Editing IPCC:AR6/WGIII/Chapter-14 (section)

=== Cross-Chapter Box 10 | Policy Attribution – Methodologies for Estimating the Macro-level Impact of Mitigation Policies on Indices of Greenhouse Gas Mitigation ===

<div id="h2-10-siblings" class="h2-siblings"></div>

'''Authors:''' Mustafa Babiker (Sudan/Saudi Arabia), Paolo Bertoldi (Italy), Christopher Bataille (Canada), Felix Creutzig (Germany), Navroz K. Dubash (India), Michael Grubb (United Kingdom), Erik Haites (Canada), Ben Hinder (United Kingdom), Janna Hoppe (Switzerland), Yong-Gun Kim (Republic of Korea), Gregory F. Nemet (the United States of America/Canada), Anthony Patt (Switzerland), Yamina Saheb (France), Raphael Slade (United Kingdom)

This report notes both a growing prevalence of mitigation policies over the past quarter century (Chapter 13), and ‘signs of progress’ including various quantified indices of GHG mitigation (Table 2.4). Even though policies implemented and planned to date are clearly insufficient for meeting the Paris long-term temperature goals, a natural question is to what extent the observed macro-level changes (global, national, sectoral, technological) can be attributed to policy developments. This Assessment Report is the first to address that question. This box describes the methods for conducting such ‘attribution analysis’ as well as its key results, focusing on the extent to which polices have affected three main types of ‘outcome indices’:

'''•''' '''GHG emissions:''' emissions volumes and trends at various levels of governance including sub- and supra-national levels, and within and across sectors.

'''•''' '''Proximate emission drivers:''' trends in the factors that drive emissions, distinguished through decomposition analyses, notably: energy/GDP intensity and carbon/energy intensity (for energy-related emissions); indices of land use such as deforestation rates (for LULUCF/AFOLU); and more sector-specific component drivers such as the floor area per capita, or passenger kilometres per capita.

'''•''' '''Technologies:''' developments in key low-carbon technologies that are likely to have a strong influence on future emissions trends, notably levels of new investment and capacity expansions, as well as technology costs, with a focus on those highlighted in Figure 2.30.

''Policy attribution'' examines the extent to which emission-relevant outcomes on these indices – charted for countries, sectors and technologies, particularly in [[IPCC:Wg3:Chapter:Chapter-2|Chapter 2]] and the sectoral chapters – may be reasonably attributed to policies implemented prior to the observed changes. Such policies include regulatory instruments such as energy efficiency programmes or technical standards and codes, carbon pricing, financial support for low-carbon energy technologies and efficiency, voluntary agreements, and regulation of land-use practices. The sectoral chapters give more detail along with some accounts of policy, while trends in mitigation policy adoption are summarised in Chapter 13.

In reviewing hundreds of scientific studies cited in this report, the impacts of adopted policies on observed outcomes were assessed. The vast majority of these studies examine particular instruments in particular contexts, as covered in the sectoral chapters and Chapter 13; only a few have appraised global impacts of policies, directly or plausibly inferred (the most significant are cited in Figure 1 in this Cross-Chapter Box). Typically, studies consider ‘mitigation policies’ to be those adopted with either a primary objective of reducing GHG emissions or emissions reductions as one among multiple objectives.

<div id="ccbox-10-1" class="Boxes_Blue-Boxes_•-Box-body"></div>

[[File:fa94f77b8fd08d8a1408bd3d8665401a IPCC_AR6_WGIII_CCBox_10_Figure_1.png]]

'''Cross-Chapter Box 10, Figure 1 | Policy impacts on key outcome indices.''' '''The figure shows the impacts of policies on three indices: proximate emission drivers, technologies and GHG emissions, including several lines of evidence on GHG abatement attributable to policies.'''

Policies differ in design, scope, and stringency, may change over time as they require amendments or new laws, and often partially overlap with other instruments. Overall, the literature indicates that policy mixes are, theoretically and empirically, more effective in reducing emissions, stimulating innovation, and inducing behavioural change than stand-alone policy instruments (Sections 5.6 and 13.7) ( [[#Rosenow--2017|Rosenow et al. 2017]] ; [[#Best--2018|Best and Burke 2018]] ; [[#Sethi--2020|Sethi et al. 2020]] ). Nevertheless, these factors complicate analysis, because they give rise to the potential for double counting emissions reductions that have been observed, and which separate studies can attribute to different policy instruments.

Efforts to attribute observed outcomes to a policy or policy mix is also greatlycomplicated by the influence of many exogenous factors, including fossil fuel prices and socio-economic conditions. Likewise, technological progress can result from both exogenous causes, such as ‘spillover’ from other sectors, and policy pressure. Further, other policies, such as fossil fuel subsidies as well as trade-related policies, can partially counteract the effect of mitigation policies by increasing the demand for energy or carbon-intensive goods and services. In some cases, policies aimed at development, energy security, or air quality have climate co-benefits, while others increase emissions.

Studies have applied a number of methods to identify the actual effects of mitigation policies in the presence of such confounding factors. These include statistical attribution methodologies, including experimental and quasi-experimental design, instrumental variable approaches, and simple correlational methods. Typically, the relevant mitigation metric is the outcome variable, while measures of policies and other factors act as explanatory variables. Other methodologies include aggregations and extrapolations

Cross-Chapter Box 10

from micro-level data evaluation, and inference from combining multiple lines of analysis, including expert opinion. Additionally, the literature contains reviews, many of them systematic in nature, that assess and aggregate multiple empirical studies.

With these considerations in mind, multiple lines of evidence, based upon the literature, support a set of high-level findings, as illustrated in Figure 1 in this Cross-Chapter Box, as follows.

'''1. GHG Emissions.''' There is robust evidence with a high level of agreement that mitigation policies have had a discernible impact on emissions. Several lines of evidence indicate that mitigation policies have led to avoided global emissions to date of several billion tonnes CO 2 -eq annually. The figure in this box shows a selection of results giving rise to this estimate.

As a starting point, one methodologically sophisticated econometric study links global mitigation policies (defined as climate laws and executive orders) to emission outcomes; it estimates emission savings of 5.9 GtCO 2 yr –1 in 2016 compared to a no-policy world ( [[#Eskander--2020|Eskander and Fankhauser 2020]] ) ( [[IPCC:Wg3:Chapter:Chapter-13#13.6.2|Section 13.6.2]] ).

A second line of evidence derives from analyses of the Kyoto Protocol. Countries which took on Kyoto Protocol targets accounted for about 24% of global emissions during the first commitment period (2008–12). The most recent robust econometric assessment ( [[#Maamoun--2019|Maamoun 2019]] ) estimates that these countries cut GHG emissions by about 7% on average over 2005–2012, rising over the period to around 12% (1.3 GtCO 2 -eq yr –1 ) ''relative to a no-Kyoto scenario.'' This is consistent with estimates of Grunewald and Martinez (2016) of about 800 MtCO 2 -eq yr –1 averaged to 2009. Developing countries’ emissions reduction projects through the CDM (defined in Article 12 of the Kyoto Protocol) were certified as growing to over 240 MtCO 2 -eq yr –1 by 2012 (UNFCC 2021c). With debates about the full

Cross-Chapter Box 10

extent of ‘additionality’, academic assessments of savings from the CDM have been slightly lower, with particular concerns around some non-energy projects ( [[#14.3.3.1|Section 14.3.3.1]] ).

A third line of evidence derives from studies that identify policy-related, absolute reductions from historical levels in particular countries and sectors through decomposition analyses (Le Quéré et al. 2019; [[#Lamb--2021|Lamb et al. 2021]] ), or evaluate the impact of particular policies, such as carbon pricing systems. From a wide range of estimates in the literature (Sections 2.8.2.2 and 13.6), many evaluations of the EU ETS suggest that it has reduced emissions by around 3% to 9% relative to unregulated firms and/or sectors ( [[#Schäfer--2019|Schäfer 2019]] ; Colmer et al. 2020), while other factors, both policy (energy efficiency and renewable support) and exogenous trends, played a larger role in the overall reductions seen ( [[#Haites--2018|Haites 2018]] ).

These findings derived from the peer-reviewed literature are also consistent with two additional sets of analysis. The first set concerns trends in emissions, drawing directly from Chapters 2, 6 and 11, showing that global annual emission growth has slowed, as evidenced by annual emission increments of 0.55 GtCO 2 -eq yr –1 between 2011 and 2019 compared to 1.014 GtCO 2 -eq yr –1 in 2000 and 2008. This suggests avoided emissions of 4–5 GtCO 2 -eq yr –1 (see also Figure 1.1d). The second set concerns emissions reductions projected by Annex I governments for 2020 in their fourth biennial reports to the UNFCCC. It is important to note that these are mostly projected annual savings from implemented policies (not ''ex-post'' evaluations), and there are considerable differences in countries’ estimation methodologies. Nevertheless, combining estimates from 38% of the total of 2,811 reported policies and measures yields an overall estimate of 3.81 GtCO 2 -eq yr –1 emission savings ( [[#UNFCCC--2020d|UNFCCC 2020d]] ).

'''2. Proximate''' '''emission drivers.''' With less overt focus on emissions, studies of trends in energy efficiency, carbon intensity, or deforestation often point to associated policies. The literature includes an increasing number of studies on demonstrable progress in developing countries. For example, South and South-East Asia have seen energy intensity in buildings improving at about 5–6% yr –1 since 2010 (Figure 2.22). In India alone, innovative programmes in efficient air conditioning, LED lighting, and industrial efficiency are reported as saving around 25 Mtoe in 2019–2020, thus leading to avoided emissions of over 150 MtCO 2 yr –1 ( [[#Malhotra--2021|Malhotra et al. 2021]] ) (Box 16.3). Likewise, reductions in deforestation rates in several South and Central American and Asian countries are at least partly attributable to ecosystem payments, land-use regulation, and internal efforts ( [[IPCC:Wg3:Chapter:Chapter-7#7.6.2|Section 7.6.2]] ). Finally, the policy-driven displacement of fossil fuel combustion by renewables in energy has led to reductions in carbon intensity in several world regions (Chapters 2 and 6).

'''3. Technologies.''' The literature indicates unambiguously that the rapid expansion of low-carbon energy technologies is substantially attributable to policy (Sections 6.7.5 and 16.5). Technology-specific adoption incentives have led to a greater use of less carbon-intensive (e.g., renewable electricity) and less energy-intensive (especially in transport and buildings) technologies. As Chapters 2 and 6 of this report note that modern renewable energy sources currently satisfy over 9% of global electricity demand, and this is largely attributable to policy. There are no global-level studies estimating the avoided emissions due to renewable energy support policies, but there are methods that have been developed to link renewable energy penetration to avoided emissions, such as that of [[#IRENA--2021|IRENA (2021)]] . Using that method, and assuming that 70% of modern renewable energy expansion has been policy induced, yields an estimate of avoided emissions of 1.3 GtCO 2 -eq yr –1 in 2019. Furthermore, observed cost reductions are the result of policy-driven capacity expansion as well as publicly funded resarch and development, in individual countries and globally. These correspond with induced effects on number of patents, ‘learning curve’ correlations with deployed capacity, and cost component and related case study analyses ( [[#Kavlak--2018|Kavlak et al. 2018]] ; [[#Nemet--2019|Nemet 2019]] ; [[#Popp--2019|Popp 2019]] ; [[#Grubb--2021|Grubb et al. 2021]] ).

<div id="14.4" class="h1-container"></div>

<span id="supplementary-means-and-mechanisms-of-implementation"></span>