Coordinating Lead Authors:

Terry Barker (UK) [SRC:6], Igor Bashmakov (Russia) [SRC:1],

Lead Authors:

Awwad Alharthi (Saudi Arabia), Markus Amann (Austria) [SRC:4], Luis Cifuentes (Chile) [SRC:2], John Drexhage (Canada), Maosheng Duan (China), Ottmar Edenhofer (Germany) [SRC:3], Brian Flannery (USA), Michael Grubb (UK) [SRC:9], Monique Hoogwijk (Netherlands) [SRC:2], Francis I. Ibitoye (Nigeria), Catrinus J. Jepma (Netherlands), William A. Pizer (USA) [SRC:1], Kenji Yamaji (Japan) [SRC:6],

Contributing Authors:

Shimon Awerbuch = (USA), Lenny Bernstein (USA), Andre Faaij (Netherlands) [SRC:7], Hitoshi Hayami (Japan), Tom Heggedal (Norway), Snorre Kverndokk (Norway) [SRC:2], John Latham (UK) [SRC:3], Axel Michaelowa (Germany) [SRC:2], David Popp (USA) [SRC:7], Peter Read (New Zealand) [SRC:1], Stefan P. Schleicher (Austria), Mike Smith (UK), Ferenc Toth (Hungary),

Review Editors:

David Hawkins (USA), Aviel Verbruggen (Belgium),

This chapter should be cited as:

Barker, T., I. Bashmakov, A. Alharthi, M. Amann, L. Cifuentes, J. Drexhage, M. Duan, O. Edenhofer, B. Flannery, M. Grubb, M. Hoogwijk, F. I. Ibitoye, C. J. Jepma, W.A. Pizer, K. Yamaji, 2007: Mitigation from a cross-sectoral perspective. In Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer (eds)], Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

11.2.1 Cross-sectoral technological options

Cross-sectoral mitigation technologies can be broken down into three categories in which the implementation of the technology:

1. occurs in parallel in more than one sector;

2. could involve interaction between sectors, or

3. could create competition among sectors for scarce resources.

Some of the technologies implemented in parallel have been discussed earlier in this report. Efficient electric motor-driven systems are used in the industrial sector ( Section 7.3.2 ) and are also a part of many of the technologies for the buildings sector, e.g. efficient heating, ventilation and air conditioning systems ( Section 6.4.5 ). Solar PV can be used in the energy sector for centralized electricity generation ( Section 4.3.3.6 ) and in the buildings sector for distributed electricity generation ( Section 6.4.7 ). Any improvement in these technologies in one sector will benefit the other sectors.

On a broad scale, information technology (IT) is imple-mented in parallel across sectors as a component of many end-use technologies, but the cumulative impact of its use has not been analyzed. For example, IT is the basis for integrating the control of various building systems, and has the potential to reduce building energy consumption ( Section 6.4.6 ). IT is also the key to the performance of hybrids and other advanced vehicle technologies ( Section 5.3.1.2 ). Smart end-use devices (household appliances, etc) could use IT to program their operation at times when electricity demand is low. This could reduce peak demand for electricity, resulting in a shift to base load generation, which is usually more efficient ( Hirst, 2006 [NPR] ). The impact of such a switch on CO₂ emissions is unknown, because it is easy to construct cases where shifts from peak load to base load would increase CO₂ emissions (e.g., natural-gas-fired peak load, but coal-fired base load). General improvements in IT, e.g. cheaper computer chips, will benefit all sectors, but applications have to be tailored to the specific end-use. Of course, the net impact of IT on greenhouse gas emissions could result either in net reductions or gains, depending for example on whether or not efficiency gains are offset by increases in production.

An example of a group of technologies that could involve interaction between sectors is gasification/hydrogen/carbon dioxide capture and storage (CCS) technology ( IPCC, 2005 [NPR] and Chapter 4 .3.6). While these technologies can be discussed separately, they are interrelated and being applied as a group enhances their CO₂-emission mitigation potential. For example, CCS can be applied as a post-combustion technology, in which case it will increase the amount of resource needed to generate a unit of heat or electricity. Using a pre-combustion approach, i.e. gasifying fossil fuels to produce hydrogen that can be used in fuel cells or directly in combustion engines, may improve overall energy efficiency. However, unless CCS is used to mitigate the CO₂ by-product from this process, the use of that hydrogen will offer only modest benefits. (See Section 5.3.1.4 for a comparison of fuel cell and hybrid vehicles.) Adding CCS would make hydrogen an energy carrier, providing a low CO₂ emission approach for transportation, buildings, or industrial applications. Implementation of fuel cells in stationary applications could provide valuable learning for vehicle application; in addition, fuel cell vehicles could provide electric power to homes and buildings ( Romeri, 2004 [NPR] ).

In the longer term, hydrogen could be manufactured by gasifying biomass – an approach which has the potential to achieve negative CO₂ emissions ( IPCC, 2005 [NPR] ) – or through electrolysis using carbon-free sources of electricity, a zero CO₂ option. In the even longer term, it may be possible to produce hydrogen by other processes, e.g. biologically, using genetically-modified organisms ( GCEP, 2005 [NPR] ). However, none of these longer-term technologies are likely to have a significant impact before 2030, the time frame for this analysis.

Biomass is an example of a cross-sectoral technology which may compete for resources. Any assessment of the use of biomass, e.g., as a source of transportation fuels, must consider competing demands from other sectors for the creation and utilization of biomass resources. Technical breakthroughs could allow biomass to make a larger future contribution to world energy needs. Such breakthroughs could also stimulate the investments required to improve biomass productivity for fuel, food and fibre. See Chapter 4 and Section 11.3 .

Another example of resource competition involves natural gas. Natural gas availability could limit the application of some short- to medium-term mitigation technology. Switching to lower carbon fuels, e.g. from coal to natural gas for electricity generation, or from gasoline or diesel to natural gas for vehicles, is a commonly cited short-term option. Because of its higher hydrogen content, natural gas is also the preferred fossil fuel for hydrogen manufacture. Discussion of these options in one sector rarely takes natural gas demand from other sectors into account.

In conclusion, there are several important interactions between technologies across sectors that are seldom taken into account. This is an area of energy system modelling that requires further investigation.

11.2.2 Ocean fertilization and other geo-engineering options

Since the TAR, a body of literature has developed on alternative, geo-engineering techniques for mitigating climate change. This section focuses on apparently promising techniques: ocean fertilization, geo-engineering methods for capturing and safely sequestering CO₂ and reducing the amount of sunlight absorbed by the earth’s atmospheric system. These options tend to be speculative and many of their environmental side-effects have yet to be assessed; detailed cost estimates have not been published; and they are without a clear institutional framework for implementation. Conventional carbon capture and storage is covered in Chapter 4 , Section 4.3.6 and the IPCC Special Report ( 2005 ) on the topic.

11.3.1 Integrated summary of sectoral emission potentials

Chapters 4 to 10 assessed the economic potential of GHG mitigation at a sectoral scale for the time frame out to 2030 (for a discussion of the different definitions of potential, see Chapter 2 ). These bottom-up estimates are derived using a variety of literature sources and various methodologies, as discussed in the underlying chapters. This section derives ranges of aggregate economic potentials for GHG mitigation over different costs (i.e. carbon prices) at year- 2000 prices.

11.3.1.2 The baseline

All mitigation potentials have to be estimated against a baseline. The main baseline scenarios used for compiling the assessments in the chapters are the SRES B2 and A1B marker scenarios ( IPCC, 2000 [NPR] ) and the World Energy Outlook 2004 (WEO 2004 ( IEA, 2004 [NPR] ). The assumed emissions in the three baseline scenarios vary in magnitude and regional distribution. The baseline scenarios B2 and WEO 2004 are comparable in the main assumptions for population, GDP and energy use. Figure 11.1 shows that the emissions are also comparable. Scenario A1B, which assumes relatively higher economic growth, shows substantially higher emissions in countries outside the OECD/EIT region.

Figure 11.1: Energy-related CO₂-only emissions per world region for the year 2030 in the World Energy Outlook, and in the SRES B2 and A1B scenarios

Source: Price et al., 2006 [NPR, ARC] .

The crude oil prices assumed in SRES B2 and WEO 2004 are of the same order of magnitude. The oil prices in the SRES scenarios vary across studies. For the MESSAGE model (B2 scenario), the price is about 25 US$/barrel ( Riahi et al., 2006 [NPR, MoS, ARC] ). In the case of the WEO 2004, for example, the oil price assumed in 2030 is 29 US$/barrel. These prices (and all other energy price assumptions) are substantially lower than those prevailing in 2006 and assumed for later projections ( IEA, 2005 [NPR] and 2006b ). The 2002 –6 rises in world energy prices are also reflected in the energy futures markets for at least another five to ten years. In fact, the rise in crude oil prices during this period, some 50 US$/barrel, is comparable to the impact of a 100 US$/tCO₂-eq increase in the price of carbon. However, it is still uncertain whether these price increases will have a significant impact on the long-term energy price trend.

Higher energy prices and further action on mitigation may reinforce each other in their impact on mitigation potential, although it is still uncertain how and to what extent. On the one hand, for instance, economies of scale may facilitate the introduction of some new technologies if supported by a higher energy price trend. On the other hand, it is also conceivable that, once some cost-effective innovation has already been triggered by higher energy prices, any further mitigation action through policies and measures may become more costly and difficult. Finally, although general energy prices rises will encourage energy efficiency, the mix of the different fuel prices is also important. Oil and gas prices have risen substantially in relation to coal prices 2002 –6, and this will encourage greater use of coal, for example in electricity generation, increasing GHG emissions.

As a rule, the SRES B2 and WEO 2004 baselines were both used for the synthesis of the emission mitigation potentials by sector. Most chapters have reported the mitigation potential for at least one of these baseline scenarios. There are a few exceptions. Chapter 5 (transportation) uses a different, more suitable, scenario ( WBCSD, 2004 [NPR] ). However, it is comparable to WEO 2004 . Chapter 6 (buildings) constructed a baseline scenario with CO₂ emissions between those of the SRES B2 and A1B marker scenarios taken from the literature (see Section 6.5 ). The agriculture and forestry sectors based their mitigation potential on changes in land use as deduced from various scenarios (including marker scenarios, see Sections 8.4.3 and 9.4.3 ). The SRES scenarios did not include enough detail for the waste sector, so Chapter 10 used the GDP and population figures from SRES A1B and the methodologies described in IPCC Guidelines 2006 (see Section 10.4.7 ).

Table 11.1 compares the emissions of the different sectoral baselines for 2004 and 2030 against a background of the end-use and point-of-emission allocation of emissions attributed to electricity use. Since the 2030 data are from studies that differ in terms of coverage and comparability, they should not be directly aggregated across the different sectors and therefore no totals across all sectors are shown in Table 11.16 . An important difference between the WEO baseline and SRES B2 is that the WEO emissions do not include all non-CO₂ GHG emissions.

Table 11.1: Overview of the global emissions for the year 2004 and the baseline emissions for all GHGs adopted for the year 2030 (in GtCO₂-eq)

	Global emissions 2004 (allocated to the end-use sector)a, c	Global emissions 2004 (point of emissions)a, b	Type of baseline usedd	Global emissions 2030 (allocated to the end-use sector)	Global emissions 2030 (point of emissions)
Energy supply	- j	12.7	WEO	- j, f)	15.8 f)
Transport	6.4	6.4	WEO	10.6 f)	10.6 f)
Buildings	9.2	3.9	Own	14.3 f)	5.9 e) f)
Industry	12.0	9.5	B2/USEPA	14.6	8.5 g)
Agriculture	6.6	6.6	B2/FAO	8.3	8.3
LULUCF/Forestryk)	5.8	5.8	Own	5.8 h)	5.8 h)
Wastei)	1.4	1.4	A1B	2.1	2.1

Notes:

^a) The emissions in the year 2004 as reported in the sectoral chapters and Chapter 1 , Figure 1.3 a/b.

^b) The allocation to point of emission means that the emissions are allocated to the sector where the emission takes place. For example, electricity emissions are allocated to the power sector. There is a difference between the sum when allocating the emissions in different ways. This is explained by the exclusion of electricity emissions from the agricultural and transport sectors due to lack of data and by the exclusion of emissions from conversion of energy as most end-use emissions are based on final energy supply.

^c) ‘Allocated to the end-use sector’ means that the emissions are allocated to the sectors that use the energy. For example, electricity emissions are allocated to the end-use sectors, mainly buildings and industry. Emissions from extraction and distribution are not included here.

^d) See text for further clarification on the type of baselines used.

^e) This figure is based on the assumption that the share of electricity-related emissions in the constructed baseline in Chapter 6 is the same as for the SRES B2 scenario. According to Price et al. 2006 [NPR, ARC] ), the electricity-related emissions amount to 59%. 59% of the baseline (14.3 GtCO2-eq) is 8.4 GtCO2-eq. The remaining emissions are allocated to the buildings sector.

^f) 2030 emissions of the F-gases are not available for the Transport, Buildings, and Energy Supply sectors.

^g) Source: Price et al., 2006 [NPR, ARC] .

^h) No baseline emissions for the year 2030 from the forestry sector are reported. See 9.4.3. On the basis of top-down models, it can be expected that the emissions in 2030 will be similar to 2004 .

ⁱ⁾ The data for waste include waste disposal, wastewater and incineration. The emissions from wastewater treatment are for the years 2005 and 2020 .

^j) The emissions from conversion losses are not included due to lack of data.

^k) Note that the peat fires and other bog fires, as mentioned in Chapter 1 , are not included here. Nor are they included in Chapters 8 and 9.

11.3.1.3 Synthesizing the potentials from Chapters 4 to 10 involving electricity

When aggregating the sectoral mitigation potentials, the links between sectors need to be considered ( Figure 11.2 ). For example, the options in electricity supply interact with those for electricity demand in the buildings and industry sectors. On the supply side, fossil-fuel electricity can be substituted by low-CO₂ or CO₂-free technologies such as renewable sources, nuclear energy, bioenergy or fossil fuel in combination with carbon capture and storage. On the demand side, the buildings and the industrial sectors have options for electricity savings. The emission reductions from these two sets of options cannot be aggregated since emission reductions in demand reduce the potential for those in supply and vice-versa.

Figure 11.2: Interaction of CO₂ mitigation measures between electricity supply and demand sectors

To overcome this problem, the following approach was adopted: The World Energy Outlook ( IEA, 2004 [NPR] ) for the year 2030 was used as the baseline. The potentials from electricity savings in the buildings and the industry sectors were estimated first. Electricity savings then reduce demand for electricity. This sequence was followed because electricity savings can be achieved at relatively low cost and their implementation can therefore be expected first. Electricity savings were converted to emission reductions using the average carbon intensity of the electricity supply in the baseline for the year 2030 . In reality, it can be expected that electricity savings would result in a relatively larger reduction in fossil-fuel electricity generation than electricity generation involving low marginal costs such as renewables and nuclear. This is because, in the operating system, low-cost generation is normally called on before high-cost generation. However, this response depends on local conditions and it is not appropriate to consider it here. However, it does imply that the emission reductions for electricity savings reported here are an under-estimate. This under-estimate becomes more pronounced with higher carbon prices, and higher marginal costs for fossil fuels.

The detailed sequence is as follows:

1. Electricity savings from the measures in the buildings and industry sector were subtracted from the baseline supply estimates to obtain the corrected electricity supply for 2030 .

2. No early withdrawal of plant or stranded assets is assumed. Low-carbon options can therefore only be applied to new electricity supply.

3. The new electricity supply required to 2030 was calculated from 1) additional new capacity between 2010 and 2030 and 2) capacity replaced in the period 2010 – 2030 after an assumed average plant lifetime of 50 years (see Chapter 4.4.3).

4. The new electricity supply required was divided between available low-carbon supply options. As the cost estimates were lowest for a fuel switch from coal to natural gas supply, it was assumed that this would take place first. In accordance with Chapter 4 it was assumed that 20% of the new coal plants required would be substituted by gas technologies.

5. An assessment was made of the prevented emissions from the other low-carbon substitution options after the fuel switch. The following technologies were taken into account: renewables (wind and geothermal), bioenergy, hydro, nuclear and CCS. It was assumed that the fossil fuel requirement in the baseline (after adjustments for the previous step) was met by these low-carbon intensive technologies. The substitution was made on the basis of relative maximum technical potential. The same breakdown as in Section 4.4.3 was used for the low-carbon options.

6. It was then possible to estimate the resulting mitigation potential for the energy sector, after savings in the end-use sectors buildings and industry.

7. For the buildings and industry sectors, the mitigation potential was broken down into emission savings resulting from less electricity use and the remainder.

8. For sectors other than energy, buildings and industry, the data given in the chapters were used for the overall aggregation.

When evaluating mitigation potential in the energy supply sector, the calculations in Chapter 4 did not subtract the electricity savings from the buildings and industry sectors (see Chapter 4 , Table 4.19 ). Adopting this order (which is not the preferred order, as explained above) implies first taking all the mitigation measures in the energy sector and then applying the electricity savings from buildings and industry sectors. This would result in different mitigation potentials for each of the sectors and mitigation measures, although the total will not change. See Appendix 11.1 for a further discussion of the methodology and details of the calculation.

In the case of the other sectors, the data given in the chapters were used for the overall aggregation. The mitigation potential for the buildings and industry sectors was broken down into emission savings for lower electricity use and the remainder, so that the potential could be re-allocated where necessary to the power sector.

11.3.2 Comparing bottom-up and top-down sectoral potentials for 2030

Table 11.5 and Figure 11.3 bring together the ranges of economic potentials synthesized from Chapters 4 to 10 , as discussed in 11.3.1, with the ranges of top-down sectoral estimates for 2030 presented in Chapter 3 . The bottom-up estimates are shown with the potentials from end-use electricity savings attributed (1) to the end-use sectors, i.e. to the buildings and industry sectors primarily responsible for the electricity use and (2) upstream, at the point of emission to the energy supply sector. The top-down ranges are provided by an analysis of the data from multi-gas studies for 2030 reported in Section 3.6 . A relationship has been estimated between the absolute reductions in total GHGs and the carbon prices required to achieve them (see Appendix 3.1). Ranges for mitigation potential have been calculated for a 68% confidence interval for carbon prices at 20 and 100 US$( 2000 )/tCO₂-eq. The ranges are shown in the last two columns of Table 11.5 .

Figure 11.3: Economic mitigation potential in different cost categories as compared to the baseline

Notes: The ranges reported in Tables 3.13 and 3.14 were used for comparison for the top-down studies. ‘High’ and ‘low’ refer to the high and low ends

of the economic potential range reported.

The ranges of bottom-up and top-down aggregate estimates of potentials overlap substantially under all cost ceilings except for the no-regrets bottom-up options. This contrasts with the comparison in the TAR, where top-down costs were higher. It is not the case that bottom-up approaches systematically generate higher abatement potentials. This change comes largely from lower costs in the top-down models, because some have introduced multi-gas abatement and have introduced more bottom-up features, such as induced technological change, which also tend to reduce costs.

Two further points can be made with regard to the comparison of bottom-up and top-down results:

1) Sector definitions differ between top-down and bottom-up approaches. The sectoral data presented here are not fully comparable. The main difference is that the electricity savings are allocated to the power sector in the top-down models compared to the end-use sectors in Table 11.3 . Both allocation approaches are presented in Table 11.5 .

2) At a sector level however, there are some systematic and striking discrepancies:

Energy supply. The top-down models indicate a higher emission reduction. This can be explained in part by differences in the mitigation options that are included in the top-down models and not included in the bottom-up approach. Examples are: reductions in extraction and distribution, reductions of other non-CO₂ emissions, and reductions through the increased use of CHP. Further, different estimates of the inertia of the substitution are expected to play a role. In bottom-up estimates, fuel substitution is assumed only after end-use savings whereas top-down models adopt a more continuous approach. Finally, the top-down estimates include the effects of energy savings in other sectors and structural changes. For example, a reduction in oil use also implies a reduction in emissions from refineries. These effects are excluded from the bottom-up estimates.

Buildings. Top-down models give estimates of reduction potentials from the buildings sector that are lower than those from bottom-up assessments. This is because the top-down models look only at responses to price signals, whereas most of the potential in the buildings sector is thought to be from ‘negative cost’ measures that would be primarily realized through other kinds of interventions (such as buildings or appliance standards). Top-down models assume that the regulatory environments of ‘reference’ and ‘abatement’ cases are similar, so that any negative cost poten- tial is either neglected or assumed to be included in baseline.

Agriculture and forestry. The estimates from bottom-up assessments were higher than those found in top-down studies, particularly at higher cost levels. These sectors are often not covered well by top-down models due to their specific character. An additional explanation is that the data from the top-down estimates include additional deforestation (negative mitigation potential) due to biomass energy plantations. This factor is not included in the bottom-up estimates.

Industry. The top-down models generate higher estimates of reduction potentials in industry than the bottom-up assessments. One of the reasons could be that top-down models allow for product substitution, which is often excluded in bottom-up sector analysis; equally, top-down models may have a greater tendency to allow for innovation over time.

The overall bottom-up potential, both at low and high carbon prices, is consistent with that of 2030 results from top-down models as reported in Chapter 3 , Section 3.6.2 for a limited set of models. For carbon prices <20 US$/tCO₂-eq, the ranges are 10–17 GtCO₂-eq/yr for bottom-up, as opposed to 9–18 GtCO₂-eq/yr for top-down studies. For carbon prices <50 US$/tCO₂-eq, the ranges are 14–25 GtCO₂-eq/yr for bottom-up versus 14–23 GtCO₂-eq/yr for top-down studies. For carbon prices <100 US$/tCO₂-eq the ranges are 16–30 GtCO₂-eq/yr and 17–26 GtCO₂-eq/yr for bottom-up and top-down respectively. As explained above, the differences between bottom-up and top-down are larger at the sector level.

11.3.3 Studies of interactions between energy supply and demand

This section looks at literature dealing specifically with the modelling of interactions between energy supply and demand. It first considers the carbon content of electricity, a crucial feature of the cross-sectoral aggregation of potentials discussed above, and then the effect of mitigation on energy prices. The studies emphasize the dependence of mitigation potentials from end-use electricity savings on the generation mix.

11.3.4 Regional cross-sectoral effects of greenhouse gas mitigation policies to 2025

Various estimates of cross-sectoral mitigation potential for specific regions have been published, usually as reports commissioned by governments. Unfortunately, however, the issue of attributing costs to cross-sectoral effects of greenhouse gas mitigation policies has not been reported extensively since the TAR, and literature on this topic is consequently sparse.

In one of the few studies to examine the sectoral effects of mitigation policies across countries, Meyer and Lutz 2002 [NPR, MoS] ), using the COMPASS model, carried out a simulation of the effects of carbon taxes or the G7 countries, which include some of the biggest energy users. The authors assumed the introduction of a carbon tax of 1US$ per ton of CO₂ in 2001 in all of these countries, rising linearly to 10 US$ in 2010, with revenues used to lower social security contributions. Table 11.6 shows the effects on output: the decline in petroleum and coal products will be highest, with the effects on construction being mild. The scale of the effects differs substantially between countries, depending on the energy intensities of the economies and the carbon content of this energy, with effects on output being much larger in US and Canadian industries.

One major cross-sectoral study (EU DG Environment, 2001 ) brings together low-cost mitigation options and shows their effects across sectors and regions. It shows how a Kyoto-style target (8% reduction of EU GHGs below 1990 /95 by 2010 ) can be achieved for the EU-15 member states with options costing less than 20 US$/tCO₂. The study assesses the direct and indirect outcomes using a top-down model (PRIMES) for energy-related CO₂ and a bottom-up model (GENESIS) for all other GHGs. The synthesis of the results is presented in Table 11.7 . This multi-gas study considers all GHGs, but assumes that the JI and CDM flexibility instruments are not used. The study shows the wide variations in cost-effective mitigation across sectors. The largest reductions compared to the 1990 /95 baselines are in the energy and energy-intensive sectors, whereas there is an increase of 25% in the transport sector compared to 1990 /95 emissions. Note also the large reductions in methane and N₂O in the achievement of the overall target as shown in the lower panel of the table. The results are, however, dominated by bottom-up energy-engineering assumptions since PRIMES is a partial-equilibrium model. Consequently, the GDP effects of the options are not provided. These potentials can be compared to those at less than 20 US$/tCO₂ in Table 11.3 above for the sectoral synthesis for the OECD. The EU potentials are similarly concentrated in the buildings sector, but with a larger share for industry, and a lower one for transport, reflecting the high existing taxes on transport fuels in the EU.

Masui et al. 2005 [ARC] ) report the effects of a tax and sectoral subsidy regime for Japan to achieve the Kyoto target by 2010, in which carbon tax revenues are used to subsidize additional investments to reduce greenhouse gases. The investment costs are shown in Table 11.8 for each sector. The table shows that about about 9 US$/tCO₂ (3,400 Japanese Yen/tC) will be required as carbon tax and most of the investment will be in energy-saving measures in the buildings sector (Residential and Commercial). The macro-economic effects for this study are reported in Section 11.4.3.4 .

( Schumacher and Sands 2006 ) ) model the response of German GHG emissions to various technology and carbon policy assumptions over the next few decades using the SGM model for Germany. Accounting for advanced technologies such as coal IGCC, NGCC, CCS, and wind power, they show that emission reductions can be achieved at substantially lower marginal abatement costs in the long run with new advanced electricity generating technologies in place. In a scenario assuming a carbon price of 50 US$/tCO₂ giving a 15% reduction of CO₂ below baseline by 2020, they show that, with the new and advanced technologies, the electricity sector would account for the largest share of emissions reductions (around 50% of total emissions reductions), followed by other (non-energy-intensive) industries and households. The effects on gross output are very uneven across sectors: energy transformation is 9% below base, but other industry, services and agriculture (and GDP) are 0.7% below base by 2050 .

The effects of different policy mixes on sectoral outcomes are shown in the US EIA 2005 [NPR, MoS] ) analysis of the National Commission on Energy Policy’s (NCEP) 2004 proposals. These involve reductions in the US emissions in GHGs of about 11% by 2025 below a reference case, including an analysis of the cap-and-trade component, (involving a safety valve limiting the maximum cost of emissions permits to US$ ( 2003 )8.50/tCO₂ through to 2025 ) and a no-safety-valve case (in which the cost rises to US$( 2003 ) 35/tCO₂ and the GHG reduction to 15% by 2025 ). The effects on CO₂ emissions by broad sector are shown in Figure 11.4 . Note that the NCEP scenario includes the cap-and-trade scheme (with a safety valve) shown separately in the figure and that the no-safety-valve scenario is additional to the NCEP scenario. The NCEP scenario includes substantial energy efficiency policies for transportation and buildings. This explains the relatively large contributions of these sectors in this scenario. The cap-and-trade schemes mainly affect the electricity sector, since the price of coal-fired generation rises relative to other generation technologies. For discussion of macro-economic estimates of mitigation costs for the US from this study and others, see Section 11.4.3.1 .

Figure 11.4: Carbon dioxide reductions by sector in the NCEP, Cap-Trade, and No-Safety Cases, 2015 and 2025

Notes: National Energy Modeling System, runs BING-ICE-CAP.D0 2100 5C BING-CAP.D0 2100 5A, and BING-NOCAP.D0 2080 5A. Note that NCEP includes technology mandates, and Cap-Trade is without technology mandates.

Source: US Energy Information Administration ( EIA)( 2005 [NPR, MoS] , p.15).

11.3.5 Portfolio analysis of mitigation options

Portfolio analysis in this context is the study of the mix of actions available to reduce emissions or adapt to climate change and to business in diversifying their investments against risk.

One issue is the allocation of GHG abatement across sectors or regions. Capros and Mantzos 2000 [NPR, MoS] ) show that, within the EU, equal percentage reductions across sectors cost more than twice as much as a least cost distribution (which can be obtained by, for example, allowing trade between sectors); see Table 11.9 . The table also shows the gains through international trading both across the EU and in Annex I, confirming the benefits reported in the TAR from a wide range of previous literature.

The reference case assumes that the Kyoto commitment is implemented separately by domestic action in each EU member state. The alternative reference case assumes that, within a member state, the overall emission reduction target of the burden-sharing agreement applies equally to each individual sector of the economy, with allocation evidently being more expensive than the least-cost approach in the reference case.

A related issue is the allocation of CO₂ emission reductions under Kyoto to sources in the EU Emissions Trading Scheme (EU ETS), as compared all non-ETS sources. ( Klepper and Peterson 2006 ) ), using a CGE model, conclude that ETS National Allocation Plans reduce the allowance price in the ETS below the implicit tax necessary for reaching the Kyoto targets in the non-ETS sectors, implying significant distortion. The limited use of CDM and JI to meet the allocations would result in a negative effect on welfare of close to 1% in 2012 relative to ‘business as usual’; this assumes that EU Member States do not import more than 50% of their required reductions and that they do not import ‘hot air’. Unrestricted trading in CDM and JI credits and allowances would result in an allocation where the Kyoto target can be met with hardly any welfare costs.

Jaccard et al. 2002 [NPR] ) evaluate the cost of climate policy in Canada. They compare the costs of achieving the Canadian Kyoto target in 2010 (using the CIMS model) for equal sector targets or one national target. According to their estimates, the electricity, residential, and commercial/institutional sectors contribute more, at lower marginal costs, to reductions when there is one national target, while the industry and transportation sectors contribute less. For example, the marginal cost for the electricity sector is about 20 US$/tCO₂-eq for the sector target and 80 US$/tCO₂-eq for the national target, while those of industrial sector are 200 and 80 US$/tCO₂-eq respectively.

Both studies illustrate a general finding that a portfolio of options which attempts to balance emission reductions across sectors with ‘equal percentage reductions’ is more costly than optimizing the policy mix for cost effectiveness.

Another aspect of mitigation options is the opportunity afforded by portfolio analysis to reduce risks and costs. Because fossil fuel prices are uncertain and variable, there are potential benefits in portfolios of energy supply sources that increase diversity so as to include, in particular, sources such as renewables and nuclear, the costs of which do not depend on fossil fuel prices. Long-standing methods from finance theory can help to quantify a new low-carbon technology’s contribution to overall risk, and to quantify costs associated with the development of a set of options for GHG mitigation and energy security. The portfolio approach differs from the traditional stand-alone cost approach in that it introduces market risk and includes inter-relationships between the costs of different technologies ( (Awerbuch, 2006, ) MITI). New technologies that diversify the generating mix and low-carbon options tend to be quantifiably more diverse than business-as-usual reliance on fossil fuels (see ( Stirling, 1994; ) 1996; Grubb et al., 2006 [SRC] ). Moreover, in contrast to the expected year-to-year variability of fossil fuel prices (which can be estimated from historic patterns), operating costs for wind, solar, nuclear and other capital-intensive non-fossil technologies are largely uncorrelated to fossil fuel prices.

Theory, supported by application, suggests that risk-optimized generating mixes will include larger shares of wind, geothermal and other fixed-cost renewables, even where these technologies cost more than gas and coal generation. Optimal mixes will also enhance energy security while simultaneously minimizing expected generating cost and risk. Awerbuch, Stirling, Jansen and Beurskens 2006 [NotFound] ) explore the limitations of the mean-variance portfolio (MVP) approach, and compare MVP optimal generating mixes to ‘maximum diversity’ mixes that also provide protection against uncertainty, ignorance and ‘surprise’. They find that the optimal mixes in both cases contain larger shares of wind energy.

These findings suggest that portfolios of cross-sector energy options that include low-carbon technologies and products will reduce risks and costs, simply because fossil fuel prices are more volatile relative to other costs, in addition to the usual benefits from diversification.

11.4.1 Measures of economic costs

Chapter 2 discusses cost concepts. Here we report, where available, the prices associated with CO₂ emissions, and negative or positive impacts on GDP, welfare and employment.

The TAR reviewed studies of climate policy interactions with the existing tax system. These interactions change the aggregate impact of a climate policy by changing the costs associated with taxes in other markets. They also point to the opportunity for climate policy – through carbon taxes or auctioned permits – to generate government revenue and, in turn, to reduce other taxes and their associated burden. The TAR pointed to this opportunity as a way to reduce climate policy costs. Since the TAR, additional studies have extended the debate ( Bach et al., 2002 [MoS] ; Roson, 2003 [MoS] ). Meanwhile, such arguments have been the basis of the UK Climate Change Levy and linked reduction in National Insurance Contributions, small auctions under the EU ETS and US NO_x Budget Program, large proposed auctions under the Regional Greenhouse Gas Initiative in the United States, and proposals in the United States, Japan, and New Zealand for carbon taxes.

11.4.2 Policy analysis of the effects of the Kyoto Protocol

Most analyses discussed in the TAR focused on national emission policies under the Kyoto Protocol in the form of an economy-wide tax or tradeable permit system. This has continued to be an active area of policy modelling since the Kyoto Protocol came into force. Global cost studies of the Kyoto Protocol since the TAR have considered more detailed implementation questions and their likely impact on overall cost. Chief among these have been the impact of the Bonn-Marrakesh agreements concerning sink budgets, the non-participation of the United States, and banking and the use of ‘hot air’ ( Manne and Richels, 2001 [NPR, ARC] ;( Böhringer, 2002; ) den Elzen and de Moor, 2002 [JoC, ARC] ;( Löschel and Zhang, 2002; ) ( Böhringer and Löschel, 2003; ) McKibbin and Wilcoxen, 2004 [MoS] ;( Klepper and Peterson, 2005 ) ).

U.S. non-participation in the Kyoto Protocol, coupled with the increase in sink budgets in Bonn and Marrakech, implies that the target for Annex B countries as a whole will likely be met with virtually no effort. In other words, excess allowances in Russian and Ukraine (referred to as ‘hot air’) roughly equal the shortfall in Europe, Japan, Canada, and other countries. However, some of these same studies emphasize that strategic behaviour by Russia and Ukraine, acting as a supply cartel and/or choosing to bank allowances until the next commitment period, will lead to a positive emission price ( (Löschel and Zhang, 2002; ) ( Böhringer and Löschel, 2003; ) ( Maeda, 2003; ) ( Klepper and Peterson, 2005 ) ). For example, ( Böhringer and Löschel 2003 ) ) use a large-scale static CGE model of the world economy to analyse the costs of Kyoto in different scenarios with and without Annex B emissions trading and U.S. participation. GDP costs of Kyoto for 2010 without US participation, with Annex B trading, but without use of ‘hot air’ are estimated at 0.03% for Annex B (without US) for a carbon price of 13 US$/tCO₂, with a 6.6% reduction in total Annex-B CO₂. Regional GDP costs are 0.05% for the EU15 and Japan, and 0.1% for Canada, with benefits of 0.2% for the European Economies in Transition and 0.4% for Russia and other countries in Eastern Europe, the Caucasus and Central Asia. Without Annex B trading, the costs are estimated at 0.08% for Annex B (without US).

National and regional studies cited below also suggest similar low or negligible costs for the ratified Kyoto Protocol for Canada, the EU and Japan compared with the estimates in the TAR, depending on the extent of trade in emission permits and CDM/JI certificates. The importance of CDM supply and other assumptions on the carbon price is shown in Figure 11.5 ( den Elzen and de Moor, 2002 [JoC, ARC] ).

Figure 11.5: Key sensitivities for the emission permit price from the FAIR model applied to the Kyoto Protocol under the Bonn-Marrakesh Accords

11.4.3 National and regional studies of responses to mitigation policies

As individual countries have begun contemplating domestic policy responses (see Chapter 13 ), an increasing number of studies have focused on more detailed national cost assessments. This increased detail includes both more careful representation of proposed and actual policy responses and more disaggregated results by sector, region, and consumer group. This detail is difficult to achieve in the context of a global model. We briefly summarize the results of studies for various countries/blocks on the basis of the literature available.

11.4.3.5 Policy studies for China

The ERI 2003 [NPR, MoS] ) report on three alternatives for China’s development to 2020 presents effects of policies that reduce CO₂ emissions in a ‘green growth’ scenario. For the same GDP growth of 7% per year, policies of accelerated economic reform, increased energy efficiency standards, higher taxes on vehicle fuels and more use of low-carbon technologies in power generation reduce the growth of CO₂ to 1.7% per year compared to 3.6% per year in an ‘ordinary effort’ scenario.

Chen 2005 [MoS, ARC] ) presents a comparison of assumed marginal abatement cost curves and GDP costs associated with various reduction efforts in China in different models (see Figure 11.6 and Table 11.12 below). Chen (p. 891) discusses the reasons for the differences, which are largely due to differences in baselines and assumptions about available technologies and substitution between fossil and non-fossil energy. GDP costs for 2010 vary: 0.2 and 1.5% reduction compared to baseline, associated with a 20% reduction in CO₂ compared to baseline. ( Garbaccio et al. 1999 ) ) consider smaller CO₂ reductions – between 5 and 15% – and find not only lower costs, but potentially positive GDP effects after only a few years owing to a double-dividend effect.

Figure 11.6: A comparison of Marginal Abatement Curves for China in 2010 from different models

Source: adapted from Chen 2005 [MoS, ARC] , p. 891)

Table 11.12: A comparison of GDP loss rates for China across models in 2010

Model	Emission reduction compared to baseline (%)	Marginal carbon abatement cost ((2000)US$/tCO2)	GDP (GNP) loss relative to baseline (%)
GLOBAL 2100	20	25	1.0
	30	50	1.9
GREEN	20	4	0.3
	30	7	0.5
Zhang’s CGE model	20	7	1.5
	30	13	2.8
China MARKAL-MACRO	20	18	0.7
	30	22	1.0
	40	35	1.7

Notes: Marginal carbon abatement costs were originally measured at 1990 prices in GLOBAL 2100, at 1985 prices in GREEN, at 1987 prices in Zhang’s CGE model, and at 1995 prices in China MARKAL-MACRO. They were converted to 2000 prices for comparison with other carbon prices in the chapter.

Source: adapted from Chen 2005 [MoS, ARC] , p.894)

11.4.4 Post-Kyoto studies

The macro-economic cost measure adopted in the literature on mitigation costs is generally GDP or gross marketed world output, excluding valuations of environmental costs and benefits.

11.4.4.1 A comparison of the macro-economic costs of mitigation to 2030 from modelling studies

Since the TAR, groups of modellers have found a reduction in expected macro-economic costs as a result of the use of multigas options (EMF21, Weyant et al., 2006 [ARC] ) (see Section 3.3.5.4 ) and because carbon prices affect technological change in the models (EMF19, IMCP) (see Section 11.5 ). Figure 11.7 summarizes the 2030 data brought together in these studies as well in as other post-TAR Category III (stabilization at around 550ppm CO₂-eq) studies covered in Chapter 3 . ^[11] The figure is in 3 parts, showing (a) the carbon prices in US$( 2000 ) by 2030 (typically a rising trend) and their effects on CO₂ emissions, (b) the effects of CO₂ abatement on GDP, and (c) the relationship between carbon prices and gross world output (GDP). All data are differences from the baseline projections for 2030 . The studies are grouped around two of the stabilization categories set out in Chapter 3 ( Table 3.5 ), with corresponding insights.

Figure 11.7: Year 2030 estimated carbon prices and gross-world-product (GDP) costs of various pathways to stabilization targets

Notes: Figure 11.7 shows, for 2030, the carbon price, CO₂ abatement relative to the baseline, and global GDP differences from baseline for five different sets of stabilization studies: EMF21 radiative forcing at 4.5 W/m² (multigas); IMCP at 550 and 450ppm (CO₂-only with induced technological change); EMF19 at 550ppm (CO₂-only with induced technological change) and 6 studies in category III included in Figure 3.24 . The results as shown exclude incomplete sets (i.e. data have to be available for all three variables shown). The EMF21 results exclude studies unsuitable for near-term analysis (e.g. substantial effects for a past year). The IMCP results exclude those from two experimental/partial studies. The breakdown into Category III and IV scenarios treats CO₂-only studies as if they also allow for cost-effective non-CO₂ multigas GHG mitigation (see Table 3.14 ). Note that prices and outputs are based on various definitions, so the figures are indicative only. The price bases in the original studies vary and have been converted to 2000 US$.

Sources: Weyant, 2004 [ARC] ; Masui et al., 2005 [ARC] ; Edenhofer et al., 2006b [NPR, SRC] , Weyant et al., 2006 [ARC] and Chapter 3 .

Category IV stabilization trajectories from 25 scenarios: In most models (24 of the 25 scenarios ^[12] ) the ‘optimal’ trajectory towards stabilization at 4.5W/m² (EMF21 studies), or the near-equivalent 550 ppm CO₂-only (IMCP and EMF19), requires abatement at less than 20% CO₂ compared to baseline by 2030, with correspondingly low-carbon prices (mostly below 20 US$/tCO₂-eq, all prices in 2000 US$). Costs are less than 0.7% global GDP, consistent with the median of 0.2% and the 10–90 percentile range –0.6 to 1.2% for the full set of scenarios given in Chapter 3 (see Figure 3.14 ). Carbon prices in the EMF21 multigas studies for 4.5W/m² by 2030 average 18 US$/tCO₂-eq, and span 1.2–26 US$/tCO₂-eq, except one at 110 US$/tCO₂-eq. Carbon prices in the corresponding 550 ppm CO₂-only studies in EMF19 average 14 US$/tCO₂ and span 3-19 US$/ tCO₂-eq, except one at 50 US$/tCO₂. Six of the IMCP 550 ppm CO₂-only models have 2030 prices in the range 7–12 US$/tCO₂, but four have low to zero prices in 2030, bringing the average to only 6 US$/tCO₂.

Category III stabilization trajectories from 12 scenarios: In 11 of the 12 post-TAR scenarios, ^[13] abatement is less than 40% of CO₂ by 2030 . Costs are below 1% GDP, consistent with the median of 0.6% and the 10–90 percentile range 0 to 2.5% for the full set in Chapter 3 , which also has a range of 18–79 US$/tCO₂-eq for carbon prices (see Figure 3.14 ). The largest comparable dataset available in this category is the IMCP 450ppm CO₂-only studies. Most of these produce a carbon price by 2030 in the range 20–45 US$/tCO₂, with one higher outlier, and a mean of 31 US$/tCO₂ (just over 110 US$/tC). The other Category III models nearly all give higher prices.

The lower estimates of costs and carbon prices for studies assessed here, in comparison with the full set of studies reported in Chapter 3 , are mainly caused by a larger share of studies that allow for enhanced technological innovation triggered by climate policies; see 11.5 below. The impact of endogenous technological change is greater for more stringent mitigation scenarios.

Figures 11.7 (a) and (c) show how the carbon prices affect CO₂ and global GDP in the models. Note that carbon prices are rising (not shown in Figure 11.7 ) – sharply for some of the higher numbers – from lower levels in 2020 and also after 2030 . Most models considered in this analysis therefore suggest that the 20–50 US$/tCO₂ cost category of the sector studies is the carbon price level which, if reached globally by 2020 – 2030, delivers trajectories compatible with subsequent stabilization at mid-category III levels. The corresponding CO₂ reduction by 2030 is 5–40% relative to baseline (which varies between studies, with higher baselines giving higher reduction percentages in 2030 ).

Figure 11.7 (b) shows the CO₂ abatement plotted against world GDP. In most studies, higher abatement is associated with higher loss of GDP. The relationships vary, and two models in particular stand out as radically different from others (E3MG and FUND). Three models in the IMCP predict GDP gains under different assumptions. ^[14] These prices and costs are largely determined by the approaches and assumptions adopted by the modellers, with GDP outcomes being strongly affected by assumptions about technology costs and change processes (see 11.5 below), the use of revenues from permits and taxes (see above), and capital stock and inertia (considered in 11.6) ( Barker et al., 2006a [NPR, MoS, SRC] ; Fischer and Morgenstern, 2006 [MoS] ).

11.4.5 Differences between models

Research has continued to focus on differences in various cost estimates between models ( Weyant, 2000 [NPR, ARC] ; Weyant, 2001 [NPR, ARC] ; Lasky, 2003 [NPR, MoS] ; Weyant, 2003 [NPR, MoS, ARC] ; Barker et al., 2006a [NPR, MoS, SRC] ; Fischer and Morgenstern, 2006 [MoS] Weyant 2001 [NPR, ARC] ) argues that the five major determinants of costs are: projections for base case GHG emissions; climate policy (flexibility, for example); substitution possibilities for producers and consumers; the rate and process of technological change; and the characterization of mitigation benefits. Turning to the base case, he notes the importance of assumptions about population and economic activity, resource availability and prices, and technology availability and costs. The key policy feature is flexibility, in other words whether trading over companies, nations, gases, and time is allowed. Substitution possibilities are governed by assumptions about the malleability of capital, economic foresight, and technology detail. Technology modelling includes assumptions about whether technological change is endogenous or exogenous, and whether technology costs drop with increasing use of technologies. Finally, mitigation benefits may be included in varying degrees in different models.

The factors accounting for differences between cost estimates can be divided into three groups: features inherent in the economies being studied (for example, high substitution possibilities at low cost), assumptions about policy (such as the use of international trading in emission permits, or the recycling of auction revenues), and simplifying assumptions chosen by the model builders to represent the economy (how many sector or regions are included in the model). The first two sets of factors can be controlled by specifying the countries and time-scales of the mitigation action, and the exact details of the policies, as in the EMF-16 studies. However, differences in modellers’ approaches and assumptions persist in the treatment of substitution and technology. The various factors can be disentangled by a meta-analysis of published finding (this may include an analysis of analyses). This technique was first used in this context by Repetto and Austin 1997 [NPR] ) in a mitigation-cost analysis of GDP costs for the US economy. Fischer and Morgenstern 2006 [MoS] ) conduct a similar meta-analysis dealing with the carbon prices (taken to be the marginal abatement costs) of achieving Kyoto targets reported by the EMF-16 studies and discussed in the TAR ( Weyant and Hill, 1999 [MoS, ARC] ).

The crucial finding of these meta-analyses is that most of the differences between models are accounted for by the modellers’ assumptions. For example, the strongest factor leading to lower carbon prices is the assumption of high substitutability between internationally-traded products. Other factors leading to lower prices include the greater disaggregation of product and regional markets. This suggests that any particular set of results about costs may well be the outcome of the particular assumptions and characterization of the problem chosen by the model builder, and these results may not be replicated by others choosing different assumptions.

Like earlier studies, the comparison of model results in Barker et al., 2006a [NPR, MoS, SRC] ) emphasizes that the uncertainty in costs estimates comes from both policy and modelling approaches as well as the baseline adopted. Uncertainty about policy is associated with the design of the abatement policies and measures (flexibility over countries, greenhouse gases and time) and with the use of carbon taxes or auctioned CO₂ permits to provide the opportunity for beneficial reforms of the tax system or incentives for low-carbon innovation. In addition, targeted reductions in fossil-fuel use resulting from climate policies can yield benefits in terms of non-climate policy e.g. reductions in local air pollution. Uncertainty about the modelling approaches is associated with the extent to which substitution is allowed in terms of backstop technology, whether the economy responds efficiently (in terms of the use of CGE models), and whether technological change is assumed to respond to carbon prices, the topic of the next section. Uncertainty about the baseline is associated with assumptions adopted for rates of technological change and economic growth, and future prices of fossil fuels.

11.5.1 Endogenous and exogenous technological development and diffusion

A major development since the TAR has been the treatment of technological change in many models as endogenous – and therefore potentially induced by climate policy – compared to previous assumptions of exogenous technological change that is unaffected by climate policies (see glossary for definitions). This section discusses the effect of the new endogenous approach on emission permit prices, carbon tax rates, GDP and/or economic welfare, and policy modelling ( Chapter 2 , Section 2.7.1 discusses the concepts and definitions, and Chapter 13 provides a broader discussion of mitigation and technology policy choices).

The TAR reported that most models make exogenous assumptions about technological change (9.4.2.3) and that there continues to be active debate about whether the rate of aggregate technological change will respond to climate policies (7.3.4.1). The TAR also reported that endogenizing technological change could shift the optimal timing of mitigation forward or backward (8.4.5). The direction depends on whether technological change is driven by R&D investments (suggesting less mitigation now and more mitigation later, when costs decline) or by accumulation of experience induced by the policies (suggesting an acceleration in mitigation to gain that experience, and lower costs, earlier). Overall, the TAR noted that differences in exogenous technology assumptions were a central determinate of differences in estimated mitigation costs and other impacts.

Table 11.13 lists the implications for modelling of exogenous and endogenous technological change ^[15] and demonstrates the challenges for research. The table shows that, at least in their simplified forms, the two types of innovation processes potentially have very different policy implications in a number of different dimensions.

The role of technology assumptions in models continues to be viewed as a critical determinant of GDP and welfare costs, and emission permit prices or carbon tax rates ( Barker et al., 2002 [SRC] ; Fischer and Morgenstern, 2006 [MoS] ). These analyses cover large numbers of modelling studies undertaken before 2000 and regard the treatment of technology as influential in reducing costs and carbon prices, but find that the cross-model results on the issue are conflicting, uncertain and weak. Since the TAR, there has been considerable focus on the role of technology, especially in top-down and hybrid modelling, in estimating the impact of mitigation policies. However, syntheses of this work tend to reveal wide differences in the theoretical approaches, and results that are strongly dependent on a wide range of assumptions adopted ( Barker et al., 2006a [NPR, MoS, SRC] ; Stern, 2006 [NPR] ), about which there is little agreement (DeCanio, 2003 ).

The approaches to modelling technological change (see Section 2.7.2.1 ), include (1) explicit investment in research and development (R&D) that increases the stock of knowledge, (2) the (typically) cost-free accumulation of applying that knowledge through ‘learning-by-doing’ (LBD); and (3) spillover effects. These approaches are in addition to simple analyses of sensitivity to cost assumptions, especially when technological change is treated as exogenous. There have been many reviews (see Clarke and Weyant, 2002 [NPR, ARC] ; Grubb et al., 2002b [MoS, SRC] ; Löschel, 2002 [MoS] ; Jaffe et al., 2003 [NPR] ; Goulder, 2004 [NPR] ; Weyant, 2004 [ARC] ; Smulders, 2005 [NPR] ; Grübler et al. 2002 [NPR, ARC] ;( Vollebergh and Kemfert, 2005; ) Clarke et al., 2006 [ARC] ; Edenhofer et al., 2006b [NPR, SRC] ; Köhler et al., 2006 [NPR, SRC] ;( Newell et al., 2006; ) Popp, 2006b [MoS, SRC] ; Sue Wing, 2006 [MoS] ; Sue Wing and Popp, 2006 [NPR, SRC] ). One feature that emerges from the studies is the considerable variety in the treatment of technological change and its relationship to economic growth. Another is the substantial reductions in costs apparent in some studies when endogenous technological change is introduced, comparable to previously estimated cost savings from ad hoc increases in the exogenous rate of technological change ( Kopp, 2001 [NPR] ) or in the modelling of advanced technologies ( Placet et al., 2004 [NPR] p. 5.2 & 8.10).

This section reviews the effect of endogenizing techno-logical change on model estimates of the costs of mitigation. It follows the majority of the literature and takes a cost-effectiveness approach to assess the costs associated with particular emission or cumulative emission goals, such as post- 2012 CO₂ reduction below 1990 levels or medium-term pathways to stabilization.

The review shows that endogenizing technological change – via R&D responses and learning-by-doing – lowers costs, perhaps substantially, relative to estimates where the path of technological change is fixed from the baseline. The degree to which costs are reduced hinges critically on assumptions about the returns from climate change R&D, spillovers (across sectors and regions) associated with climate change R&D, crowding-out associated with climate change R&D, and (in models with learning-by-doing) assumed learning rates. Table 11.14 shows the policies that have been modelled to induce technological change, and how they have been introduced into the models.

The policies are in two groups: effects through R&D expendi-ture, and those through learning-by-doing.Unfortunately, our empirical understanding of these phenomena over long periods of time is no better than our ability to forecast exogenous rates of change. As Popp 2006b [MoS, SRC] ) notes, none of the ETC models he reviews use empirical estimates of technological change to calibrate the models because, until recently, there were few empirical studies of innovation and environmental policy. So although we are confident that mitigation costs will be lower than those predicted by models assuming historically-based, exogenous rates of technological change, views continue to differ about how much lower they will be.

11.5.2 Effects of modelling sectoral technologies on estimated mitigation costs

The Energy Modelling Forum conducted a comparative study (EMF19) with the aim of determining how models for global climate change policy analyses represent current and potential future energy technologies, and technological change. The study assesses how assumptions about technology development – whether endogenous or exogenous – might affect estimates of aggregate costs for a 550 ppm CO₂ concentration stabilization target. The modellers emphasize the detailed representations for one or more technologies within integrated frameworks. Weyant 2004 [ARC] ) summarizes the results, which indicate low GDP costs and a wide range of estimated carbon tax rates hinging on assumptions about baseline emission growth, as well as technology developments with regard to carbon capture, nuclear, renewables, and end-use efficiency. Figure 11.8 shows that the carbon tax rates are very low before 2050, with all models indicating values below about 14 US$/tCO₂ to 2030 . Six of the nine models generate values below 27 US$/tCO₂ by 2050 . By comparison, the EU ETS price of carbon reached nearly 35 US$/tCO₂ in August 2005 and again in April 2006 .

Perhaps more revealing in the EMF-19 study are the specific features chosen by various modelling teams in their respective papers. Six teams focused on carbon capture and storage ( Edmonds et al., 2004 [ARC] ;( Kurosawa, 2004; ) McFarland et al., 2004 [MoS] ; Riahi et al., 2004 [ARC] ; Sands, 2004 [MoS] ; Smekens-Ramirez Morales, 2004 [MoS] ), one on nuclear ( Mori and Saito, 2004 [MoS] ), one on renewables ( van Vuuren et al., 2004 [MoS, ARC] ), two on end-use efficiency ( Akimoto et al., 2004 [MoS, SRC] ;( Hanson and Laitner, 2004 ) ), and one on an unspecified carbon-free technology ( Manne and Richels, 2004 [ARC] ). The impacts associated with varying technology assumptions within a given model ranged from a net economic gain, to substantial cuts in the cost of stabilization, to almost no effect on the cost of stabilization.

Figure 11.8: Carbon price projections for the 550 ppm CO₂-only stabilization scenario

Source: Weyant 2004 [ARC] ).

Despite the wide range of results, they suggest some overarching conclusions ( Weyant, 2004 [ARC] ). First, technological development, however and under whatever policy it unfolds, is a (if not the) critical factor determining the long-term costs and benefits of mitigation. Second, there is no obvious silver bullet: a variety of technologies may be important depending on local circumstances in the future, and a portfolio of investments will be necessary to achieve significant mitigation at lower costs. Third, major technology shifts like carbon capture, advanced nuclear, and hydrogen require a long transition as learning-by-doing accumulates and markets expand so that they tend to play a more significant role in the second half of the century. By contrast, end-use efficiency may provide major opportunities in the shorter term.

11.5.3 The costs of mitigation with and without endogenous technological change

Modellers have pursued two broad approaches to endogenizing technological change, usually independently of each other: explicit modelling of R&D activities that contribute to a knowledge stock and reduce costs, and the accumulation of knowledge through learning-by-doing. Sijm 2004 [NPR, MoS] and Edenhofer et al. 2006b [NPR, SRC] ) provide detailed comparative assessments of different implementations of both approaches with a focus on mitigation costs when endogenous technology effects are ‘switched on’. Their syntheses provide a useful window for understanding the variation in results and how policies might induce technological change.

In his review, Sijm 2004 [NPR, MoS] ) distinguishes top-down models that mostly focus on explicit R&D effects, and bottom-up models that focus mostly on LBD effects. Among the top-down models, which are described in Table 11.15 , he finds considerable variation in the effect of including Endogenous Technological Change (ETC). While some models find a large reduction in mitigation costs (e.g. Popp, 2006a [NPR, SRC] ), some find small impacts (e.g. Nordhaus, 2002 [NPR, MoS] ). These differences can be attributed to:

• the extent of substitution allowed of low-carbon fuels for high-carbon fuels. When this factor is included, the reduction in costs is more pronounced, and the higher it is, the greater the reduction.
• the degree of ‘crowding-out’ associated with energy R&D expenditures. If new energy R&D is assumed to be in addition to existing R&D, this will generate larger reductions in mitigation than if new energy R&D is assumed to lead to a reduction in R&D elsewhere.
• the approach to spillover. In addition to justifying higher rates of return from R&D, spillover implies that the market outcome with too little investment could be improved by policy intervention.
• the degree of differentiation among R&D activities, the assumed rates of return from those activities, and the capacity of R&D activities to lower costs for low-carbon technologies.
• the rate of learning if LBD is included. Higher rates imply larger reductions in mitigation costs with ETC included.

The first point is that the way low-carbon and high-carbon energy are treated in the models –whether as complements or substitutes – is critical is determining the flexibility of the model to low-carbon innovation and costs of mitigation. Models that do not allow high levels of substitution between low-carbon and high-carbon energy ( (Goulder and Mathai, 2000; ) Nordhaus, 2002 [NPR, MoS] ; Popp, 2006b [MoS, SRC] ) indicate that R&D has less impact than those that do, e.g. by introducing a carbon-free backstop technology ( (Gerlagh and Lise, 2005; ) Popp, 2006b [MoS, SRC] ). Similar results are found more widely for LBD and R&D models: the more substitution possibilities allowed in the models, the lower the costs ( Edenhofer et al., 2006a [NPR, SRC] , p.104).

When providing evidence to support the second point – the studies of induced R&D effects via the stock of knowledge – Goulder and Schneider 1999 [PoC, JoC] ( Goulder and Mathai 2000 ) Nordhaus 2002 [NPR, MoS] Buonanno et al. 2003 [JoC] and Popp 2004 [MoS, SRC] ) differ considerably about the extent of crowding-out. In other words, does R&D have an above-average rate of return and does an increase in R&D to support the carbon-saving technologies come from ordinary production activities (no crowding-out), or equally valuable R&D in other areas (crowding-out)? Nordhaus 2002 [NPR, MoS] ) assumes complete crowding-out in which carbon-saving R&D has a social rate of return that is four times the private rate of return but, because it is assumed that it replaces other equally valuable R&D activities, it costs four times as much as conventional investment. At the other extreme, Buonanno et al. 2003 [JoC] ) consider spillovers that lead to similarly high social rates of return, but without the higher opportunity costs. Not surprisingly, Nordhaus finds very modest mitigation cost savings and Buonanno et al. find enormous savings. In general, induced technological change in a general-equilibrium framework has its own opportunity costs, which may reduce the potential for cost reduction in CGE models substantially.

Popp 2006b [MoS, SRC] ), in turn, suggests on the basis of the empirical evidence that half of the R&D spending on energy in the 1970 s and 1980 s took place at the expense of other R&D. Something between full and partial crowding-out appears more recently in ( Gerlagh and Lise 2005 ) ). Goulder and Matthai ( 2000 ) provide an example of the importance for cost reduction of parameters describing returns from R&D and capacity for innovation. They compare both R&D as new knowledge and learning-by-doing (LBD), finding a 29% reduction in the marginal costs with R&D by 2050 and 39% with LBD. As they note, however, this reflects the calibration of their model to a 30% cost saving based on Manne and Richels’ assumptions ( 1992 ). The model results simply reflect the choice of calibrated parameter values.

By contrast with the results for top-down models, Sijm 2004 [NPR, MoS] ) finds considerably more consistency among bottom-up models, where the effects of learning-by-doing typically reduce costs by 20% to 40% over the next half century, and by 60% to 80% over the next century. Importantly, however, these numbers are relative to a static technology alternative. To demonstrate the influence of this assumption, van Vuuren et al. 2004 [MoS, ARC] ) run their model without a carbon constraint, but with learning, to identify a baseline level of technological change. Their approach roughly halves the estimated effect of ETC on mitigation.

The variations in the estimated effects of learning on costs in bottom-up models are driven primarily by variations in the assumed rate of learning (in other words, the extent of the reduction in costs for each doubling of installed capacity). Estimates of these rates vary, depending on whether they are assumed or econometrically estimated, and whether they derive from expert elicitation or historical studies. These learning rates vary between four leading models by as much as a factor of two for a given technology, as shown in Table 11.16 .

The modelling of LBD is beset with problems. Model solutions become more complex because costs can fall indefinitely, depending on the extent of the market. Avoidance of multiple solutions typically requires the modeller to constrain the penetration of new technologies, making one element of the cost reduction effectively exogenous. Since many low-carbon technologies are compared with mature energy technologies early in the learning process, it becomes inevitable that their adoption spreads and that they eventually take over as carbon prices fall. Finally, the approach often assumes that diffusion and accompanying R&D are cost-free, although the investments required for the technologies with high learning rates may be comparable with those that are replaced.

In addition, the measurement of learning rates poses econometric problems. It is difficult to separate the effects of time trends, economies of scale and R&D from those of LBD ( (Isoard and Soria, 2001 ) ) and different functional forms and data periods yield different estimates, so the learning rates may be more uncertain than suggested by their treatment in the models. When controls for the effects of other variables are included, such as crowding-out effects, the influence of LBD in some models may become very small compared to the effect of R&D ( Köhler et al., 2006 [NPR, SRC] ; Popp, 2006c [MoS, SRC] ).

A second survey of ETC effects on aggregate mitigation costs comes from the Innovation Modelling Comparison Project (IMCP) ( Edenhofer et al., 2006b [NPR, SRC] ). Rather than reviewing previous results, the IMCP engaged modelling teams to report results for specific concentration scenarios and, in particular, with and without ETC. Like the van Vuuren et al. 2004 [MoS, ARC] ) study noted earlier, the IMCP creates a baseline technology path with ETC but without an explicit climate policy. This baseline technology path can then be either fixed, as autonomous technological change, or allowed to change in response to the climate policy.

Table 11.15 also summarizes the treatment of technological change in the IMCP models; in principle, the wide range of approaches provides additional confidence in the results when common patterns emerge. Like Sijm 2004 [NPR, MoS] Edenhofer et al. 2006b [NPR, SRC] ) find that, while ETC reduces mitigation costs, there continues to be a wide range of quantitative results: some are close to zero and others generate substantial reductions in costs.

Figure 11.9 shows the effects of introducing ETC into the models for the 550 and 450 ppm CO₂ stabilization scenarios 2000 – 2100 . The reductions in carbon prices and GDP are substantial for many studies in both stabilization cases when ETC is introduced. The effects on CO₂ reductions show that including ETC in the models leads to earlier reductions in emissions. It should be noted that the reduction of costs in IMCP models is not mainly driven by LBD. The assumptions about the crowding-out of conventional R&D by low-carbon R&D and the availability of mitigation options (models have different sets of options) are more important factors determining costs and mitigation profiles than LBD ( Edenhofer et al., 2006b [NPR, SRC] , p.101–104). One major research challenge is to test the influence of these aspects of ETC on current technologies by econometric and backcasting methods, fitting the models to historical data.

Figure 11.9: Averaged effects of including ETC on carbon tax rates, CO₂ emissions and GDP: 9 global models 2000 – 2100 for the 450 ppm and 550 ppm CO₂-only stabilization scenarios

Notes: The figures show the simple averages of results from 9 global models 2000 – 2100 for (a) carbon tax rates and emission permit prices in US$( 2000 )/tCO₂, (b) changes in CO₂ (% difference from base) and (c) changes in global GDP (% difference from base). The results are shown with and without endogenous technological change. The grey background lines show the range from the individual models for 450 ppm with ETC. See source for details of models.

Source: adapted from Edenhofer et al. 2006b [NPR, SRC] ).

Figure 11.9 emphasizes the range and the uncertainty of the results for induced technological change ^[16] from climate policies. The potential of ETC to reduce mitigation costs varies remarkably between different model types. For a 450 ppm CO₂-only concentration stabilization level at the upper end of the range, including ETC in the model reduces mitigation costs by about 90%, but at the lower end it makes no difference ( Edenhofer et al. 2006b [NPR, SRC] , p. 74). The averages also somewhat exaggerate the effects of ETC because there are other assumptions that affect the costs, as evident in a meta-analysis of the macro-economic costs of mitigation undertaken for the UK Treasury’s Stern Review ( Barker et al., 2006b [SRC] ). An example is the use of tax/permit revenues, as discussed in 11.4.4 above. This study combines the IMCP results on costs with earlier data on post-SRES scenarios ( Repetto and Austin, 1997 [NPR] ; Morita et al., 2000 [ARC] ) so that the effects of other assumptions can be identified. The average effects of including ETC in the IMCP models by 2030 for pathways to 550 and 450 ppm CO₂-only are reduced from 1.1 and 2.7% of global GDP compared to baseline, as shown in Figure 11.9 , to 0.4 and 1.3% respectively using the full equation of the meta-analysis, which allows for individual model outliers, time and scenario effects as well as the approaches and assumptions adopted by the modellers. In other words, allowing for technologies to respond to climate policies reduces the GDP costs of Category III stabilization, as estimated by the IMCP models, by 1.3% by 2030 . Costs across models of 2.1% without ETC, but allowing for emissions trading and backstop technologies, are reduced to 0.8% GDP by 2030 with ETC. The ETC effects become more pronounced in the reduction of costs for later years and as the stabilization targets become more stringent, partly due to the associated extra increases in the required carbon prices.

Edenhofer et al. conclude that the results for effects of ETC depend on:

• baseline effects: baseline assumptions about the role of technology that generate relatively low emission scenarios can leave little opportunity for further ETC effects;
• the assumption of the inefficient use of resources in the baseline (distinct from market failure associated with greenhouse gas emissions and climate change): this provides opportunities for policy to improve otherwise inefficient private decisions and may even raise welfare. Spillovers were an example of this in the Sijm (2004) discussion; some simulations also include inefficient energy investment decisions.
• how the investment decision is modelled: recursive savings decisions, as opposed to foresight and intertemporal opti-mization, provide less opportunity for investment and R&D to expand. In the Sijm (2004) context, less responsiveness in aggregate investment and R&D implies more crowding-out.
• the modelling of substitution towards a backstop techno-logy (such as a carbon-free energy source available at constant, albeit initially high, marginal cost): this can substantially affect the results. For example, if investment in the technology is endogenous and involves learning-by-doing, costs can fall dramatically. Popp (2006a, p.168) goes further, and shows that the addition of a backstop technology by itself can have a larger effect on mitigation costs than the addition of LBD. These results are also confirmed by the IMCP study (Edenhofer et al., 2006b, p.214). However, investment in backstop technologies requires time-consistent policies (Montgomery and Smith, 2006). It is therefore debatable to what extent the indicated potential for cost reduction can be realized under real-world conditions where a global, long-term and time-consistent climate policy has yet to be implemented.

11.5.4 Modelling policies that induce technological change

Most of the studies discussed so far consider only how endogenous technological change affects the cost associated with correcting the market failure of damaging GHG emissions through market-based approaches to carbon taxes and/or emissions trading schemes. However, when spillovers from low-carbon innovation are introduced into the modelling of ETC, for example where the social rate of return exceeds the private rate of return from R&D because innovators cannot capture all the benefits of their investment, there is a second market failure. This implies that at least two instruments should be included for policy optimization ( Clarke and Weyant, 2002 [NPR, ARC] , p.332; ( Fischer, 2003; ) ( Jaffe et al., 2005 ) ). Even without the spillover effect, however, the advantage of models with endogenous technological change is their potential to model the effect of technology policy, distinct from mitigation policy, or in tandem. As discussed in Chapter 13 , there has been increasing interest in such policies.

Surprising, few models have explored this question of mitigation versus technology policies, and they have focused instead on the cost assessments reviewed above. Those studies that have looked at this question find that technology policies alone tend to have smaller impacts on emissions than mitigation policies ( Nordhaus, 2002 [NPR, MoS] ; Fischer and Newell, 2004 [NPR] ; Popp, 2006b [MoS, SRC] ;( Yang and Nordhaus, 2006 ) ). In other words, it is more important to encourage the use of technologies than to encourage their development. On the other hand, with the existence of spillovers, technology policies alone may lead to larger welfare gains ( Otto et al., 2006 [NPR] ). However, the same study points out that an even better policy (in terms of improving welfare) is to fix the R&D market failure throughout the economy. Given the difficulty in correcting the economy-wide market failure (e.g., through more effective patent protection or significantly increased government spending on research), it may be unrealistic to expect successful correction within the narrow area of energy R&D. This is true despite our ability to model such results.

However, this does open up the possibility of portfolios of policies utilizing some of the revenues from emission permit auctions to provide incentives for low-carbon technological innovation. An example is the approach of Masui et al. 2005 [ARC] ) for Japan discussed in 11.3.4. Weber et al. 2005 [MoS] ), using a long-run calibrated global growth model, conclude that ‘…increasing the fraction of carbon taxes recycled into subsidizing investments in mitigation technologies not only reduces global warming, but also enhances economic growth by freeing business resources, which are then available for investments in human and physical capital’ (p. 321).

Unlike the studies that assess the effects of technology and mitigation policies on emissions and welfare in a simulation model, Popp 2002 [SRC] ) examines the empirical effect of both energy prices and government spending on US patent activities in 11 energy technologies in the period 1970 – 1998 . He finds that while energy prices have a swift and significant effect on shifting the mix of patents towards energy-related activities, government-sponsored energy R&D has no significant effect. While not addressing efforts to encourage private-sector R&D, this work casts doubt on the effectiveness of government-sponsored low-GHG research by itself as a mitigation option.

11.6.1 Structural trends in the transition

Most studies suggest that GHG mitigation shifts over time from energy efficiency improvements to the decarbonization of supply. This is the clear trend in the global scenarios survey in Chapter 3 ( Figure 3.23 ), and also in the time-path plots of energy against carbon intensity changes in the models in the IMCP studies ( Edenhofer et al., 2006b [NPR, SRC] ). It is also true of the national long-term studies surveyed in Chapter 3 ( Table 3.7 ); of the detailed sectoral assessments of Chapters 4 – 10 ; and the IEA’s ETP study ( IEA, 2006a [NPR] ). In the first quarter of this century, the majority of global emission savings are associated with end-use savings in buildings and, to a lesser degree, in industry and transport. Moreover, despite important savings in electricity use in these sectors, economies in mitigation scenarios tend to become more electrified ( Edmonds et al., 2006 [ARC] ). In the second quarter of this century, the degree of decarbonization of supplies starts to dominate efficiency savings as a result of a mix of strategies including CCS and diverse low-carbon energy sources. In the IEA study, the power sector consists of more than 50% non-fossil generation by 2050, and half of the remainder is made up of coal plants with CCS. The power sector still tends to dominate emission savings by 2030, even at lower carbon prices (see also Table 11.5 ), but obviously the degree of decarbonization is less.

There are two reasons for these trends. First, there are strong indications in the literature that improvements in energy efficiency with current technologies have greater potential at lower cost (see Chapters 5 – 7 ). This is apparent from the sectoral assessments summarized in Table 11.3 , where energy efficiency accounts for nearly all the potential available at negative cost (particularly in buildings), and at least as much as the potential available from switching to lower carbon fuels and technologies in energy supply, for costs in the range up to 20 US$/tCO₂ -eq. The second reason is that most models assume some inertia in the capital stock and diffusion of supply-side technologies, but not of many demand-side technologies. This slows down the penetration of low-carbon supply sources even when carbon prices rise enough (or when costs fall sufficiently) to make them economic. Some end-use technologies (such as appliances or vehicles) do have a capital lifetime that is much shorter than major supply-side investments; but there are very important caveats to this, as discussed below.

For the analysis of transitions during the first quarter of this century, then, most of the relevant modelling literature emphasizes, for stabilization between 650 and 550 ppm CO₂-eq (categories III and IV in table 3.5 ), energy supply and other sectors such as forestry in which mitigation potentials are dominated by long-lifetime, medium-cost options.

11.6.2 Carbon prices by 2030 and after in global stabilization studies

Many analyses in this report emphasize that efficient mitigation will require a mix of incentives: regulatory measures to overcome barriers to energy efficiency; funding and other support for innovation; and carbon prices to improve the economic attractiveness of energy efficiency and of low-carbon sources, and to provide incentives for low-carbon innovation and CCS. Most of the regulatory and R&D measures are sector-specific and are discussed in the respective sectoral chapters (4–10). Some implications of innovation processes are discussed below. Most global models focus on the additional costs of mitigation in the form of shadow prices or marginal costs, and the resulting changes that would be delivered by carbon prices. The carbon prices reached by 2030 are discussed in Section 11.4.4 above. The levels and trends in these prices are crucial to the transition processes.

The time trend of carbon prices after 2030 is important but specific to each model. Some models maintain a constant rate of price increase that largely reflects the discount rate employed (they establish an emissions time-path to reflect this). Two models in the EMF studies, for example, assume increases in carbon prices of about 5.5% per year and over 6% per year that are constant throughout the century. In this approach, carbon prices roughly treble over the period 2030 – 2050, and every two decades thereafter. Two models in the IMCP studies also use constant, and much lower, growth rates for prices that vary with the stabilization constraint. Edenhofer et al. 2006b [NPR, SRC] ) find that real carbon prices for stabilization targets rise with time in the early years for all models, with some models showing a decline in the optimal price after 2050 due to the accumulated effects of LBD and positive spillovers on economic growth. In these cases, a high-price policy in the earlier years may generate innovation that provides benefits in later years. In all these models, the rates of change frequently reflect intrinsic model parameters (notably the discount rate) and do not depend much on the stabilization target, which is reached by adjusting the starting carbon price instead. However, most but not all models with endogenous technical change have rates of carbon price increase that decline over time, and two models actually result in carbon price falls as technological systems mature.

A carbon price that rises over time is a natural feature of an efficient trajectory towards stabilization. The macro-economic cost depends on the average mitigation cost, which tends to rise more slowly and may decline with technical progress. The Stern review illustrates and explains scenarios in which rising carbon prices accompany declining average costs over time ( Stern, 2006 [NPR] ).

11.6.3 Price levels required for deep mid-century emission reductions: the wider evidence

Several other lines of evidence shed light on the carbon prices required to deliver transitions to deep mid-century CO₂ reductions. By contrast with the rising prices in most ‘optimal’ stabilization trajectories, some global models have been run with constant prices. Perhaps the most extensive, the IEA-ETP ( 2006a ) study (MAP scenario), returns global CO₂ emissions roughly equal to 2005 levels by 2050 (more than halving emissions from reference). This is consistent with a trajectory towards category III stabilization at around 550 ppm CO₂-eq, with carbon prices rising to 24 US,-eq, with carbon prices rising to 24 US,)$/tCO₂ ($87/tC) by 2030 and then remaining fixed. The IEA study emphasizes the combination of end-use efficiency in buildings, industry and transport, together with the decarbonization of power generation as indicated. In other global studies that report sectoral results, the power sector dominates emission savings even in the weaker category IV scenarios. Some other models with detailed energy sectors do not force a constantly rising price or display periods of relatively stable prices along with stable or declining emissions. ^[17]

The carbon price results are consistent with the technology cost analyses in Chapter 4 . These suggest that price levels in the 20–50 US$/tCO₂ range should make both CCS and a diversity of low-carbon power generation technologies economic on a global scale, with correspondingly large reduction potential attributed to the power sector in this cost range ( Table 11.8 Newell, Jaffe & Stavins 2006 [NotFound] ) focus on the economics of CCS at prices within this range, noting that the carbon prices required will depend not only on CCS technology but also upon the reference alternative. ( Schumacher and Sands 2006 ) ) also focus on CCS but, in the context of the German energy system, conclude that a similar range is critical ‘for CCS as well as advanced wind technologies to play a major role’ (p. 3941). Riahi et al. 2004 [ARC] ) project coal-based CCS costs up to 53 US$/tCO2. Corresponding reductions may accrue, whether a carbon price is considered to be implemented directly or as the incentive from certified emission reduction (CER) credits. ( Shrestha 2004 ) ) projects that ‘business-as-usual’ shares of coal in power generation by 2025 will be 46%, 78% and 85% in Vietnam, Sri Lanka and Thailand respectively, but an effective CER price of 20 US$/tCO₂ from 2006 onwards would reduce the share of coal to 18%, 0% and 45% respectively in the three countries by 2025 . Natural gas and, to a lesser extent renewables, oil and electricity imports are the main beneficiaries.

The sectoral results from Chapters 4 – 10 ( Table 11.3 ) suggest that carbon prices in the range 20–50 US$/tCO₂-eq could deliver substantial emission reductions from most sectors. Of the total potential identified below 100 US$/tCO₂-eq across all sectors, more than 80% is estimated to be economic at a cost below 50 US$/tCO₂-eq. Moreover, the lowest proportions are for agriculture (56%) and forestry (76%). Of the main sectors for which carbon cap-and-trade is being applied or considered at present, costs below 50 US$/tCO₂-eq account for 90% of the identified potential in energy supply and 86% in industry, whereas the proportion of the total below 20 US$/tCO₂-eq is about half (52%) and a quarter (27%) respectively for these sectors. This underlines the conclusion that carbon prices in the 20–50 US$/tCO₂-eq range would be critical to securing major changes in these principal industrial emitting sectors.

The bottom-up estimates of emission reductions available at less than 50 US$/tCO₂-eq for the total energy sector (supply, buildings, industry and transport) span 11.5–15 GtCO₂-eq/yr ( Table 11.3 , full range). This is strikingly similar to the range of CO₂ reductions by 2030 that global top-down studies consider to be necessary for trajectories consistent with stabilization in the Category III range ( Figure 11.7 (a), in the range 25–40% reduction of CO₂ which, against the central baseline projection of 37–40 GtCO₂-eq (WEO/A2) for energy-related emissions that is used for the bottom-up estimates, equates to 10–16 GtCO₂-eq. Incidentally, this also equates to global emissions in 2030 that are roughly at present levels).

The capital stock lifetime of industrial and forestry systems (discussed further below) means that it takes some decades for the impact of a given carbon price to work its way through in terms of delivered reductions. ^[18] The assessment of timing is complicated by the fact that most global stabilization studies model a steadily rising price with ‘perfect foresight’. However, Figure 11.7 (c) confirms that almost all models project prices of at least 20 US$/tCO₂-eq by 2030, and some breach the 50 US$/tCO₂-eq level earlier in that decade, as might be expected in order to secure the required reductions by 2030 . Applying the same statistical framing as Chapter 3 , the analysis of price trends confirms that global carbon prices in more than 80% of the Category III stabilization studies cross within the range $20–50/tCO₂-eq during the decade 2020 –30. These diverse strands of evidence therefore suggest a high level of confidence that carbon prices of 20–50 US$/tCO₂-eq (75–185 US$/tC-eq) reached globally in 2020 – 2030 and sustained or increased thereafter would deliver deep emission reductions by mid-century consistent with stabilization at about 550 ppm CO₂-eq (category III levels). To depict the impact in the models, such prices would have to be implemented in a stable and predictable manner and all investors would need to plan accordingly, at the discount rates embodied in the models.

Carbon prices at these levels would deliver these changes by largely decarbonizing the world’s electricity systems, by providing a substantial incentive for additional energy efficiency and, if extended to land use, by providing major incentives to halt deforestation and reward afforestation. ^[19] By comparison, prices in the EU ETS in 2005 peaked close to 30 euros (about 40 US$)/tCO₂. Transition scenarios for non-energy sectors (in particular agriculture and deforestation) are reported in the respective sectoral chapters and in some of the multi-gas studies in Chapter 3 .

Particularly in models that embody some economies of scale/ learning-by-doing, prices maintained at such levels largely decarbonize the power sector over a period of decades. Some of the models display a second period with a similar pattern, later and at higher prices, as fuel cell-based transport matures and diffuses. In integrated Category III scenarios, such scenarios can also deliver more potential abatement in the transport sector (at a higher cost), partly because several of the low-carbon transport technologies depend on the prior availability of low-carbon electricity. Assumptions about the availability of petroleum and the costs of carbon-based ‘backstop’ liquid fuels also tend to be very important considerations in terms of the associated net costs ( Edmonds et al., 2004 [ARC] ; Edenhofer et al., 2006b [NPR, SRC] ; Hedenus et al., 2006 [NPR] ).

The price in the 20 to 50 US$/tCO₂ range required to deliver such changes – and answers to the questions of whether and by how much further carbon prices might need to rise in the longer term – depend upon developments in three other main areas: the contribution of voluntary and regulatory measures associated with energy efficiency; the extent and impact of complementary policies associated with innovation and infrastructure; and the credibility, stability and conviction that the private sector attributes to the price-based measures. We consider each in turn.

11.6.4 Complementary measures for deep emission reductions

The sectoral and multi-gas studies indicate that substantial emission savings are still available at low cost (< 20 US$/tCO₂), particularly from buildings ( Chapter 6 ) and end-use efficiencies in a number of industrial sectors ( Table 7.8 ); many governments are therefore already well embarked upon policies to exploit these low-cost opportunities. The IEAs World Energy Outlook ( IEA, 2006b [NPR] , Part 2) estimates that such measures could contribute a 16% reduction below the reference level by 2030 . This would be an important contribution, but clearly insufficient to get close to halting or reversing global emissions growth in the absence of price-based measures.

Innovation will also be crucial for deep reductions by mid-century and in the longer term. Some of the technologies required to deliver ongoing emission reductions out to 2050 are already commercialized, but others (such as CCS) are not (see sector chapters). Deeper emission reductions will get more and more difficult over time without accelerated innovation bringing down costs, and increasing the diversity, of low-carbon options. Achieving the mitigation scenarios indicated therefore requires adequate progress in a range of relevant industries based on low-carbon technology ( Weyant, 2004 [ARC] ; IEA, 2006b [NPR] ). Chapter 2 has laid out the basic principles for low-carbon innovation and Chapter 3 the long-term role of technologies in stabilization scenarios. The sectoral chapters discussed the specific technologies and Section 11.5 covered the post-TAR modelling of induced technological change. This section briefly assesses the insights relating to innovation that are relevant to transitions in the second quarter of this century.

The conceptual relationship between such innovation investments and measures relating to carbon pricing is sketched out in Figure 11.10 . Most low-carbon technologies (at least for supply) are currently much more expensive than carbon-based fuels. As R&D, investment and associated learning accumulates, their costs will decline, and market scale may grow. Rising carbon prices bring forward the time when they become competitive (indeed, many such technologies might never become competitive without carbon pricing). The faster the rise in carbon prices particularly if industry can project such increases with confidence in a clear and stable policy environment the sooner such technologies will become competitive and the greater the overall economic returns from the initial learning investment.

Figure 11.10: Relationship between learning investments and carbon prices

Notes: The figure illustrates cost relationships for new low-carbon technology as experience and scale build over time. Initially introduction is characterized by relatively high current costs and a very small market share, and requires a high unit rate for learning investment. With increasing scale and learning, costs move towards existing, higher-carbon sources, the costs of which may also be declining, but more slowly. Rising carbon prices over time bring forward the time when the new technology may be competitive without additional support, and may greatly magnify the economic returns from the initial learning investment.

Source: Adapted with authors permission from ( Neuhoff 2004 ) ).

However, the literature also emphasizes that carbon pricing alone is insufficient. ( Sanden and Azar 2005 ) ) argue that carbon cap-and-trade is important for diffusion picking technologies from the shelf but insufficient for innovation replenishing the shelf. Foxon 2003 [NPR, MoS] ) emphasizes the interaction of environmental and knowledge market failures, arguing that this creates systemic obstacles that require government action beyond simply fixing the two market failures (of climate damages and technology spillovers) independently. There is therefore general consensus in the literature that, whilst emission reduction (including pricing) mechanisms are a necessary component for delivering such innovation, they are not sufficient: efficient innovation requires even more government action.

This underlines the complexity of measures required to drive adequate innovation. On the basis of four general lessons from US technology policy, Alic et al. 2003 [NPR] ) derive various specific conclusions for action. ^[20] They break them down into direct R&D funding, support for deployment, and support for education and training. However, they also underline that technology policies alone cannot adequately respond to global climate change. They must be complemented by regulatory and/or energy pricing policies that create incentives for innovation and adoption of improved or alternative technologies the technological response will depend critically on environmental and energy policies as well as technology polices.

Philibert 2005 [JoC] ) places climate technology policy in the context of the wider experience of US, European and IEA technology programmes and present initiatives, and discusses explicitly the international dimensions associated with globalization, export credit, diffusion, standards and explicit technology negotiations. Grubb 2004 [SRC] ) outlines at least six different possible forms of international technology-oriented agreements that could, in principle, help to foster global moves towards lower-carbon energy structures (see Chapter 13 ).

The common theme in all these studies is the need for multiple and mutually supporting policies that combine technology push and pull across the various stages of the innovation chain, so as to foster more effective innovation and more rapid diffusion of low-carbon technologies, both nationally (the tax and subsidy regime for Japan discussed in 11.4.3.4, for example) and internationally. Most studies also emphasize the need for feedback enabling policy to learn from experience and experimentation using learning-by-doing in the process of policy development itself.

11.6.5 Capital stock and inertia determinants of transitions in the second quarter of the century

The scope for change, and the rate of transition, will be constrained by the inertia of the relevant systems. The IPCC SAR Summary for Policymakers noted that ‘the choice of abatement paths involves balancing the economic risks of rapid abatement now (that premature capital stock retirement will later be proved unnecessary) against the corresponding risk of delay (that more rapid reduction will then be required, necessitating premature retirement of future capital stock).’ Capital stock is therefore a central consideration.

The time scales of stock turnover vary enormously between different economic sectors, but appear to be very long for most greenhouse-gas emitting sectors. Typical investment time scales are several decades for forestry, coal mining and transporting facilities, oil & gas production, refineries, and power generation. On the demand side, observed time scales for typical industrial stock using energy are estimated to range from decades to a century ( Worrell and Biermans, 2005 [ARC] ; see Table 11.17 ). The time scales for other end-use infrastructure (e.g. processes, building stock, roads and rail) may be even longer, though components (such as heaters, cars) may have considerably faster turnover.

However, Lempert et al. 2002 [NPR] ) caution against overly simplistic interpretations of nameplate lifetimes, emphasizing that they ‘are not significant drivers [of retirement decisions] in the absence of policy or market incentives’ and that ‘capital has no fixed cycle’. This can be crucial to rates of decarbonization. A study of the US paper industry found that ‘an increase in the rate of capital turnover is the most important factor in permanently changing carbon emission profiles and energy efficiency’ ( (Davidsdottir and Ruth, 2004 ) ). Similarly, emission reductions in the UK power sector were largely driven by the retirement of old, inefficient coal plant during the 1990 s, through sulphur regulations which meant plant owners were faced with the choice of either retrofitting stock or retiring it ( Eyre, 2001 [JoC] ). Such micro-level ‘tipping points’ at which investment decisions need to be taken may offer ongoing opportunities for lower cost abatement.

Energy system inertia provides another dimension to the time scales involved. It has taken at least 50 years for each major energy source to move from 1% penetration to a major position in global supplies. Such long time scales – and the even longer periods associated with interactions between systems – imply that, for stabilization, higher inertia brings forward the date at which abatement must begin to start meeting any given constraint, and lowers the subsequent emissions trajectory ( Ha-Duong et al., 1997 [JoC, SRC] ). In the context of stabilization at 550 ppm CO₂ van Vuuren et al. 2004 [MoS, ARC] ( and Schwoon and Tol 2006 ) ) demonstrate that higher inertia in the energy system brings forward mitigation. ^[21]

However, beyond a certain point, inertia can also dramatically increase the cost of stabilization, particularly when infrastructure constraints are likely to limit the growth of new industries more than established ones. Manne & Richels 2004 [NotFound] ) illustrate that if global total contributions from new (renewable) power sources are limited to 1% by 2010 and treble each decade thereafter, the world has little choice other than to continue expanding carbon-intensive power systems out to around 2030 . This feature appears to drive their finding of high costs for 450 ppm CO₂ stabilization, since much of this stock then has to be retired in subsequent decades to meet the constraint. This pattern contrasts sharply with some other studies, such the MIT study ( McFarland et al., 2004 [MoS] ) that states an opposing time profile based partly upon the rapid deployment of natural gas plant, including CCS. Crassous et al. 2006 [MoS, ARC] ) also find high costs by assuming that long-lived infrastructure construction continues without foresight over the century. If low-carbon transport technologies do not become available quickly enough, the economy is squeezed as carbon controls tighten. They also show that the early adoption of appropriate infrastructure avoids this squeeze and allows lower costs for carbon control. Drawing partly on more sociological literature, and the systems innovation literature ( (Unruh, 2002 ) ), tends to support a view that we are now ‘locked in’ to carbon-intensive systems, with profound implications: ‘Carbon lock-in arises through technological, organizational, social and institutional co-evolution ... due to the self-referential nature of [this process], escape conditions are unlikely to be generated internally.’

Lock-in is less of a problem for new investment in rapidly developing countries where the CDM is currently the principal economic incentive to decarbonize new investments. The ( Shrestha 2004 ) ) study cited above illustrates how the structure of power sectors could be radically different depending upon the value of Certified Emission Reduction (CER) units. Their finding that an effective CER price of 20 US$/tCO₂ from 2006 onwards could drive a radical switch of investment from new coal plants and primarily to natural gas and renewables in the three Asian countries studied would not only represent a large saving in CO₂ emission, but a totally different capital endowment that would sustain far lower emission trajectories after 2030 . Again, this supports the conclusion that carbon prices of this order play a very important role, with their potential to forestall the construction of carbon-intensive stock in developing countries. Diverse policies that deter investment in long-lived carbon-intensive infrastructure and reward low-carbon investment may maintain options for low stabilization levels in Category I and II at lower costs.

At a global scale, van Vuuren et al. 2004 [MoS, ARC] ) present a systematic set of results showing the effects of different time profiles for carbon prices in studies that combine the representation of inertia and induced innovation. A carbon price that rises linearly to 82 US$/tCO₂ by 2030 reduces emissions by 40% by 2030 if the tax is introduced in 2020 and raised sharply, but by 55% if it is introduced in 2000 and increased more slowly. van Vuuren et al. do not describe the impact on subsequent trajectories, but clearly the capital stock endowment differs substantially. Moreover, Lecocq et al. 1998 [ARC] ) demonstrate that, in the face of uncertainty, an efficient approach may include greater effort directed at reducing investment in longer-lived carbon intensive infrastructure, over and above the incentives of any uniform carbon price.

Chapter 3 ( Section 3.6 ) emphasizes the importance of ‘hedging’ strategies based upon sequential ‘act-then-learn’ decision-making. Mitigation over the next couple of decades that would be consistent with enabling stabilization at lower levels (Categories I, II or III) does not irrevocably commit the world to such levels. The major numerical addition to the literature in this vein appears to be that of ( Mori 2006 ) ). Using the MARIA model, he analyses optimal strategies to limit the global temperature increase to 2.5 ºC given uncertainty about climate sensitivity in the range of 1.5-4.5 ºC per doubling of CO₂-equivalent. When there is no uncertainty, only the above-average sensitivities require significant mitigation in the next few decades. In the context of uncertainty, however, the optimal strategy is to keep global emissions relatively constant at present levels until the uncertainty is resolved, after which they may rise or decline depending upon the findings.

11.6.6 Investment and incentive stability

The longevity of capital stock, projections of rapidly growing global emissions under ‘business-as-usual’, and the importance of industrial scale and learning in low-carbon technology industries all illustrate the central role of investment in relation to the climate change problem. As discussed in Chapter 4 , the IEA 2004 [NPR] ) estimates that about US$20 trillion will be invested in energy supplies up to 2030, half to two-thirds of which is associated with power generation.

Several major studies shed light upon the investment implications of low-carbon scenarios over the next few decades. The World Bank 2006 [Ambiguous] ) estimates that to ‘significantly de-carbonize power production’ would require incremental investments of ‘up to’ US $40bn per year globally, of which about US$30bn per year would be in non-OECD countries. However, in a comprehensive scenario, this would be offset by the reduced investment requirements resulting from improved end-use efficiency. The IEA WEO ( 2006b ) ‘alternative policy scenario’ estimates that an increased investment of US$2.4 trillion in improved efficiency would be more than offset by US$3 trillion savings in supply investments. The more aggressive IEA ‘Map’ scenario ( IEA, 2006a [NPR] ), that returns emissions to 2005 levels by 2050 (and is consistent with trajectories towards stabilization between 550 and 650 ppm CO₂-eq) as discussed above, reflects greater impact as a result of switching investment from more to less carbon-intensive paths. Investments across renewables, nuclear and CCS are projected of US$7.9 trillion, US$4.5 trillion of which is offset directly by the reduced investment required in fossil-fuel power plants. Most of the rest is offset by the reduced need for transmission and distribution investment and fuel savings arising from increased energy efficiency. The net additional cost for the Map scenario out to 2050 is only US$100 billion, about 0.5% of total projected sector investments.

Because the net cost estimates arise from balancing supply and demand, there is considerable uncertainty. The World Bank figure for incremental low-carbon power generation costs, for example, is much higher, at close to 10% of projected total investment costs, but does not fully offset these against end-use savings, or co-benefits. It is clear that low-carbon paths consistent with the IEA Map result of returning global CO₂ emissions to present levels involve a large redirection of investment, but the net additional cost based on this limited set of studies is likely to be less than 5–10% of the total investment requirements, and may be negligible. The studies collectively emphasize that the choice of path over the next few decades will have profound implications for the structure of capital stock, and its carbon intensity, well into the second half of this century and even beyond.

Much of this investment will come from the private sector. However, the associated literature emphasizes that current signals are inadequate and that the effectiveness of carbon pricing depends critically upon its credibility and predictability. For example, the perceived uncertainty with respect to the EU ETS after 2012 deters companies from investing on the basis of price. The credit agency Standard Poor's, deters companies from investing on the basis of price. The credit agency Standard Poor's,) state that ‘this uncertainty has and will result in delays to investment decisions’. Sullivan and Blyth 2006 [NPR] ) analyse the economics of investment in conditions of uncertainty and concur that the perceived uncertainties make it optimal for companies to defer investment and to keep old power plants running instead. This could even increase emissions. Consequently, the ‘electricity or carbon prices required to stimulate investment in low-carbon technology may be higher than expected...’ due to the uncertainties. This underlines the present gap between the modelling abstraction of perfect foresight, and the real-world uncertainties. The costs of mitigation will be reduced only to the extent that governments can make clear and credible commitments about future carbon controls that are sufficient for the private sector to see as ‘bankable’ in project investment appraisals.

11.6.7 Some generic features of long-term national studies

Finally, the rapidly growing number of national goals and strategies oriented towards securing ambitious CO₂ reduction goals, typically by 60–80% below present levels in industrialized countries, are relevant to the understanding of low-carbon transitions for the first half of this century. Some quantitative findings from some long-term national modelling studies have been summarized in Chapter 3 , and some shorter-term studies earlier in this chapter. ^[22] Additional studies of long-term mitigation in developing countries are beginning to emerge (e.g. Jiang and Hu, 2006 [ARC] ; Shukla et al., 2006 [MoS, ARC] ). The range and number of national analyses, scenarios and strategies devoted to mitigation targets is beyond the scope of this section but, in general, they suggest that there are a number of common ‘high-level’ features that underpin some main messages of the academic literature in terms of the need for a combination of:

• innovation-related action on all fronts, both R&D and market-based learning-by-doing stimulated by a variety of instruments;
• measures that establish a long-term, stable and predictable price for carbon to encourage lower carbon investment, particularly but not exclusively in power sector investments;
• measures that span the range of non-CO₂ gases so as to capture the ‘low-hanging fruit’ across the economy;
• measures relating to long-lived capital stock, especially buildings and energy infrastructure;
• institution- and option-building including considerations relating both to system structures, and policy experimentation with review processes to learn which are the most effective and efficient policies in delivering such radical long-term changes as knowledge about climate impacts accumulates.

11.7.1 The nature and importance of spillover

Spillover effects of mitigation in a cross-sectoral perspective are the effects that mitigation policies and measures in one country or group of countries have on sectors in other countries. (Inter-generational consequences, which are the effects of actions taken by the present generation on future generations, are covered in Chapter 2 .) Spillover effects are an important element in the evaluation of environmental policies in economies globally linked through trade, foreign direct investment, technology transfer and information. Due to spillover effects, it is difficult to determine precisely the net mitigation potential for sectors and regions, and the effects of policies. An added complication is that the effects may be displaced over time. The measurement of the effects is also complex because effects are often indirect and secondary, although they can also accumulate to make local or regional mitigation action either ineffective or the source of global transformation. Much of the literature recognizes the existence of spillover effects. However, uncertainty and disagreement about time scale, cost, technology development, modelling approaches, policy and investment pathways lead to uncertainty about their extent and therefore the overall mitigation potentials.

The same spillover effect will be seen differently depending on the point of view adopted. Multiple differences between regions and nations imply differing, and perhaps contradictory views, about mitigation policies and their implementation. These differences emanate from the diverse and sometimes distinct natural endowments and social structures of those regions, as well as differences in the financial ability to cope with the costs that may be incurred as a result of the implementation of these policies. Methodologies that are developed for market-based industrialized economies may not be completely relevant for the economies of developing countries.

Some researchers who use general-equilibrium models (e.g. Babiker, 2005 [ARC] ) conclude that spillover will, given certain assumptions, render mitigation action ineffective or worse if it is confined to Annex I countries. Other researchers (e.g. Grubb et al., 2002a [JoC, SRC] ; Sijm et al., 2004 [NPR, MoS] ) argue that spillovers from Annex I action, implemented via induced technological change, could have substantial effects on sustainable development, with emission intensities from developing countries being a fraction of what they would be otherwise. ‘However, no global models yet exist that could credibly quantify directly the process of global diffusion of induced technological change.’ ( Grubb et al., 2002b [MoS, SRC] , p.303). It is important to empha- size the uncertainties in estimating spillover effects, as well as uncertainties in estimating potential mitigation costs and benefits. In the modelling of spillovers through international trade, researchers rely on different approaches (bottom-up or top-down, for example), different assumptions (perfect/imperfect or ‘Armington’ substitution) and estimates of parameters when signs and magnitudes are disputed. Many of the models used to estimate the costs of mitigation focus on substitution effects and set aside information, policy and political spillovers, as well as the induced development and diffusion of technologies.

11.7.2 Carbon leakage

Carbon leakage is defined as the increase in CO₂ emissions outside the countries taking domestic mitigation action divided by the reduction in the emissions of these countries. It has been demonstrated that an increase in local fossil fuel prices resulting, for example, from mitigation policies may lead to the re-allocation of production to regions with less stringent mitigation rules (or with no rules at all), leading to higher emissions in those regions and therefore to carbon leakage. Furthermore, a decrease in global fossil fuel demand and resulting lower fossil fuel prices may lead to increased fossil fuel consumption in non-mitigating countries and therefore to carbon leakage as well. However, the investment climate in many developing countries may be such that they are not ready yet to take advantage of such leakage. Different emission constraints in different regions may also affect the technology choice and emission profiles in regions with fewer or no constraints because of the spillover of learning (this is discussed in Section 11.7.6 ).

Since the TAR, the literature has extended earlier-equilibrium analysis to include effects of trade liberalization and increasing returns in energy-intensive industries. A new empirical literature has also developed. The literature on carbon leakage since the TAR has introduced a new dimension to the analysis of the subject: the potential carbon leakage from projects intended for developing countries to help them reduce carbon emissions. One example is ( Gundimeda 2004 ) ) in the case of India (discussed in Section 11.7.3 below).

11.7.3 Spillover impact on sustainable development via the Kyoto mechanisms and compensation

The Kyoto mechanisms may also result in spillover effects that offset their additionality. ( Gundimeda 2004 ) ) considers how the clean development mechanism (CDM) might work in India. (The CDM is considered in detail in Chapter 13 .) The paper examines the effects of CDM projects involving land-use change and forestry on the livelihoods of the rural poor. It concludes that, for CDM to be sustainable and to result in sustainable development for local people, three important criteria must be met: (1) in sequestration projects, local use of forestry (as firewood, for example) should also be an integral part of the project (2) management of the common lands by the rural poor through proper design of the rules for the sustenance of user groups; and (3) ensuring that the maximum revenue from carbon sequestration is channelled to the rural poor. ‘Otherwise CDM would just result in either [carbon] leakage [e.g. through unplanned use of forestry for firewood] ... or have negative welfare implications for the poor’ (p. 329).

Kemfert 2002 [MoS] ) considers the spillover and competitiveness effects of the Kyoto mechanisms used separately (CDM, CDM with sinks, joint implementation (JI) and emissions trading (ET)) using a general-equilibrium model – WIAGEM – with Kyoto-style action (including the USA) continuing until 2050 . The study shows the full welfare effect (% difference from business as usual) in 2050, broken down into the effects of domestic action, competitiveness and spillovers. It is notable that the mechanisms have a very small impact on welfare. At most, as an outlier, there is a 0.7% increase for countries in transition (REC) for emissions trading and a 0.1% decrease for the EU-15 for joint implementation. The CDM is found to improve welfare most in developing countries. However, the model does not include induced technological change or environmental co-benefits and it assumes full employment in all countries. If the CDM is assumed to result in more technological development, a more productive use of labour or an improvement in air or water quality, then the environmental and welfare effects in non-Annex I countries will be much larger than those reported.

( Böhringer and Rutherford 2004 ) ) use a CGE model to assess the implications of UNFCCC articles 4.8 and 4.9 dealing with compensation. They conclude that ‘spillover effects are an important consequence of multilateral carbon abatement policies. Emission mitigation by individual developed regions may not only significantly affect development and performance in non-abating developing countries, but may also cause large changes in the economic costs of emission abatement for other industrialized nations.’ They estimate that the US should pay OPEC and Mexico estimated compensation of 0.7 billion US$ annually to offset the adverse impacts on these regions and that the EU should pay the same amount to the US to account for the positive spillover.

11.7.4 Impact of mitigation action on competitiveness (trade, investment, labour, sector structure)

The international competitiveness of economies and sectors is affected by mitigation action (see surveys by ( Boltho 1996 ) Adams 1997 [NPR] and Barker and Köhler 1998 [NPR, SRC] )). In the long run, exchange rates change to compensate for a persistent loss of national competitiveness, but this is a general effect and particular sectors can become more or less competitive. In the short run, the higher costs of fossil fuels lead to a loss in sectoral price competitiveness, especially in energy-intensive industries. The effects of domestic mitigation action on a region’s international competitiveness are broken down by the literature into the effects on price and non-price competitiveness. This section covers price competitiveness, while technological spillover effects are discussed in Section 11.7.6 below.

In general, energy efficiency policies intended for GHG mitigation will tend to improve competitiveness (see Section 11.6.3 above). ( Zhang and Baranzini 2004 ) ) have reviewed empirical studies on the effects of Annex 1 action on international competitiveness. They conclude that ‘empirical studies on existing carbon/energy taxes seem to indicate that competitive losses are not significant’. They therefore support the conclusions of the TAR, namely that ‘reported effects on international competitiveness are very small and that at the firm and sector level, given well-designed policies, there will not be a significant loss of competitiveness from tax-based policies to achieve targets similar to those of the Kyoto Protocol.’ (p.589). Baron and ECONEnergy ( 1997 ) looked at carbon prices similar to those expected to be necessary to implement the Kyoto Protocol (see 11.4.3.3). They report a static analysis of the cost increases from a tax of 27 US$/tCO₂ on four energy-intensive sectors in 9 OECD economies (iron and steel, other metals, paper and pulp, and chemicals). Average cost increases are very low – less than about 3% for most country sectors studied – with higher cost increases in Canada (all 4 sectors), Australia (both metal sectors) and Belgium (iron and steel).

However, action by Annex I governments (the EU, Denmark, Norway, Sweden, UK) have generally exempted or provided special treatment for energy-intensive industries. Babiker et al. 2003 [ARC] ) suggest that this is a potentially expensive way of maintaining competitiveness, and recommend a tax and subsidy scheme instead. One reason for such exemptions being expensive is that, for a given target, non-exempt sectors require a higher tax rate, with mitigation at higher cost.

The impact of mitigation policies on trade within a region and between regions as a result of spillover is linked through capital flows from one country to another (within a region) or from one region to another, as individual investors and firms look for a higher rate of return on their investments which are considered by the receiving countries to be Foreign Direct Investment (FDI). Different market regulations and the flow of goods and services are influenced by mitigation policies, and the resulting spillover make ‘measuring the welfare cost of climate change policies a real challenge, raising difficult issues of micro- and macro-economics: cost-benefit analysis on the one hand, foreign trade and international specialization on the second hand’ ( Bernard and Vielle, 2003 [MoS] ). Partly for these reasons, the literature is sparse and the effects of different mitigation policies on FDI, trade, investment and labour market development within and between regions and any spillover effects are important areas for further research.

11.7.5 Effect of mitigation on energy prices

As discussed in 11.7.2, perhaps one of the most important ways in which spillovers from mitigation action in one region affect the others is through their effect on world energy prices. When a region reduces its fossil fuel demand as a result of mitigation policy, it will reduce the world demand for that commodity and so put downward pressure on the prices. Depending on responses from producers of fossil fuels, oil, gas or coal prices may fall, leading to losses of revenue for the producers, and lower import costs for the consumers. Demand for alternative, low-carbon fuels may increase. Three distinct spillover effects have been identified for non-mitigating countries. First, income for producers of fossil fuels will decline as the quantity sold is reduced, causing welfare losses and unemployment along with associated problems. Second, consuming nations will face lower prices for imported energy and may reduce subsidies or allow domestic energy prices to fall so that they tend to consume more, leading to carbon leakage as discussed above. Third, those non-mitigating countries producing low-carbon or alternative fuels will see an increase in demand and prices, with potentially positive effects on the markets for bioenergy.

11.7.6 Technological spillover

Mitigation action may lead to more advances in mitigation technologies. Transfer of these technologies, typically from industrialized nations to developing countries, is another avenue for spillover effects. However, as discussed in Chapter 2 , effective transfer implies that developing countries have an active role in both the development and the adaptation of the technologies. The transfer also implies changes in flows of capital, production and trade between regions.

Sijm et al. 2004 [NPR, MoS] ) assess the spillover effects of technological change. They divide the literature into two groups, depending on their ‘top-down’ or ‘bottom-up’ approach to modelling. (See the discussion on the topic in Section 11.3 above.) Most top-down modelling studies omit the effect or show it playing a minor role. The authors argue that the potential beneficial effect of technology transfer to developing countries arising from technological development brought about by Annex I action may be substantial for energy-intensive industries, but has so far not been quantified in a reliable manner. ‘Even in a world of pricing CO₂ emissions, there is a good chance that net spillover effects are positive given the unexploited no-regret potentials and the technology and know-how transfer by foreign trade and educational impulses from Annex I countries to Non-Annex I countries.’ (p. 179).

However, results from bottom-up and top-down models are strongly influenced by assumptions and data transformations and that lead to high levels of uncertainty. ‘Innovation and technical progress are only portrayed superficially in the predominant environmental economic top-down models, and that the assumption of perfect factor substitution does not correctly mirror actual production conditions in many energy-intensive production sectors. Bottom-up models, on the other hand, neglect macroeconomic interdependencies between the modelled sector and the general economy.’ ( (Lutz et al., 2005 ) ). The effects of spillovers combined with learning-by-doing are explored specifically using bottom-up models by Barreto and Kypreos 2002 [MoS] ) using MARKAL, and by Barreto and Klaassen 2004 [MoS] ) using ERIS. They find that, owing to the presence of spillovers, the imposition of emission constraints in the Annex I region may induce technological change and, hence, emission reductions in the non-Annex I region even when the latter region does not face emission constraints itself.

The existence of spillover effects also changes the theoretical conclusions in the economics literature. In the pure competition-equilibrium model, the most efficient policy is an equal rate of carbon tax for every sector and region. ( Rosendahl 2004 ) ) shows that, for maximum efficiency with spillovers and learning-by-doing, the carbon tax should be higher in those sectors and regions with the highest potential for technological progress. This is a general argument for stronger mitigation in those sectors and countries where technological progress is most likely to be accelerated by higher taxes on carbon use. In a game-theory context, with the shared benefits of R&D improving energy efficiencies, ( Kemfert 2004, ) p. 463) finds that ‘full cooperation on climate control and technological improvements benefits all nations in comparison to a unilateral strategy.’

Although the technologies for CO₂ reduction in the electricity sectors are accessible, their dissemination still faces some challenges, especially in economies with low purchasing power and educational levels ( (Kumar et al., 2003 ) ). An additional issue is that technology sharing by the fossil-fuel energy suppliers has been severely limited to date, probably due to the industrial organization of coal, oil and gas production, which is dominated by a few large private and state companies. Unlike, for example, new IT technologies, which quickly become industry standards, newly developed energy-related technology providing a competitive advantage generally becomes available to competitors slowly. However, modelling of the spillovers and the evolution of technologies, as well as structural changes in corporate management, require a better understanding of knowledge production and the knowledge transfer process within and between industries, and of the role and efficiency of transfer institutions such as universities, technology transfer centres and consultancy companies ( Haag and Liedl, 2001 [NPR] ).

11.8.1 Interaction between GHG mitigation and air pollution control

Many of the traditional air pollutants and GHGs have common sources. Their emissions interact in the atmosphere and, separately or jointly, they cause a variety of environmental effects at the local, regional and global scales. Since the TAR, a wealth of new literature has pointed out that capturing synergies and avoiding trade-offs when addressing the two problems simultaneously through a single set of technologies or policy measures offers potentially large cost reductions and additional benefits.

However, there are important differences at the temporal and spatial scales between air pollution control and climate change effects. Benefits from reduced air pollution are more certain; they occur earlier, and closer to the places where measures are taken, while climate impact is long-term and global. These mismatches of scales are mirrored by a separation of the current scientific and policy frameworks that address these problems ( Swart et al., 2004 [JoC, SRC] ; Rypdal et al., 2005 [ARC] ).

Since the TAR, numerous studies have identified a variety of co-benefits of greenhouse gas mitigation on air pollution for industrialized and developing countries. In many cases, when measured using standard economic techniques, the health and environmental benefits add up to substantial fractions of the direct mitigation costs. More recent studies have found that decarbonization strategies generate significant direct cost savings because of reduced air pollution costs, highlighting the urgency of an integrated approach for greenhouse gas mitigation and air pollution control strategies.

11.8.2 Impacts of GHG mitigation on employment

A number of studies point out that investments in greenhouse gas mitigation could have a greater impact on employment than investments in conventional technologies. The net impact on employment in Europe in the manufacturing and construction industries of a 1% annual improvement in energy efficiency has been shown to induce a positive effect on total employment ( Jeeninga et al., 1999 [NPR, MoS] ). The effect has been shown to be substantially positive, even after taking into account all direct and indirect macro-economic factors such as the reduced consumption of energy, impact on energy prices, reduced VAT, etc. (European Commission, 2003 ) The strongest effects are seen in the area of semi-skilled labour in the building trades, which also accounts for the strongest regional policy effects. Furthermore, the European Commission ( 2005 ) estimates that a 20% saving on present energy consumption in the European Union by 2020 has the potential to create, directly or indirectly, up to one million new jobs in Europe.

Meyer and Lutz 2002 [NPR, MoS] ) use the COMPASS model to study the carbon taxes for the G7 countries. They find that recycling revenues via social security contributions increases employment by nearly 1% by 2010 in France and Germany, but much less in US and Japan. Bach et al. 2002 [MoS] ), using the models PANTHA RHEI and LEAN, find that the modest ecological tax reform enacted in Germany in 1999 – 2003 increased employment by 0.1 to 0.6% by 2010 . This is as much as 250,000 additional jobs. There is also a 2–2.5% reduction in CO₂ emissions and a negligible effect on GDP. The labour intensity of renewable energy sources has been estimated to be approximately 10 times higher in Poland than that of traditional coal power (0.1–0.9 jobs/GWh compared to 0.01–0.1 jobs/GWh). Given this assumption, government targets for renewable energy would create 30,000 new jobs by 2010 ( Jeeninga et al., 1999 [NPR, MoS] ).

In a study of climate policies for California, Hanemann et al. 2006 [NPR] ) report small increases in employment for a package of measures focusing on the tightening of regulations affecting emissions.

11.8.3 Impacts of GHG mitigation on energy security

Since the TAR, new literature has addressed the question of energy security and climate change, especially following the rapid increases and fluctuations in commodity prices, particularly oil, in the period 2004 - 2006 . The concept of energy security is usually understood to be an issue of the reliability of energy supplies that is illustrated by the exposure of oil im-porters to world market prices ( Bauen, 2006 [MoS] ) and, as Sullivan and Blyth 2006 [NPR] ) point out, the reliability of electricity systems given the growing penetration of intermittant renewables, which may require back-up generation capacity (but see UKERC, 2006 [NPR] ).

The possibilities of synergies and trade-offs between mitigation actions and energy security are very specific to national circumstances, particularly the relevant fuel mixes as a result of evolving energy markets, the sectors being targeted and energy consumption trends ( (Turton and Barreto, 2006 ) ). The transportation sector, in particular, is characterized by strong synergies relating to energy supply: measures replacing oil with domestic biofuels reduce both emissions and reliance on oil imports. Mitigation action for the electricity sector may lead to synergies with energy security. For example, a more decentralized system based on new renewable generation may reduce gas imports. Alternatively, there may be trade-offs. For example, security reasons may lead countries to increase their dependence on internal reserves of coal rather than relying on natural gas imports ( (Kuik, 2003 ) ).

Whether in the form of synergies or trade-offs, there is a growing recognition of the critical linkages that exist between climate change and energy security, and the fact that energy prices still have yet to reflect these ‘externalities’ effectively ( Bauen, 2006 [MoS] ). The inability to manage either one of these threats could result in significant economic and social costs ( (Turton and Barreto, 2006 ) ). Measures that successfully address both issues therefore have the potential to provide signficant social and economic benefits. In conclusion, it seems likely that climate change and energy security pressures will become more acute as international development proceeds. Public policies to address either of these issues can take many forms and their combination makes the effects uncertain, implying a gap in understanding their synergies and trade-offs ( Blyth and Lefevre, 2004 [NPR, ARC] ).

11.8.4 Summary

The recent literature has produced an increasing understanding of the interactions between greenhouse gas mitigation and other policy areas. Numerous studies have identified a wide range of co-benefits and quantified them for industrialized and developing countries. However, the literature does not (as yet) provide a complete picture that includes all the different types of co-benefits needed for a comprehensive assessment. Nevertheless, even the co-benefits quantified at present can make up substantial fractions of, or under specific conditions even exceed, direct mitigation costs.

Beyond the recognition of co-benefits, the realization of potential synergies and avoidance of trade-offs requires an integrated approach that considers a single set of technologies or policy measures in order to simultaneously address all relevant areas. There are practical examples of targeted programmes for pinpointing co-benefits and identifying those policy measures that offer most potential for capturing possible synergies.

In the case of low-income countries, the consideration of potential synergies between GHG mitigation and other policy objectives could be even more important than in high-income countries. At present, climate change policies are often still relatively marginal issues in these countries compared to issues such as poverty eradication, food supply, the provision of energy services, employment, transportation and local environmental quality. Accelerated and sustainable development could therefore become a common interest for both local and global communities ( Criqui et al., 2003 [Ambiguous] ).

11.9.1 Sectoral mitigation action: links to climate change and adaptation