Climate Change 2001:
Working Group III: Mitigation
Other reports in this collection

8.2.2 Domestic Policy Instruments and Net Mitigation Costs

Tapping the technical abatement potentials requires setting up new incentive structures (taxes, emissions trading, technical standards, voluntary agreements, subsidies) for production and consumption, i.e. climate policies. In the following, empirical models that measure net mitigation costs of climate policies are reviewed in order to disentangle the reasons why certain policy packages have similar or different outcomes in various countries. As a first step, the results are presented at an aggregated level; then the impact of measures meant to mitigate the sectoral and distributional consequences of climate policies is examined. Finally, in a third step, the ancillary benefits from the joint reduction of carbon emissions and other pollutants are considered to complete the picture. Aggregate Assessment of Revenue-raising Instruments

Introducing a carbon tax (or auctioned tradable permits) provides an incentive to change the technology over the short and long term. Such policies generate tax revenues and the way these revenues are used has major impacts on the social costs of the climate policy. The reason is that these revenues are, in principle, available to offset some or all of the costs of the mitigation policy. When emission targets go beyond the negative cost potentials, there is a general agreement among economists (see Chapters 6 and 7) that if standards are used (or if emissions permits are allocated for free) the resultant social cost is higher than the total abatement expenditures. Producers pass part of the marginal abatement cost on to consumers through higher selling prices, which implies a loss of consumer surplus. If the elasticity of supply is quite high, this might lead to a net loss of producer surplus. However, if the elasticity of supply is fairly low, overall (or net) producer surplus can rise when policies cause a restriction in output, because the policy-generated rents per unit of production enjoyed by producers more than compensate for the net decrease in sales.

In the 1990s there was considerable interest in how revenue-neutral carbon taxes may mitigate this effect on the economy by enabling the government to cut the marginal rates of pre-existing taxes, such as income, payroll, and sales taxes. The possibility is a double dividend policy (Pearce, 1992), by both (1) improving the environment and (2) offsetting at least part of the welfare losses of climate policies by reducing the costs of the tax system (see the discussion in Chapter 7). The same mechanism occurs when nationally auctioned permits are used; for simplicity, the term carbon tax is used in the rest of this chapter, except when the distinction between these two instruments is necessary.

The starting point in a discussion of a double dividend is how expensive it is to raise government income, that is, how big is the marginal cost of funds (MCF). A high MCF gives more scope for a double dividend than a small MCF in the economy. This arises because the parameters that determine the magnitude of the double-dividend (see Chapter 7) are:

Figure 8.3: Carbon taxes and the costs of environmental policies.

When the revenues of carbon taxes are returned in a lump-sum fashion to households and firms, the tax-interaction effect is systematically higher than the revenue-recycling effect. Also the net cost of climate policy is higher than its gross cost (while lower than that with a no-tax policy, see A1 and A2 in Figure 8.3). However, it is possible to improve this result by targetting tax revenues to cuts in the most distortionary taxes; this can yield either a weak or a strong form of double dividend (Goulder, 1995a). The weak double dividend occurs as long as there is a revenue-recycling effect due to the swap between carbon taxes and the most distortionary taxes. Mitigation costs are systematically lower when revenues are recycled this way than when they are returned lump sum. The strong double dividend is more difficult to obtain. It requires that the (beneficial) revenue-recycling effect more than offset the combination of the primary cost and the tax-interaction effect. In this case, the net cost of abatement is negative (at least within some range). As discussed in Chapter 7, this is possible if, prior to the introduction of the mitigation policy, the tax system is already highly inefficient along non-environmental dimensions. In terms of Figure 8.3, the revenue-recycling effect is represented by the downward shift from curve A1 to curve A2 or A3. If the shift is from A1 to A2, the weak double dividend occurs, but not the strong double dividend. If the shift is from A1 to A3, not only does the weak double dividend occur, but the strong double dividend is realized as well, since the net costs are negative within a range.

While the weak form of double dividend enjoys broad support from theoretical and numerical studies, the strong double dividend hypothesis is less broadly supported and more controversial. Indeed, reaching an economical dividend is impossible when the economy is at full employment and if all other taxation is optimal (abstracting for the environmental externality). Therefore, it may be argued that the double dividend accrues from the tax reform, independently of the climate policy. However, empirical models capture the fact that, in the real world, a carbon tax or auctioned emissions permits will not be implemented after the enforcement of an optimal fiscal reform. To the contrary, introducing a new tax may be a sine qua non condition to the fiscal reform. For a given carbon tax revenue, models help interpret the best way to recycle this revenue.

Specific features of the tax systems and markets of the production factors (labour, capital, and energy) ultimately determine the presence or absence of a strong double dividend. For example, a double dividend is likely if production factors are very distorted by prior taxation or specific market conditions, if there is a problem of trade-balance because of the import of fossil energy, or if consumer choice is highly distorted because of tax-deducible spending provisions (Parry and Bento, 2000).

Empirical studies try to gauge the impact of these many determinants and to understand why the effects of a given recycling strategy (reducing payroll, personal income, corporate income, investment income, or expenditure taxes) differ from one country to another.

Other reports in this collection