Many gases and many measures: Choice of targets and selection of measures in climate policy
Abstract
This report discusses the economic impacts of taking a comprehensive approach to climate policy. A comprehensive climate policy implies that all relevant sources, sinks and reservoirs of climate gases are covered. From the outset, this widens the policy options, and leads to a higher degree of flexibility when it comes to the implementation of measures. This study provides a model-based analysis of how the costs of climate policy may be affected.
Inclusion of many greenhouse gases requires that one have to pay attention to two problems which need not be considered when analysing emission control of carbon dioxide (CO2) alone. The first is how to include all relevant measures appropriately, and the second is how to aggregate different greenhouse gases. The problem of including relevant measures does not become critical when focusing on CO2-emissions, because nearly all of these emissions are attached the use of fossil fuels. Being a commodity subject to market transactions, the use of fossil fuels, and thereby CO2-emissions, can be regulated by means of charges. For other gases, this is not equally evident. Emission of e.g. methane (CH4) may be subject to ‘non-economic’ factors, and it might therefore be possible to find less costly measures to reduce these emissions than those being implemented as a response to charges.
Emissions of three greenhouse gases, CO2, CH4 and N2O (nitrous oxide), are considered. The measures include enhancement of carbon sinks by forest management, reduction of emissions of CH4 from landfills, reduction of N2O-emissions from production of fertilisers and from fertilisation in agriculture, henceforth called direct measures, and charges on fossil fuels. Cost functions for each of these measures are estimated on the basic of other studies, and explicitly included in the model.
The problem of aggregation is usually solved by the use of global warming potentials (GWP). By this measure, one attempts to express the change in radiative forcing due to the emission of a unit of a greenhouse gas relative to that of CO2-emissions. That is, the GWPs aim at expressing the emissions in terms of ‘CO2-equivalents’. GWPs are, however, highly inaccurate. This is due to the fact that the life-time of different greenhouse gases in the atmosphere varies considerably. The life-time of methane (CH4) is approximately 12 years, between 100 and 150 years for CO2 and approximately 50 000 years for chlorfluor carbons. Hence, the GWP for a gas is strongly conditioned upon the time-horizon over which the integrals are taken.
One way to solve this problem is to calculate radiative forcing directly by means of a dynamic model. Radiative forcing is an expression for the warming effect of a change in atmospheric concentrations. According to the IPCC (1994), a doubling of the atmospheric concentrations of greenhouse gases in the atmosphere results in a radiative forcing of 4. Model studies indicate that the global average surface air temperature thereby increases between 1.5 and 4.5 ° C, with 2.5 ° C as a "best estimate" (IPCC, 1996b). Application of radiative forcing requires that also the economic model is dynamic, and that the target for climate policy sets limits for radiative forcing, or concentrations of greenhouse gases, at a given future point in time rather than emissions. This has large impacts on the choice of optimal policy. This report addresses this issue by comparing the optimal policy under emission control with that of a control of concentrations.
Some main results are presented in table 1. The calculations should be considered as illustrations, only. Indications of how e.g. GDP is affected by climate policy should not be taken as a quantitative estimate since, in particular, the long-term properties of the model are not realistic. Due to the properties of the solution, we also had to limit the time horizon to 50 years. However, the results may give indications on the relative importance of gases and measures for a climate policy, and how to change the relative emphasise of different gases over time. The reference case, to which all the alternative runs of the model are compared, assumes that no initiatives to reduce emissions of greenhouse gases are taken. According to the calculations, the cost of a 10 percent target for emission reductions is minimised if CO2 accounts for nearly 85 percent of the total emission reductions. CH4 accounts for 7 percent and N2O for 9 percent. The contribution from other measures than a carbon charge on the use of energy is substantial under emission control. The least costly way to reduce greenhouse gas emissions by 10 percent is to apply approximately equal emphasis on carbon charges and other, direct measures where forest management is of major importance. The level of the charge at this target for emission reductions were approximately 6 percent of the basic energy price. This corresponds to an admissible abatement cost for direct measures at 1 øre/kg CO2.
Table 1 Summary of the main numeric results
Emission
reductions
Targets on
adiative forcing of 5
10
percent
30
percent
Static allocation of
initial abatement
Dynamic allocation of
initial abatement
t = 0
Average charge (pct)
6
27
1.9
0.7
Abatement cost
CO2(NOK/kg)
0.01
0.04
0.003
0.001
CH4(NOK/kg)
0.17
0.93
0.063
0.091
N2O (NOK/kg)
2.25
13.78
0.939
0.023
T = 50
Average charge (pct)
-
-
638
768
Abatement cost
-
-
CO2 (NOK/kg)
-
-
0.05
0.01
CH4 (NOK/kg)
-
-
110.28
157.05
N2O (NOK/kg)
-
-
18.24
4.52
for forest management, 17 øre/kg CH4 for reducing methane emissions from landfills, and 225 øre/kg N2O for reducing emissions from production and use of fertilisers.
With more ambitious targets, the contribution from charges increases. If the emissions are to be reduced by 30 percent relative to the reference alternative, charges contribute more than 70 percent of the reductions. With this target, the charge amounts to 27 percent of the basic energy price, and the admissible direct abatement costs for CO2, CH4 and N2O measures are 4, 93 and 1378 øre/kg, respectively. According to the calculations, CO2 is becoming slightly more important when the target for reductions in greenhouse gas emissions is increased. In other words, to the extent that these calculations are representative, one is slightly better off in terms of abatement costs by concentrating on CO2-emissions alone.
A shift of policy from emission targets towards targets on radiative forcing causes a substantial shift in cost minimising behaviour. Firstly, the general timing of policy becomes crucial. For example, a 50 percent reduction of greenhouse gas emissions relative to the reference path results in a forcing of 8.5 relative to the present forcing over the next 50 years. This requires an emission charge equal to 64 percent of the basic energy price. When timing the policy according to optimality criteria, a forcing of 5 might be achieved in 50 years by starting with a charge at 1.9 percent of the basic energy price ‘today’. A charge rate of 64 percent is not reached before 30 years, but after that, one has to tighten up the policy considerably. After 50 years, the charge is about 640 percent of the basic energy price.
Secondly, the relative emphasise of the different gases under a concentration target is changed compared with the optimal composition under emission control. This is due to the different life-times of greenhouse gases. Short-lived gases, such as methane, give a quicker response in the atmosphere than long-lived gases, such as carbon dioxide. Thus, one may reach a target on radiative forcing earlier, or alternatively postpone the efforts to reduce emissions, if putting more weight on the abatement of methane. Compared with a policy which allocates measures according to the marginal cost of emissions in terms of GWP, an optimal long-term allocation of measures would put more weight on reducing the emissions of methane, because there is an economic yield in delaying abatement costs. By following this advice, the initial carbon charge would decrease from 1.9 percent to 0.7 percent of the basic energy price for a target on radiative forcing at 5 in year 50. In the terminal year, the charge has increased to 768 percent. Admissible abatement cost for reducing the emissions of methane from landfills increases initially from 6.3 øre/kg, when measures are allocated according to marginal costs measured in GWP, to 9.1 øre/kg under optimal allocation. At t = 50 they reach 157.05 NOK/kg. Abatement costs related to forest management and emissions of N2O are comparably small when allocating the abatement of different gases optimally.
The aim of the discussion in this report is to point out some properties embedded in the comprehensive approach, which may change the conception of optimal climate policy. There is no doubt that CO2-emissions will remain the most important issue in climate policy. Nevertheless, inclusion of other gases may provide an opportunity to expand the options and thereby reduce costs. If the targets are set on concentrations rather than emissions, availability of measures to reduce the emissions of methane may cause a substantial reduction in the abatement costs. Moreover, the calculations indicate quite clearly that it may be worthwhile to search for measures as supplements to CO2-charges in the design of climate policy.