Energy and Environmental Taxation: Theory and Practice within the EU

Transcription

1 Energy and Environmental Taxation: Theory and Practice within the EU An IREF report by Miroslav Zajíček [1], Pierre Garello [2], Markéta Grušáková [3], Karel Zeman [4] Abstract: Environmental tax reforms have a history of almost two decades and were viewed as a way to the better world the double-dividend theory predicted. Much political capital has been invested in policies leading to environmental tax reforms on European and national levels from 1992 (the year of the first EU-wide energy and carbon tax proposal) till today. We compare shares of environmental taxes on GDP and overall tax revenues in the EU from 1995 till 2010 to identify the real impact of such efforts. Contrary to rhetoric, we find that environmental taxes importance in fact declined in the last decade in most European countries with very few exceptions (notably Denmark). We also identify reasons for such a surprising development. Among these, rising energy prices (due to other regulatory policies) and introduction of new environmental and economic policies such as tradable permits schemes or feed-in-tariff schemes for promotion of renewable energy sources are the most important. JEL Classification: H23, H39, D62, Q58 Keywords: Ecotax, Environmental Tax, Externalities, Pigouvian Tax, Redistribution, Tradable Permits

2 List of Abbreviations: CO 2 EC ET(s) ETR ETS EU EU 15 EU 12 EU SDS ITR GDP GHGs OECD PURPA RES USA Carbon dioxide emissions European Commission Environmental tax(es) Environmental Tax Reform Emission Trading Scheme European Union Old Member States New Member States EU Sustainable Development Strategy Implicit tax rate Gross Domestic Product Greenhouse Gases Organization for Economic Cooperation and Development Public Utility Regulatory Policies Act Renewable energy sources United States of America

3 I. Introduction Environmental Tax Reforms (ETRs) gained increasing political momentum in the 90s [5]. Their basic idea was to shift the tax burden from labour and capital towards the use of natural resources and of (supposedly) environmentally harmful activities. These ideas were clearly stated and summarized in the White Paper on Growth, Competitiveness and Employment (1993) published by the European Commission and later elaborated in a series of documents; the Sustainable Development Strategy (2001). The asserted rationale behind the reform was labelled in the political arena and also sometimes in economic literature as the double dividend theory. The double dividend theory was based on the idea that it is advantageous for the economy to change the structure of overall taxation by shifting the taxation from taxing "goods" (work, savings, investment) towards taxing "bads" (such as pollution, congestion, and later carbon). [6] By making such a shift, it was believed that governments could achieve their environmental goals at no or minimal costs by simply shifting weights among existing taxes or by creating new ones. In other words, politicians came to believe they held the possibility of a free lunch and the prospect of a zero-cost environmental policy became irresistible. This alone should have raised some suspicion since there is no such thing as a free lunch in real life, but surprisingly enough the double dividend theory received at the time the support of several prominent experts. To cite just a few: Irwin and Liroff (1974), Pearce (1991) he actually introduced the terminology, Weizsäcker and Jesinghaus (1992) and Repetto et. al. (1992). However, and not surprisingly, those claims sparked off an academic debate and some assumptions of the double dividend theory were soon undermined (Bovenberg and de Mooij, 1994; Goulder, 1995; Parry, 1994) [7]. Still, the debate is far from over and in some sense there has been a kind of resurrection of the double dividend theory in past years although with a slight different emphasis on market instruments such as emission trading schemes as opposed to direct taxation of carbon [8]. Environmental Tax Reforms (ETRs) gained increasing political momentum in the 90s. The asserted rationale behind the reform was labelled in the political arena and also sometimes in economic literature as the double dividend theory. The double dividend theory was based on the idea that it is advantageous for the economy to change the structure of overall taxation by shifting the taxation from taxing "goods" (work, savings, investment) towards taxing "bads" (such as pollution, congestion, and later carbon). 1

4 politicians came to believe they held the possibility of a free lunch and the prospect of a zero-cost environmental policy became irresistible. In the first half of the 90s, the share of environmental taxes on total tax revenues rose by more than ten percent. But the political machine was already grinding. In the first half of the 90s, the share of environmental taxes on total tax revenues rose by more than ten percent. Thus, in 1995, one could easily expect, at least judging from the political rhetoric, that within a decade or two, the whole tax system which evolved after World War II would be turned upside down regardless of academic merits of the double dividend theory. Is it what we observe today, more than fifteen years later? The answer is clearly no. As we will see, we live in a world that is completely different from what the proponents of the double dividend theory suggested; indeed, a world quite similar to the one we were in the mid 90s at least from the point of view of the structure of taxes levied upon us. The report starts with a short overview of the legal framework for environmental taxes within the EU, and then continues with the classification and description of environmental taxes used currently by EU member states. Then we provide an overview of the role of environmental taxes from 1995 to 2008 as parts of the general tax policies both for old member states (EU 15) and new member states (EU 12) and explore the outcomes of environmental tax reforms in European countries. In the next section we explain why the role of environmental taxes decreased within the last decade. We conclude discussing whether or not the expectations vested in the introduction of environmental taxes and environmental tax reforms were justified. II. European Legal Framework Prior to the 2003 Energy taxation directive (see below for details), the only EU legislation setting minimum levels for the taxation of energy products was the Mineral Oils Directive dealing with oil products used for transport or heating and natural gas used for heating. In 1992 the system for a harmonized taxation of mineral oils was established via two directives: Directive 92/81/EC (harmonized structure of excise duties on mineral oils) and Directive 92/82/EC (harmonized tax rates of excise duties on mineral oils). 2

5 As has been the case for many pieces of European legislation, the Directives left room for member states to adopt different exemptions and tax rate reductions, which resulted in more than 100 special provisions. The first proposals to promote the use of environmental taxation schemes submitted by the European Commission predate even the publication of the 1993 White paper [9]. The EC proposed the first EU-wide energy and carbon tax in 1992 [10]. While all Member States agreed on the principle of taxation as an instrument to combat climate change and politicians acclaimed the double dividend theory, the negotiations never resulted in any agreement. The proposal was rejected in 1994 [11]. Therefore, the EC changed tactic and a new scheme was introduced in 1997 [12] that was primarily based, not on environmental considerations, but instead on the strengthening of the internal market. In the own words of the Commission: The aim was now no longer to introduce a new totally harmonized EU CO 2 /energy tax but, more pragmatically, to extend and improve the existing framework for the Member States taxation of mineral oils to cover all energy products sold on the Internal Market [13]. As is clear from the citation, not only the whole reasoning changed, but the Commission also changed its approach. Instead of proposing a new and completely harmonized tax, the prevailing idea was to harmonize levels and broaden the bases of taxes already in existence within the member states. Despite those changes in tactics and approach, the proposal was still opposed by some countries, notably Spain. It is only after the introduction of a series of amendments, changes, of more flexibility for member states and after the lowering of the minimum rates of taxation, that the proposal achieved the overall support from member states. On 27 October 2003, the Council of Ministers adopted Directive 2003/96/EC designed to restructure the Community framework for the taxation of energy products and electricity. In a sense, the Directive was an overhaul in the history of environmental taxation in Europe. Till the Directive, the Community coverage was limited to mineral oil products. But now, the coverage of the Community framework was widened to include other energy products such as natural gas, coal, and electricity. The Directive also increased the minimum rates of taxation for mineral oils and introduced new The EC proposed the first EU-wide energy and carbon tax in While all Member States agreed on the principle of taxation as an instrument to combat climate change and politicians acclaimed the double dividend theory, the negotiations never resulted in any agreement. 3

6

7 Table 1 Minimum rates in the Energy taxation directive motor fuels [16] Prior to the Directive Leaded petrol ( /1,000l) Unleaded petrol ( /1,000l) Gas oil / diesel ( /1,000l) Kerosene ( /1,000l) LPG ( /1,000l) Natural Gas ( /1,000 kg) ( /GJ) Current ,6 Table 2 Minimum rates in the Energy taxation directive heating fuels and electricity [17] Prior to the Directive Gas oil / diesel ( /1,000l) Heavy Fuel Oil / diesel ( /1,000l) Kerosene ( /1,000l) LPG ( /1,000l) Natural Gas ( /1,000 kg) ( /GJ) Coal and Coke ( /GJ) Electricity ( /MWh) Current ,3 0,3 1 No substantial development of positive law in the area of energy taxation took place since the adoption of this Energy Taxation Directive. However, changes will come sooner or later as indicated in the Green Paper on market-based instruments for environmental and policy purposes (COM(2007) 140 final). The document provides (at least in theory) the scope for restructuring the Energy Taxation Directive to better reflect the EU energy and climate policy and to reflect all other regulatory and economic changes that took place since the adoption of the Energy Taxation Directive. However, the enthusiasm from the early 90s for environmental taxes seems to be gone. III. Classification of environmental taxes In order to organize the debate properly, it is necessary to define what we understand by environmental taxes. Such taxes can be classified in four categories: Energy taxes; 5

8 Energy taxes are by far the most important part of all environmental taxes and amount to approximately 75% of the whole package. most member states (old and new alike) derive more than 90% of all revenues from energy taxes from transport fuel taxes. The high share of taxes derived from the transport use of fuels has been sometimes attributed to the structure of the Energy Tax Directive (2003/96/EC) in which the minimum rates for petrol are the highest among all energy products serviced by the Directive. However, this misses the point. The real question should be: why are the minimum rates for petrol set in the Directive so high? Transport taxes; Pollution taxes and Resource taxes [18]. Energy taxes are by far the most important part of all environmental taxes and amount to approximately 75% of the whole package. Energy taxes include taxes on both transport and stationary use of energy products (incl. heating) with transport fuel taxes being predominant. They are also the oldest ones [19]. However, as usual with energy taxes, their relative importance varies considerably among not only states but also energy products and energy uses. The fuel taxes constitute the biggest part of all energy taxes. Considering old member states, the relative share of fuel tax revenue from all energy taxes revenues reaches more than 90% in Ireland, Greece, Portugal and the UK. On the other side of the spectrum we find countries where that share makes for only about 50% of all revenues from energy taxes. This is the case in Denmark, Sweden and the Netherlands, where the difference can be assigned mostly to taxes on natural gas and electricity. In new member states, the situation is more homogeneous. Almost all new Member states levy about 90% of all energy taxes on transport fuels with only three exceptions Cyprus, Poland and Estonia with about 80% mainly due to the fact that Poland and Estonia levied taxes on electricity using rates above the minimum required by the Directive. The reason for such homogeneity among new member states is that the Directive 2004/74/EC allowed for exemptions in minimum tax rates on electricity, natural gas, and coal for all EU12 countries. Distinguishing now between transport and non-transport use of fuel one can see that most member states (old and new alike) derive more than 90% of all revenues from energy taxes from transport fuel taxes. The only exceptions are Romania (with almost 20 % share of nontransport use fuel taxes), Italy, Sweden, Cyprus (with about 13 %) and Germany and Denmark (with more than 10 % share of non-transport use fuel taxes) mostly due to high taxes on gas and fuel oils for heating. 6

9 The high share of taxes derived from the transport use of fuels has been sometimes attributed to the structure of the Energy Tax Directive (2003/96/EC) in which the minimum rates for petrol are the highest among all energy products serviced by the Directive. However, this misses the point. The real question should be: why are the minimum rates for petrol set in the Directive so high? The same question should apply for gas oil which can be used for both transportation and non-transportation purposes. The rates for gas oil are substantially different between its uses high when used for transportation purposes, low when used for nontransportation purposes. Once we assume that the burning process of the same fuel is about the same regardless of its use, the only plausible conclusion is that fuel taxes are not closely connected to their environmental impact. After energy taxes, transport taxes are second in importance among environment taxes with approximately 25% of the whole pie. Only one country in the EU has a share of transport taxes on energy taxes actually decreasing: Belgium. This is due to the introduction of the federal contribution on electricity and natural gas. A similar change took place in Estonia following its ETR in 2008 and the related introduction of a tax on electricity. The more countries introduce and increase taxes on electricity and natural gas, the lower the share of fuel taxes will be. After energy taxes, transport taxes are second in importance among environment taxes with approximately 25% of the whole pie. However, there are countries within the EU where the share of those taxes are substantially bigger than average. Among these are Cyprus, Malta and Ireland where transport taxes cover almost a half of all revenues derived from environmental taxes. Then comes Denmark with a third of all revenues from environmental taxes to come from transport taxes. Pollution and resource taxes are usually marginal and not very important and usually amount to approximately 5% of the total. Again, in some countries even these taxes can be important. An example of it is once more Denmark, where a third of all revenues from environmental taxes come from high hydrocarbon tax. 7

10 IV. The results of ETRs in the 90s and 00s overall shares of environmental taxes revenues have fallen within the last decade to 2,4 % of GDP and 6,1% of total tax revenues. This means, that shares on total tax revenues are back at their levels of the early 90s. The fall is noticeable especially after 2004, which is quite ironic just after the Directive entered into force. At the beginning of the 90s, it appeared that the role of energy taxes was set to rise over the coming years. This development was also supported by the political momentum with regard to ETRs; a momentum that we already described earlier. Indeed, data shows that, over the period , environmental taxes as a share of total fiscal revenue increased while taxes on personal and corporate income declined slightly, indicating a modest shift in tax policy. However, things started to change from 1995 on. In what follows, we look at the evolution of the ratio of energy taxes to GDP and to total tax revenues for both EU 15 (old member states) and EU 12 (new member states). Despite the Directive, results vary substantially from nation to nation. Still, one can identify several general patterns that are valid across states. Environmental taxes usually increased in all countries in absolute terms. Proceeds from the introduction or increase of environmental taxes were sometimes (not always) used to finance cuts on labour tax. And above all, we find many rate reductions and refund schemes to protect producers from rising input costs. Also, several new member states took steps in direction of the ETRs, notably Slovenia in 1997, the Czech Republic in 2006 and Estonia in However, overall shares of environmental taxes revenues have fallen within the last decade to 2,4 % of GDP and 6,1% of total tax revenues. This means, that shares on total tax revenues are back at their levels of the early 90s. The fall is noticeable especially after 2004, which is quite ironic just after the Directive entered into force. The high-water mark for environmental taxes was achieved in 1999 when they peaked with 2,9 % share of GDP and 7% of total tax revenues. However, within this general picture, there are big differences among Member States. In what follows we analyze the data in greater detail. First, development within EU15 (old member states) will be analyzed before moving to the study of EU 12 (new member states). Graph 1 shows shares of environmental taxes with respect to GDP for EU 15, graph 2 shows the same variable for EU 12. 8

11 Graph 1 ET revenues in % of GDP, EU15 [20] Graph 2 ET revenues in % of GDP, EU12 [21] 9

12 In only three EU 15 countries and three EU 12 countries the share of environment tax revenues is higher in 2008 than it was in These are Denmark, the Netherlands and Austria for EU 15 and Estonia [22], Latvia and Poland for EU 12; clearly not the largest economies of the EU. Denmark actually comes first in terms of share of environmental taxation relative to GDP, followed by the Netherlands and then Sweden. Among EU 12 countries, the highest level of environmental taxes on GDP is observed in Bulgaria [23] followed by Slovenia and Hungary on second and third places. Note that to see Denmark with the highest ratio of environmental taxes revenues to GDP should note entirely come as a surprise since Denmark is the EU country with the highest level of taxation [24] relatively to GDP (48,2 % for 2008) followed by Sweden (47,1 % for 2008). At the opposite end lies Romania with a ratio of tax revenues to GDP at only 28 % and a ratio of environmental taxes to GDP at only 1,8 %; the second lowest number after Lithuania with 1,7 %. It is also to be noticed that the overall level of taxation is higher in the EU 15 in comparison with the EU 12 [25], which is mirrored by the fact that the level of environmental taxes on GDP is also lower in EU 12 (2,41 %) compared to EU 15 (2,73 %). If the share of GDP informs us about the importance of environmental taxes relative to the size of the economy, this measure however can overstate the role of environmental taxes in the general fiscal policy of the country. As was just pointed out, a high environmental tax-to-gdp ratio does not necessarily mean that the country is placing a higher focus on these types of taxes as opposed to taxing labour or capital, but rather that we are dealing with a country where the fiscal burden is generally high. To show the relative (political) importance of environmental taxes to the system one should compare environmental taxes revenues with total tax revenues. This is what will be done next, again distinguishing between old and new member states. Graph 3 shows shares of environmental taxes with respect to total tax revenues for EU 15, graph 4 does the same for EU

13 Graph 3 ET revenues in % of total taxation, EU15 [26] Graph 4 ET revenues in % of total taxation, EU12 [27] 11

14 The countries with positive differences between the years 1995 and 2008 are: Denmark, the Netherlands, Austria and Sweden, for the group EU 15, and Estonia, Latvia, Poland and Slovakia out of the group of EU12. The shares on total taxation are in both groups almost the same 6,67 % for old member states and 6,97 % for new member states. Interestingly, the order between both groups is reversed from what we obtained comparing the tax-to-gdp ratio. This is likely to be due to the fact that ETs are by their construction more complicated to evade from than taxes on labour or, even more surely, on capital. Another complementary explanation could be that the overall energy intensity is lower in old member states than in new member states. This could also explain why lower tax rates in new member states tend to raise the same amount of revenues than higher rates in old member states. And it also may be the rationale behind the passage of two directives in April 2004 amending the original Directive 2003/96/EC. These are the Directive (2004/74/EC) giving the possibility for the Czech Republic, Estonia, Latvia, Lithuania, Hungary, Malta, Poland, Slovenia and Slovakia to apply temporary exemptions or reductions in the levels of minimal taxation of energy products, and the Directive 2004/75/EC that gives the same possibility to Cyprus. Cyprus, along with Poland, was granted further exemptions with regard to environmental taxation in the Treaty of Accession that provided for transitional arrangements and other special features. The highest levels of environmental taxes with respect to total taxation are found in Denmark, the Netherlands and Ireland. For the new member states it is found in Bulgaria followed by Poland and Estonia. From what we have presented so far, it is clear that most EU member states are similar in their use of environmental taxes with only few outstanding countries that diverge from general patterns, especially Denmark, Sweden and the Netherlands from the pack made of old member states and Slovenia and Bulgaria from the pack of the new member states. Denmark levies three individual taxes framing the energy taxation mix: most EU member states are similar in their use of environmental taxes with only few outstanding countries that diverge from general patterns, especially Denmark, Sweden and the Netherlands and Slovenia and Bulgaria. Denmark was the first country in the world to implement a CO 2 taxation scheme, closely followed by Sweden in the early 1990s. Despite the fact that EC attempts to levy the European carbon tax failed in 1994, some countries, including the Netherlands, Germany, and the U.K., soon followed the Nordic lead. The energy tax (based on the energy content of the 12

15 fuel, levied on fossil fuels, oil products, and coal, the only exemption being natural gas where the energy content is not taken into account), The CO 2 tax (introduced in 1992 for details see below), and The sulfur tax (introduced in 1996, levied on all fossil fuels with a sulfur content exceeding 0.05% based on weight). Denmark was the first country in the world to implement a CO 2 taxation scheme, closely followed by Sweden in the early 1990s [28]. That was the time when the first proposals to promote the use of environmental taxation schemes and overall European carbon taxes were submitted by the European Commission (1992 see above). Despite the fact that EC attempts to levy the European carbon tax failed in 1994, some countries, including the Netherlands, Germany, and the U.K., soon followed the Nordic lead. As an example of such a development we can use data from Denmark see Table 3. Table 3 Development of energy and CO 2 tax rates for different users and uses [29] Household and service sector Industry heating Industry light process Industry heavy process Light Fuel Oil /1,000 l /1,000 l /1,000 l /1,000 l ,2 239,2 18,3 1, ,3 268,3 24,6 1, ,5 286,5 24,6 1,1 Heavy fuel oil /ton /ton /ton /ton ,7 1, ,5 304,5 29,2 1, ,8 324,8 29,2 1,3 Natural Gas /1,000 m 3 /1,000 m 3 /1,000 m 3 /1,000 m ,3 31,3 14,9 0, ,2 244,2 20,1 0, ,8 305,8 20,1 0,9 The 1996 tax reform created a rather complex system of energy and carbon tax differentiation for industry in Denmark [30]. As a part of that system, the industrial companies can reduce their CO 2 tax burden by entering into voluntary agreements with the government. The tax on electricity consumption can be used as an example of the energy tax used in Denmark. Again, the tax on electricity consumption in Denmark belongs to the oldest taxes of its kind in Europe and in the world. It was first levied on all electricity consumption in 1977 and has been used ever since. The tax is levied on consumption 13

16 regardless of where or how electricity is generated (domestically, abroad or conventional or renewable energy sources). As for the magnitude see the table 4. Table 4 Energy and CO 2 taxes levied on electricity [31] Heating purposes Other purposes Industry Electricity /MWh /MWh /MWh ,8 62,5 8, ,6 85,3 13, ,8 89,5 8,6 Thus, the tax rate on electricity in Denmark exceeds by up to 100 times the minimum excise duty stated in the Directive. The same magnitude applies also for Sweden and the Netherlands (the other two rather exceptional countries as noticed earlier). However, not even this could explain Denmark s position as a leader in the area of environmental taxation. The energy taxes constitute only about 40% of all state revenues coming from environmental taxes. Together with transport taxes they constitute about 70 % of the total. Energy and transport taxes alone would make Denmark number 1 in terms of tax-to-gdp ratio among EU 27, but what makes Denmark really special are its pollution and resource taxes that contribute the other 30 %. The difference between EU average composition and Denmark s is clearly visible on the two pie diagrams [32]. Graph 5: Comparing the structure of environmental taxes: EU average and Denmark [33] 14

17 The per-unit taxes have some counter-cyclical qualities: when the price of fuel increases, the share of the tax in the overall price decreases, and vice versa. As for Slovenia, the irony of the environmental tax reform lies in the fact that Slovenia used to be number one in terms of share of environmental taxes on overall tax revenues (12.2%) and also in terms of tax-to-gdp ratio among EU 12 (4.5%). This was before accession, in With such numbers Slovenia would have ranked second among EU 27 as far as tax-to-gdp ratio is concerned, and as number one according to the share of ETs on the tax revenues. However, as a result of the ETR of 1997, both tax-to-gdp ratio and the share of ETs on the entire tax revenues felt substantially from 4.5 % and 12.2% to 3 % and 8 % respectively. V. Reasons for the decrease in weight of environmental taxes The per-unit taxes have some counter-cyclical qualities: when the price of fuel increases, the share of the tax in the overall price decreases, and vice versa. Given the political momentum for ETRs, it is fair to ask what went wrong for environmental taxes. Remember that the rationale behind the double dividend theory was that the taxes to be abolished or reduced (taxes on capital and labour) were distortionary while the taxes to be increased or introduced (taxes on energy products and other bads ) were virtuous. It was expected that new taxes on bads would generate health, ecological and other benefits by lowering consumption of bads, and elimination of old taxes would generate another set of benefits stemming from the elimination of losses caused by distortionary effects. There were even some attempts (mostly on US data) to compute the magnitude of those dividends with highly favourable results: "the costs of a carbon tax may be largely and perhaps even fully offset by taking advantage of its efficiency value and using the revenues to cut existing taxes that discourage capital formation or labour supply." [34] It all looked like a win-win situation. But still, ETRs went out of fashion in the first decade of this century. In this section, we will suggest plausible reasons for this very unexpected development. The reasons for the decline of environmental taxes can be summarized as follows: The very construction of environmental taxes; Increasing overall energy efficiency; Political unpopularity of an increase in environmental taxation in the era of increasing 15

18 energy prices; and Introduction of other policies namely emission trading schemes, renewable energy sources subsidies, road charges/tolls etc. First, environmental taxes are typically imposed on the sale of a given commodity/service on a per-unit basis. Very few of them are levied ad valorem (unlike other types of taxes such as VAT). The per-unit taxes have some counter-cyclical qualities: when the price of fuel increases, the share of the tax in the overall price decreases, and vice versa. To see the mechanism, it is possible to look at the data of percentage of taxes in premium unleaded (95RON) gasoline prices compared with the development of the crude oil price. Graph 6: Share of taxes in premium unleaded (95RON) gasoline prices and price of crude oil [35] 16

19 Second, because environmental taxes are established on a per-unit base, any increase in overall energy efficiency leads to lower amounts of tax revenues. As evidenced by data (see graph 7 and graph 8), every EU 27 country has experienced a considerable improvement in terms of energy efficiency, the effect being stronger for EU 12 than for old member states. The average improvement in energy efficiency for countries in EU 15 was approximately 23%. It was almost three times higher for EU 12. However, the energy intensity [37] is still almost three times higher for EU 12 than for EU 15 (445,93 as opposed to 156,37). Graph 7 Energy intensity, EU15 [38] 17

20 Graph 8 Energy intensity, EU12 [39] One can then logically expect that the effect of increased energy efficiency will be stronger in EU 12 than EU 15. However, if anything can be deduced from the data, it is that the decline in the weight of environmental taxes was a bit larger in EU 15. Why is it so? To get an idea, one can look at Bulgaria that had the highest improvement in energy efficiency of all countries studied. Yet, Bulgaria remains one of those countries in which the weight of environmental taxes actually increased from its 1995 level. It must be that the effect of higher energy intensity is offset, in reality, by other factors mainly by the economic growth itself. When an economy is growing, energy tends to be used more efficiently per unit of output but the amount of output is growing so fast that the amount of energy used increases despite all efficiency improvements. This reasoning lies behind the fact that energy consumption is rising with GDP growth although admittedly not as fast. Third, from a political point of view, it is highly unpopular to tax more heavily consumption goods, which prices are already increasing. Inversely, it is easier to increase tax rates when the price of the taxed good is decreasing. The data support this explanation. 18

21 Graph 9 Tax rates on unleaded gasoline and oil prices [40] As can be clearly seen, the sharp increases in tax rates on gasoline end around 2002, when prices of oil started to climb to their 2008 peak (reaching 140 USD/barrel). Clearly, oil and fuel taxes cover just a part of environmental taxation issue, but the same logic works for other taxes as well. Furthermore, because the price of other energy products closely correlate to oil price, it is not surprising that the appetite for new tax and increases in existing taxes on other energy products diminished considerably during the last decade. Those ETRs that were enacted within that decade could be described as half-hearted (the Czech Republic, with its environmental tax reform of 2008 and subsequent development, providing a good illustration). At the height of energy product prices, some voices within the EU 27 were even heard calling for a cut in energy taxes [41]. Fourth, it took a long time to get the Energy taxation directive accepted, and by the end of the negotiations, what survived was a watered down version of the proposal submitted in 1997 by Mario Monti. It is indeed difficult to compare the final version of the Directive with the original 1992 proposals for a pan-european carbon tax. The resulting compromise of 2003 pleased almost no one and in the meantime new pet projects for green coalitions appeared on the agenda. Hence, by the time they were introduced, the ETRs movement from the 90s had lost its momentum, which explains the surprisingly quiet development of environmental taxation from 1995 till the present day documented in the previous sections. 19

22 In the EU, the main competing green project seems to be the emission trading scheme for carbon. The idea of trading emission allowances originated in the USA where several states adopted such schemes not trading permits to emit CO 2, but to emit other pollutants such as SO 2 [42]. In the run-up to the 1997 Kyoto Protocol Americans injected this idea into the Kyoto protocol flexible mechanisms (1997) despite European unwillingness to accept it at the beginning. One can even say that the EU was a leading sceptic towards international ETS in Greenhouse Gases (GHGs). However, as is well known, after few years of time Europeans had changed their opinion, now embracing this particular flexible mechanism while Americans rejected the Kyoto protocol as a whole [43]. Hence, by the end of the 90s, the emission trading scheme had gained political momentum except in the US-- precisely when ETRs lost their momentum. This seems to be no coincidence. Thus, the combination of new political pet projects (EU ETS and others) [52] combined with a sharp increase in prices of oil, oil products, natural gas and electricity put on hold any discussion about environmental tax reforms at least based on environmental justifications. Accordingly, the EC commissioned several documents in which it showed a new enthusiasm with respect to emission trading (as opposed to environmental taxation) as a main tool to achieve results similar to those that were targeted in first proposals of the environmental taxation directive. These include: The Green paper on emission trading schemes (March 2000) [44] and the Green Paper on market-based instruments for environment and related policy purposes (March 2007) [45] ; The Sixth Environmental Action Program of the EU under the name Environment 2010: Our Future, Our Choice. It was published in 2002 as a joint decision of the European Parliament and the Council (Nr. 1600/2002/EC) [46] ; Communication from the Commission to the Council and the European Parliament Progress Report on the Sustainable Development Strategy 2007, COM/2007/0642 [47] ; 20

23 White paper Adapting to climate change: towards a European framework for action, COM(2009) 147, The 2000 Green paper became a base for the development of the Directive 2003/87/EC on greenhouse gas emission trading [48]. The Directive 2003/87/EC was passed in October 2003, that is, more or less at the same time as the Energy taxation directive (six months later to be precise), and its passing was much smoother and faster (instead of 11 years separating the first proposal from the final directive in the case of the Energy taxation directive, it took this time only two years to get to a final directive on emission trading from its first proposal as Communication 581/2001 in October 2001). From its very beginning the EU ETS (EU energy trading scheme) was supposed to be a corner stone of the EU climate change abatement policy and a main tool for the EU as a whole (EU 15 at that time) and individual member states to simultaneously comply with both their emission reduction commitments under the Kyoto protocol and also their own targets originating from the Green Paper of EU ETS started its operation on 1 January 2005 with a pilot phase from [49]. The second phase started in 2008 and will run until 2012 with the third phase to follow in 2013 for a period of eight years till The scheme incorporates about 60% of all CO 2 emitters and the scope of the system is getting broader all the time [50]. The same emitters are also subject to the Energy taxation directive. In a sense, it means a double taxation of the same set of products which is also not a plus from a political point of view, since the introduction of the EU ETS had an immediate impact on electricity prices all over Europe see graph 10. Graph 10 The effect of ETS introduction on electricity prices in EEX [51] 21

24 Thus, the combination of new political pet projects (EU ETS and others) [52] combined with a sharp increase in prices of oil, oil products, natural gas and electricity put on hold any discussion about environmental tax reforms at least based on environmental justifications. If there will be any attempt to introduce new taxes on energy products or to increase existing taxes on energy products their main rationale will be to increase fiscal revenues. In any case, enthusiasm for ETRs is over and there is no sign of its revival any time soon. VI. Official justifications behind ETRs and tax harmonization and real achievements Having established that ETRs lost their momentum and understood why, it is still interesting to see whether they have achieved at least some of the goals they pledged to achieve and that also served as a rationale behind the reforms (albeit on limited scale as we have seen already). There were several justifications lying behind the environmental tax reform movement and environmental tax harmonization: This target had in fact three dimensions: harmonization among member states, among energy sources and among different uses of a given energy product. On all counts the policies introduced to achieve them failed. Reduction of tax distortions between countries and products Creation of incentives to use energy efficiently Compensation for harmful effects of energy consumption and production Reduction of import dependency, higher energy security Reduction of CO 2 emissions and support of RES Trade-off between taxing energy and labour It is no place here to come back on the theoretical debate. As was mentioned earlier, that debate took place in the first half of the 90s and, in our opinion, is now closed, with the defeat of the double dividend proponents. Interestingly, ETRs that were passed recently were not even mentioning the double dividend theory. In what follows we will instead investigate whether the results actually achieved by the ETRs have been in accordance the minimum taxes for gas oil used as a source of fuel for transportation are about 15 times higher than the minimum taxes for stationary uses of the same fuel. 22

25 with justifications and expectations. There is no sign for a convergence in taxation of different energy products. the way environmental taxes are constructed might give a false impression of convergence or reduction of distortions. Reduction in distortions This target had in fact three dimensions: harmonization among member states, among energy sources and among different uses of a given energy product. On all counts the policies introduced to achieve them failed. We will go through these points in reverse order. If burning fossil fuel causes external effects in terms of worsening the environment especially in terms of the production of CO 2 (or any other pollutant) then it should be highly irrelevant where the particular fuel has been burnt. So, if such environmental considerations were primary reasons for the reform, then the use of the fuel or energy product should not matter for the tax levy. Or, to be more precise, if environmental considerations other than CO 2 emissions were the reason for imposing environmental taxes then burning of fossil fuels for industrial uses and big scale electricity/heat production should be taxed less and not more because big industrial power plants, power plants of energy companies or big heating plants for the supply of heat into the district heating systems (often using cogeneration of heat and electricity) produce less pollutants per unit of burnt fossil fuel than distributed energy production in households or small sites. The reasons are straightforward higher temperatures in big producing sites, are more optimized in terms of operational, thermal and energy efficiency, developed technology to clean residue of combustion etc. But this is not the logic followed by existing environmental taxes. Enough is to look at the Directive to realize that other considerations were at play while setting minimum rates for particular types of energy product and particular use of an energy product. There are striking differences between taxes imposed on fuels used for transportation and fuels used for non-transportation purposes, mainly for heating, and also between taxes imposed on industrial and domestic use of the same energy product. For instance, the minimum taxes for gas oil used as a source of fuel for transportation are about 15 times higher than the minimum taxes for stationary uses of the same 23

26 fuel. The reasoning that high taxes on transport fuels can be motivated by the existence of other negative externalities related to the transport sector (accidents, noise, and congestion) does not hold water. One would need to argue that other negative externalities caused by transportation (accidents, noise, and congestion) are 15 times more costly from a social point of view than externalities from burning of the fuel itself. Either it is true, then why do we have energy taxes or environmental taxes in the first place since the associated externality is relatively negligible, or it is not true, and we must conclude that other considerations are at play (mostly social ones and also the need to finance the road infrastructure or/and the need to fill the gaps in other governmental revenues, i.e., fiscal considerations). Also, there is no sign for a convergence in taxation of different energy products. As a matter of fact, if such convergence were seriously targeted, one would need first to find a common denominator over which one can levy the various taxes on different energy products. But so far, no such common denominator has been offered. It could have been, for instance, the carbon content of each energy product (if the overall goal is to limit CO 2 emissions), but no tax (with the exception of the carbon tax, which is not general and does not exist in most states) has been constructed on such a basis. Moreover, since burning a particular fuel has almost the same effect regardless of the place where it is burnt in Europe, convergence of taxation among energy products requires that taxes be the same or almost the same everywhere in Europe. As we will see below, there is no evidence that we are moving in that direction. If no convergence is observed at the European level, at least one could find convergence- -according to a common denominator--within some of the member states. The data, however, does not provide any support for this happening either (see Annex 3). Before we can present and analyze the data used to study the convergence among member states, a short and easy statistical digression is in order because the way environmental taxes are constructed might give a false impression of convergence or reduction of distortions. If the tax is levied per-unit and the price of the taxed good is increasing then the share of tax on the total price decreases for all levels of imposed tax. Mathematically it appears that different tax rates converge. An example is given below in a table. The effect is stronger when the share of the tax in the overall price is smaller. 24

27 Table 5 Mathematical convergence of different level of tax levied per unit Price net of tax Price plus tax in percentages (levy = 5) Price plus tax in percentages (levy = 4) Price plus tax in percentages (levy = 3) Price plus tax in percentages (levy = 2) Price plus tax in percentages (levy = 1) 150% 125% 117% 113% 110% 108% 107% 106% 106% 105% 140% 120% 113% 110% 108% 107% 106% 105% 104% 104% 130% 115% 110% 108% 106% 105% 104% 104% 103% 103% 120% 110% 107% 105% 104% 103% 103% 103% 102% 102% 110% 105% 103% 103% 102% 102% 101% 101% 101% 101% Average 130% 115% 110% 108% 106% 105% 104% 104% 103% 103% Standard deviation (SD) 16% 8% 5% 4% 3% 3% 2% 2% 2% 2% Average/SD 0,122 0,069 0,048 0,037 0,030 0,025 0,022 0,019 0,017 0,015 Graph 11 Mathematical convergence of different level of tax levied per unit 25

28 Clearly, although the levies have remained the same and no convergence took place, a careless study can easily conclude to the contrary. The next table shows actual data on convergence among several types of energy products [53]. Further details are provided in Annex 1. Graph 12 Convergence of taxation levels for different energy commodities among European countries [54] 26

29 Creation of incentives to use energy more efficiently True, everything else equal, when the price of a good increases consumers usually buy less of it. But to reduce energy consumption by means of taxation is nonetheless trickier than it seems. Taxing the use of energy products may or may not achieve such a goal and the net effects of a tax are in fact very hard to quantify and measure. There are good reasons behind that. People tend to react in three ways when, due to taxation, energy and energy products become more expensive: We can see from the data that no overall convergence of energy taxes has taken place among European countries within the last decade. They substitute other products for the taxed ones, They lower their use of the taxed products They consume less of other products especially when the price elasticity of the taxed product is low, which is, at least in the short term, the case for energy and most energy products. Contrary to what is often believed, it is possible to substitute energy with other products mostly with some kind of capital goods (for instance more fuel efficient cars [56] or more insulation). However, there are several effects coming with this kind of substitution. While using more capital goods as a substitute for energy someone had to produce these capital goods and use more energy to produce them. The net effect is likely to be positive in a sense that less energy is consumed in total [57]. In any way, when taking into account the production of the capital good, the global result in terms of environment and energy consumption is (at best) much less shining than expected. For almost all other energy products even divergence seems to be the case. Besides, the fiscal policy can lead industries to transfer their energy intensive production to other countries with more lenient energy and environmental policies as a way to lower their overall production costs. Because of these transfers one might have at home the impression that the policy is working while this is in fact just a statistical illusion from a global point of view [58],[59]. 27

30 although on the face of it the higher the costs of energy the less of it is used, the truth is very likely that the total cost of saving the energy via artificially high prices due to environmental taxation will be much higher than expected. Pigou suggested that a special kind of tax be introduced to force individuals to take into account the effects that some of their actions have on third parties. Substituting capital goods for energy leads not only to higher costs of capital goods (because of higher demand for capital and higher cost associated to the production of capital) but also to higher overall costs, at least if we keep technology constant. Indeed, if the additional capital good was not used before the taxation was introduced it must have been because it was more expensive and other combinations of capital and other inputs were cheaper; otherwise the producer would have done it. Finally, because the overall costs are higher than before, expenditures on other goods and services have to be lowered [60]. The same reasoning applies not only to environmental taxes but also to other measures introduced to promote a more efficient use of energy such as tax credits for investments in renewable energy or above-market-value feed-in tariffs for renewable electricity, funding for a smart electric grid, upgrading government vehicles to be more energy efficient, subsidizing of insulation, ban on old-fashioned bulbs [61] etc. Other types of substitutions are likely to occur such as more energy efficient cars for more secure cars. One way to make a car more energy efficient is by reducing its weight so, under the general direct pressure to produce more efficient cars in terms of mileage and CO 2 emissions and the indirect pressure caused by high fuel prices in Europe, the cars in Europe are getting lighter (less steel, narrower metal plates, no spare wheel etc.) and less secure in terms of life and health protection of passengers. To put it bluntly, it is not unlikely that more people die on European streets because of environmental taxes and other environmental regulations. Taxing energy use in transportation also leads to less transportation and therefore less division of labour and lower overall economic growth [62]. Higher taxes on transportation-- and resulting lower division of labour--are probably one of several reasons why European economies are not growing as fast as the US economy. For data comparing US and European taxes on transport fuel see Annex 2 of this report. 28

31 Hence, although on the face of it the higher the costs of energy the less of it is used, the truth is very likely that the total cost of saving the energy via artificially high prices due to environmental taxation will be much higher than expected. So much so that the strategy could prove inefficient at least from a global viewpoint. Compensation for harmful effects To understand this idea brought in support of energy and environmental taxation, the best is to go back in time to Cambridge in the years immediately following World War I. In 1920, an famous economist by the name of A.C. Pigou suggested that a special kind of tax be introduced to force individuals to take into account the effects that some of their actions have on third parties (today, these effects are known as externalities and Pigou s suggested solution as the Pigovian taxes, however Pigou did not use these terms; they evolved later on after World War II). In Pigou s terminology, the effect of the tax is to equalize the private cost of the action with the cost to society. This approach bears close resemblance to the polluter-paysprinciple promoted by the OECD (OECD, 1972 and OECD 1992) and placed as a cornerstone of EU environmental policy. Half a century later, Pigou s idea was given a new impetus with the work of Baumol and Oates (1971). Although not identical, the rationales of Pigou and Baumol-Oates were very close to each other as were the suggested solution: a uniform tax levied on the polluter or polluting substance. This approach bears close resemblance to the polluter-paysprinciple promoted by the OECD (OECD, 1972 and OECD 1992) and placed as a cornerstone of EU environmental policy (Treaty of Rome consolidated version, 2006). Taxes were also seen as more efficient tool to reduce externalities than direct regulation (Surrey, 1973). Despite of its success in academic circles, politicians and policy makers did not favoured the Pigovian tax for many years. A change finally occurred with the White paper (1993) and proposed ETRs that embraced Pigovian taxes overwhelmingly for the first time. However, there are several obstacles for these taxes to reach the promised lands of economic nirvana. 29

32 Taxes do not entirely destroy the system; they pollute it with false noises that generate distortions bigger than the ones created by existing externalities they were supposed to mitigate. the very reason why some economists supported a Pigovian tax in the first place was precisely that neither government nor anyone else knows who should be compensated and by how much. This is a real Catch 22 for Pigovian taxes they can be a successful solution only and only in situations where they are not needed. Firstly, Pigou (and many others after him till the early 90s) did not consider the possibility that such taxes could themselves introduce new distortions in the allocation of economic resources. However, research done in the 90s indicated that distortion caused by environmental taxes can be as important as distortions caused by alternative policy instruments (such as direct regulation or trading permits). Pigovian taxes can cause distortions in many ways but the most important distortions are due to a feature of that tax that paradoxically is often perceived to be an advantage: the constant value of the tax rate, at least over a sufficiently long period of time. The argument for a fix rate is usually built on predictability of such a tax: Having an unchanging price for carbon offers a certainty to businesses and the public. Yet this highly praised feature is also the biggest problem. For the Pigovian tax to provide the right incentives to economic agents, it has to convey at any point in time updated and reliable information on the state of the system (preferences, technologies, resources, cost of externalities, etc.) But if the tax is set to be constant for a long time (year or two, or even more) then, even if it were set correctly at the moment of imposition an heroic assumption--, within weeks or months its value will not longer signal what the economic system would need. It starts to work as a regulated price (a tariff) in the system that is, it starts to emit false information (just as the regulation it was substituted for). Let us remember the fate of socialism. Socialism was an attempt to stabilize prices for a long time and it did not work because, by stabilizing prices it destroyed the only system of information dissemination available to economic agents [63]. Taxes do not entirely destroy the system; they pollute it with false noises that generate distortions bigger than the ones created by existing externalities they were supposed to mitigate. 30

33 Secondly, the polluter-pays principle suggests that those harmed by the polluting activity will be compensated. Otherwise to whom would the polluter be paying? It even suggests that the victims of the externality receive damages that exactly match the harm: Just as one party to a transaction pays a price to the other party so that the latter agrees to transfer his/her property rights, the producer of an externality would pay a tax to get the right to create that externality. But this is never the case and will never be. Indeed, the very reason why some economists supported a Pigovian tax in the first place was precisely that neither government nor anyone else knows who should be compensated and by how much. Otherwise it would have been easy to design other types of remedies (including individually tailored regulation) or institutions to tackle the problem. In other words, if the price system does not work properly in those externality cases it is precisely because of our (at least temporary) inability to set up an institutional framework for it to function (a viable compensation scheme or a system of property rights). In no way does the Pigovian taxation provide a solution to this problem. If governments succeeded in setting the Pigovian tax right and at the same time succeeded in compensating affected parties correctly, then no Pigovian tax would ever be needed. The solution would have been known already otherwise the government would not be able to succeed. This is a real Catch 22 for Pigovian taxes they can be a successful solution only and only in situations where they are not needed. one should bear in mind the possibility that the primary reason for the existence of energy taxes (and, later on, environmental taxes) were fiscal considerations. the reason why taxes were levied on certain goods was the low elasticity of the demand for and supply of these goods. Because taxes tackle only the producer s part of the problem, not the problem of those affected by the production, they can change the behaviour of producers of harm, but not of those who are harmed. Absent any compensation, the latter will demand for instance stricter standards regarding the power plant construction. Stricter standards cost them close to nothing (everybody pays the cost increments in market prices, not only those who benefit from the higher standard) but their overall costs are substantial. 31

34 Could it be then that appeals to the double-dividend theory or to the polluter-pays principle be just a way to put some descent clothes on what is nothing more than the eternal quest from our government for more tax revenues? the data offers quite a different story from the one told by proponents of ETRs, with no influence of environmental taxes whatsoever on the rise of energy product imports. Thirdly, one should bear in mind the possibility that the primary reason for the existence of energy taxes (and, later on, environmental taxes) were fiscal considerations. As early as 1927, Frank Ramsey set formally some basic principles of efficient indirect taxation (that had long been known from smart governments), the main principle being to impose taxes on those goods with the most inelastic supply and demand. Doing so will minimize the indirect costs of taxation (what economists call the deadweight loss) since it will induce little change in allocation of resources. Despite some additions to the theory by other economists (notably Mirlees), the principle is surprisingly robust. In other words, the reason why taxes were levied on certain goods was the low elasticity of the demand for and supply of these goods. In the Middle Ages, salt was such a good. Today tobacco, alcohol or energy products are in that category. But, if elasticities of supply and demand are low then even high taxes will not change the behaviour of economic agents much, contrarily to what proponents of the double dividend theory or Pigovian taxes want. Could it be then that appeals to the double-dividend theory or to the polluterpays principle be just a way to put some descent clothes on what is nothing more than the eternal quest from our government for more tax revenues? Regardless of any practical or theoretical obstacles to the proper functioning of a Pigovian taxes, the current policies in Europe show little care for the theoretical principles set by Pigou, Baumol and Oates. As already noticed, rates, bases and scope of taxes are driven by fiscal, political and social considerations. Environmental taxes discriminate between users, energy products, uses of the products and member states. There are myriads of tax rebates, special provisions and exemptions, all of which complicating the whole system and dragging it away from any possible theoretical justification. Import dependency reduction The notion of energy dependency has little economic merit [64], but it is nonetheless interesting to investigate whether ETRs had any significant impact on the amount of energy products imported to the EU 27 in the last decade, that is, whether the policy of imposing environmental taxes achieved what it was officially designed for. We can look at 32

35 the data on coal, natural gas and oil and compare the shares of overall imports in overall consumption at different points in time. Clearly, the data offers quite a different story from the one told by proponents of ETRs, with no influence of environmental taxes whatsoever on the rise of energy product imports. The share of solid fuel imports (relative to total consumption) was 41% in 2007, up from 25% in The share of coal in gross inland energy consumption [65] has declined over time, from 20 % in 1995 to 18 % in 2007 due to increasing use of natural gas, mainly in power generation and space heating. The share of natural gas imports was 60,3 % in 2007, up from 45,2 % in The share of natural gas in gross inland energy consumption has grown over time, from 21 % in 1995 to 24 % in 2007 due to increasing use of natural gas, mainly in power generation and space heating. The share of crude oil imports was 82,7 % in 2007, up from 76 % in The share of crude oil in gross inland energy consumption has decreased over time from 39 % in 1995 to 36 % in years The overall share of fossil fuels imports on primary energy consumption was 53,1 % in 2007, up from 45 % in Graph 13-1 Energy dependence of the EU: Net natural gas imports as a % of primary gas consumption [66] 33

36 Graph 13-2 Energy dependence of the EU: Net crude oil imports as a % of primary crude oil consumption Graph 13-3 Energy dependence of the EU: Net hard coal imports as a % of primary solid-fuels consumption 34

37 Graph 13-4 Energy dependence of the EU: Net natural gas, crude oil and hard coal imports as a % of primary energy consumption If, there seems to be a slight slow-down in rise of energy dependency after 2006 this is not very consistent with claimed effects of the ETRs the slow-down materializes seven years after the environmental taxes peaked. Moreover, the slow-down seems to be more a result of high prices of crude oil and natural gas than a result of new energy taxes imposed on their consumption. Reduction of CO 2 and support of RES To find out whether there is any influence of energy taxes (and especially environmental taxes) on environment-friendly sources of energy one can look at the data on the production of electricity from so called Renewable energy sources (RES) [67]. 35