When less is more: limits to international transfers under Article 6 of the Paris Agreement

ABSTRACT

International carbon markets can be an important tool in achieving countries’ mitigation targets under the Paris Agreement, but they are subject to a number of environmental integrity risks. An important risk is that some countries have mitigation targets that correspond to higher levels of emissions than independent projections of their likely emissions. If such ‘hot air’ can be transferred to other countries, it could increase aggregated emissions and create a perverse incentive for countries not to enhance the ambition of future mitigation targets. Limits to international transfers of mitigation outcomes have been proposed to address this risk. This article proposes a typology for such limits, explores key design options, and tests different types of limits in the context of 15 countries. Our analysis indicates that limits to international transfers could, if designed appropriately, prevent most of the hot air contained in current mitigation targets from being transferred, but also involve trade-offs between different policy objectives. Given the risks from international transfer of hot air and the uncertainty over whether other approaches will be effective in ensuring environmental integrity, we recommend that countries take a cautious approach and pursue a portfolio of approaches to ensure environmental integrity, in which case limits could provide for additional safeguards.

Key policy insights

• Limits to international transfers involve trade-offs between different policy objectives, in particular reducing the risk that countries transfer hot air and enabling participation in carbon markets.

• Under ‘relative’ limits a country may transfer mitigation outcomes to the extent that its actual emissions are below the limit. Relative limits derived from historical emissions data have significant limitations, and none of the tested approaches was found to be effective for all countries. Relative limits based on emission projections could be a more valid approach, although they are also technically and politically challenging.

• Under ‘absolute’ limits a country could only issue, transfer or acquire a certain amount of mitigation outcomes. Absolute limits set at sufficiently low levels could prevent countries from transferring large amounts of hot air, but are bluntly applicable to all countries, whether or not they have hot air.

KEYWORDS:

• Paris Agreement
• emissions trading
• environmental integrity
• carbon markets

1. Introduction

Article 6 of the Paris Agreement allows countries to use international carbon markets to achieve their mitigation targets communicated in nationally determined contributions (NDCs), hereinafter referred to as ‘NDC targets’. Article 6.2 allows countries to use ‘internationally transferred mitigation outcomes’ (ITMOs) – i.e. climate change mitigation achieved in one country but claimed by another – to achieve their NDC targets. Article 6.4 establishes a new crediting mechanism under international supervision that could be used for similar purposes.

Carbon markets are considered a key tool to reduce greenhouse gas (GHG) emissions (IPCC, 2014). They can reduce the cost of achieving mitigation targets by providing flexibility in how and where emissions are reduced and could thereby facilitate the adoption of more ambitious mitigation targets. Yet international carbon markets involve a number of environmental integrity risks (Schneider & La Hoz Theuer, 2018): if not designed and implemented appropriately, they could result in greater GHG emissions than if they were not employed. The Paris Agreement therefore requires Parties to ensure environmental integrity when engaging in international transfers of mitigation outcomes.

A key risk to environmental integrity concerns international transfers from countries with weak mitigation targets. Under the 1997 Kyoto Protocol, some countries had mitigation targets which did not require the country to take any mitigation action, creating surplus units that were often referred to as ‘hot air’ (Boehringer, 2000; Boehringer, Moslener, & Sturm, 2007; Brandt & Svendsen, 2002; den Elzen, Roelfsema, & Slingerland, 2009; IPCC, 2001, sec. 6.3.1; Paltsev, 2000; Victor, Nakicenovic, & Victor, 1998). Independent assessments of current NDC targets suggest that this situation may also arise under the Paris Agreement, since the mitigation targets of several countries could correspond to higher levels of emissions than the projection of their likely emissions level with the policies in place at the time of setting the target (CAT, 2018; den Elzen et al., 2016; Meinshausen & Alexander, 2016). These countries could thus appear to generate emission reductions (relative to their targets), without generating any actual emission reductions.

Hot air could pose several risks to environmental integrity under Article 6. If the rules under Article 6 were to allow countries to transfer hot air to other countries, then global GHG emissions could end up higher than they would have been if the countries’ NDC targets were achieved without such transfers. This is because the transfer would allow the acquiring country to increase its emissions above its NDC target, while the transferring country would not need to engage in corresponding emission reductions to achieve its NDC target. Allowing countries to transfer hot air could also create a perverse incentive for transferring countries not to enhance the ambition of mitigation targets in future NDCs, in order to accrue higher benefits from international transfers. Countries with hot air, moreover, may have less incentives to ensure environmental integrity, as they would achieve their NDC target even if they engage in transfers that do not represent actual mitigation outcomes (Kollmuss, Schneider, & Zhezherin, 2015; Schneider & La Hoz Theuer, 2018).

Parties to the Paris Agreement are currently negotiating the rules for the implementation of Article 6. An important and controversial issue is whether and how international rules should promote environmental integrity. Several Parties have proposed that international transfers under Article 6 be subject to quantitative limits. Such limits are proposed to address environmental integrity concerns but also to pursue other policy objectives, such as ensuring that a minimum portion of mitigation action takes place domestically (see, e.g. AOSIS, 2017; Arab Group, 2017; Brazil, 2014, 2016, 2017; LMDC, 2017; Venezuela, 2017).

While limits to international transfers are an important topic in the ongoing negotiations, research on this topic is sparse. This article explores how limits could be established and assesses the implications of different types of limits, with a focus on whether and how they mitigate the risk of international transfer of hot air. The article draws on relevant literature, submissions by countries, and the experience with limits under the United Nations Framework Convention on Climate Change (UNFCCC) and the Kyoto Protocol. In addition, the implications of different types of limits are tested in the context of 15 countries, using data on historical emissions, information from NDCs, and independently-established emission projections.

The article employs specific terminology and makes several assumptions. Article 6.2 of the Paris Agreement does not specify what an ITMO is, nor how transfers should take place. ITMOs could be international units that are transferred between electronic registries or they could be amounts that are reported by countries for accounting purposes (Schneider, Füssler, Kohli, et al., 2017). It is assumed here that ITMOs are amounts reported by countries. For simplicity, the term ‘ITMOs’ is used to refer to transfers both of mitigation outcomes generated under Article 6.2 and of emission reductions resulting from the Article 6.4 mechanism. It is further assumed that ITMOs are expressed in tonnes of carbon dioxide equivalent (tCO₂e), noting that the findings would also hold if this was not the case.

No agreed definition of ‘hot air’ exists. This paper adopts the approach typically employed in the literature where hot air is defined in the context of mitigation targets that countries over-achieve without pursuing further mitigation actions (Boehringer, 2000; den Elzen & de Moor, 2002; Kollmuss et al., 2015; Schneider & La Hoz Theuer, 2018; UNFCCC, n.d.; Victor, Nakićenović, & Victor, 2001). Throughout this paper, hot air is understood as the difference between the NDC target level and the projection of the likely emissions level with the policies in place at the time of setting the target. We use the projection with policies in place at the time of setting the target because countries formulate their targets based on known circumstances at that point in time.

2. General design options for limits to international transfers

Limits could be pursued to achieve different policy objectives. They could be established with the aim to prevent the transfer of hot air, or to reduce disincentives for transferring countries to increase ambition in future NDCs. They could also be introduced in order to address the risk of ‘over-selling’, i.e. that countries transfer so many mitigation outcomes that they can no longer achieve their NDC target. Finally, limits could also be pursued to ensure that a minimum portion of mitigation effort takes place domestically – which is often referred to as ‘supplementarity’.

Limits could be applied not only to international transfers in the context of Article 6, but also in the context of ‘banking’ or ‘carry-over’ of units. Although this latter aspect is not further explored in this paper, similar considerations would apply.

Limits to international transfers were employed under the Kyoto Protocol, albeit without addressing the risk of transfer of hot air. A principle of ‘supplementarity’ was established – that participation in market mechanisms should be ‘supplemental’ to domestic action – but never operationalized in the form of a quantitative limit. To prevent over-selling, a ‘commitment period reserve’ was introduced, requiring that each Party maintain a minimum reserve of units that cannot be transferred internationally (Yamin & Depledge, 2004).

The Doha Amendment to the Kyoto Protocol includes a provision that aims specifically at addressing the risk of transfer of hot air: Article 3.7ter establishes that countries’ assigned amounts (i.e. their budget of emissions under the Kyoto Protocol) in the period 2013–2020 cannot exceed each country’s average annual emissions over the period 2008–2010. This provision thus implicitly ensures that countries’ targets in the second commitment period correspond, at a minimum, to the average reported emissions for 2008–2010. In so doing, this provision reduces the amount of hot air that countries could transfer (Chen, Gütschow, Vieweg, Macey, & Schaeffer, 2013). The Doha Amendment, however, has not yet entered into force.

In the ongoing negotiations, Brazil proposes limits to international transfers which could address the risk of transfers of hot air. Other countries and groups of countries – such as the Alliance of Small Island Developing States, the Arab Group, Like Minded Developing Countries and Venezuela – propose limits akin to the supplementarity principle and to the commitment period reserve, among others. This section explores various design options for limits, taking into account the various policy objectives mentioned above. Section 3 then tests different options for limits, focusing specifically on the objective of preventing transfer of hot air.

2.1. Relative and absolute limits

Limits could be established in a number of ways. Here, two main types are distinguished. Under a relative limit, a country would be allowed to transfer ITMOs to the extent that its actual emissions in the target year or period are below a specified limit. This type of limit is referred to here as a relative because it would allow the country to transfer any amount of ITMOs – as long as it reduces emissions respectively below the limit. Relative limits can reduce the risk of transferring hot air, as well as avoid perverse incentives for countries not to enhance the ambition of future NDCs.

To prevent transfers of hot air, relative limits would ideally be set exactly at the projection of the likely emissions level with the policies in place at the time of setting the target, as shown in Figure 1. In this case, the limit would prevent transfers of hot air while still enabling transfers that do result from abatement action. In practice, however, establishing emission projections is both technically and politically challenging: they are based on assumptions about future developments and therefore involve considerable uncertainties. No internationally agreed method exists to establish emission projections, and estimations often diverge considerably between authors, depending on the assumptions and methods used (Rogelj et al., 2017). Policy-makers could therefore also consider deriving relative limits from other parameters, such as average historical emissions, as proposed by Brazil (Brazil, 2014, 2017).

Figure 1. Relative limit based on emissions projections in 2030. Note: The figure illustrates the application of a relative limit for a country with an NDC target for 2030 (black square) that is less stringent than the projected emissions (blue line). The country thus has hot air (red arrow). The country implements mitigation actions which bring its emissions (black dashed line) below the emissions projection. In this example, the relative limit (orange line) for 2030 is set exactly at the level of the projection of the likely emissions level with the policies in place at the time of setting the target. The amount of ITMOs the country is allowed to transfer in 2030 corresponds to the reduction of emissions below the limit (green arrow).

Under an absolute limit, a country could issue, transfer or acquire only a certain absolute (or fixed) number of ITMOs – e.g. a fixed percentage of reported emissions in a given year. Absolute limits would restrict international transfers from all countries, irrespective of environmental integrity risks. They would, therefore, help contain rather than address the risk of transferring hot air. Absolute limits could also be employed to reduce the risk of over-selling or to operationalize the principle of supplementarity.

2.2. Applicability to transferring or acquiring countries

Limits could be applied to transferring and/or acquiring countries, as well as to different groups of countries. Limits placed on transferring countries could help prevent hot air transfers, reduce perverse incentives not to enhance the ambition of future NDCs, and prevent over-selling. Limits placed on acquiring countries could help ensure supplementarity but may do little to address the hot air risk, because transferring countries could still in principle transfer all of the hot air contained in their NDCs.

Limits could also apply to all countries, or only to some countries. Limits could, for example, not apply to least developed countries or to small island developing states, as these country groups are differentiated under the Paris Agreement.

2.3. Applicability to types of transfers

Limits could apply to all international transfers under the Paris Agreement or only to some types of transfers. Limits could, for example, not apply to ITMO transfers that are backed by mechanisms that may involve lower environmental integrity risks. International linkages between emissions trading systems (ETSs), for example, could have limited environmental integrity risks because jurisdictions with ambitious ETS caps are unlikely to link to an ETS that is much less ambitious or over-allocated (Ranson & Stavins, 2016). Although crediting mechanisms could also ensure environmental integrity, the available analyses indicate that they face considerable challenges and uncertainty in this regard (Cames et al., 2017; Erickson, Lazarus, & Spalding-Fecher, 2014; Gillenwater, 2012; Kollmuss et al., 2015; Schneider, 2009). Yet distinguishing ITMO transfers backed by linkages of ETSs from other types of transfers could be technically and politically challenging. It would require international agreement on a definition of ETSs and a method to determine how many ITMOs were transferred through the linkage of two ETSs.

2.4. Methods for establishing the limit

The level of limits could be established in a number of ways. Limits could be based on different parameters, such as historical GHG emissions or the historical GHG emissions intensity per gross domestic product (GDP), a GHG emissions level corresponding to the NDC target, or actual GHG emissions in the target year or period. The parameters could be determined using different reference periods, with various lengths and starting points. Finally, several methods could be applied to calculate the level of the limit, such as simple percentages, average values or an extrapolation of trends.

2.5. Timing and point of application

Internationally agreed limits are only effective if they are adhered to when countries account for their NDCs. When and how limits could be applied would depend on the rules under the Paris Agreement. An ongoing application of limits (e.g. at the point of issuance in transferring countries and/or at the point of use by acquiring countries) could be implemented if ITMOs were international units that are tracked through registries. If ITMOs were amounts reported for the purpose of accounting for international transfers through ‘corresponding adjustments’, limits could be applied ex-post, when countries account for their NDCs. In this latter case, it would be the responsibility of countries to ensure that they engage in transfers in a manner consistent with the limit.

Other design features are also relevant, such as considerations on when to establish and assess limits. These are not further considered here.

3. Testing different options for establishing limits

3.1. Approach and methodology

To understand the suitability and implications of the various options for establishing limits, different methods for determining limits are tested in the context of 15 countries that represent a variety of circumstances. The different options for establishing limits are assessed against two criteria: whether and how they address the environmental integrity risk of international transfers of hot air, and whether and how they allow countries to transfer ITMOs that result from actual mitigation action.

To assess these implications, the GHG emissions levels corresponding to the NDC targets are compared with independently-established emissions projections in order to assess whether and how much hot air is contained in NDC targets. For each type of limit, it is then assessed whether and how the country could engage in international transfers and the extent to which this would involve hot air. The analysis assesses the implications for the year 2030, which is used as target year by all selected countries. It is important to note that this analysis is not an assessment of the ambition of individual NDC targets, which would have to take into account equity and development considerations as well as other country circumstances.

The data used for the analysis is derived from two main sources. Information on historical GHG emissions, on emissions projections, and on the quantification of countries’ NDC targets is drawn from the Climate Action Tracker (CAT) project (CAT, 2015). The data from CAT is used because it provides independent, coherently-applied emissions projections derived from country-specific information. We use the 2015 emissions projections, as they approximately reflect the policies that were in place when countries set their targets. Where the CAT data includes a range for the level of NDC targets and/or for emission projections, the average value is used. Data gaps were filled using data from CAT (2017) and WRI (2017), making adjustments to ensure time series consistency. Some approaches make use of historical and projected GDP values; these were drawn from USDA-ERS (2017).

3.2. Relative limits

Relative limits could be established with the objective to prevent the transfer of hot air while enabling the transfer of mitigation outcomes that are additional to those implemented at the time when the target is set. In this case, relative limits set exactly at the level of emissions projections with policies in place at the time of setting the target would – theoretically – best achieve both objectives (see section 2.1 above). As establishing emissions projections is both technically and politically challenging, several alternative approaches for establishing relative limits are tested, with the aim of understanding whether they could be good approximations of emissions projections with policies at the time of setting the target.

In total eight alternative approaches are tested to each of the 15 countries. These include relative limits based on historical GHG emissions, as proposed by Brazil, or based on historical GHG emissions per gross domestic product (GDP). In addition to using average historical data to establish the limit, the extrapolation of historical trends is also tested, using the least squares method. In all scenarios, a historical reference period is used to calculate the average values or extrapolate the trend, though with different lengths. For average historical data, three-year and five-year periods are tested; for the extrapolation of historical trends, five-year and ten-year periods are tested.

Table 1 shows the results of the analysis. For each country and for each approach to establish relative limits, the table presents the deviation of the relative limit from the average emissions projections in 2030. The smaller the deviation, the better the approach approximates the emissions projections. A negative value denotes a situation where the relative limit lies below (i.e. is more stringent than) the projections. Information on the range of the projections by CAT is also provided, highlighting the uncertainty surrounding the projections. Finally, Table 1 also presents the relative difference between the NDC target level and the emissions projections; positive values mean that the NDC target level is above the projection (i.e. contains hot air). Figure 2 illustrates the results of the different approaches for selected countries.

Figure 2. Results for relative limits for selected countries.

Table 1. Deviation of relative limits from average emission projections in 2030.

Country	Range of emissions projection	NDC compared to avg. projection	Deviation of relative limits from average emission projections in 2030
			Historical GHG emissions				Historical GHG emissions per GDP
			Average		Trend		Average		Trend
			3 yrs	5 yrs	5 yrs	10 yrs	3 yrs	5 yrs	5 yrs	10 yrs
Argentina	+/−14%	−6%	−35%	−35%	−39%	−22%	+2%	+9%	−214%	−212%
Brazil	Single point	−5%	−23%	−25%	+10%	+8%	+6%	+8%	−21%	−31%
China	+/−4%	+10%	−24%	−28%	+54%	+49%	+123%	+130%	−23%	−72%
Ethiopia	Single point	−40%	−70%	−70%	−61%	−54%	−5%	+4%	−272%	−281%
EU	+/−7%	−13%	+16%	+18%	−30%	−24%	+54%	+57%	−11%	−37%
Gambia	Single point	−23%	−42%	−46%	+30%	+13%	+30%	+25%	+129%	+92%
India	+/−1%	+5%	−54%	−56%	−13%	−16%	+74%	+79%	−10%	−14%
Indonesia	Single point	+21%	−36%	−39%	+17%	−6%	+66%	+68%	+29%	−36%
Japan	+/−5%	−11%	+8%	+6%	+43%	−6%	+25%	+23%	+56%	−3%
New Zealand	Single point	−37%	−13%	−13%	−10%	−21%	+37%	+39%	+2%	−21%
Norway	Single point	−52%	+2%	+2%	−5%	−8%	+45%	+45%	+15%	−11%
Peru	Single point	−22%	−43%	−44%	−20%	−21%	+18%	+23%	−64%	−96%
Russia	+/−1%	+17%	−14%	−15%	+4%	−2%	+10%	+10%	−1%	−68%
South Africa	Single point	−44%	−39%	−39%	−29%	−22%	−11%	−10%	−29%	−43%
South Korea	+/−8%	−27%	−8%	−12%	+63%	+34%	+47%	+45%	+65%	+11%
	Standard deviation		0.24	0.24	0.35	0.25	0.35	0.36	1.01	0.90

Source: Own calculations based on data by CAT (2015) and by USDA-ERS (2017). Figures in red indicate hot air.

Table 1 and Figure 2 show that none of the analysed approaches is effective for all countries. The calculated relative limits often differ significantly from the emission projections, leading to high standard deviations for all approaches for relative limits. In many cases, the difference between the relative limit and the emissions projection is larger than the difference between the NDC target and the emissions projection. This implies that none of the tested approaches achieves the objective of preventing the transfers of hot air for all countries while enabling ITMO transfers that result from mitigation actions. It was also not possible to identify groups of countries, such as developed or developing countries, for which a particular approach would consistently achieve these objectives. The results, however, differ between the different approaches.

Relative limits based on the average of historical GHG emissions would imply that countries can only transfer ITMOs if they are on a decreasing emissions pathway. That would prevent the transfer of all hot air contained in the NDC targets of the tested countries. Yet most countries have increasing emission trends. This type of limit could make it difficult for these countries to engage in international transfers, especially if they would have to reduce emissions far below their NDC target before being able to transfer ITMOs (see example of India in Panel A of Figure 2). Nonetheless, the standard deviations in Table 1 indicate that relative limits based on the average of historical GHG emissions performed best in approximating emission projections among the countries and options tested. No significant differences in the results were found between 3-year and 5-year reference periods. Panel A in Figure 2 illustrates this limit for a country with stable emissions (Norway) as well as for countries with increasing emissions (Russia and India).

Relative limits based on trends of historical GHG emissions are good approximations for emissions projections for the few countries where the rate of increase or decrease in emissions is expected to stay stable over time, such as in the case of Russia (see panel B in Figure 2). Yet this approach quickly loses accuracy when countries’ rates change over time – which is the case for most other countries analyzed here, such as China. The time period used to calculate the trend was found to have a very significant impact for some countries, such as for Japan. Somewhat surprisingly, for the tested countries, the standard deviation indicates that relative limits based on trends of GHG emissions fared overall worse in approximating emissions projections than relative limits based on averages of GHG emissions.

Relative limits based on the average of historical emissions intensity (i.e. GHG emissions per GDP), as shown in panel C, are close to the emission projections for a few countries, such as Brazil, but lie far above the projections for most countries – as the emissions intensity is expected to decrease for most of the countries. This is of little consequence for countries such as New Zealand, whose targets are more stringent than the emission projection. It would, however, allow for most or all of the hot air to be transferred for other countries, such as in the case of Indonesia.

Relative limits based on trends of historical emissions intensity, as shown in panel D, could potentially reflect that the emissions intensity is decreasing for most countries. Yet the suitability of this approach depends on how countries’ rate of emissions intensity changes over time. In India, for example, the rate of decrease is expected to stay relatively constant until 2030. In Argentina, the rate of decrease is expected to become less prominent over time, causing this type of limit to fall below the emissions projection. In Gambia, a change in the trend causes this type of limit to lie far above the projections. The standard deviation indicates that limits based on trends of historical emissions intensity are the worst performing approach among all relative limits tested in terms of their suitability as an approximation of emissions projections.

3.3. Absolute limits

Six approaches for establishing absolute limits are tested. These include absolute limits based on GHG emissions, as well as absolute limits based on NDC target levels. For limits based on the NDC target level, the reference period is the NDC target year, i.e. 2030. For limits based on GHG emissions, two options are tested: GHG emissions in the three years preceding the communication of the NDC target, with a time gap to account for data availability (2010–2012); and projected GHG emissions in the three years preceding the target year with a time gap to account for data availability (i.e. 2025–2027). For illustrative purposes, the implications are assessed for fixed percentages of 1% and 5%, noting that any other values could be used.

Table 2 outlines the results for each country and each approach for establishing absolute limits. The upper section of the table presents how much hot air is estimated to be contained in each NDC target and contrasts this amount with the volume (in MtCO₂e) that each country would be able to transfer internationally under different absolute limits. The bottom section of Table 2 summarizes the results for each of the approaches.

Table 2. Transferrable volumes of ITMOs under different absolute limits in 2030 (MtCO₂e).

	Amount of hot air	Limits based on average 2010–2012 emissions		Limits based on pre-target year emissions		Limits based on NDC target level
		1%	5%	1%	5%	1%	5%
Argentina	−33	3	17	5	23	5	24
Brazil	−71	10	50	12	59	12	61
China	1393	106	530	144	722	153	763
Ethiopia	−125	1	5	2	8	2	9
EU	−514	46	232	39	195	35	174
Gambia	−1	0	0	0	0	0	0
India	285	25	124	49	245	57	285
Indonesia	230	7	35	11	57	13	66
Japan	−129	13	65	12	60	11	54
New Zealand	−31	1	4	1	3	1	3
Norway	−27	1	3	0	1	0	1
Peru	−30	1	4	1	5	1	5
Russia	435	23	113	29	143	31	154
South Africa	−419	6	29	5	26	5	26
South Korea	−196	7	34	5	27	5	27
Aggregated amount of hot air (100%)	2344
Amount of hot air that can be transferred		160	802	233	1167	253	1267
Percentage of hot air that can be transferred		7%	34%	10%	50%	11%	54%
Percentage of hot air that is prevented from transfer		93%	66%	90%	50%	89%	46%

Source: Own calculations based on data by CAT (2015).

Note: negative values for the amount of hot air denote situations where the country’s NDC target is more stringent than the projection of the likely emissions with the policies in place at the time of setting the target. Figures in red indicate hot air. In the bottom section of the table, the volume of hot air and the ‘amount of hot air that could be transferred’ correspond to the sum from the four countries with hot air. Cells marked with a ‘0’ are rounded-down figures.

The results in Table 2 illustrate that absolute limits would contain the transfer of hot air from countries such as Russia and China, but also restrict transfers from countries with NDC targets more stringent than BAU, such as South Africa. The table also indicates that, in containing hot air transfers, the threshold used (i.e. 1% or 5%) plays a more important role than the parameter employed to derive the limit. A 1% limit on any of the parameters, for example, would prevent about 90% of the hot air from being transferred – whereas a 5% limit would allow up to 54% of the hot air to be transferred.

4. Discussion and conclusions

In considering limits to address the risks from international transfers of hot air, policy-makers may have to balance various policy objectives. Those include reducing the risk that countries engage in the transfer of hot air, facilitating participation in international transfers (in particular for countries that effectively engage in mitigation action), and avoiding disincentives for countries to increase the ambition of future NDCs targets. Practical considerations – such as methodological challenges and the political feasibility of different options – are also relevant, among others. Ultimately, the setting of limits is a policy choice that must balance these policy objectives.

The analysis in this article showed that approaches for establishing relative limits based on historical emissions data have significant limitations. Among all approaches tested here, none was found to be effective for all countries: in some instances, the limits were far below the projection of the likely emissions with the policies in place at the time of setting the target, in others they were far above. Among the countries and options tested, relative limits based on historical GHG emissions were most effective in preventing the transfer of hot air. This is because this type of limit prevents transfers of hot air from all countries with increasing GHG emission trends. While the analysis in this article is limited to 15 countries, data by Meinshausen and Alexander (2016) on current NDC targets and emissions projections for 196 countries indicates that most countries with hot air have increasing emission trends – which means that this type of limit would prevent nearly all hot air in current NDC targets from being transferred. This type of limit would, however, only allow countries with decreasing emissions pathways to transfer ITMOs. While this could be seen to be consistent with the long-term goals of the Paris Agreement, it could prevent many developing countries from engaging in international carbon market mechanisms – even if their targets do not contain hot air. This, in turn, could make it difficult to agree on such an approach under the consensus-based decision-making process under the UNFCCC.

Given the challenges with approaches based on historical emissions data, relative limits based on emission projections may provide a more valid approach if relative limits are to be considered, although they also face technical and political challenges. Any limits based on emission projections would ideally be based on an internationally agreed methodology which would include provisions to ensure consistency and comparability between countries, as well as means to address the uncertainty of key variables such as economic growth.

Absolute limits set at sufficiently low levels could prevent an individual country from transferring large amounts of hot air. The analysis showed that the threshold used (e.g. 1% or 5%) is the main factor in the effectiveness of absolute limits in preventing transfers of hot air, with the lower of the two tested thresholds being significantly more effective in containing (albeit not fully preventing) transfers of hot air. Further advantages of absolute limits include that they are simple and provide ex-ante certainty on the volume of permissible transfers. A key disadvantage is that they are bluntly applicable to all countries – whether or not there is a risk of transfer of hot air – and could thus affect countries’ ability to engage in international transfers. This, in turn, could potentially increase the costs of mitigating climate change if, without such limits, countries were to engage in larger amounts of transfers that are backed by actual emission reductions. The ability to transfer is impacted mainly by the threshold applied. For example, if countries representing half of global GHG emissions were transferring countries, a 1% limit would imply a total global potential for ITMO transfers of about 250 million tCO₂e in 2030. Other possible impacts of absolute limits may also be important to consider. Absolute limits on the basis of NDC target levels, for example, could generate perverse incentives for countries not to enhance the ambition of future NDC targets. Similarly, limits based on pre-target year emissions could provide disincentives for countries to over-achieve their NDC targets, as this would affect the limit applicable to the subsequent period.

For both relative and absolute limits, a key political challenge is that countries may be reluctant to agree to international rules that might restrict their ability to engage in international transfers. This holds for all approaches but may be particularly relevant for the context of international linking of ETSs where the amount of transfers is driven by the regulated entities. As ITMO transfers that are backed by linking of ETSs may also pose lower environmental integrity risks (see section 2.5), policy makers could consider exempting such transfers from any limits. While this increases complexity, it could help balance the policy objectives of promoting environmental integrity and enabling participation in international carbon markets. It may also be important to set limits in a way that minimizes the possibility of manipulation that could result in inflated limit levels. The methodology should therefore be internationally agreed and be applied consistently across countries.

The analysis has a few limitations. An important limitation is that there are significant uncertainties in emissions projections and challenges in interpreting NDC targets. For these reasons, the results for individual countries are also uncertain. Another limitation is that the analysis was conducted only for 15 countries and only four of them are estimated to have hot air. Other data and assessments of NDC targets, however, confirm that the ambition of NDC targets varies strongly and that hot air could be included in a large number of NDCs (den Elzen et al., 2016; Meinshausen & Alexander, 2016). Whether and how much hot air a country has varies also considerably between the different assessments. This confirms that establishing emission projections and interpreting NDC targets is not straightforward and that limits derived from emission projections may be both technically and politically difficult to implement.

Whether or not limits are ultimately needed also depends on whether other policy approaches can effectively address environmental integrity risks under Article 6. Several approaches could be pursued (Schneider & La Hoz Theuer, 2018): First, environmental integrity risks could also be addressed by ensuring that ITMOs represent actual abatement action, such as through international guidance and oversight on the implementation of carbon market mechanisms. The experience from existing carbon market mechanisms suggests, however, that ensuring quality can be both technically and politically challenging (Cames et al., 2017; Kollmuss et al., 2015; Schneider, 2009). Second, environmental integrity risks might also be reduced by facilitating progression in the ambition of NDC targets. Yet this could prove to be difficult, as NDCs are self-determined by Parties. Third, countries could pursue approaches outside the context of the UNFCCC and the Paris Agreement, such as forming ‘carbon clubs’, i.e. groups of countries that apply uniform environmental integrity principles or provisions. These clubs, however, can only address environmental integrity within members of the club, and rely on the willingness of club members to ensure environmental integrity, even when circumstances change (Schneider, Füssler, La Hoz Theuer, et al., 2017). Finally, a key question is whether buyer countries will acquire ITMOs from countries that have hot air and whether and how they can identify the situations where acquiring ITMOs involves environmental integrity risks. In the early years of the Kyoto Protocol, for example, several countries declared that they would not purchase hot air. In later years, however, several countries purchased units that originated from countries with hot air and that were unlikely to be backed by actual emission reductions (Kollmuss et al., 2015; Tuerk, Fazekas, Schreiber, Frieden, & Wolf, 2013). Any declared intentions by countries to uphold environmental integrity would thus need to survive the test of time, which may be difficult in the context of changing political climates and rising carbon prices.

In promoting environmental integrity under the Paris Agreement, the balance between centralized international oversight on the one side and flexibility for countries and market participants on the other has always been a controversial point of discussion. This is particularly the case with the bottom-up nature of the Paris Agreement. In the light of past experiences with the Kyoto Protocol and the potentially significant volume of hot air contained in current NDCs (see Table 2), we recommend that policy-makers carefully consider how to address the risk of hot air under the Paris Agreement.

Given the uncertainty of whether other approaches to promote environmental integrity will be effective, we recommend that Parties take a cautious approach and consider establishing limits for those types of transfers that pose higher environmental integrity risks. Given the challenges identified with relative limits, absolute limits may be a simpler option to pursue.

Acknowledgement

We thank Sophie Closson, Andrew Howard, Olivier Kassi, Anja Kollmuss, Michael Lazarus, M. J. Mace, and Konrad Raeschke-Kessler for review comments and inputs on the Working Paper that formed the basis for this article. We thank the NewClimate Institute and Malte Meinshausen for providing data on nationally determined contributions and emission projections. This work was supported by the Federal Public Service Health, Food Chain Safety andEnvironment of the government of Belgium.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Stephanie La Hoz Theuer http://orcid.org/0000-0001-5606-655X

Lambert Schneider http://orcid.org/0000-0002-5704-0965

Additional information

Funding

This work was supported by Federal Public Service Health, Food Chain Safety and Environment of the government of Belgium.