Environmental taxation in open economies: unilateralism or partial harmonization.

• Author: Cremer, Helmuth

1. Introduction

   There are two international dimensions to environmental policies, each raising a specific type of question. First, when pollution is worldwide (e.g., emissions of greenhouse gases like C[O.sub.2]), we are effectively dealing with a global and essentially pure public good, namely environmental quality. (1) Different countries contribute to this public good or, more precisely, they contribute toward its degradation through their emissions. As long as there is no supranational government, one has a framework that resembles that of the voluntary provision of public goods. The problem here is that individual countries do not have the right incentives to take the welfare of the other countries into consideration. Their cost/benefit calculus does not account for the full cost of the emissions imposed on the rest of the world. Consequently, one can expect (noncooperative) national policies to lead to an excessive level of emissions. (2)

   The remedies that countries adopt to combat pollution introduce a second international dimension of their own. For example, when France unilaterally taxes domestic producers in order to entice them to cut their emissions, the price of domestic products will increase and consumers may turn to imported substitutes. This effect is, of course, neither intended nor in general positive for the French economy. In addition, it mitigates the environmental benefits of the policy because the nontaxed foreign producers can be expected to use dirtier technologies. This is indicative of the fact that unilateral environmental policies, regardless of the global or local character of pollution, are not immune to the phenomenon of fiscal competition. When tax bases are mobile, the capacity of an individual country to levy taxes is reduced. This problem arises for most forms of taxation, including for environmental levies. The fiscal competition aspect may then tempt an individual country to cut its environmental taxes in order to enhance the competitiveness of its economy as long as there is some degree of mobility of goods and/or factors. (3)

   The two dimensions of international environmental policies have thus far been studied independently of one another--at least in a general equilibrium framework. (4) Each dimension alone suggests a reason for the environmental quality to be inefficiently low. One may then be tempted to argue that, when the two elements are put together, they can only reinforce one another. However, the inefficiencies due to global nature of externalities and tax competition are not simply additive. We present a model where the two problems are accounted for simultaneously, thus providing a single framework to study the complexities that are brought about by their interaction. In our model, emissions vary with the level of output and the polluting technology employed. This feature captures the different roles that output and emission taxes may play in financing government expenditures and combating emissions--the two instruments do not collapse to one. This provides an appropriate framework to study a number of questions regarding the design of environmental policies. Does economic integrations, and the potential for tax competition it induces, necessarily lead to a decline in environmental quality? And if so, how should one go about remedying this problem? Will there be a targeting of tax instruments with output taxes used for financing public goods and tax competition and emission taxes solely for combating emissions? (5) Of particular importance, in this respect, is the implications of a policy of partial harmonization; that is, harmonization in only one instrument. We shall examine, in particular, if harmonization of output taxes intended to avoid tax competition are neutralized by an adjustment of other taxes, like those that are directly imposed on emissions. (6)

   Certain aspects of the questions we are raising have been studied in the literature. There is a huge literature on environmental dumping, which compares cooperative and noncooperative outcomes under trade; see Ulph (1997) for a survey. However, as a rule, this literature does not distinguish between emission and output taxes, ignores the question of the public good provision, and is cast in terms of competition between imperfectly competitive firms. In yet another approach, Antweiler, Copeland, and Taylor (2001) do distinguish between scale of output and the intensity of polluting technologies in determining emissions, but their concern is not tax competition and public good provision. (7)

   We consider a simple model of commodity tax competition. There are two identical countries whose inhabitants consume three goods: one publicly provided locally and two privately provided goods. The publicly provided good is nonpolluting. One of the privately provided goods is the numeraire good, which is also nonpolluting. The other privately provided good is polluting. Every consumer has an endowment of the numeraire good, some of which he consumes, spending the rest to purchase the polluting good and pay taxes. Production technologies are identical in both countries. The publicly provided good is produced at a constant average and marginal cost.

   Pollution (C[O.sub.2], S[O.sub.2], etc.) is global and a by-product of production. The polluting good may be produced in different ways. Each procedure entails a different resource cost and a different emission level. Emissions are beneficial in that a higher level of emission reduces the private (per unit) production costs of polluting goods. That is, the production costs of polluting goods are inversely related to their emissions. This is to capture the fact that technologies that cut emissions are more expensive to employ. Firms producing the polluting good operate in a competitive environment. The good is produced by an industry that is comprised of a fixed but sufficiently large number of identical firms. It is produced, for a given unit cost of production, by a linear technology subject to constant returns to scale. (8)

   Each country provides the publicly provided good to its own residents only. The polluting good is produced and consumed in both countries. Prior to economic integration, there is no trade between the two countries. Upon integration, residents of each country will be able to purchase the polluting good from the foreign as well as the home country. While the physical characteristics of the home- and foreign-produced goods are identical, consumers have a preference for purchasing the home-produced goods. We model this by assuming that consumers experience a certain disutility when they consume one unit of the foreign-produced good. The extent of the disutility differs across consumers. Individuals have otherwise identical quasi-linear preferences.

   There are two (distortionary) tax instruments: commodity and emission taxes. These are origin based. Thus, each country levies a certain tax on each unit of the (polluting) consumption good that its firms produce and sell regardless of where the purchasers come from. Second, to combat pollution, the country imposes another tax per unit of emissions on (home) firms.

   Within this framework, we characterize second-best commodity and emission tax rates. Next, we characterize the equilibrium values of commodity and emission taxes in closed and open economies. We show that the formula for the emission tax remains the same in open as in closed economies. On the other hand, the equilibrium value of the commodity tax changes and includes a negative term due to tax competition. The targeting principle applies; emission taxes are used only for the purpose of combating emissions and commodity taxes for tax competition. (9) The intuition for this result must be seen in the availability of output taxes for tax competition. As far as competitiveness in terms of prices is concerned, whatever the government of any particular country can do via emission taxes, it can also do via output taxes. However, the latter tax does not affect production decisions, but the former does. There is thus no reason for the government to want to use emission taxes and distort its production decisions. We show that lower prices lead to increased aggregate emissions, but that the firms will adopt less (or same) polluting technologies as the emission tax increases (or remains the same). Overall welfare declines. Finally, we show that partially harmonizing commodity taxes, above their unrestricted Nash equilibrium value, can potentially hurt as well as improve the pollution technology, overall quality of the environment, and welfare. The three attributes move positively together. On the other hand, harmonizing of emission taxes above their Nash equilibrium values appears to always lead to improvements in the environment and welfare via adoption of cleaner technologies.

2. The Model

   Consider two identical countries, A and B, whose inhabitants consume three goods: one publicly provided (locally) and two privately provided goods. The publicly provided good, G, is nonpolluting. One of the privately provided goods is the numeraire good, which is also nonpolluting. The other privately provided good, x, is polluting. Every consumer has an endowment of m units of the numeraire good, some of which he consumes, spending the rest to purchase the polluting good and pay his taxes. Production technologies are identical in both countries. The publicly provided good is produced at a constant average and marginal cost, which we can normalize at one.

   Pollution is global and a by-product of production. The polluting good may be produced in different ways. Each procedure entails a different resource cost and a different emission level. (10) Specifically, assume that the resource cost of producing one unit of output C([e.sub.i]), where [e.sub.i] (i = A, B) denotes emission per unit of output in country i, is a continuously differentiable, decreasing, and...