Comparability of effort

Under the Paris Agreement, governments have agreed to hold global mean warming to well below 2°C and pursue efforts to limit warming to 1.5°C. In their nationally determined contributions (NDCs), governments put forward their commitments to achieve this global goal. 

Although there are no agreed guidelines on what would constitute a fair level of contribution to the global effort, beyond the general understanding of it to reflect the “highest possible ambition” and “common but differentiated responsibilities and respective capabilities, in the light of different national circumstances” (Paris Agreement, Article 4.3), governments are expected to provide some justification of their proposed efforts.

The Paris Agreement envisages an iterative approach to updating and progressing NDCs, in which individual government efforts are to be regularly revised informed by a regular global stock-taking process.

The Climate Action Tracker (CAT) provides a transparent way of comparing NDCs [1] with the many interpretations of what is “fair.”  We hope that it helps governments, the media and observers to interpret the commitments of countries under the Paris Agreement.  

Our rating system evaluates the emissions levels resulting from emissions reductions commitments against effort sharing benchmarks for each country. Whenever current policies are not in line with the emissions commitments, we also give a second rating to that country based on the emissions levels resulting from its current policy projections.

Method to rate level of effort

There is no single, agreed framework on what is a fair contribution to global efforts.  Assessing what is fair depends on the viewpoint and interests of governments. Many consider it fair that those who have made a bigger contribution to the problem - or have a higher capability to act - do more. Others argue that purely economic metrics should be applied with emission reductions made where they are cheapest.      

In our effort sharing assessment, we have compiled a wide range of literature on what different researchers from many perspectives would consider a “fair” contribution to greenhouse gas reductions.

The effort-sharing studies in the CAT’s database include over 40 studies used by the IPCC (chapter 6 of WG III and Höhne et al. (2013)) plus additional analyses the CAT has performed to complete the dataset. They cover very different viewpoints of what could be fair, including considerations of equity such as historical responsibility, capability, and equality. We take into account results from studies that are compatible with the former 2°C goal, as well as the 1.5°C limit in the Paris Agreement, to cover the full range of perspectives and historical developments of the long-term temperature goals.

Instead of deciding what is fair, we construct a Fair Share range for each country from the range of fairness estimates from the literature. 

We then divide the Fair Share range into three sections: 

Each section corresponds to the temperature outcomes that would result if all other governments were to put forward emissions reduction commitments with the same relative ambition level.  

For example, if all governments were to put forward “insufficient” commitments at the least ambitious end of their Fair Share range (maximum fair emissions), warming would more likely than not exceed 2°C and be below 3°C by the end of the century.  

An “insufficient” rating therefore means that although the commitment could be considered fair by some approaches, it is not sufficient to hold warming below 2°C, much less 1.5°C, unless others do substantially more.   

If all governments were to put forward commitments within the “2°C compatible” category, warming could be held below 2°C with a likely probability (66% or greater), but not “well below 2°C” or below 1.5°C.

If all governments put forward “1.5°C Paris Agreement compatible” commitments at the most ambitious of their Fair Share range (minimum fair emissions), warming would be held well below 2°C and limited to 1.5°C.
  

Figure 1. The Climate Action Tracker’s rating system

If a government’s commitment results in emissions above the Fair Share range, we would rate it either “Critically Insufficient” (black) or “Highly Insufficient” (red):

  • If all governments were to follow a “critically insufficient” ambition level, warming would exceed 4°C.
  • If all governments were to follow a “highly insufficient” ambition level, warming reach above 3°C and below 4°C.  Commitments with emissions levels in these two categories are not in line with any interpretation of what is “fair” and are not compatible with a 1.5°C or 2°C pathway.

In contrast, if a commitment results in emissions levels below a country’s full Fair Share range, it is rated “role model” (dark green). Such a proposal is even more ambitious than any interpretations of what is considered “fair.”  This means this country could be seen as doing more than its fair share to limit warming to 1.5°C.

Figure 2 shows an example of a country’s Fair Share ranges for 2020, 2025, 2030 and 2050, for which the NDC target falls into the “Highly insufficient” category. As shown in the example, the Fair Share range is usually large: in this case, for 2030, the range covers emissions similar to 2000 levels to less than zero.  

The “less than zero” means that in some interpretations of what is fair, this government would have no emissions allowances left in 2030 and would have to have fully phased out its emissions or compensate its remaining emissions with reductions elsewhere, for example through supporting emissions reductions in other countries.

Figure 2. Example illustration of the ratings (example Singapore) with the Fair Share ranges for 2020, 2025, 2030 and 2050 shown in coloured bars representing the CAT rating categories, and the country’s emissions projections and NDC commitment shown in comparison.

If a country is at the border between two categories, several additional elements are taken into account for the rating:
  • Unconditional / conditional:  some governments’ commitments are conditional on what other countries commit do or on international financing. Others have made unconditional commitments, and some have not clearly stated whether their commitment is conditional or not.  We always rate the unconditional commitment by default, but we also show the conditional commitment in the figures. If a government only provided a commitment that is conditional, we rate it one category lower than an unconditional commitment of the same stringency. If a government only provided a commitment that is conditional and is at the border of two categories, we rate it in the lower category. In such cases, more information and clarity on the conditionality of the commitment or an additional unconditional commitment could improve the rating.
  • Details of the commitment: some governments’ commitments are based only on an announcement by e.g. the country’s President, while other commitments are accompanied by detailed reports and/or legislation and decisions on implementation and monitoring. If a country is at the border between two categories and if the details of the commitment were not available or if the assessment required us to make many additional assumptions, we rate it in the lower category. In such cases, more information could improve the rating.
  • LULUCF accounting:  governments propose different accounting methods for emissions of the LULUCF sector in their NDCs. If a country is at the border of two categories and its intended accounting methods for the LULUCF sector are not specified in detail in the NDC, but could make a substantial difference in the target emissions levels according to our assessment, we rate it in the lower category. In such cases, more information and clarity on the LULUCF accounting could improve the rating.

The CAT rates all governments’ NDCs against their fair share contribution towards reducing emissions from fossil fuel combustion, industry, agriculture and waste sources—in effect on their contribution towards long-term decarbonisation—excluding LULUCF. The reasoning behind this approach can be found here.

For some countries, we give a special rating because we are unable to rate their NDC, following the guidelines below:

  • Countries that withdraw an NDC, or formally declare their intent to withdraw their NDC, we rate “Critically insufficient”. The removal of a previous pledged contribution is interpreted as an intent to proceed with unconstrained emissions. This is not in line with a fair contribution to global emissions reductions.
  • Countries that have not submitted an NDC are rated either “Critically insufficient”, where non-submission reflects active disengagement with the Paris Agreement, or “Not rated”.
  • Countries that have submitted an NDC, but have failed to include an emissions reduction target, we rate taking into account their history of commitments under the UNFCCC, their current policy emission projections and other relevant circumstances in the country.

Taking all published sharing approaches into account

For each country and year we show the ranges that result from seven specific effort sharing categories summarised in Figure 3, based on the definitions used in the IPCC report (chapter 6 of WG III).

 

Figure 3. Categories of effort sharing approaches (Höhne, den Elzen, & Escalante, 2014). Note: cost effectiveness is a concept included in the capability/costs category, but isn’t a stand alone category.  

Each category puts an emphasis on one particular aspect of effort-sharing and can therefore result in (very) different outcomes from the other categories:

  • Responsibility: emissions reductions below a reference are determined by the level of a country’s historical emissions. This was first proposed by Brazil in the Kyoto Protocol negotiations (UNFCCC, 1997)
  • Capability/Need: emissions reductions below a reference are determined by a country’s level of economic capability, often measured by GDP/capita or the Human Development Index.
  • Equality: emissions per capita converge to, or immediately reach, the same level for all countries, e.g. (Chakravarty et al., 2009; GCI, 2005)
  • Equal cumulative per capita emissions: emissions need to be reduced so that cumulative emissions per capita reach the same level, e.g. (Pan, Teng, & Wang, 2013; WBGU, 2009)
  • Responsibility/capability/need: a range of studies have explicitly used responsibility and capability as the basis for distributing emissions reductions e.g. (Paul Baer, Athanasiou, Kartha, & Kemp-Benedict, 2009; Winkler, Jayaraman, et al., 2011)
  • Capability/cost: a range of studies use equal costs or welfare loss per GDP as a basis. This is essentially a combination of mitigation potential and capability.
  • Staged: a suite of studies have proposed or have analysed approaches where countries take differentiated commitments in various stages. Categorisation to a stage and the respective commitments are determined by indicators using many equity principles, e.g.(Michel G J den Elzen & Meinshausen, 2005; Höhne, Gardiner, Gilbert, Hagemann, & Moltmann, 2008).

Figure 4. Detailed results of effort sharing categories and how they are used to construct the Fair Share range for Singapore in 2050. The left hand graph shows the range of emission levels expected for Singapore under each of the seven different sharing categories (number of data points included for each category in brackets) for both the 2°C and 1.5°C temperature limits. The coloured bar shows the translation into the six rating categories used by CAT (critically insufficient, highly insufficient, insufficient, 2°C compatible, 1.5°C Paris Agreement compatible, and role model). The “Insufficient” to “1.5°C Paris Agreement compatible” range represents the full fair share range of a country, excluding outliers (the categories with the least ambitious and the most ambitious approaches).

Using the detailed representation of the Fair Share range, which pops-up when hovering the cursor on each of the country’s effort sharing bars on the CAT country pages (see Figure 4 for an example), users can discern the emissions allowances of an effort sharing category that they consider fair.  

The “insufficient” to “1.5°C Paris Agreement compatible” range represents the full Fair Share range of a country, excluding the highest and the lowest values of the full sample of equity studies, which represent the outliers for this country. This means that the top and bottom end of the Fair Share range are defined by the second highest/lowest categories. To eliminate extreme outliers for each category, we only consider values within the 10th to 90th percentile of all the values included in that category.

As an example, for Singapore the bottom end of the range (most ambitious) is defined by the approaches based on capability. These approaches assume that Singapore, used as our example, has so much capability that it should actually have negative emissions in 2050 (if necessary compensating its real emissions with allowances from elsewhere).]

The staged effort sharing results, (as well as the ones under the equality category) define the top (least ambitious) end of the range for Singapore in 2050  (Figure 4).  The staged approach, among others, assumes that costs to reduce emissions in Singapore are high and that larger emissions reductions would have to be made by other countries for overall compatibility with a 1.5°C or 2°C pathway.

For further methodological details on our approach to calculate the full Fair Share range for each country refer to the detailed methodology section.

Changes compared to the 2015 methodology

To evaluate the compatibility of targets with the Paris Agreement’s long-term temperature goal (LTTG), we updated the CAT rating system methodology in 2017. This update covered not only the inclusion of the Paris Agreement long-term temperature goal in the rating system, but also a data update to reflect the most up-to date, country-level, emissions data, and the introduction of new rating categories. For further information on how the 2017 methodology update affects our rating system click here.

Detailed methodology

Input data

We include raw data from over 40 studies (see list of studies in table format in Annex). Additionally, we take into account the following sources:

  • Calculations run by Climate Analytics with the PRIMAP Equity tool (a detailed description is available here)
  • Calculations run by NewClimate Institute with the EVOC tool (a detailed description is available in Annex B of this external report)

The first step is to filter the data. The algorithm applies the following filters:

  • The GHG concentration levels of 450ppm (for studies consistent with 2°C  warming) and 400ppm (for studies consistent with 1.5°C warming).
  • Best estimate (rather than high/low)
  • Years 2020, 2025, 2030, 2050

A few studies are excluded, either if they are too old or not all data inconsistencies could be clarified (see Annex).

Harmonisation to base year

Independently of the differences in emissions scope (i.e. sectoral coverage), the studies may have different values for historical emissions, i.e. emissions before a study’s starting year (the year where the study’s effort sharing calculations start). This is due to the fact that the underlying models often use different data sources for historical emissions. The differences can be substantial, especially for countries where the share of non-CO2 emissions is high.

To account for these differences, we scale the whole time series of each study and scenario upwards or downwards by one factor, so that the emissions in the study’s base year have the same emissions value as the standard CAT dataset. For countries not covered by the Climate Action Tracker, historical data from the PRIMAP database serves as a common base level.

Where this option leads to drastic absolute differences between the original and the harmonised value of over 100%, we scale the whole time series by an absolute amount of emissions, rather than by a multiplicative factor. For two thirds of our countries, differences are not very large and we do not use this method.  For the countries where we do find large discrepancies, this is usually the case for just a few studies (e.g. for Costa Rica and Bhutan it happens in 2 out of 17 studies for 2°C scenarios and 5 out of 14 studies for 1.5°C scenarios while for Brazil it is 4 out of 27 for 2°C scenarios and 7 out of 17 for 1.5°C degree scenarios).

This happens mostly for developing countries with large data uncertainty in the historical data. An example where the historical data in those cases varies strongly from the CAT or PRIMAP dataset is the Gambia, where the historical emissions according to the CAT are around 2 MtCO2e/a in 2010, whereas they are 6.8 MtCO2e in Baer et al (2008), or Ethiopia, where CAT estimates emissions of around 100 MtCO2e in 2010, while the EVOC tool calculates historical emissions of almost 240 MtCO2e in the same year.

Calculation of the factor to differentiate between insufficient and 1.5°C or 2°C compatible rating

A commitment in the upper part of the Fair Share range for one country implies that other countries have to reduce to below the upper part of the range in order to compensate and to jointly meet the required global emissions pathway.  The maximum of country A does not generally reflect the same burden-sharing approach as the maximum of country B, and the sum of all maxima would be above the level required to stay on a 1.5°C or 2°C compatible pathway.

Proposing an emissions reduction target in the upper part of a country’s Fair Share range thus does not necessarily make this proposal “sufficient”.  Even though the target could be considered “fair” by some measures, it would not be enough to reach 1.5°C or 2°C unless other countries did more.  The Climate Action Tracker introduces an “insufficient” category to reflect this condition. An “insufficient” rating means that the proposal would only be 1.5°C or 2°C compatible, if other countries moved to their more ambitious end of their Fair Share range.

We calculate the share of the insufficient, 2°C and 1.5°C categories in the Fair Share range in the following way:

1. We calculate emissions levels consistent with a global Fair Share range, determined by:
  • A global effort sharing best case scenario: where all countries choose to reduce emissions to the very bottom of their range, which is numerically equivalent to the lowest 10th percentile of all countries’ Fair Share ranges, and necessarily below the target scenario.
  • A global effort sharing worst case scenario: where all countries choose to reduce emissions only to the top of their equity range, which is numerically equivalent to the highest 10th percentile of all countries’ equity ranges, and necessarily above the target scenario.
2. In a next step, the 1.5°C and 2°C compatible emissions pathways used for our global aggregation methodology are overlaid with the global Fair Share range to determine the intersection between the global Fair Share range and the target scenario.
3. We apply the relative level of the global Fair Share range that corresponds to the 1.5°C and 2°C compatible global pathways to each country´s Fair Share range in order to determine the minimal emissions reduction level that would be required in order to make sure that the global target is met without relying on other countries making a comparably bigger effort to reduce emissions.

Footnotes

[1] For those countries which have not yet ratified the Paris Agreement, the CAT rates the country’s INDC.

[2] The 2oC compatible limit refers to scenarios that hold warming below a 2°C increase above the global mean pre-industrial climate with a likely (66%) or greater probability over the entire 21st century.

[2] The 1.5°C Paris Agreement compatible refers to scenarios that hold the global mean increase in warming well below 2°C and limit this warming to 1.5° C. This means that peak 21st century warming is significantly lower than in the 2°C compatible scenarios, with warming by the end of the 21st century is 1.5°C or below with about 50% probability. 

Annex: List of studies used in the analysis

The table below shows, which studies have been considered and which have been included in the data harmonisation exercise.

Study name

Included in harmonsiation

001 (den Elzen, Beltran, Hof, van Ruijven, & van Vliet, 2013)

yes

002 (Van Vuuren, Isaac, Den Elzen, Stehfest, & Van Vliet, 2010)

yes

003 (Hof & Den Elzen, 2010)

yes

004 (Knopf et al., 2009)

yes

005 (den Elzen, Höhne, & Moltmann, 2008)

yes

007 (Hof & Den Elzen, 2010)

yes

008 (den Elzen, Lucas, & van Vuuren, 2008)

yes

009 (van Vuuren et al., 2009)

yes

010 (den Elzen, Höhne, Brouns, Winkler, & Ott, 2007)

yes

011 (den Elzen & Meinshausen, 2006)

yes

012 (den Elzen, Lucas, & van Vuuren, 2005)

yes

013 (Den Elzen & Lucas, 2005)

yes

014 (Criqui et al., 2003)

yes

015 (Berk & den Elzen, 2001)

no, outdated

016 (Kuntsi-Reunanen & Luukkanen, 2006)

no, different stabilisation level

017 (Winkler, Letete, & Marquard, 2011)

yes

018 (Chakravarty et al., 2009)

yes

019 (Bows & Anderson, 2008)

yes

021 (Vaillancourt & Waaub, 2004)

no, different stabilisation level

022 (Miketa & Schrattenholzer, 2006)

no, different stabilisation level

023 (Bode, 2004)

no, different stabilisation level

025 (Böhringer & Welsch, 2006)

yes

026 (Groenenberg, Blok, & van der Sluijs, 2004)

no, different stabilisation level

027 (WBGU, 2009)

yes

028 (Knopf, Kowarsch, Lüken, Edenhofer, & Luderer, 2012)

no, data inconsistencies

031 (Nabel et al., 2011)

yes

032 (Peterson & Klepper, 2007)

yes

034 (Onigkeit, Anger, & Brouns, 2009)

yes

035 (Jacoby, Babiker, Paltsev, & Reilly, 2008)

yes

036 (Edenhofer et al., 2010)

yes

037 (Höhne & Moltmann, 2009)

yes

038 (Höhne & Moltmann, 2008)

yes

042 (Jayaraman, Kanitkar, & Dsouza, 2011)

yes

043 NIES (Kriegel et al., 2014)

yes

044 PNNL (Kriegel et al., 2014)

yes

045 PBL (Kriegel et al., 2014)

yes

046 IIASA (Kriegel et al., 2014)

yes

047 PIK (Kriegel et al., 2014)

yes

048 FEEM (Kriegel et al., 2014)

yes

049 (Kober, Zwaan, & Rösler, 2012)

no, data inconsistencies

050 (Baer et al., 2008)

yes

051 (Pan, Teng, & Wang, 2014)

yes

052 (Pan, Teng, & Wang, 2013)*

yes

057 (Robiou du Pont et al., 2016)

yes

Approaches for equal cumulative emissions per capita from this study were excluded from out equity ranges due to data inconsistencies.