The other curve in Figure 1 is a measure of the marginal damages caused by the emissions. These would consist of increased incidence of respiratory diseases, damage to property, crops, water bodies etc. The marginal damage curve MD is drawn assuming that, at each stage, individuals respond to any increase in the emissions so as to minimize the impact it has on them. The shape of this curve is less clear but it is commonly assumed that, as the environment improves (i.e., emissions are reduced), the marginal value of a further improvement declines. Hence the shape drawn in Figure 1.
The "optimal" level of emissions is given by OE, which is the point at which the marginal cost of abatement and the marginal damage are equal. An increase in emissions from that point would entail damage to the environment greater than the savings in abatement cost; and a decrease in emissions would imply a cost of abatement in excess of the damage reduced.
How is such an optimum to be achieved? One method would be to estimate the two curves and then calculate OE. Then each polluter would be given a quota of emissions such that OE was achieved in total. This would involve detailed regulation and control, increasing with the number of polluters, as well as a detailed knowledge of the costs of abatement in each industry. Furthermore, there could be complaints from polluters who were not being treated equally, i.e., some polluters were being required to make bigger reductions in order to satisfy the conditions of the MAC curve, which is the least-cost-of-abatement curve. In practice, regulators are unable to enforce such a solution and so, when going for direct controls they tend to impose the same technology requirements on all polluters. However, this is no longer a least cost solution, even assuming that the reduction from OA to OE is achieved. Furthermore it does not provide incentives for continued research on methods for reducing emissions, as there are no rewards for further reductions beyond those specified in the regulation. For both these reasons, therefore, it is not a least cost solution.
The economically optimal answer could be obtained by imposing a tax of EE* on each unit of emissions. With such a tax, each polluter would reduce emissions to the point where its MAC was equal to EE*, and thereby achieve the optimal reduction at least cost. There would also be a continuing incentive to reduce emissions as these would be rewarded by a savings in tax.
The same solution could also be achieved by giving a payment (subsidy) to each polluter of EE* for each unit of pollution reduced from the present level[2]. Then polluters would adjust emissions to equate the subsidy to MAC because the subsidy is the implicit price received per unit of emissions reduced.
Finally the same solution could be obtained if the authority issued emissions permits equal to OE and allowed them to be traded in a competitive market. The market would ensure that each polluter purchased permits to the point where the price was equal to its MAC, which then also equal MD.
Generally the choice between taxes and subsidies is resolved in favor of taxes. This is because:
The choice between taxes and permits depends largely on how one perceives the uncertainty associated with these instruments. Since neither the MAC nor the MD curves are known with any precision, a tax would require estimating the value of EE* while a permit system would require estimating the value of OE. If the tax is estimated wrongly, the level of pollution reduction will be "incorrect" - too high or too low. However, the maximum cost impact on industry is not uncertain; as the charge is known, as is the current level of pollution. On the other hand if the level of permits is wrong, the cost impact on industry could be substantial as it will be forced to meet the new target in reductions. The argument is not resolved, but there are more applications with taxes than there are with permits. Permits also require fairly sophisticated market structures where prices are fixed and trades made. These conditions are not always set up, even in developed countries.
In view of these constraints any "real world" application of environmental taxes will be quite far removed from the ideal model outlined above. The main forms that such taxes take have been summarized by Opschoor and Voss [1989][3] as:
In addition, there are other instruments which will not be dealt with in depth in this paper, namely:
Effluent charges have the advantage that they are as closely related to the environmental damage as possible. They are usually levied at the regional level. Unfortunately, in many cases they are not practical for the reasons given above. That is why there are few examples for effluent charges as such. In air pollution, there are very few (see next section). For water pollution, waste and noise, there are more but even when they are imposed, they are always accompanied by other direct regulations. It is important to note also that such effluent charges as have been imposed have not made any significant use of the marginal damages. In other words they are not 'optimal' charges in the sense described above. In fact, in a number of cases the link between the charge and emissions is not that strong (see next section).Product charges are more common, and are applied when emissions cannot reasonably be monitored directly. They have the disadvantage that they do not capture all the linkages between the production process and the pollution. Thus, two enterprises, one of which has very low emissions, because it uses the latest technology, and the other of which is extremely polluting, will pay the same charge per unit of output. There are ways to get around this. One is to give a rebate, depending on the technology being used. A second is to offer a subsidy for the installation of emissions abating equipment. Product charges have been used for beverage containers, fuels, plastic bags, feedstock, fertilizers and pesticides. As with emissions taxes, there are varying degrees to which the charge is linked to the pollution level. This depends partly on the stage in the production/consumption cycle that the tax is levied. The advantage of placing the tax at an early stage in the product cycle is the savings in administration. The disadvantage is that it does not capture the environmental damages that arise in the later stages. For example, a tax on pesticides at the manufacturing stage will not provide incentives to avoid poor practices in storage and application. As with effluent charges, the link between the charge and the pollution that cases the environmental damage can be weak or strong, depending on how the tax is structured.
User charges are levied for the collective treatment of pollution. Almost exclusively they have been applied to water pollution related problems. The important point to note about such charges is that they have to be closely related to any effluent or product charges that are in place. For example the incentive effects of an effluent charge would be negated if enterprises could dispose of their waste water in a public sewerage system without regard to its pollution content. Even if the sewerage authority was obliged to clean up the water to a certain standard, the social costs of doing so would be higher than necessary.
Tax differentiation involves a reduction in the taxes levied on the more environmentally friendly activities at the expense of the less friendly ones. As such it is a special case of subsidies, as the government is giving up some revenue to foster a certain improvement in environmental quality. Tax differentiation has been used to encourage the shift to unleaded gasoline, and to encourage energy efficiency in vehicles.
Although the case against earmarking is a strong one, there are also arguments on the other side. First, the support for environmental taxes may only be forthcoming if the revenues are to be allocated for specific purposes. In this case it would be better to accept earmarking as the 'price' of having such a tax in the first place. However, there is a more positive argument. One can think of some environmental taxes as charges, for the use of specific environmental services, such as rivers, air sheds, land resources etc., that belong to the nation. Interpreting them in this way, it is appropriate that the government should charge those who utilize or degrade these resources and use the revenue to maintain the quality of that service. This would certainly justify direct expenditures on an earmarked basis for the maintenance of water resources, or protection of waste sites from extensive environmental degradation out of the taxes collected. It is less clear whether it could justify the financing of abatement investments in the activities that were generating the pollution in the first place. However, if that is the least costly way of reducing the damage then, arguably, such payments should be acceptable as a category of expenditure from the charge.
Earmarking can play an important role for funding environmental protection in a declining economy. A huge budget deficit tends to encourage very short-sighted decisions regarding resource allocation. The lack of a longer term view in the budget discussions makes it inherently difficult to promote environmental interests in government budgeting. That fuels the desire of environmental policy makers to secure a separate fund. However, the experience with setting up such a fund in Hungary revealed a potential danger of aiming at such independence from the budget[4] in a period of massive institutional restructuring. This danger is that the countries in transition still do not have an established, detailed division of tasks and responsibilities among the central government, local governments, and polluters. The language of related legislations is rather general and vague, if not conflicting. Therefore, it is possible to designate too many tasks to be financed from the separate fund which may lead to a drastic cut back in budget allocations to the environment administration.
The use of the term 'charge' or 'tax' thus reveals important differences in principle. A tax is properly something that is part of the general revenue, either at the local, regional, or national level from which earmarking is not appropriate. On the other hand a charge is something that should be used to finance specific activities related to the resource that is being charged. In the context of environmental taxes and charges there are no established guidelines that can be used to dictate which items should be treated as taxes and which as charges. One suggestion would be to treat all collection of revenues from effluent fees as charges, as they are related to specific environmental services. On the other hand, revenues collected from product charges should be treated as taxes, because of their indirect relationship to the environmental service. This is, admittedly, a somewhat arbitrary division, but it does have the advantage of making a clear distinction, based on the closeness of the link between the source of the revenue and the treatment of the revenue.
One must note, however, that the incentive effect and stable revenue raising potential are not always complementary properties of an environmental charge. We are likely to end up addressing different environmental problems with a different design of economic instrument if the objective is providing stable revenue rather than creating incentive to achieve a desired level of pollution. Stable and predictable revenue sources are broadly based taxes on activities with a low price elasticity. However, that translates into small incentive effects. Therefore, designing an actual environmental charge with the potential of 'double dividend' is not a simple task.
Another way of looking at the same issue is to speak of making changes in the tax system that are 'fiscally neutral'. The term fiscal neutrality is a complex one and has several interpretations which are not relevant to this discussion. The interpretation being used here is to make changes in the other taxes in the system so that the net impact on the economy is unchanged when an environmental tax is introduced. One way in which this could be measured is to hold real total tax receipts constant. The first point to note is that this would be a legitimate goal only if the environmental fee is in fact a tax, and not a charge. In the case of a charge, the additional revenues are earmarked and fiscal restructuring is not relevant. If, however, one is dealing with an environmental tax, then allowance for changes in other taxes so as to leave the net tax burden constant could be of enormous importance. It would permit higher rates of such taxes to be imposed, and it would reduce the overall tax inefficiency of the system by allowing the most economically costly taxes to be reduced. The only area where this discussion has taken place is with respect to carbon taxes. Although none of the actual taxes imposed have been accompanied by observable changes in the tax structure, the analysis of impacts of such taxes has looked at the macroeconomic consequences of reducing, for example VAT, by an amount equal to the yield of the carbon tax.
In economies in transition, it is particularly important to look at such possibilities, as the tax system has not been fully cast, and there is more scope for trading-off one kind of tax against another than there is in systems with entrenched positions and vested interests in keeping the status quo. One hindering force, nevertheless, is present in the countries with high budget deficits. In these countries, revenue neutrality is not an acceptable goal for fiscal policy; they want increased revenue.
If a quantity based tax is used, it is very important to make sure that some mechanism exists for indexation to inflation. The impact of environmental taxes are often lost because of erosion as a result of inflation (see next two sections). The institutional limitations on this arise if any changes in the tax rates have to be subject to parliamentary or legislative approval. In that event, changes are more difficult to make and the erosion on grounds of inflation is greater. If, however, the authority to index rates is given to an administrative body, the problem can be substantially resolved with little difficulty.
If, on the other hand, the tax is based on products, it will have to be integrated into the main tax system. The need for simplicity of collection will dictate that an established tax base be used for the tax. Since this is operated by the finance authority of the country, it would only make sense for the same authority to be the one that collects and administers the environmental tax. The same applies where tax differentiation is used as the relevant instrument. In both cases, the revenue collecting ministry will have to allocate a budget to the spending ministries, on the basis of nationally identified needs. Where the instrument is a user charge, it is appropriate that the authority that provides the environmental service be the one that collects the fee. The fee should, moreover, cover the costs of providing the service. Whatever that fee is, should also be taken into account when fixing the environmental charge or product tax (see below).
Much of the institutional debate on which authority has the responsibility for environmental taxation/charges hinges on the issue of what role the legislative body has in the process. Where approval by such a body is essential for any change in the tax or charge, the process becomes more politicized, and there is less scope for basing the charge on cost and benefit considerations. If, on the other hand, the charge can be authorized by a regulatory body such as the environmental agency, the chances of implementing an economically efficient charge are greater. In the context of environmental regulation, it may be possible to delegate the authority for an environmental charge to the regulatory authority if it can show convincingly that it will do so according to clear and acceptable principles.
With the exception of the water charge system in the Netherlands, there are no cases that have been examined where the user charges and the effluent charges have been demonstrably tied to one another. Yet if this is not the case, the regulation will not be efficient. The coordination of different charges and fees in waste water management is especially important for countries in transition while they are developing the financial scheme for waste water management. Properly designed systems of charges allow one to move toward "self financing" arrangements.
Effluent charges should also be set so that account is taken of the possibility of converting waste from one form to another (e.g., from solid to liquid or gaseous). If all disposal methods bear their full costs, the choice can be left to the polluter. However, as this is not generally the case, charges have to be set so as to take account of the linkages between the different forms of waste. If more encouragement is required for converting from liquid to solid waste, charges should be raised for liquid waste, and vice versa. In most countries, rates seem to be set independently of each other.
2. The subsidy could also be paid from a level of emissions different from that existing without regulation. In fact one of the problems with a subsidy solution is that the level from which the subsidy is paid (the benchmark) is not defined.
3. There is another form of environmental tax mentioned: the administrative charge which is a fee to be paid for some monitoring and administrative activities to the environmental agencies. This instrument is clearly not so important in terms of incentives for direct pollution reduction. They may have, however, important revenue raising potential for financing local monitoring and enforcement activities for environmental protection. Although, these activities are crucial to effective environmental policy, it is difficult to secure funding for them in the yearly budget negotiations, especially under severe budget deficit.
4. That aim is often called as establishing 'self financing' environmental protection
