The Y factor for Climate Change abatement – A method to rank options beyond abatement costs

The tools available to translate climate targets into abatement actions are mainly based on costs and technical feasibility. Options for greenhouse gas abatement span all sectors, all countries, and involve a huge variety of technologies. The reasons for abatement to be realized, or not, are diverse and complex. In particular, the po- litical discussion why many affordable options do not materialize is naïve and ad hoc. Here we show the Y factor , an approach for a quick scan of abatement options against a set of prominent abatement barriers. We define 12 factors which capture a broad set of barriers related to 1) costs and financing, 2) multi-actor complexity, 3) physical interdependencies and 4) behaviour. We rank 24 abatement options using an explicit, but coarse scoring for these barriers. We show that all abatement options have implementation barriers, many of which may well drastically impact their implementation beyond what would be expected from their abatement costs. The analysis implies that the relation between abatement costs and the barriers withholding implementation is not straightforward and calls for a deliberate policy debate on prioritization and policy intervention. The Y factor structures such a policy debate.


Introduction
Clear policy targets are key to drive societal change. The global challenge to reduce greenhouse gases to the respective targets for meeting the Paris 1.5 • goal is unprecedented. Options for greenhouse gas abatement span all sectors, all countries, and involve a huge variety of technologies (Pachauri et al., 2015). Despite the fact that the targets may be achievable with known technologies (Capros et al., 2014), immense changes are required in large-scale interconnected infrastructure systems that form the backbone for the economy. It is a political question to establish emission reduction targets for countries, sectors and particular technologies. Energy and climate policies are key to bring emission reduction targets within reach. A key input to the policy process are marginal abatement cost (MAC) curves which capture marginal costs for abatement options (in Euro/ton CO 2 (equivalent) saved) together with abatement potential (Jackson, 1991); they feed into low-costs technically feasible scenarios (Blok et al., 1993).
The strength of a MAC curve is its elegance: it is an overview of the options to reduce emissions sorted on abatement cost; MAC curves have been widely adopted (Barthel et al., 2006;Eide et al., 2011;Eory et al., 2018;ESMAP, 2012;Nauclér and Enkvist, 2009). It is easy to misinterpret MAC curves and use them for prioritizing actions (Ward, 2014). Low cost abatement options appear desirable but there may be sufficient reason to assume they are not the easiest to implement. MAC curves have been criticized for a lack of transparency, the fact that important uncertainties (such as the discount rates (Kesicki and Strachan, 2011)) are often hidden, for the lack of dynamics, for the inability to show how options and sectors interact, and for ignoring possible ancillary benefits (Kesicki and Ekins, 2012). Furthermore, ranking options at negative abatement costs is conceptually problematic (Taylor, 2012). Despite the fact that scenario studies using integrated assessment models (IAMs) incorporate many complexities, a focus on (lowest) cost abatement paths prevails and the concept of feasibility receives little attention (Anderson and Jewell, 2019). The reasons for abatement to be realized or not are diverse (Intergovernmental Panel on Climate Change, 2014) and depend on complex socio-economic dynamics (Edwards, 2011). This requires complementary tools alongside MAC curves and IAMs.
We propose the Y factor which shows why it may be difficult for abatement options to materialize by means of a quick scan of options against a variety of implementation barriers. The Y factor scores abatement options on 12 barriers in four groups: 1) costs and financing, 2) multi-actor complexity, 3) physical interdependencies, and 4) behaviour. The quick scan enables a coarse, but fully transparent ranking of abatement options and provides policy makers with a broad set of indicators to deepen the policy debate. In section 2 we present the Y factor, including our approach for its development, application and validation. In section 3 we present a global Y factor abatement curve and discuss the findings in terms of the curve's robustness and usability. In section 4 we conclude the article and derive policy implications.

Methodology: the Y factor
The Y factor was first developed, tested and theoretically grounded. Afterwards, it was applied to a prominent set of abatement options. Finally, the Y factor was used with policy makers.

The Y factor
The Y factor is determined by scoring an abatement option on each factor (i.e. assigning 0, 1 or 2); each abatement option can, therefore, have a total score between 0 and 24. Table 1 gives an overview of the factors, and how the factors and scores are defined. Though the different barriers are not necessarily comparable, the application of the Y factor captures why (Y) an abatement option may be difficult to materialize (in addition to abatement cost levels).
Barriers for abatement are represented in Y factors generic enough to be applicable to all sectors, while being explicit and operational enough that they can be scored on a 0-2 scale. While most of the sector-specific barriers as described by the IPCC (Intergovernmental Panel on Climate Change, 2014) are captured, Y factors are typically formulated in more practical terms (Arensman, 2018). This is essential for the Y factor to be effective in comparing abatement options across sectors and for a broad set of barriers. The Y factors are rooted in a socio-technical systems perspective (Bijker et al., 1987): realising abatement options implies changes in large-scale socio-technical systems, where technologies are used differently and/or are replaced with others, thereby avoiding greenhouse gas emissions. This perspective translates into these four categories: 1) Abatement options that replace (or introduce) technologiesbeing more efficient or using different energy sources/carriersrequire investments (Dixit and Pindyck, 1994). This leads to barriers on costs and financing, for which the significance of an investment, the payback time and issues related to financing are well known criteria. The factors are framed to the investor so that it is not the absolute sum which is relevant, but whether it is significant for who is paying. 2) Decisions in socio-technical systems are made by a wide variety of individual actors, which are not all able to decide unilaterally; they operate in stakeholder networks (Bruijn and Herder, 2009). Multi-actor complexity is caused by key decisions depending on others, by the influence of others with opposing values and interests (Weisbuch et al., 2008) (including public acceptance issues), and by the institutions determining the roles and responsibilities which, in turn, may enable or hamper actors' abatement decisions (Williamson, 1998). 3) Abatement implies intervening in existing technological systems (Hughes, 1987). This feeds the category of physical interdepencies (Herder et al., 2008), which captures well-known barriers related to whether changes are required in embedded physical systems, such as infrastructures, whether current technical operations need to be interrupted, and whether the technology has been fully proven. 4) Abatement relates to actors' behaviour whose rationality is bounded, affecting which abatement options are considered and when (Foxon, 2006). This implies, behaviour is limited to the knowledge about abatement options, and to when opportunities arise that put abatement decisions on the table. Additionally, abating can rely on changes in behavioural patterns (Hesselink and Chappin, 2019).

Table 1
List of Y factors categories, Y factors, values and definitions. Each abatement measure scores either 0, 1, or 2 on each of the factor. Scoring 0 suggests that there is no barrier for the factor in question; scoring 1 suggests a possible barrier; scoring 2 suggests a significant barrier.

Development of the Y factor definitions
A preliminary version of the Y factor, illustrated with a preliminary scoring was presented in (Chappin, 2016). Factor definitions were systematically reviewed and theoretically grounded and revised. This resulted in a Y factor that allows for a coarse scoring of 0 (no barrier), 1 (possible barrier), or 2 (significant barrier) for an abatement option on the basis of 12 factors. Final definitions (as revised for this manuscript) are in Table 1 in the main text.
The sector-specific barriers mentioned in the IPCC report from five sectors (Intergovernmental Panel on Climate Change, 2014) and theory on transition and system innovation (e.g. (Geels, 2002;Hekkert et al., 2007;Loorbach and Kemp, 2005); see also sections 2.3 and 2.4 of (Arensman, 2018)) underpin the four Y factor categories: a broad diversity of sector-specific and conceptual barriers in the literature, relate to these categories, such as range anxiety (that typically applies to electric vehicles) linking to the behaviour category.
Detailed case studies for four distinct abatement options by means of three to four semi-structured interviews per case with Dutch experts from business, government and academia demonstrated the relevance of all individual Y factors (section 3.5 and chapter 4 in (Arensman, 2018)); this led to further refinement of factor definitions (section 5.3 of (Arensman, 2018)).

Application of the Y factor
We developed scores for each abatement option on the basis of a broad set of recent literature and validated these scores with sector experts. A complete overview of sources for the initial scoring are in chapter 2 of (Soana, 2018) (publicly accessible). The scores were validated with sector experts from research institutes and universities in various countries (names and affiliations mentioned on page 46 of ( Soana, 2018)) by means of semi-structured interviews. Two to three sector experts from in total five countries (Germany, Italy, the Netherlands, the UK, and Mexico, further details in chapter 4 of (Soana, 2018)) validated 80% of the scores, the remander was validated by one expert. 8% of the scores are significantly uncertain. We concluded that the scoring is robust against the selection of experts.

Use of the Y factor results
The Y curve is presented through an interactive website (http://eml ab.tudelft.nl/yfactor/) that invites discussion amongst climate policy makers in terms of the arguments supporting the scoring, the need and desired direction for climate policy and the robustness of the Y factor curve (Swart, 2019). Focus groups with 6 senior and 3 junior climate policymakers and strategists from the Netherlands confirmed the ability of the interactive Y factor website to generate discussion on prioritization of abatement options beyond abatement costs, and the ability to discuss key elements needed in the climate psolicy debate (chapter 5 of (Swart, 2019)).

Results and discussion
We present the Y factor curve in Fig. 1. The curve was developed and  Cars full hybrid 7 Cars plug-in hybrid 8 Clinker substitution by fly ash 9 Coal CCS new built 10 Coal CCS retrofit 11 Composting new waste 12 Cropland nutrient management 13 Electricity from landfill gas 14 Energy efficiency 1 iron & steel 15 Geothermal 16 Grassland management 17 High penetration wind 18 Lighting switch to LED (residential) 19 Nuclear energy 20 PV panels homes 21 Reduced deforestation (agriculture) 22 Reduced deforestation (timber harvesting) 23 Residential appliances 24 Small hydro validated for 24 abatement options (Soana, 2018) from McKinsey's most recent MAC curve (Nauclér and Enkvist, 2009), listed in Table 2. We deliberately selected options from this curve to be able to illustrate similarities between McKinsey's MAC curve and the Y factor curve. We focus on effective measures; our selection contains primarily high abatement potential options from various sectors. Fig. 2 illustrates the correlation between the barriers identified with the Y factor method, and the abatement costs and potentials originating from (Nauclér and Enkvist, 2009). The curve essentially has a global scope, capturing the most appropriate values for the factors world-wide. Where more specific choices were in order, simply because the values depend on local rules, regulations, institutions or culture, we chose the conditions in Europe and the Netherlands to underpin our choices. We interpret the results from Figs. 1 and 2 (section 3.1), discuss the robustness of these results (section 3.2) and elaborate on their usability (section 3.3).

Interpretation of the Y curve
In general, abatement options score relatively high (at an average score somewhat above 10) which confirms thatbesides abatement cost emission reductions are hard to achieve. The options with the lowest number of barriers are on the left in Fig. 1. LED lighting, full hybrid vehicles are options that are penetrating in the market in the last decade and they are cost effective. The options with the largest number of barriers are on the right. Coal CCS new built and retrofit are options that have shown to be very hard to implement (Viebahn and Chappin, 2018) and also have the highest abatement costs in the set of options except for battery electric vehicles.
For the options in the middle, flatter portion of the curve, the relation between abatement costs and barriers withholding implementation is less straightforward. Prime examples of cost-effective abatement options with significant barriers are residential appliances and bioethanol lignocellulosic. 15 out of 24 abatement options have at least one significant barrier (represented in the scoring with a 2), withholding implementation. The other 9 abatement options all have possible barriers (represented in the scoring with a 1). For the latter options, the number of possible barriers range from 2 to 10. For two of the 24 abatement options (Coal CCS retrofit and Bioethanol lignocellulosic) barriers were identified for all 12 factors.

Robustness of the Y curve
We have shown that the Y factor scores are robust. In order to develop unambiguous scoring, a clear definition and appropriate scoping of abatement options are required (Arensman, 2018). Rigorous scoping is necessary because geographic/spatial aspects, economic and social conditions influence some of the Y factor values. The ranking provided by the curve does not equate to political prioritization; it provides insight in the number and significance of the implementation barriers that could be expected. It is robust in the latter context. The ranking's edges, i.e. the low (left) and high (right) side are particularly stable against individual differences in barriers due to regional conditions or cultural factors. For the middle, flatter part of the curve, the ranking is more sensitive to changes in the individual scores. These sensitivities in the Y curve translate to crucial elements in the policy debate (see also section 3.3): i.e. 1) to apply focus in the debate on options with barriers that are addressed by exisiting or easily achievable policies, and/or 2) to consider whether it may be more appropriate to deal with multiple possible barriers (scores of 1) or to aim to irradicate fewer significant ones instead (scores of 2).
We have shown that scores are stable in time, except for when technologies or economic and social conditions evolve significantly (Arensman, 2018) (e.g. cost reductions for offshore wind or with new fixtures and better quality of LED lights). The scores are also robust against many uncertainties known to affect abatement: many do not affect the scoring because the uncertainty fits within the definition of the factors themselves or within the range covered in a particular score (e.g. developments in costs may reduce payback time, but it may stay within one of the ranges in Table 1). The scores may only change due to significant technological improvements, developments in infrastructure, in institutions, or in public acceptance, which would be reasons to revisit the curve.

Usability and limitations of the Y curve
The focus groups confirmed the interpretation of results and analysis of robustness above: climate policymakers desire tools that go beyond merely abatement costs and potentials and the Y factor tool can be used to develop a better understanding of the implementation complexities. The first contribution to the policy process is the initial ranking that is provided by the Y factor scores. It enables a quick analysis in terms of 'if we want to implement this abatement option, we need to deal with these barriers'. The interactive website of the Y factor fuels an adequate climate policy debate (see page 71 of (Swart, 2019)) by readily providing a systematic overview of Y factor definitions, the scores on each of the barriers, and arguments for those scores, together with abatement costs and potentials that are of (Nauclér and Enkvist, 2009). The Y curve may help the discussion by an initial focus on attractive options in terms of the number and significance of barriers (in addition to the two MACC dimensions -abatement potentials and costs), before moving to the detailed context and content of particular abatement options.
The Y factor intends to structure the policy process and provide important arguments in the policy debate. The Y factors do not include all systemic effects of each option, as aiming to include all those effects would obfuscate the overall gist of the method: being fully transparent while enabling a coarse scoring of abatement options against a wide variety of barriers. Furthermore, the different abatement options tap into different value systems and these may differ between cultures and affect political choices. Much of this is beyond the scope of the Y factor. Nevertheless, within the context of the Y factor, scores can be further specified to particular geographic areas or socio-economic conditions.
Subsequently, the robustness of the Y scores (including the resulting ranking of abatement options) can be further explored by varying factors' weights, which is possible with the interactive website. Reasons to do so could be related to cultural or locational aspects, for instance where in some regions (and possibly sectors) financial aspects may be more dominant than in others. Applying factor weights for other purposes than checking robustness is a normative approach to using the Y factor, which is beyond the scope of this paper. Additional (e.g. cultural or economic) theory would be needed in order to validate such choices.

Conclusions and policy implications
The Y factor is a tool for a quick scan that illustrates which barriers may deem significant for a wide range of abatement options. This enables a rich discussion on prioritization of abatement options and provides an overview of key elements needed in the policy debate; we have shown it provides structure for such a debate (Swart, 2019). The analysis deliberately steps away from a focus on low-cost solutions and would, therefore, be useful alongside MAC curves and IAM results. A key asset is that the method is fully transparent: all data is in view and this provokes a discussion on the arguments supporting the Y factor scores, in particular when used interactively (Swart, 2019). The Y factor does not explicitly mention energy and climate policies but it fits to the policy process (Swart, 2019). The Y factor may inspire policies that, if effective, alter individual scores in specific regions or under specific conditions. Examples are economic policies that could affect costs and financing factors, rules and regulations that could affect multi-actor complexity factors, innovation policies that could affect factors related to physical interdependences and information policies that could affect behavioural factors. At the end of the day, we hope that such policies resolve key barriers hampering the abatement urgently needed for the Paris 1.5 • goal to stay within reach.
CRediT authorship contribution statement E.J.L. Chappin: Writing -original draft, has written the paper, developed the Y factor concept, has developed the preliminary set of definitions and a preliminary scoring, has improved the Y factor website, has developed the final Y factor definitions, and has supervised all the steps of the work by the other authors. M. Soana: Writing -review & editing, has performed the Y factor scoring as presented in the paper with literature review and has performed the validation by means of semi-strsuctured interviews with sector experts, Writing -review & editing, has performed the Y factor scoring as presented in the paper with literature review and has performed the validation by means of semi-. C.E.C. Arensman: has developed the theoretical and practical underpinning of the Y factors with literature review and semi-structured interviews with Dutch experts, has developed the theoretical and practical underpinning of the Y factors with literature review and semistructured interviews with Dutch experts. F. Swart: has developed the preliminary version of the interactive website and has tested the applicability of the Y factor curve by means of focus groups with climate policymakers, Writing -original draft, has written the paper, developed the Y factor concept, has developed the preliminary set of definitions and a preliminary scoring, has improved the Y factor website, has developed the final Y factor definitions, and has supervised all the steps of the work by the other authors, has developed the preliminary version of the interactive website and has tested the applicability of the Y factor curve by means of focus groups with climate policymakers.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.