective of whether they belong to the North or South of Europe or
whether they belong to the former “Eastern” or Western sphere, see table 2.
It is, however, again, not easy to say what the effect of the share of taxes in
GDP should be on the share of environmental taxes in GDP. One might say
that with low taxes in general as a share of GDP one would expect any
particular tax to be a small share of GDP. This line of reasoning appears to
assume that there is some logic by which environmental taxes should be
some proportion of total taxes. One such reason would be if the motivation
for environmental taxes were not primarily to correct for environmental
externalities but rather exploit a tax base. If instead we were to believe that
environmental taxes were decided simply on the logic of what is needed in
order to attain a livable ambient quality, then one would simply expect
environmental taxes to be a larger share of total taxes in countries with a
bad environmental situation.
Table 2. Taxes as share of GDP
% range Countries
20-22 USA, Canada, Greece, Switzerland, Turkey, Australia,
26-29 Germany, Iceland, Spain, Finland. Poland, New Zealand
31-33 Ireland, Portugal, Slovakia, Denmark, Czech. R, Hungary
34-35 Norway, Austria, Sweden, UK
39-43 France, Italy, Luxembourg, Netherlands, Belgium
Source: World Development Report Indicators 2002, Figures are for 2000 or latest available year.
Countries are in order by ascending percentage.
Figure 2. Environmental and total tax shares in GDP
As can be seen by
comparing table 2 and figure
2, there is a certain amount
of correspondence in the
sense that the non-European
OECD countries like the US,
Canada and Australia do
appear to rely much less on
taxes in general and as a
means to improving the

128
environment. Within Europe however the correlation is less clear and there
are some notable exceptions: For instance France and Belgium with large
public sectors and a high tax share of GDP still have a moderate or small
share of environmental taxes. At the other end of the scale are Denmark,
Finland and particularly Greece with small total tax shares in GDP but large
shares of environmental taxes. The correlation coefficient between the two
tax shares is 0.47 and the countries clearly below the regression line for
environmental taxes in figure 2 are the US, N Zealand and Belgium while
Denmark and Turkey are clearly above.
A comparison of gasoline tax rates
When we interpret the environmental tax revenues above we should
bear in mind that most of these revenues are from the transportation sector:
fuel taxes and vehicle or road related taxes. The latter are more complicated
to compare since they are typically defined in different ways in different
countries (vehicle taxes may be levied at purchase, for imports, or yearly
and they may be differentiated by distinct characteristics such as weight or
horsepower, making inter-country comparison difficult). We will therefore
in this section concentrate on the comparison of gasoline taxes.
In table 3, we have chosen to show the average rate of taxation on
gasoline. This indicator is a weighted average reflecting the varying
composition of fuels with different octane (premium and regular) in the
proportions actually used in each country. It is expressed in international
cents converted by purchasing power parity. This is a useful way to provide
an indicator of the actual burden the tax places on the representative
motorist. This makes it attractive as an indicator of the intensity or strength
of a policy instrument. The reader should be aware, however, that the
comparisons would be different if we had used market exchange rates. The
difference for most high-income countries is small but for some low-income
countries it is more substantial. In our sample of countries, the US would
have had a somewhat higher tax and the Eastern European countries
considerably lower values.
As we can see in table 3 the average for Western Europe is 67 cents
per liter, which is clearly high compared to all the non-European countries
and most notably with respect to the USA, countries such as Japan and
Australia being intermediate. Within Western Europe variation is limited

129
although it can still be quite significant considering that many are actually
neighboring countries. Looking at gasoline prices and taxes in the 1970s
and 80s there was a wide divergence within Europe and several important
countries such as Germany and the UK had low taxes and prices, (Angelier
and Sterner, 1990). Looking at the figures today, all the major EU
economies, Germany, France, Italy, UK, Belgium, Holland now have high
taxes and are fairly well harmonized at quite a high level (70-86 cents/l).
This reflects a fairly long and conscious effort in countries such as the UK
where the “fuel tax escalator” has implied a pre-announced, long-run
program of fuel tax increases. Some of the smaller and more peripheral
economies now have somewhat lower values. This actually includes some
of the countries that earlier were renowned for having high taxes such as
Denmark. When comparing these policies one should however keep in mind
that Denmark has a set of very draconian vehicle taxes so that motoring as a
whole is very heavily taxed in that country.
Table 3 Gasoline taxes in cents/liter in selected countries
Western European Gas Tax Eastern European Gas tax
Italy 86 Hungary 131
UK 86 Czech republic 120
Netherlands 85 Poland 97
France 76 Average 116
Belgium 75
Finland 74 Non European
Germany 71 Japan 38
Norway 67 Australia 34
Portugal 67 New Zealand 32
Sweden 66 Canada 25
Denmark 64 Mexico 11
Spain 63 USA 10
Greece 59 Average 25
Austria 58
Ireland 56
Luxembourg 46
Switzerland 45
Average 67
Source: IEA, 2000
The main outliers are Luxemburg and Switzerland. Luxemburg is, as
mentioned, a very special case where they actually appear to be consciously
profiteering on the tax difference and attracting motorists from surrounding

130
countries to fuel their cars thereby giving Luxemburg high revenues
through a low tax rate. This cannot be interpreted as environmental or non-
environmental policy; it appears to be a case of very simple local tax
maximization without regard for any other principles. The case of
Switzerland is quite distinct and they are currently using regulations and
advanced road pricing to deter the transit traffic that is a considerable local
environmental problem as well as causing considerable costs to this alp
country where road maintenance is problematic.
In other countries such as Austria and Ireland that have low gasoline
tax rates, the fairly high (or intermediate) environmental tax revenues are
explained by the fact that these countries have sizeable road or vehicle
taxes. In countries such as Italy and the UK with high gasoline taxes these
other road and vehicle taxes appear to be relatively lower.
The most eye-catching and perhaps politically interesting
comparison is that between the US and Europe. Fuel taxes are very small in
the US compared to the European average – and even to the lowest of tax
rates in Europe. This is clearly correlated to higher fuel use. In the US
annual gas consumption per capita is at 1300 liters. The only European
country with such levels is Luxembourg while most countries are less than a
third (Germany 360, France 240, UK 360, Italy 300); see also figures 3 and
4. Similar differences appear to apply to a number of other forms of fossil
energy since the aggregate emissions of carbon dioxide per capita are
considerably higher in the US, 5.5 tons, compared to 2.5 tons in Europe. It
seems clear that if the EU had followed a similar tax policy to that in the
US, then aggregate carbon emissions would have been substantially higher.
Figure 3-4. Price and Demand for gasoline in selected countries
0
100
200
300
400
500
1960 1970 1980 1990 2000
Price of gasoline
0
200
400
600
800
1000
1200
1400
1600
1960 1980 2000
Gasoline per capita
USA
CANADA
AUSTRALIA
UK
ITALY

131
Sterner (2002)
There is little evidence that the variation in fuel taxes would be due
to local factors, such as ecosystem attributes, N, introduced in our model.
Given the complexity in vehicle related externalities, there is actually little
scope for fuel taxes to act as efficient Pigou taxes. Environmental effects
are usually either local or international. Neither coincides well with national
territory and thus they are not necessarily well dealt with by national taxes.
Still, it is interesting to note that Parry and Small (2002) estimate optimal
gasoline tax to be 27 cents per liter in the US and 35 cents per liter in the
United Kingdom. This would imply that the US is well below and the UK
well above its optimal gasoline tax rates.
In order to cast some light on the importance of the level of gasoline taxes
for the total share of environmental tax revenues in GDP, we plot one
against the other as in figure 5. The correlation is 0.6 and the main outlier is
Denmark which has an environmental tax share of almost 5% but moderate
gas taxes. This is largely explained by other high taxes on vehicles and
energy. The four countries with very high gas taxes but low or moderate
shares are eastern European countries with low gasoline consumption
levels.
Figure 5. The tax on gasoline and total share of environmental taxes
0,00
0,50
1,00
1,50
2,00
2,50
3,00
3,50
4,00
4,50
5,00
020 40 60 80 100 120 140
Gasoline Tax level
Share of Envir. taxes in total taxes
Sources: OECD 2002 and IEA 2000.
US
Czeck. Hung
Turkey
Poland
Denmark

132
A Comparison of other environmental taxes
Although fuel taxes have the largest impact on revenues, there are a
large number of other taxes or fees that contribute to abatement at the
regional, national and local levels. The diversity in this area is quite
considerable and we cannot cover all countries or taxes systematically.
Instead we focus on some striking examples at different levels.
Carbon taxes
We might expect greater harmonization between countries in taxes
targeting global pollution since variations in damage cost are smaller.
Greenhouse gas emissions are of course the most conspicuous such
pollutant. Although fuel taxes have an effect on carbon emissions they only
affect the transportation sector which, from the viewpoint of climate policy
is very incomplete. There are at present really no countries that have true
broad-based carbon taxes. Sweden and Norway have sizeable carbon taxes
for some sectors6. These are regularly updated and the Swedish tax level is
now 0.63 SEK/kg CO2 (October 2002). This corresponds to over 200
USD/ton of carbon7, which is clearly high compared to the levels that have
been discussed internationally. There are however numerous exceptions to
the tax: Production of electricity is exempted; non-energy use (for instance
coal in the steel industry) is exempted; air freight and international ship
bunkers are exempted; industrial use has a 65% reduction as well as a cap
of 1.2% of total revenue that is particularly important for the energy-
intensive industries. There has recently been a parliamentary investigation
that has revisited this difficult and contentious area, (SOU 2003:38). One of
the problems they highlight is that EU legislation makes it illegal for
Sweden to single out some sectors of business for preferential treatment.
This makes it impossible to treat “energy-intensive” industries from other
firms. This also applies to the industries within the energy sector (power
plants etc) that used to be national and often publicly owned. Deregulation
of this industry and EU membership has thus introduced new layers of
complexity here. Their suggestions include a general lowering of energy
6 Denmark, Finland, Italy and the Netherlands also have some form of carbon taxes, see
Stavins (2002).
7 At a current exchange rate of 8.5 SEK/USD this gives just over 74 USD/ton CO2 and a
conversion of 44/12 from tons of CO2 to tons of C gives 270 USD/ton C.

133
taxes for business and an overall cap of 0,7% of total revenue (unless this
implies that the tax falls below the EU minimum).
In spite of the tax reliefs, the tax has had some effects: Since the
creation of a Swedish carbon tax in 1990 the annual use of biomass,
particularly forest residues, in district heating has risen very rapidly and
biomass is clearly dominant in this sector today. However, with the current
formulation of this tax there is no incentive for biomass in transportation
fuels. An even more poignant example is the fact that in Norway, thanks to
the carbon tax, large-scale storage of CO2 has started in industrial aquifers.
On the Sleipner Vest gas field, CO2 rich natural gas is being stripped for its
CO2 and this is being injected into a saline aquifer at about 1000 m depth.
Such techniques are sometimes used to enhance oil recovery. It is however
unique that this is being done for environmental reasons and as a result of
the carbon tax. The captured and injected quantity is roughly 1 million tons
CO2 per year, which corresponds to a non-trivial 3% of Norway’s total CO2
emissions.
Sulphur taxes
Acid rain has been one of the most contentious forms of pollution in
Europe over the last decades since it is regional in impact, emissions are
very skewed depending on dependence on coal and oil and the sensitivity to
the acidity depends on the sensitivity of soils (the “N” in our model). In
addition to regulating contents, a number of European countries use
differentiated energy or fuel taxation to encourage a reduction in the use of
sulphur. The Scandinavian countries Sweden, Norway and Denmark have
high taxes (3400, 2100 and 1300 USD/ton respectively) in comparison to
which the taxes in Italy, France, Switzerland, Spain (actually a regional tax
in Galicia) and Finland (only diesel) are quite small (all below 50 USD/ton).
These tax levels can be compared to the US permit prices, which tend to be
in the range of 100-200 USD/ton. While all taxes should, in principle, have
some effect it is unlikely to be significant when they are so low. Even for
the high-tax countries, the effect is hard to disentangle from the effect of
other concurrent policies. For Sweden, SEPA (1997) judges that the tax was
responsible for 30% of the reduction between 1989 and 1995. Notice that a
large part of the reduction in sulphur use in Sweden predates the tax. In
1970, emissions were over 900 ktons, by 1980 it was reduced to 500 ktons
and in 1990 it had dropped to 136 ktons. It is only the last, but perhaps most

134
difficult, reduction down to 66 ktons in 1999 that has been (partly) caused
by the tax. The tax applies to sulphur actually emitted but is levied on fuels
based on their sulphur content. There is then a rebate for sulphur removed
by filters etc. Hammar and Löfgren (2001) find that the sulphur tax appears
to have been important in reducing the actual sulphur content of fuel oil
below the legal limits. Sweden had in some years stricter limits than the EU,
for instance 0.2% for light fuel oil already in 1976 but actual sulphur
content was lower still. In 1991 the actual figure was 0.08% and in 1994
0.058% when the legal limit was still 0.2% and this must be mainly due to
the tax. Sulphur emission rates have fallen dramatically and are today very
low in the high tax countries compared to other countries. Emission rates of
8 kg S/cap in Sweden and Norway are less than half the corresponding
value in the UK, France or Germany and about one eighth of US figures.
The reason for the strong policies in Scandinavia is that local ecosystems
are very sensitive to acidification; see further Sterner (2002).
One more country deserves special mention when it comes to
sulphur taxation and that is Poland. It is presumably the Eastern European
country that most consistently has applied quite tough economic
instruments in order to reduce pollution and to finance the abatement of
pollution (the fees are not strictly fiscal taxes but paid into funds). Poland
has a sulphur tax of over 80 USD/ton which also appears to be collected in a
very consistent manner which may not apply to all the other Eastern
European countries.8 Considering the low-income level and high pollution
levels this tax must clearly be considered a sizeable tax, higher than in
many West European countries and almost on par with permit prices in the
US. The tax is thought of as successful both through its incentive effect and
through the funds for abatement investments that it generates.
Nitrogen taxes and charges
Nitrogen oxides have a very similar acidifying effect as sulphur but
are much harder to measure. Most of the NOx is formed by the effect of
high temperature on atmospheric nitrogen and does not (like sulphur) come
from an impurity in the fuel. This implies that NOx emission cannot be
8 According to Stavins (2002) the following sulphur taxes in USD/ton applied to a
selection of Eastern European countries: Bulgaria 0.02, Czech R. 30 for permitted and 45
for excess emissions, Hungary 2.4, Estonia 2 for permitted and 95 for excess emissions,
Lithuania 46 and Slovakia 33 USD/ton.

135
easily predicted and thus complicated and expensive monitoring equipment
is required. As in the case of sulphur, the US uses permits and several
European countries have tax schemes. According to Stavins (2002)
Bulgaria, Czechoslovakia, Estonia, France, Hungary, Italy, Lithuania,
Poland, Slovakia all have some form of tax or fee. For the major countries
the level is around 30-100 USD/ton and most typically they are levied on
the basis of estimated pollution (which implies little incentive for real
abatement). However the funds are often used to promote abatement, which
of course should lead to some reduction (although generally not in an
effective way).
The only country that appears to base payments exclusively on
measured emissions and that also charges a very high fee is Sweden, where
a refunded emission payment of 5 000 USD/ton is used, see Sterner and
Höglund (2000). The design of the charge has been unique. It is on the one
hand a very high charge - more than 200 times higher than the French
charge. On the other hand its proceeds are distributed back to the polluting
companies in relation to the amount of energy produced by the specific
plant. This means that the polluting industry as a whole does not pay
anything to society – and it is presumably this fact that has made the charge
politically feasible. The design mechanism was partly chosen due to the fact
that only large combustion plants are obliged to pay the charge. This was
based partly on the high costs of metering which (together with abatement
costs) were considered unreasonable for smaller plants. Now if a tax were
applied to only a subsection of some industry then this subsection would be
unfairly disadvantaged compared to other firms in the same industry. In this
case, if the tax were applied only to the large plants, large companies would
have an incentive to set up several small combustion plants instead of one
big one and this is typically not desirable (from any viewpoint including
emissions of NOx and other pollutants). As the system has developed and
turned out to be effective, (emission rates have fallen by about 40% in six
years) costs for abatement and metering have fallen and the criterion for
inclusion has been lowered twice: in 1996 plants producing at least 40 GWh
useful energy per year were included and in 1997 the boundary was lowered
to 25 GWh. In 1998, 400 plants were subject to the charge, compared to 200
plants in 1992.

136
Taxes related to lead content of fuels
From the viewpoint of the local environment and particularly human
health, the most important issue is of course not at all fossil carbon but
various toxic substances found in fuels or formed upon combustion. One of
the most crucial environmental and health improvements has been the
elimination of lead from gasoline. The US was the first country to phase out
lead from gasoline in the early to mid 1980s. They used tradable permits for
the lead content of gasoline. Other countries have not copied this scheme
although most have or are phasing out lead from gasoline, in many cases a
decade or even two after the US. Scandinavia, Germany and Austria have
already phased out lead completely but some other European countries lag
behind. The EU Council of Ministers banned leaded fuel use after 2000
(with some exemptions to 2005).
Most European countries chose a mixture of policy instruments
including fuel specifications and mandatory catalysts for new cars and in
some cases subsidizing them on older cars (e.g. Sweden). This created a
demand for unleaded gas since catalysts become defunct if used with leaded
gasoline. To speed up the transition many countries then used differential
taxation of leaded and unleaded gasoline or other instruments such as
subsidies. The tax differential between leaded and unleaded in Sweden is a
case in point. Sweden raised the tax differential on leaded gasoline
drastically to about 8 US cents/liter. With a tax wedge much larger than the
production cost differential, both retailers and consumers were anxious to
switch. The same occurred in several countries such as Denmark where
unleaded gas was about 7% cheaper resulting in a 50% market share even
before the cars with catalysts had any market share (and 100% unleaded
when these cars had only 25%). In this case the lower tax appears to have
had a large effect but this varied between countries. See Löfgren and
Hammar (2000) for an econometric analysis of the phase out of lead in the
EU showing the importance of tax differentials, income levels and catalysts.
The general conclusion is that the phase out of lead was speedy because
politicians used powerful policy instruments. According to Kågeson (1993)
the instruments used in Europe corresponded to a shadow price of 200-500
USD/kg of lead. The environmental tax revenues from these taxes are
impossible to isolate since they were intertwined with fuel taxes. Frequently
they were formulated as rebates from gas taxes for lead free fuel.
Furthermore the “revenues” would have been very transitory as the tax on

137
lead was often low at first and then raised as the share of leaded fuel
dwindled.
Other local environmental taxes
In many European countries, (notably France, Germany and the
Netherlands) water was one of the earliest examples of a resource for which
policy makers have used something of an ecosystem approach (the water
basins often being self-governing). It has also been fairly common to use
some form of taxation or pricing to cover numerous aspects such as
scarcity, supply cost and the costs caused by effluents. The diversity of
these schemes is however such that it is hard to provide any numerical
comparisons. Typically the fees are refunded to the sector to provide
finance for water treatment plants or similar use, see Stavins (2002) and the
references therein.
Several European (as well as non-European) countries have waste
taxes or differentiated charges for municipal waste management. In several
countries the driving force is the lack of space (and/or of local acceptance)
for landfills. Consequently a number of countries have particular taxes on
landfills (sometimes combined with other policies concerning incineration).
According to Stavins (2002), such fees may be found in the UK, Denmark,
Netherlands and Finland at rates that range from USD 3/ton for inert waste
to several hundred USD for hazardous waste, with typical figures of 15-50
USD/ton of typical landfill waste. Estonia, Latvia and Poland also have
such fees although with somewhat lower rates.
In several countries such as Sweden, a number of municipalities are
now charging for waste on a “per kg” or “per bag” basis. Sterner and
Bartelings (1999) analyze waste disposal, recycling and composting in a
municipality in southwest Sweden, Varberg, which in 1994 introduced a
weight-based billing system for household waste charging 1 SEK/kg (.18
USD/kg) of waste and at the same time, recycling centers were set up and a
“green shopping” campaign was launched. This led to a significant
reduction in waste collected (35% within a couple of years). The article
shows however, that economic incentives, although important, are not the
only driving force behind the observed reduction in municipal waste: Given
the proper infrastructure that facilitates recycling, people are willing to

138
invest more time than can be motivated purely by savings on their waste
management bill!
Waste management fees are often used in combination with specific
taxes on such products as batteries and tires that have particular
environmental effects. In many cases these may be combined with deposit-
refund systems, user charge systems or there may be other means for
earmarking the revenues for clean up, abatement or waste management.
Also information campaigns, eco-labeling, liability schemes and even
permit trading (such as the UK system for trade in packaging waste
recovery notes) are being used in this area. Taxes are applied to a broad
range of items such as chlorinated solvents in Denmark and Norway to
batteries in Sweden, non-refillable containers in Sweden and Finland,
ozone-depleting substances in Denmark, VOCs in Switzerland and France.
In most of these cases the tax revenues generated by these schemes were
insignificant.
Concluding comments
The broad range of this subject makes it hard to produce firm and
encompassing conclusions regarding the role of environmental taxes in the
overall environmental policy mix in different countries. Still, environmental
taxation does appear to play a somewhat more prominent role in Europe
than in other continents. This is partly but very imperfectly reflected in
higher environmental tax revenue as a share of GDP. Environmental taxes
should however be separated into two categories. In the first we have taxes
related to energy and transportation and motivated at least partly by global
concerns such as climate change. In this category we have considerable
revenues already and in principle an even larger potential if taxation were to
be used as the primary way of achieving global goals for climate. The
elasticity of energy use to price is limited but not insignificant which means
that high taxes will be needed and large revenues generated in order for
ambitious goals to be achieved. The global nature of this environmental
problem causes numerous problems of coordination between countries that
are largely motivated by concerns of industrial competitiveness. A second
concern may be if the incidence of the taxes is regressive or hurts politically
influential groups. The European experience shows that a number of steps
have been taken to reduce the competitive and regressive effects of taxation

139
by modifying the tax schedules or through the use of the tax proceeds, see
further Ekins (1999), Sterner (2002) and Stavins (2002).
The other category of environmental taxes is somewhat less visible
both to ministries of finance and thus to the political world and to the
researcher. They generate fairly insignificant revenue. On the other hand,
one of the reasons for the low revenue is precisely the tax base erosion,
which from a fiscal point of view may be unsatisfactory but from an
environmental viewpoint is very desirable. The diversity of these taxes is
considerable, they are applied to various chemical pollutants and their
administration ranges from pure taxes to product charges, local user fees,
and noncompliance fees. Furthermore, they include elements of refunding
and other aspects that make them hard to compare. On the whole, however,
there is considerable evidence that they are sometimes efficient instruments
for environmental improvement.
References
Angelier, Jean-Pierre, and Thomas Sterner. 1990. "Tax harmonization for
petroleum products in the EC." Energy Policy. July 1990.
Baumol, William. J. and Wallace E. Oates. 1988. The Theory of
Environmental Policy. Cambridge: Cambridge University Press.
Bohm, Peter and C.S. Russell. 1997. “Comparative Analysis of Alternative
Policy Instruments.” In The Economics of Environmental Protection:
Theory and Demand Revelation. Ed. Peter Bohm. Cheltenham: Edward
Elgar, pp. 48-113.
Dijkstra, B.R. 1999. The Political Economy of Environmental Policy. A
Public Choice Approach to Market Instruments. Cheltenham: Edward
Elgar.
Ekins, Paul. 1999. “European Environmental Taxes and Charges: Recent
Experience, Issues and Trends.” Ecological Economics. 31(1): pp. 39-62.
Hammar, Henrik and Åsa Löfgren. 2001. “The Determinants of Sulfur
Emissions from Oil Consumption in Swedish Manufacturing Industry,
1976-1995.” The Energy Journal. 22(2): pp. 107-126.

140
International Energy Agency. 2000. “Energy Prices and Taxes.” Quarterly
statistics 2000, Paris: OECD.
IRS. 2002. (US Internal Revenue Service) “Table 21--Federal Excise Taxes
Reported to or Collected by the Internal Revenue Service.”
http://www.irs.gov/taxstats/index.html. (November 10, 2002).
Kågeson, Per. 1993. “Getting the Prices Right. A European Scheme for
Making Transport Pay its True Costs.” Transport and Environment 93(6).
Löfgren, Åsa and Henrik Hammar. 2000). ‘The phase-out of leaded
gasoline in the EU: a successful failure?.” Transportation Research, Part D
Transport and Environment. (5): pp. 419-431.
Nordic Council of Ministers. 1999. “The Use of Economic Instruments in
Nordic Environmental Policy 1997-98.” TemaNord Copenhagen. 1999:524.
Oates, Wallace E. 1998. ”Environmental Policy in the European
Community: Harmonization or National Standards?” Empirica. (25): pp. 1-
13.
OECD. 2002. “Environmentally Related Taxes database”.
http://www.oecd.org/oecd/pages/home/displaygeneral/0,3380,EN-
document-471-14-no-1-3016-0,FF.html (November 10, 2002).
Panayotou, Theodore. 1998. Instruments of Change – Motivating and
Financing Sustainable Development. UNEP.
Parry, Ian W.H. and Kenneth A. Small. 2002. “Does Britain or the United
States Have the Right Gasoline Tax?” Discussion Paper 02-12. Resources
for the Future. Washington D.C.
Russell, C.S. and P.T. Powell. 1996. Choosing Environmental Policy Tools:
Theoretical Cautions and Practical Considerations. Inter-American
Development Bank.
Schlegelmilch, Kai. 1998. Energy Taxation in the EU and some member
States:Looking for Opportunities Ahead. Wuppertal Institute for Climate,
Environment and Energy.

141
SEPA. 1997. (Swedish Environmental Protection Agency) “Environmental
Taxes in Sweden – Economic Instruments of Environmental Policy.”
Report 4745, Stockholm.
SOU 2003:38, “Svåra Skatter.” Green Paper by the Swedish Committee on
Energy Taxation of the Business Sector. Statens Offentliga Utredningar.
Stockholm.
Stavins, Robert N. 2002. "Experience with Market-Based Environmental
Policy Instruments." Handbook of Environmental Economics. Eds. Karl-
Göran Mäler and Jeffrey Vincent. Amsterdam: Elsevier Science.
Sterner, Thomas. 2002. Policy Instruments for Environmental and Natural
Resource Management. RFF press in collaboration with the World Bank
and Sida: Washington DC.
Sterner, Thomas. and Heleen Bartelings. 1999. “Household Waste
Management in a Swedish Municipality.” Environmental and Resource
Economics. June 13(4).
Sterner, Thomas and Lena Höglund. 2000. “Output-Based Refunding of
Emission Payments: Theory, Distribution of Costs, and International
Experience.” Discussion Paper 00-29. Resources for the Future:
Washington D.C.
Tietenberg, Thomas H. 1990. “Economic Instruments for Environmental
Regulation.” Oxford Review of Economic Policy. 6(1): pp. 17-33.
US EPA. 2001. “The United States Experience with Economic Incentives
for Protecting the Environment.” Report EPA-240-R-01-001. Washington
DC.
Wheeler, David et al. 2000. “Greening Industry: New Roles for
Communities, Markets and Governments.” World Bank Policy Research
Report. New York: Oxford University Press.
World Development Report Indicators. 2002. World Bank: Washington DC.

142
Xepapadeas, Anastasios. 1997. Advanced Principles in Environmental
Policy. Edward Elgar.