The distributional effects of Environmental Tax Reform on the European Union

(1)

1

The distributional effects of Environmental Tax

Reform on the European Union

Faculty of Economics and Business (FEB), University of Amsterdam (UvA)

Student: Paulius Virbalas Supervisor: Ieva Rozentale BSc in Economics

(2)

2

1. Introduction

The increasingly growing concerns about environmental problems and their impact on humanity have brought new dimension into the process of economic policy formulation. It has been generally accepted by now that potential costs from the global climate change and its consequences on our planet can be disastrous for future generations unless an immediate action is taken to drastically reduce greenhouse gas emissions (UNCED, 1992). Consequently, the last two decades have been marked by an intensive search for economic instruments enabling to achieve this objective.

One of such policy options is to impose environmental tax on products having negative effect on environment. It has attracted a huge attention within the European Union in particular, where the concept of Environmental Tax Reform (ETR) has emerged. The underlying idea of the reform is that the inclusion of environmental damage in the pricing system would stimulate the shift towards an economy based on renewable energy sources as well as less polluting means of production. On the other side, revenues collected from environmental taxes would be injected into economy to compensate those negatively affected by higher price level resulting from an increase in energy price. Ideally, the implementation of ETR, would allow achieving social, environmental and economic improvements at the same time (EEA, 2005).

Despite the optimistic prospects expressed in the original European Environmental Agency (EEA) proposal, the implementation of the suggested reform across the EU has been relatively slow, as indicated by the fact that the share of revenue from environmental taxes has not increased since 2002 (Eurostat, 2011). The main reason of such result has been a high political resistance towards implementation of major ETR, which was even further intensified by the pressures brought by European debt crisis (OECD, 2010). One of the central causes of this resistance is the uncertainty regarding distributional impacts of ETR. In particular, there is a huge concern that the heaviest burden from such reform will fall on people with the lowest levels of income, since their budget share spent on essentials such as electricity or heating is larger. While quite often such claim is made overlooking the fact that tax revenue would be recycled, even in the case of redistribution policies, the final effects are not clear and seem to be largely dependent on a variety of specific circumstances (EEA, 2011). Consequently, the main research question can be stated as: what would be the distributional effects of environmental tax reform in the European Union? Answering this research question implies the objective of understanding better which groups win and which lose from the reform in terms of real income. Realizing this, the policy responses aimed at equal distribution can be constructed as well as the likelihood of a grand-scale ETR implementation in the near future can be evaluated.

There is a substantial amount of literature analyzing distributional effects of environmental policies in specific countries, however there is very little research discussing this issue on the EU level. The only scientific attempt to model the distributional impacts of ETR on the EU scale was conducted by Cambridge Econometrics consultancy team. The results from their modelling simulations were published by European Environmental Agency in 2011 (EEA, 2011). The research question of this thesis is answered by careful examination and interpretation of those results based on distributional principles established by the analysis of earlier literature.

(4)

4

The thesis is structured as follows: firstly the historical development of ETR concept as well as the environmental taxes across the EU are presented. Next thesis approach is described. After that, a literature review, describing the main scientific conclusions that have been advanced till now, follows in order to prepare the background for the main analysis. Chapter 5, describes the mechanics of E3ME model used in EEA report, by means of which the changes in real income for household groups were estimated. Chapter 6 uses the results from EEA publication to model the distributional trends of ETR and answer the main research question. Finally, the summary and conclusions are provided.

2. Introduction to EU Environmental Taxes

The international efforts to include environmental concerns into economic policy formulation started with the United Nations Conference on Human Environment in 1972. This trend has gained momentum since the Earth Summit hosted by Rio de Janeiro in 1992, as a result of which the Kyoto Protocol was signed in 1997. This international agreement committed its State Parties to reduce greenhouse gas (GHG) emissions, based on the recognition that global warming exists and man-made CO2 emissions have caused it. In the case of the EU, it meant that by the year of 2020, the levels of GHG have to be cut by 20% with respect to the reference year 1990. The methods however, by means of which this target will be achieved, were left freely to choose upon the Union itself (EEA, 2010). European Commission coordinates the environmental policy within the EU, while the European Environmental Agency (EEA) is responsible for research, provision of independent information and policy proposals. In 2005, EEA published an extensive report calling for Environmental Tax Reform, by means of which environmental quality and economic growth could be improved simultaneously. According to the EEA, Environmental Tax Reform (ETR) refers to “changes in the national tax system where the burden of taxes shifts from economic functions such as labor (personal income tax), capital (corporate income tax) and consumption (VAT and other indirect taxes), to activities that lead to environmental damage and unsustainable use of natural resources” ( EEA, 2005, p. 84).

In principle, ETR consists of two main pillars:

1) To make environmental damaging activities more costly in order to stimulate changes in consumption and production patterns;

2) To recycle the revenues raised from environmental taxes and use it for the creation of positive economic as well as social effects;

The revenue recycling component allows governments to compensate the groups that are most negatively affected by an increase in energy prices. In other words, to make ETR politically acceptable and socially just, it has to be ensured that the effects of ETR will be distributed uniformly across the society. This point has been emphasized by European Commission as well, which stated that considerations on how specific groups will be affected have to be made ahead of any actual changes in policy or regulation (EC, 2009). Distributional concerns have become an especially important topic in evaluating the feasibility of the harmonized ETR implementation across the EU, since there is a significant concern that as a result of such reform a disproportionate burden would fall on the most vulnerable groups. (EEA, 2011). In particular, it is argued that persons living from pensions or unemployed benefits not only would loose from higher prices but benefit nothing from revenue recycling mechanism since they do not pay income tax. Thus, it is the aim of this thesis to find out more about the validity of such statements.

(5)

5

In the following paragraphs, environmental taxation across the EU is briefly reviewed. By the definition provided by Eurostat “an environmental tax is one whose tax base is a physical unit (or a proxy of it) of something that has proven to have a negative effects on environment” (Eurostat, 2001, p.11). All environmental taxes are divided into four main categories: energy, transportation, pollution and resources. It is important to note that according to Eurostat classification, taxes on CO2 emissions

are included under energy tax category and not pollution, in order to make international comparisons feasible.

On average, 75% of all environmental taxes revenue is raised from energy taxes, 20% from transport, and the remaining 5% from pollution and resource combined. The figure below represents the distribution of the tax categories in each member state separately.

During the period of 1990-2010 the environmental tax share as of total tax revenue fluctuated in the range from 6% to 10%, while as a share of GDP converged to 3% level. The tax revenue to GDP ratio was the highest in the Netherlands and Denmark, namely 4%, while the lowest in Lithuania and Romania, equal to 2%.

The environmental tax revenue recycling mechanism varies across countries, yet most often it is directed towards reduction in personal income tax rate or social security contributions. In addition, small part of revenue is usually dedicated to stimulate eco-innovation as well. For example, according to the ETR implemented in Germany in 1998, 90% of tax revenue was devoted to public pension scheme which resulted in lower social security contribution rates for employers and employees (Knigge & Gorlach, 2005). The remaining 10% was used to promote research and investment in renewable energy. Other approaches include reductions in corporate tax rates or increases in tax free allowances – the practice applied in the Netherlands and Ireland. Finally, most of the countries have chosen to introduce some compensation measures for the most negatively affected industries, especially those competing in the international markets (Hoerner & Bosquet, 2001). Finally, it should be noted that while all member states perform some form of environmental policy, the more elaborate environmental tax reform has been implemented only in Sweden, Denmark, the Netherlands and Germany. Other countries have introduced environmental taxation but revenue recycling component is usually missing. (IEEP, 2013).

(6)

6

3. Thesis Approach.

I have chosen to carry out a thought experiment to explore the potential consequences on income redistribution arising from a grand-scale Environmental Tax Reform (ETR) in the European Union. Since the scientific literature focusing on such effects on the EU scale is only in its starting phase, I have aimed to establish some general patterns by means of which more understanding could be gained of how to conduct further research on this subject.

In order to answer the main research question, firstly literature review is presented. In this section the main scientific conclusions from partial environmental tax reforms across the EU member states are described. They serve as the source for my considerations in the analysis part.

In order to have some empirical basis to form my interpretations, I have chosen the estimates of distributional effects published by EEA as the foundation block for the analysis part. Accordingly, chapter 5 describes the mechanics of the model used by EEA in order to better understand the causes of the estimation results. Next, the analysis part follows in which the results for different household groups are presented. Based on the principles deducted from literature review, the conclusions by EEA are complemented by my interpretations. I provide some limitations of the model and describe how they affect the estimation results. After determining the main distributional patterns as well as their causes, I consider the possible policy options to ensure a more equal income distribution. Finally, some thought is given to evaluate the overall acceptability of ETR.

4. Literature Review

As a result of significantly increased environmental awareness in 1970s, a great number of economic studies has emerged analyzing the effectiveness of environmental tax as a policy tool to mitigate unsustainable use of resources and global climate warming (Mishan, 1971), (Thompson and Batchelder, 1974). However, the problem of distributional impacts attracted more attention only in 1990s when Pearson and Smith (1991) published their study opposing environmental tax on energy in the UK. It was argued that, while such a tax may bring overall welfare improvements, it is socially unjust, since the poorest group of households would be forced to cut their consumption drastically and bear a disproportionate financial burden, while the behavior of the richest class would be merely affected.

This claim seems to have remained valid till now-a-days, since most of the research focusing on the distributional impacts of energy taxes, report that such taxes tend to be income regressive. Barker and Kohler (1998), for example, in their study based on dynamic macroeconomic models, covering eleven EU countries, concluded that energy taxes affect the lowest earners most severely, since energy is a necessity good. Symon et al. (2002) confirmed the same results for Germany, France and Italy and to a lesser extent in Spain. Yet, it should be noted, that due to complexity issues, most of the studies ignore behavioral changes resulting from higher energy prices, by inclusion of which, the regressive impacts become considerably milder (OECD, 2010). In the words of Schlegelmich, there are numerous possibilities for households to reduce their burden imposed by the energy tax by simply changing simple behavioral patterns (Schlegelmilch, 2003).

(7)

7

Moreover, the same distributional pattern do not necessarily hold for other environmental tax categories. A clear illustration is the transport sector. Empirical evidence suggests that motor fuel tax is progressive at the low end of income scale but regressive at the high (Blow and Crawford, 1997), (Smith, 1992), (Johnstone and Alavalapati, 1998). The main reason for such results, according to Leipprand et al. (2007), is the fact that the lowest-income households often do not have cars and therefore are less affected by increase in fuel taxes. In their analysis covering five EU member states, they found that the highest relative burden is put on the middle- income groups.

The results presented above, however, do not take into account another crucial component of ETR, namely revenue recycling policies. That the final distributional impacts essentially depend on the chosen revenue recycling mechanism was proven by numerous other studies, theoretical as well as empirical. Among the purely theoretical models mostly based on general equilibrium analysis, Parry (1995) found that the benefits from using pollution tax revenues to reduce labor taxes offset the costs of higher price level. Similar results were presented by Assouline (2009) in the theoretical distributional analysis with heterogeneous workers. He concluded, that even in the presence of many worker classes (heterogeneous labor) an environmental tax reform can be designed in a way that no worker class remains harmed. The study emphasizes, that in majority of the cases a progressive tax would serve the purpose better than equal tax cut in the labor tax rates. Within the class of empirical researches, West and Williams (2004) found that an initially regressive gasoline taxes under certain tax system adjustments can end up being progressive. This is in line with the findings of Peter at al. (2007) who compared the effects of environmental taxes across EU countries. When redistribution policies were neglected, the study confirmed generally regressive impacts on households. However, when tax design was included into the analysis, the negative effects were neutralized.

Such results therefore induced research aimed at finding the most effective tax revenue recycling policies, which can be divided into two broad categories: taxation and compensation policies.

On the taxation side, Leipprand et al. (2008) proposed to mitigate regressive effects by imposing higher tax rates on goods and services mostly used by high-income groups, such as airline tickets. This view was supported by Aasness and Larson (2002) who argue that taxing differently luxury transport modes from low polluting public transport modes would be a sound policy from environmental as well as social point of view. Among other approaches, tax exemptions for most severely affected industries was widely discussed, however this reduces incentives for firms to be more efficient in energy use. The more environmentally sound option would be to apply progressive taxation according to energy consumption levels (GBG, 2008).

As far as compensation policies are concerned, McNally and Mabey (1999) proposed to introduce lump-sum payments to compensate low-income households for any losses caused by higher energy taxes. They also suggested that a part of the collected tax revenues, should be used to launch a program aimed at improving home insulation systems starting from the lowest earners. Similar approach was proposed and implemented in the Netherlands, where until 2003, 15% of the environmental tax revenue were used to reward private households who installed more energy-efficient heating systems, however no priority was given for lowest earner, who most probably were unable to participate in such program. (EEA, 2011). Different compensation concept, named as eco-bonus was introduced by Bach et al. (1994). In this case, households would receive refunds from the government according to their energy consumption levels.

To sum up, without revenue recycling policies, environmental taxes have undesirable consequences for income distribution. Especially it is the case with energy taxation which has been found to be income regressive. However, by properly constructing tax revenue recycling mechanism through taxation or compensation policies, the distributional differences can be mitigated or even eliminated.

(8)

8

5. Description of the Model

With the aim to analyze the distributional impacts of the major ETR on the EU scale, the distributional estimates published by EEA in their technical report are used (EEA, 2010). The estimates were obtained by ‘Energy-Environment-Economy Model of Europe’ (E3ME) model constructed by Cambridge Econometrics Consultancy Company (Cambridge Econometrics, 2009). This grand-scale econometric model is able to incorporate the dynamic relationship between level of environmental emissions, energy market and economic processes in 42 different sectors across all members of the EU. Consequently, by means of this modeling exercise it is possible to estimate the changes in prices and income resulting from lower levels of greenhouse gas emissions, which are assumed to be reduced by imposition of CO2 tax.

The format of ETR used in this model is constructed as follows: a carbon tax is implemented at the rate that a 20% reduction target by 2020 would be met, according to the Kyoto Protocol. Tax revenue collected from businesses is devoted to reduce employer’s social security contributions (SSC), while the revenues received from households are used to cut income tax rate.

The mechanics of the model are summarized briefly below:

The ETR is implemented. As a result of a more expensive CO2, the price of energy and fuels

increases. This in turn raises prices within the industries, in which production is energy intensive. Depending on the degree of international competition, part of the tax is passed on to consumers, that translates in higher prices for goods whose production is energy intensive. Since those goods are part of consumer basket, the overall price level increases as well. Equivalently, purchasing power of the nominal wage decreases. On the other side, reduction in income tax rate affects positively net wage which leads to higher labor supply. Furthermore, higher net wage allows consumer to spend more, as a result of which aggregate demand increases. In addition, reduction in employer’s SSC stimulates labor demand as firms face lower labor costs. Finally, evaluating all the effects, the disposable income for each classification group is calculated for the year 2020, by subtracting labor taxes from real income, which consists of wages, pensions and other state benefits. By comparing the results against baseline scenario without ETR package, the percentage changes in disposable income for each socio-economic group can be estimated.

For illustrative purposes, the simplified graphical representation is provided below.

The free market price P0 causes too high level of pollution, so that environmental tax is introduced.

This leads to lower damage for environment, as production of polluting good decreases from Q0 to Q1,

however, at the expense of higher price for consumers. On a positive note, government receives tax revenue equal to a red colored area, which can be used to correct distortions in the labor market.

(9)

9

By reducing labor tax rates in the form of income tax and social security contributions, the efficiency gain is realized marked by a blue area. Lower income tax rate increases net wage and stimulates labor supply, while reduction in social security contributions allows market gross wage to decrease, which induces higher labor demand. The latter effects are demonstrated in the figure below.

The final effects on real income are ambiguous, since it is possible that loses from higher prices outweigh the benefits from lower labor taxes. The determination of final outcome depends on variety of factors specific to each country.

As it was noted, the analysis is modeled with several scenarios. Firstly, the baseline scenario is run without ETR, against which different scenarios including ETR will be compared. Since international oil price is an exogenous variable in this model, two different baseline scenarios are run, one with high oil price and another with low one, all other input values remaining identical. This is done due to the hardy predictable nature of future oil price which is an essential factor in designing any environmental policy. Below, all scenarios used in this thesis are shortly described.

• BSL (Baseline Scenario with no ETR, Low oil price = 60$ per barrel). Baseline scenario was modeled to match official forecasts of employment level, energy trends, GHG emissions and other economic variables published by the European Commission. Other statistical information, such as households spending survey data, was used from Eurostat. The scenario is run till 2020, consistent with the period for the CO2 target to achieve.

• BSH (Baseline Scenario with no ETR, High oil price = 120$ per barrel). Apart from high oil prices, all input variables are the same as in BSL.

• LS1: (Scenario when ETR is implemented, Low oil price). CO2 tax is introduced, 20%

reduction in CO2 emissions target is met. Tax revenue translates in corresponding decrease in

SSC and income tax rates.

• HS1: (Scenario when ETR is implemented, High oil price). CO2 tax is introduced, 20%

reduction in CO2 emissions target is met. Tax revenue translates in corresponding decrease in

SSC and income tax rates.

• HS2: (Scenario when ETR is implemented, High oil price). CO2tax is introduced, 20%

reduction in CO2 emissions target is met. 90% of tax revenue translates in corresponding

decrease in SSC and income tax rates. The remaining 10% is invested in eco-innovation, which in this exercise is composed of three elements: investment in alternative energy research, efficient house insulation system and vehicle efficiency.

(10)

10

The model calculates the changes in real income for households, according to their income, employment status and living area in each of the three scenarios, namely LS1, HS1 and HS2. Scenario LS2 is absent in this thesis following the original report by EEA. Yet, it is worth to note that from a perspective of comparative purposes, it would not add significant insights. The described three scenarios are enough to evaluate all the distributional differences arising from changes exogenous variables, which in this case are the design of ETR and international oil price. The results under LS1 show how much change in real income ETR produces for households in the case of low oil price. By comparing, LS1 with HS1, the impact of oil price can be evaluated, while by comparing HS2 with HS1, the effectiveness of two different ETR packages can be analyzed.

6. Interpretation of the Results

In this chapter the main analysis of the problem is being done. Subchapter 6.1 deals with income distribution classified by expenditure groups, 6.2 by socio-economic groups, while part 6.3 analyses such effects according to population density. Each subchapter starts with the table of distributional estimates from EEA report. The results are used for the establishment of more general distributional patterns resulting from ETR reform. In addition, each section offers policy measures, by means of which distributional differences could be mitigated.

6.1. Expenditure groups

The table given below shows the percentage change in real income from the appropriate baseline scenario for each expenditure group in 2020 for the EU-27 as a whole. The first quintile represents the lowest-income class.

First of all, it can be observed that ETR on the EU scale leads to a Pareto-improvement in any scenario, as every group gains from the reform. It is worth to note however, that while on average the effect is positive, it is likely that within each group there are households whose income decrease. This seems to be especially the case within the lowest income class in which the amount of energy consumed is found to be widely distributed about the mean of that particular group (OECD, 2006; Seret & Johnston, 2003). However, in view of the aim to establish only general patterns without delving too deep into complexities, the following analysis concentrates solely on differences between the groups and not within.

(11)

11

Results under the main scenario (LS1) show, that it is the middle income class who gains relatively the least from ETR. The explanation for this is that households belonging to this group on average spend the largest part of their expenditure for transport and heating fuels combined and therefore loose the most from higher prices of fuel. Groups with lower income are less affected by changes in transport fuel prices as they are observed to travel more by public transport than cars (Ekins et al., 2011). On the contrast, the richest class has the lowest share spent on fuels from all groups and therefore gains relatively the most from ETR. The validity of this argumentation is supported by the fact, that an observed distributional pattern is very similar to the one found by previous studies on transport taxation. Namely, it was observed that the tax burden is progressive at the low end of income scale but becomes regressive at the high.

The same distributional trend just in smaller magnitude exists under HS1 scenario modeled with high international oil price. The smaller overall gains are attributed to the fact that in in the presence of already high fuel price there is much less space to increase its price even more by imposing an environmental tax. As a result, tax revenue is lower than in the case of low oil price. This implies that positive effects resulting from revenue recycling mechanism become smaller as well.

On the contrast, in HS2 scenario which assumes that 10% of tax revenue is invested in eco-innovation, the distributional effects are regressive, benefiting the lowest earners the least. The change in distributional pattern from standard scenarios is attributed to the two main reasons. Firstly, one of the implications of investment in a low-carbon technology is an increase in vehicle efficiency. This effects is relatively more advantageous for upper income groups as they have higher travelling expenses. Secondly, richer classes are in better position to grasp the benefits of eco-innovation since they have more available income to apply those inventions in their house heating systems or other energy-intensive equipment. Despite the appearance of regressivity in the distributional gains among income groups, ETR package including investment in eco-innovation is preferred to the standard ETR assumed in HS1 scenario for all expenditure groups. This is due to the fact that government uses part of the tax revenue to renovate house insulation systems and public transport infrastructure. In the longer run it brings additional positive effects for all households in the form of lower expenses. Overall, looking at the EU-27 as a whole, the distributional impacts of ETR vary with each income group. However, the differences are relatively small, not exceeding 0.3%, so that by incorporating additional policies into ETR package such as tax exemptions or allowances for the most negatively affected, the disproportionate effects can be eliminated. After all, the ETR used in this modeling exercise is only a proxy of an actual ETR.

To evaluate the attractiveness of the major ETR, it is important to look at how each country is affected separately. Since countries within the EU differ in terms of economic structure as well as environmental pollution levels, the same ETR package is likely to produce different results. If effectiveness of ETR is expected to diverge a lot across the EU, to reach a political consensus on the common ETR implementation becomes a hardly possible task to achieve. The figure presented in the next page shows that such problem indeed exists.

(12)

12

The graph shows distributional patterns under the main LS1 scenario for each EU member individually. While distributional differences within each country are relatively small, usually within 1% band, the variation between countries is significantly larger. For example, in Slovakia, ETR is estimated to benefit all income groups by approximately 6% while in Greece no group gains more than 0.5% including the lowest income class, which actually looses from ETR. Such divergence in the distributional effects raises important problems for a grand-scale ETR implementation on the EU level. To explain these differences is beyond the scope of this thesis, therefore only some general insights are provided below.

The ETR seems to be the most unfavorable to Greece, Spain and Ireland. In those countries, according to this model, the distributional impacts of ETR will be regressive. Moreover, the lowest income class is even expected to experience a decrease in real income. One of the possible reasons of such phenomenon is that these countries have faced one on the highest unemployment levels across the EU in the last years. This suggests that persons belonging to the lowest income quintile are more likely to be unemployed and benefit nothing from reductions in income tax rate (Eurostat, 2013).

On the contrast, in countries such as the Netherlands, Austria or the United Kingdom, where unemployment rates are relatively lower and public transport is a convenient substitute for car, it is no longer the lowest income group who is relatively the worst affected but the middle one.

(13)

13

While it is not applicable for all countries, it can be observed that on average, older member states gain a little less from ETR than the states which have joined the EU from 2000 onwards. Since partial ETR’s are already implemented in older member states, part of positive effects resulting tax shifts have been already realized. On the contrast, newer member state have significantly less experience with ETR and therefore have more space for a potential gain (Ekins, 2009). It has to be noted however, that from political point of view, new member states generally face lower standards of living and therefore ETR is more likely to face more political resistance than in older states.

Another factor substantially influencing the success of ETR is international trade patterns of each country. If energy-intensive goods constitute a large share of a total export volume, higher energy prices resulting from ETR impose a severe disadvantage for export levels. This seems to be the case in Spain and Greece, where it was observed that due to an increase in carbon tax rate, their export volume decreased substantially. Consequently, this had a negative effect on national income and employment level (Speck, 2009). On the other hand, countries competing in international markets within labor intensive industries, were able to increased their competitive position due to lower costs caused by a decrease in social security contributions for employers.

The structure of labor market plays a key role as well. For example, countries, in which wage levels are subject to strict agreements between worker unions and government, cuts in labor taxes cannot produce their full positive impacts on employment and national income since the wages remain unaffected. Such reason is given by Speck (2009) to explain why all income groups in Sweden are expected to gain very little from further ETR in terms of real income. Reasoning in a similar way, differences in labor taxation system as well as labor law have important consequences for the effectiveness revenue recycling too.

The observed differences suggest, that the most efficient ETR design is subject to specific characteristics of each member state. For example, countries in which energy-intensive sectors play a large role in foreign trade, in order to avoid loss of competitiveness, need to devote higher share of tax revenue to subsidize those industries. Alternatively, internationally competing industries could be granted tax exemptions in exchange for meeting energy efficiency or carbon savings targets. Such practice is already applied in the UK, Denmark and the Netherlands. In any event, less tax revenue is left to compensate distributional inequalities between households. Consequently, this constitutes ETR package characterized by higher emphasis on the sectorial level.

On the contrast, countries more relying on labor-intensive industries benefit the most by focusing on reductions in labor tax rates. In the standard scenario, reductions in social security contributions for employers allow them to expand production and increase efficiency resulting from economies of scale. On the other part, cuts in income tax rate benefit workers in the form of higher net wage. If direct compensation measures for inactive population are added as well, such ETR package becomes an attractive option on an individual level. However, as most of the revenue is used to correct distributional inequalities among households, such ETR is less favorable to energy-intensive industries.

For the countries having highly regulated labor markets however, tax revenue recycling through labor market is less effective form to tackle distributional concerns. Energy tax allowances or credits, as well as financial support for eco-innovation are more beneficial components to include in the ETR package.

The considerations above therefore indicate that depending on particular features to each country, the distributional patterns as well as overall effectiveness of ETR vary substantially. It follows then, that ETR package should include some components which are obligatory to implement for all members, while also containing flexibility clause, so that each country could address its national distributional concerns in the most efficient manner.

(14)

14

6.2. Socio – Economic Groups

The table given below shows the percentage change in real income from the appropriate baseline scenario for each socio-economic group in 2020 for the EU-27 as a whole.

The results show that ETR widens the division between those who have a job on the one side and those who rely on non-labor income on the other side. This comes from the fact that households relying on non-labor income do not benefit anything from cuts in income tax rate (Bovenberg, 1999). These groups however, still experience an increase in real income since in this modeling exercise it is assumed that income transfers from government, such as pensions or unemployment benefits, are linked to average wage level. If such transfers are linked to consumer price index instead or are fixed in nominal value, the groups relying on them are affected negatively (EEA, 2011). Such phenomenon inevitably raises concerns that the distributional impacts of ETR are socially unjust, since it disadvantages the most vulnerable groups. This issue becomes even more sensitive in the presence of the ageing European population (EC, 2014).

In order therefore to make ETR acceptable for all groups, the compensation measures should be introduced. The most direct solution is to provide lump-sum transfers for groups loosing from the reform. Retired people in particular are in need of such policy, as they usually have the least opportunity to increase their income or change consumption habits. In the case of unemployed people, a reduction in real value of unemployment benefits provides greater incentives to accept job offers, so that a positive effect on employment level can be expected. Alternative approach is to set the consumption of energy floor below which heating is taxed at a lower tariff. This policy is an attractive option as it induces households to save energy in the form of monetary incentives. Overall, both policies help to mitigate unequal distributional effects, yet it is unlikely that such division would be entirely eliminated.

On a final note, it is important to mention that revenue recycling policies in the form of lower social security contributions for employers combined with cuts in income tax rates reduce an employment loss caused by labor taxes. In other words, due to lower SSC firms have more funds to hire additional workers, while lower income tax rates induce unemployed persons to accept more job offers since their net wage increases. As a result, ETR provides more favorable conditions for unemployed to get a job and receive higher income. For example, it is estimated that the ETR implemented in Germany between 1998 and 2003 was responsible for the creation of 250.000 additional jobs (Knigge & Gorlach, 2005). Throughout this exercise however, it was assumed that persons do no not change their employment status during the period examined, as a result of which, gains for unemployed were underestimated.

(15)

15

6.3. Population Density

The table given below shows the percentage change in disposable income from the corresponding baseline scenario for households living in urban and rural areas separately. The changes are calculated for the year 2020 for the EU-27 as a whole.

Analysis of the distributional effects according to living area provides important insights as well. The results in the table confirm the conclusion reached in earlier literature, namely that people living in rural areas gain less than urban population (Speck et al., 2006, Wier et al., 2005). This result holds under all scenarios.

The difference is attributed to different transportation and heating patterns between urban and rural households. Previous research on this problem has found that usually persons living in a rural area have larger houses and less efficient heating insulation systems than urban population. Both factors result in higher sensitiveness for an increase in energy prices (Leipprand, 2007).

As far as transportation is concerned, it is often the case that people living in rural areas have higher marginal fuel consumption ratio since they need to travel longer distances to reach workplaces or other destinations concentrated in the city (McNally and Mabey, 1999). This problem becomes especially visible if infrastructure of public transport connecting rural and urban areas is not well developed. If this is the case, rural residents are left with almost no choice to substitute traveling by car to other means of transport as opposed to persons living in the cities. Consequently, under the main scenario ETR poses a threat to widen inequality between rural and urban population, the problem which has been already addressed as a very important one (WHO, 2010).

As the results from H2 scenario demonstrate, when part of the tax revenue is recycled through investment in more efficient insulation and heating systems, as well as vehicle efficiency, the gap between changes in real income of rural and urban households becomes significantly smaller. This suggests that by spending more of the tax revenue in this way, especially if priority is given for rural population, ETR can secure support from urban as well as rural groups. Another approach to ensure an equal distribution would be to establish compensation schemes directly aimed at rural inhabitants. For instance, government can commit itself to cover part of transport fuel expenses for those who need to travel long distances to reach their working place in the cities. In addition, part of the tax revenue should be invested in public transport development between rural and urban areas, as it is the most efficient solution in the longer run from economical as well as environmental point of view.

(16)

16

7. Summary and Conclusions

Firstly a historical background of ETR as well as its application throughout the European Union was provided. This was followed by a literature review which served as the basis to explain the distributional patterns in the analysis part. Next chapter described the model used by EEA to generate distributional estimates of ETR. After this, the main interpretation part followed in which results published in EEA report were reviewed and contributed my interpretations in order to establish general principles governing income distribution caused by ETR. Finally, policy options to ensure a distributional equality were considered.

Evaluating the results from different scenarios, it can be concluded that in the case of high international oil prices, there is less scope to grasp the benefits from ETR and therefore the overall attractiveness of ETR becomes significantly smaller than in the presence of low oil prices. The comparison between different scenarios also reveals that ETR package including investment in eco-innovation produces the most equal distribution of benefits, suggesting an important insight for policy makers.

Analysis of distributional effects between households reveals that at the EU-27 level ETR would produce positive change in real income for all socio-economic groups, however the amount of gain differs according to income and employment status as well as living area. Without direct compensation measures, ETR constitutes a shift of tax burden from those having a work to unemployed, inactive and retired, since they face an increase in general price level but do not benefit from cuts in income tax rate. Consequently, the ETR is most likely to be opposed by those groups as socially unjust. As a solution, double energy tax tariff could be introduced. An advantageous tax rate would be applied to households with low energy consumption. An alternative approach would be to apply lower tax rate for disabled persons, retired or other vulnerable groups within the society.

The problem of unequal distribution arises when such effects are evaluated from regional perspective. In all scenarios, households living in rural area on average gain less than urban groups as a result of higher fuel and heating expenses. This in turn threatens to widen an already existing territorial inequality within some members of the EU and consequently reduces attractiveness of the major ETR, at least in the form used throughout this modeling exercise. Yet, by adjusting ETR design, the distributional differences in terms of real income can be greatly reduced. Compensations to rural households for their traveling-to-work expense or improvement in public transport connectivity between rural and urban areas would be an examples of such adjustments.

As far as classification based on income is concerned, on the EU average, the middle income group would have the smallest positive effect in income due to the largest budget share devoted on fuels. It is should be noted however, that in some countries ETR imposes the greatest burden on the lowest earners even producing a negative change in real income. The variation in distributional trend across the countries is a result of different patterns in consumption and income earning, as well as different levels in fuel and heating efficiency.

In addition, substantial differences of overall effectiveness among member states have been found. This suggests that in order to construct an acceptable ETR for all member states of the EU, more research has to be done to gain sufficient understanding of the factors causing these differences. Only this way, it is possible to design ETR which would be economically and socially sound. Nevertheless, it is clear that ETR definitely has to have some flexibility, so that each country could make slight adjustments according to their national policy and economic structure.

(17)

17

For instance, countries identified by more internationally competing energy-intensive industries are more concerned to preserve competitiveness and therefore seek to apply less aggressive energy tax rates. On the contrast, countries relying on labor-intensive export have higher incentives to spend more tax revenue through labor market channels.

In general, it is obvious, that some undesirable distributional patterns are likely to occur as a result of ETR implementation within the EU, however in most cases the differences are small enough to be mitigated by specific redistribution and compensation policies. While the concept of ETR probably has not yet matured to be applied widely in practice, due to still a lot of uncertainty surrounding its final effects, with more research and political will devoted to improve its design, ETR can become a successful example demonstrating how environmental and economic benefits can be achieved at the same time.

Reference List

Aasness, J. and Larson, E.R. (2002), “Distributional and environmental effects of taxes on

transportation”, Statistics Norway Research Department, Discussion Paper No 321, Oslo, Norway.

Assouline, M., C. (2009), “Environmental Tax and the Distribution of Income with Heterogeneous

Workers”, Paris School of Economics, Paris.

Bach, S., Kohlhaas, M., Praetorius, B. (1994), “Ecological Tax Reform even if Germany has to go

alone”, Economic Bulletin, 31, pp.32-41.

Barker, T. and Kohler, J. (1998), “Equity and Ecotax reform in the EU”, Fiscal Studies 19(4), pp. 375-402.

Blow, L. and Crawford, I. (1997), “The Distributional Effects of Taxes on Private Motoring”, Commentary no. 65, Institute for Fiscal Studies, London.

Bovenberg, A. (1999), “Green Tax Reforms and the Double Dividend”, International Tax and Public Finance, 6(3), pp.421-443.

Chewpreecha, U. and Pollitt, H. (2009), “Extending the E3ME Model to 2050”, Cambridge Econometrics, Cambridge.

Ekins, P. and Speck S. (2009), “Environmental Tax Reform: resolving the conflict between economic

growth and environment”, Oxford University Press, Oxford.

Ekins, P. and Speck S. (2011), “Environmental Tax Reform: A Policy for Green Growth”, Oxford University Press, Oxford.

European Commission (2009), “Impact Assessment Guidelines”, SEC (2009) 92, Brussels. Belgium. European Commission (2014), “The 2014 Ageing Report”, European Economy 8/2014, Brussels. Belgium.

European Environmental Agency (2005), “Market-based Instruments for Environmental Policy in

Europe”, Technical report No 8/2005, Copenhagen, Denmark.

European Environmental Agency (2010), “Tracking Progress towards Kyoto and 2020 targets in

Europe”, EEA report No /2010, Copenhagen, Denmark.

European Environmental Agency (2011), “Environmental Tax Reform in Europe”, Technical report No 16/2011, Copenhagen, Denmark.

(18)

18

Fullerton, D., Leicester, A., Smith, S. (2008), “Environmental Taxes”, National Bureau of Economics Research, Working Paper 14197, Cambridge.

Institute for European Environmental Policy (2010), “Evaluation of Environmental Tax Reforms”, Brussels, Belgium.

Green Budget Germany (2004), “Green Budget Germany demands for ecotax and fiscal reform in

Germany from 2005”, ECOTAX, GBG Memorandum.

Johnstone, N. and Alavalapati, J. (1998), “The Distributional Effects of Environmental Tax Reform”, International Institute for Environment and Development, Discussion Paper 98-101.

Hoerner. J. and Bosquet, B. (2001), “Environmental Tax Reform: The European Experience”, Center for a Sustainable Economy, Washington, US.

Knigge, M. and Gorlach, B. (2005), “Effects of Germany’s Ecological Tax Reforms on Employment

and Technological Innovation”, Institute for International and European Environmental Policy,

Berlin, Germany.

Leipprand, A., Gavalyugova, N., Meyer-Ohlendorf, N., Blobel D. and Persoon, A. (2007), “Links

between Social and Environmental Pillars of Sustainable Development”, Ecologic, September 2007.

McNally, R. and Mabey, N. (1999), “The Distributional Impact of Ecological Tax Reform”, WWF UK.

Mishan, E., J. (1971), “The Postwar Literature on Externalities: An interpretative Essay”, Journal of Economic Literature, 9 (1), pp. 1-28.

OECD (1995), “The Political Economy of Environmentally Related Taxes”, Organization for Economic Co-operation and Development, Paris.

Parry, I. (1995), “Pollution Taxes and Revenue Recycling”, Journal of Environmental Economics and Management 29, pp. 64-77.

Pearson, M. and Smith, S. (1991), “The European Carbon Tax: An assessment of the European

Commission’s proposals”, London: Institute for Fiscal Studies.

Pye, S., Skinner, I., Leipprand, A. (2008), “Addressing the Social Dimensions of Environmental

Policy”, EAE Energy and Environment, Ecologic University of Westminster for the European

Commission.

Smith, S. (1992), “The Distributional Consequences of Taxes on Energy and the Carbon Content of

Fuels”, European Economics, Special Edition no. 1, pp.241-68

Speck, S. (1999), “Energy and carbon taxes and their distributional implications”, Energy Policy, 27, pp. 659-667.

Symons, E., Speck, S., Proops, J. (2000), “The effects of pollution and energy taxes across the

European Income Distribution”, Keele Department of Economics, Discussion Paper 2000/05.

Thompson, E. and Batchelder, R. (1974), “On Taxation and the Control of Externalities”, The American Economic Review 64 (3), pp. 467-471.

United Nations (1992), “United Nations Framework Convention on Climate Change”, New York City, US.

WHO and European Commission (2010), “Rural Poverty and Health Systems in Europe”, WHO Regional Office for Europe, Copenhagen, Denmark.

(19)

19

The distributional effects of Environmental Tax Reform on the European Union