The cost-effective domestic allocation of emission cuts is given at point O in Figure 7, where the curves for the net marginal social cost of abatement in the two sectors intersect. This allocation minimises the total social cost of meeting the target for domestic emission reduction, accounting for the social costs of carbon leakage as well as for trade in emission rights in both sectors. As the figure is drawn, the optimal allocation lies to the right of point E, implying that society should accept a higher gross marginal abatement cost in the ETS sector than in the non-ETS sector. This means that ETS companies should still only receive a partial credit for the ETS allowance price against the domestic carbon tax. More precisely, the tax credit should be equal to pN + MSCLETS - MSCLNon-ETS per tonne of CO2e emitted from the ETS sector to implement the optimal allocation O in Figure 7. In other words, the larger the rate of carbon leakage (and consequently the higher the marginal social cost of leakage) in the ETS sector relative to the leakage rate in the non-ETS sector, the larger the optimal tax credit for (part of) the ETS allowance price becomes, which is intuitive.
With a different combination of allowance prices and marginal social leakage costs, the optimal allocation could lie to the left of point E, in which case ETS companies should be granted a credit greater than 100% for the ETS allowance price or, alternatively, should receive a 100% credit for the allowance price and pay a lower rate of carbon tax than emitters in the non-ETS sector.
Note that the leakage rate a in the ETS sector also measures the fall in foreign emissions generated by a one-tonne increase in the excess domestic demand for ETS allowances, i.e., a measures the fall in foreign emissions that would materialise if the domestic government decided to purchase one ETS emission allowance and withdraw it from the market. Hence, the amount MSCLETS = a∙c also measures the marginal social benefit from the government’s purchase of an ETS allowance. If this marginal benefit exceeds the allowance price pq, the government could generate a domestic welfare gain by purchasing ETS allowances until its marginal willingness to pay for foreign emission cuts (i.e., c) falls to a level where a∙c becomes equal to the allowance price. On the other hand, if MSCLETS = a∙c is initially lower than the allowance price, the government should sell emission allowances in the ETS market (so long as it holds such allowances allocated to it by the EU) until MSCLETS and the allowance price are driven into line. Theoretically, it could thus be argued that a rational government should intervene in the ETS allowance market to ensure that MSCLETS = pq. If a similar arbitrage behaviour in the non-ETS sector were to ensure that MSCLNon-ETS = pN, the optimal allocation of emission reductions would be given by point E in Figure 7, again validating the conventional recommendation that gross marginal abatement costs be equalised across sectors. Once more, this would require a 100% credit for the ETS allowance price against the domestic carbon tax bill of ETS companies.
In practice, governments rarely behave in such a rational manner, and even if they tried to do so, the lack of an integrated and liquid market for emission rights in the non-ETS sector would probably prevent an equalisation of the marginal social cost of leakage and the emission allowance price in that sector.
Realistically, we would, therefore, expect the cost-effective allocation of emission reductions to deviate from point E in Figure 7. This, in turn, implies that the optimal domestic carbon tax credit for the ETS allowance price will deviate from 100%, i.e., that the optimal total carbon prices in the ETS sector and the non-ETS sector will differ from each other.
3.4 Optimal unilateral climate policy in an EU frontrunner country: Summary
The analysis in this section can be summarised as follows: The marginal social cost of abatement in an economic sector includes the technical marginal abatement cost plus the marginal social cost of carbon leakage from the sector minus the international price of emission rights in the sector. The criterion for an optimal unilateral climate policy is that the marginal social costs of abatement should be equalised across sectors. This means that a uniform carbon price across sectors is not necessarily the optimal setting. If the domestic government does not worry about carbon leakage, a uniform carbon price is cost-effective only if the international price of emission rights is the same in the ETS and the non-ETS sector. If the government is concerned about leakage, a uniform carbon price for all private emitters is only optimal if the rates of carbon leakage as well as the prices of emission rights are the same across the two sectors, which is unlikely to be the case.
A relatively high carbon leakage rate and a relatively low international price of emission rights in a given sector work in favour of having a relatively low total carbon price in that sector. A priori this theoretical insight does not tell us unambiguously whether the ETS sector should have a higher or a lower carbon price than the non-ETS sector. However, we found that if the government cannot or does not wish to trade emission rights for the non-ETS sector and is relatively unconcerned about carbon leakage, the total carbon price in the ETS sector should generally exceed the carbon tax rate levied on the non-ETS sector, i.e., some amount of domestic carbon tax should be imposed on the ETS sector on top of the ETS allowance price. The reason is that the allowance price makes it more expensive for society to emit CO2e from the ETS sector. In effect, the EU co-finances a part of the domestic cost of abating emissions from the ETS sector since the domestic economy saves the expense on an ETS allowance when emissions from that sector are cut by one tonne. Ceteris paribus, this makes it more attractive to reduce emissions from the ETS sector and consequently warrants a higher carbon price in the sector.
3.5 Caveats: The importance of the objective for climate policy
The policy recommendations above depend to a large extent on the specific target for domestic climate policy. This can be illustrated by comparing climate policy targets in Norway and Denmark. According to the Norwegian Climate Act, the total domestic CO2e emissions in Norway must be cut by 55% by 2030 compared to 1990, to live up to the country’s commitment under the Paris Agreement. Norway co-operates with the EU on climate policy and participates in the ETS even though Norway is not a member of the EU. The Norwegian ETS sector’s contribution to the fulfilment of Norway’s obligations under the Kyoto Protocol, which preceded the Paris Agreement, was determined in negotiations with the EU. It was decided that Norway should be given a credit for emission reductions in the ETS sector corresponding to the reduction over time in the amount of ETS allowances allocated to the country under the ETS. In other words, under the Kyoto Protocol in effect until 2020, the contribution of the Norwegian ETS sector to fulfilment of the country’s target for total emission reductions did not depend on the actual emissions from the sector but rather on the politically determined quantity of ETS allowances allocated to Norway by the EU.
The contribution of the Norwegian ETS sector to the country’s 2030 climate target has yet to be negotiated with the EU, but it is expected that it will be determined in much the same way as under the Kyoto Protocol. In such a situation where a cut in the actual emissions from the ETS sector does not contribute to the fulfilment of the country’s national climate policy target, there is no case for imposing a domestic carbon tax in the ETS sector on top of the allowance price. Instead, the carbon tax should only apply to the non-ETS sector.
By contrast, according to the Danish Climate Act, which requires a 70% cut in total domestic emissions by 2030 relative to 1990, the ETS sector’s contribution to fulfilling this target is measured by the reduction of actual emissions from the sector, as assumed in our analysis above. Under such a specification of the policy target, the carbon price for the ETS sector should, in fact, include some amount of domestic carbon taxation when policy makers strive to secure domestic cost-effectiveness, as we have seen.
This conclusion assumes that the goal of domestic climate policy is to maximise national welfare rather than securing cost-effectiveness throughout the EU as a whole. This approach could be defended by the fact that a frontrunner country voluntarily incurs higher abatement costs than those required for fulfilment of its obligations towards the EU, and so it may legitimately seek to minimise the national cost of moving ahead of the EU. It can also be argued that a frontrunner country does not necessarily provide a viable example to other countries if the frontrunner mainly achieves its ambitious target for emissions reduction by forcing the production of CO2e-intensive goods abroad rather than by reducing emissions per unit of domestic output. From this perspective, the frontrunner’s carbon tax scheme should account for the risk of carbon leakage in line with the policy design illustrated in Figure 7.
On the other hand, one could adopt the “Kantian” approach that a frontrunner country seeking to lead by example should pursue a climate policy similar to one it would like adopted by the whole of the EU. Presumably, this would be a policy securing a cost-effective abatement of total EU emissions, i.e., a climate policy ensuring an equalisation of the total carbon price between the ETS and the non-ETS sector. Such an approach to national climate policy in an EU frontrunner country should not discriminate between the two sectors.
The choice between these two approaches is obviously a normative issue to be decided, in the end, by policy makers.