WBGU (2011) argues that for the type oftransformative change necessary to avoid (permanent- ly) trespassing planetary boundaries it is inevitable to internalise the external effects of carbon pollution into economic decision making. They argue for carbon pricing as a central building block ofthe regulatory framework to avoid dangerous climate change. However, WBGU also states that carbon pricing alone is not sufficient to redeem the various types of market failures that have led to the current unsustainable global socio-economic system. Grubb, Hourcade and Neuhoff (2014) go even one step further and opine that a global carbon price is a “false god” and argue that it is critical to complement it with policies that more immediately direct user behaviour and that actively govern or even steer innovation and strategic infrastructure development. The above assessment ofthetransformativepotentialofthe EU ETS resonates well with this verdict: Market-based mitigation instruments can put a price on carbon, but this alone is not suf- ficient to induce the required change in unsustainable practices. Therefore, even if the flaws on the implementational level can be dealt with, the ETS will need to be integrated in a wider policy mix. It can be an effective instrument that puts economic pressure on current unsustainable prac- tices, if prices are high and stable enough and free allocation of permits is repealed. As such the ETS can function as a ‘motor of destabilisation’ of unsustainable economic practices, but it fails to be a ‘motor of innovation’ (Kivimaa and Kern 2015). As a technology neutral policy it does not provide a protected space for emerging more sustainable technologies. The EU ETS’s track record for spurring innovation is poor. More targeted policies are necessary to complement the ETS and make up for these deficiencies.
the EU could relinquish most (if not all) decisional powers to an independent body that would manage the EU ETS, potentially including full control over the cap and/or price. In this context, the indepen- dent institution would have a mandate legislated by the EU specifying its objectives, such as achieving a certain emission target at least costs. However, this institution would be entitled to choose its instru- ments for interference. Such approaches to delegation most closely resemble the case of independent central banks having signiﬁcant discretion over money supply, while being guided by some core objec- tives such as price stability, which were politically set at their inception. 7 Within this basic framework, many ofthe options that have been put forward to reform the EU ETS can be compared. For example, if the existing EU institutions remain the main entity in charge of managing the market, various options exist. The status quo, where the EU ETS remains a pure quantity mechanism, is the ﬁrst possibility. Set I, with reform options such as a revision ofthe linear reduction factor as suggested by the European Com- mission ( 2014d ), is located in the top-left area ofthe EU ETS Reform Space. This is due to the limited changes in terms of delegation and explicit price certainty. Set II is located with set I, as adding another sector such as transport to the EU ETS, where short-term demand for permits is likely to be more predictable, could have a smoothing effect on price development but remains uncertain. Set III represents a set of options available to exert greater control over the emission price and could be located somewhere between a soft and a hard price collar depending on the speciﬁc design. The soft price collar entails less price certainty and also implies more quantity certainty than a hard price cor- ridor. When implemented by the EU, this option lies in the upper-middle area ofthe EU ETS Reform Space. The hard price collar, on the other hand, is placed in the upper-right corner as it offers stronger price certainty by implementing a strict price ﬂoor and ceiling. Both options lie below the status quo in terms of level of delegation, as the price collar would need to be deﬁned in legislated rules in addition to the existing framework.
Emission Trading North underlined the necessity of a consistent treatment of new entrants (new installa- tions), ofthe replacement of existing installations by new ones and of capacity extensions. In order to avoid incentives that could lead to higher environmental im- pacts and to assure economic efﬁ ciency, new entrants or new installations should not be privileged over the modernisation or modiﬁ cation of existing facilities or over an increase in production. Therefore, the majority of Emission Trading North participants and the expert advisory board favoured a strict buy-in solution for newcomers: any increase in absolute emissions – no matter whether due to an increase in production, ca- pacity extension, newly built plants or the substitution of existing ones – has to be covered by buying the re- sulting deﬁ cit of allowances on the market. As the as- pect of new entrants and shut-downs was regarded as a symmetrical problem of paying or receiving a price for a newly introduced input factor, it was consequent- ly argued that any operator shutting down his facility in a given compliance period should be allowed to keep the total allocated quantity until the end of that period. While the German NAP chose exactly the opposite strategy (end of allocation at shut-down, new entrants receive allowances for free), it maintains a certain consistency by making the transfer of allowances to replaced plants possible and treating capacity exten- sions as new entrants if there has been a technological investment for the extension. However, it should be noted that productivity increases in existing facilities balancing shut-downs of other facilities and leading to lower speciﬁ c and absolute emissions are not favoured by these rules as they could have been, compared to other emission reduction measures.
On the downside, the displacement of JI in the energy and industry sectors by the EU ETS might perhaps backﬁre. It is as yet unclear in how far JI in general will contribute to technology transfer and efﬁciency improvements in the host countries, but it just might. Conversely, the current NAPs are very generous and it seems that the same will hold for the period 2008-2012. The EU ETS’s stimulus for invest- ments will therefore probably be limited. As a consequence, an instrument that might have contributed to increasing en- ergy efﬁciency and emission reductions in the Accession Countries has been replaced by an instrument that at the moment seems rather likely not to contribute, unless alloca- tion becomes much more stringent in the future.
processes are based on horizontal direct and indirect interaction between equal partners as opposed to a hierarchical relationship in which the EU could use its leverage, for example in the form of conditional- ity (Schimmelfennig/Sedelmeier 2004), to coerce a policy change abroad (Knill/Tosun 2009). The US also never ratified the Kyoto Protocol with its binding commitments. Horizontal interaction includes demand- driven processes in which extra-EU actors actively search for lessons from EU policy and the EU remains passive, as well as supply-driven processes in which EU actors actively promote lessons drawn from EU policies. This demarcates the limits of what would be labeled EU external governance. Instances in which extra-EU jurisdictions spontaneously take over EU policy solely based on their own decision and as a result of externalities of EU policy are not considered in the scope of external governance. They are however a substantive part of policy diffusion studies. The concept of EU external governance involves some purpose- ful involvement ofthe EU in the diffusion process and can be described as institutionally backed cases of active regulatory export. External influence through networks and diplomatic intergovernmental relations are included in governance because the EU actively engages in these cases (Lavenex 2011). This paper mainly focuses on learning processes since – as shown below – this was found the dominant transfer mechanism in the case of GHG emissionstrading from the EU to North American states and provinces. The in-depth analysis ofthe micro level processes ofthe transfer of (elements of) the EU ETS contributes to understanding the conditions under which the EU can effectively affect and promote policy developments beyond its borders.
Linking of different trading regimes
22. Problems not only occur when linking an absolute and a rate-based permit trading regime, but also when linking trading regimes that have different monitoring, accounting and enforcement systems. If, for example, one regime does not include adequate monitoring ofemissions, a source could sell unqualified allowances resulting from inaccurate GHG monitoring to others, undermining the environmental integrity ofthe regime (Mullins and Haites 2001). Another problem that is often mentioned is, that if penalties are not comparable across linked systems, non-compliance is likely to be exported to the system with the lowest penalty level. But as Baron and Bygrave (2002) note there are also other factors such as certainty of penalties, other sanctions (e.g. loss of access to market) and registries that might not allow over-selling, so that such a problem may be less evident. Pressure toward harmonisation might stem from competitive disadvantages for firms in trading regimes with higher penalties. In addition, systems with high penalties may not be willing to link to those with low penalties. As both systems gain from linking though this can also be an incentive for penalties that are acceptable to all and for stringent compliance controls. Altogether, standardised monitoring, accounting and enforcement system simplify linking of different permit trading regimes. The standardisation of at least accounting rules is the subject of section 3 below. The CATEP workshops have also shown that wherever new trading schemes are developed monitoring and accounting ofemissions and tracking permits as well as the institutional requirements (registries etc.) are acknowledged as important issues that have to be solved (Blachowicz 2002, Burkhardt 2003, Jilkova et al. 2002). The features ofthe existing and planned different trading regimes are summarised in Table 1 in Annex A at the end of this paper.
damage country is worse off under trading while the low damage country is better off under trading. The aggregate social costs are lower under linking than under autarky.
In the absence of spillovers, the spillover effect vanishes. Moreover, from Corollary 3, we have seen that the aggregate ERT corresponds to that under autarky if α = 0 which is why also the environmental damage effect disappears, leaving the strategic price effect and the efficiency gains from trade to be non-zero. The efficiency gains from trade leads both countries to favour linking. I argued after Proposition 3 that the low damage country tends to set a smaller ERT under linking relative to autarky. Therefore the strategic price effect of this country is negative, causing the low damage country to unambiguously consent to linking. For the high damage country, the strategic price effect is positive and outweighs the efficiency gains from trade. The reason is that under the symmetry assumption regarding the abatement costs, the difference between the ERT under linking and under autarky exactly corresponds to the difference between actual abatement and the ERT under autarky. Therefore, the efficiency gains from trade can not compensate for the strategic price effect and the high damage country will always disagree with linking. In aggregate, the strategic price effects of both countries cancel out, implying the difference in aggregate social costs to be negative due to the efficiency gains from trade. Thus, linking enhances the welfare.
A first insight on the correlation of CO 2 pricing and product prices may be gained from a simple graphical illustration ofthe development of CO 2 and of output prices, e.g. placing CO 2 price trends on the x-axes and product price trends on the y-axes of a chart. While this approach seems straightforward for the electricity sector (with information on daily or hourly spot market prices available), it provides a major challenge for sectors other than electricity. No single price index exists for products that are much less homogenous than electricity. Rather, for most industries, each product has its own market price that may or may not be affected by the EU ETS. Optimally, each of these prices would need to be correlated to CO 2 prices in order to assess the impact ofthe EU ETS. Such an analysis would, thus, require collecting a vast amount of product price data, which most often is not publicly available. Alternatively, aggregate price indices may serve as proxies. However, these data are often reported on an annual basis only and would not provide sufficient data points to perform a correlation analysis. For some products, quarterly and/or monthly data may be available. Apart from the problems associated with the collection of prices on manufacturing products, a correlation analysis would face another important set of challenges. First and foremost, a simple correlation of product prices and CO 2 costs would most likely not provide statistically significant conclusions. A number of simultaneous reactions may lead to changes in product prices which would not allow to single out a CO 2 price based effect. A time series regression would perform well if the main explanatory variables were included. This implies that prices for all variable inputs to production (energy prices, intermediate input prices etc.) would need to be included. In addition, changes in product prices may be driven by developments outside ofthe EU ETS system boundary. High steel demand from China, for example, affects international steel prices and provides a parallel and exogenous driver of product prices in the European Union that would need to be separated from CO 2 based price effects.
Emissionstrading and emissions taxes are usually understood as alternatives for pollution control. Considerable efforts have been invested to study under which conditions quantity controls should be preferred to price controls and vice versa (see, for example, Weitzman, 1974; Milliman and Prince, 1989; Goulder et al., 1999; Montero, 2002; Fischer et al., 2003; Krysiak, 2008). This literature seems to imply that combining emissionstrading and emissions taxes is redundant at best and inefficient at worst. The underlying assumption is that both policies are exclusively meant to address a single pollution problem only. However, real-world policy-making may be more complex. Policy instruments often pursue multiple objectives. In this case, the classic Tinbergen (1952) rule applies: the number of policy instruments has to equal the number of policy objectives. In other words, the simultaneous attainment of multiple objectives under a policy mix may justify a certain degree of inefficiency with respect to pollution control. Consequently, pragmatic analyses of a policy mix need to take multiple objectives into account. They have to be based on a careful evaluation of welfare gains and losses ofthe policy mix in a multi-objective world. Such evaluation first of all requires a proper understanding of interactions in the policy mix. What is their actual impact on the efficiency of pollution control? And what are corresponding policy recommendations that are appropriate in a multi-objective world? These questions have received little attention so far.
Non-participating countries may think it is unfair ofthe EU to unilaterally enforce an emission cap that affects their trade. However, other Annex 1 countries have little reason to complain as they have been asked by the UNFCCC to take action together with the IMO. It is their inaction that is forcing the EU to establish a regional scheme. Non-Annex 1 countries may have better grounds for complaining. However, one should in this respect remember that the ships covered by the scheme are in most cases involved in dedicated trade between non-Annex 1 countries and Member States ofthe EU, and those ships cause as much pollution on their return voyage as on their trips to Europe. It is also important in this context to recall that in 2004, the eight largest Chinese ports accounted for over a quarter ofthe world’s container traffic (ISL, 2005). This trade is a result of commercial deals which result in both consumer and producer surpluses. Finally, one should note that it is proposed that the majority ofthe proceeds from the scheme be recycled to the industry on the basis of GT kilometres sailed. Therefore the net burden for the average ship will be small, although high-emitting ships will, of course, have to pay considerably more than they receive back.
2.3. The Role of Hot-Air
In addition to the project-based mechanism, the Kyoto-Protocol allows the transfer of AAUs (Assigned Amount Units under the Kyoto Protocol) between Annex B countries. As far as trade between countries with a binding cap is concerned, this option is of minor importance since the project credits are perfect substitutes and can in many cases be obtained at lower prices. This is not the case for countries, which do have a cap that is above their expected business-as-usual emissions in 2012. These excess emission rights are called hot-air. The countries with hot-air are mainly the countries ofthe Former Soviet Union and to a smaller degree the Eastern European countries. In an extreme scenario where these countries sell all their hot- air, most models, including DART (Klepper and Peterson 2003) predict that the excess supply of allowances is so large that the carbon price falls to zero. Thus, the Kyoto targets can be reached at zero cost, however without any emission reduction in the Annex B countries. Such a scenario is not very likely though. Different studies have estimated that it is optimal for the hot-air countries to restrict their sales of hot-air to around 40% (Haites and Seres 2004, Klepper and Peterson 2003). If some ofthe hot-air is supplied on the market, the use of CDM and JI credits will be reduced and international carbon prices will fall.
To answer these questions, we develop an analytical model of a credit-based system that takes potential overreporting of emission reductions into account. We focus on the optimal de- cisions ofthe regulator, given the impossibility to fully enforce compliance, under asymmetric information on reduction costs and heterogeneous veriﬁability of projects. We show, given a limited monitoring budget, a rational regulator will completely refrain from monitoring those projects that are most diﬃcult to verify. For larger monitoring budgets, the optimal moni- toring strategy can be discontinuous, featuring a jump within the set of projects with lower veriﬁability. Furthermore, for those projects in full compliance, the monitoring pressure reduces with increasing veriﬁability ofthe projects. For cases with intermediate veriﬁability, optimal monitoring pressure is ambiguous. For these levels of veriﬁability, we identify conditions for which monitoring pressure is either at its maximum or for which there exists a ‘U-shaped’-style monitoring policy.
by the European Commission has been necessary to ensure the effectiveness ofthe system and sufficient consistency. Following a review ofthe system (ETS Review), serious changes were agreed upon and will be implemented with the start ofthe third trading period in 2013, including a longer trading period, increased auctioning, and the establishment of an EU-wide cap. In this connection, changes have not been implemented piecemeal, but rather consis- tently and in a single package. This is particularly true with regard to the newly implemented directive regarding renewable energy as well as the decision on effort sharing in sectors currently not covered by theemissionstrading system. With these changes, the aim is to reduce greenhouse gas emissions by 2020 in Europe to 20% or 30% below their 1990 levels, depending on the result of international ne- gotiations.
The widespread use of quantitative limits on the use of offsets from linked credit systems appears to be motivated—at least in part—by the desire to establish domestic abatement incentives. Table 3 illustrates that most systems place relatively strict limits on offset use. Except for New Zealand, which allows for unrestricted use of credits from Kyoto programs, no system allows more than 20 percent ofemissions to be covered with offsets, and few allow more than 10 percent ofemissions to be covered. Although this doubtless reflects the perception that offsets may not represent real and permanent emissions reductions, it is also motivated by the stated goal of many systems of achieving domestic emissions reductions. A good example is provided by the recent decision by the EU ETS to cease accepting new offsets from CDM projects located outside of Least Developed Countries (LDCs). This change in policy is due to Article 11a(4-5) of Directive 2003/87/EC, which was triggered automatically by the absence of a world climate agreement by the end of 2010. The decision effectively excludes China and India from selling CERs to the European market.
On 15 th of July 2015, the European Commission released its proposal for a reform ofthe EU EmissionsTrading System (EU ETS). The reasons for a rather thorough reform are obvious: First, the mechanism of supply and demand is not working properly as the market for emissions allowances currently exhibits a surplus of more than one year’s emissions, thus causing the price for carbon to be very low and as a consequence postponing investment decisions into carbon reducing tech- nologies. Second, the EU ETS needs major adjustments regarding the long-term perspective of a possible carbon path that is relevant for low-carbon investment decisions under risk. Third, the EU ETS needs to be aligned with the 2030 frame- work for climate and energy policy, which in turn has to be updated in view ofthe Paris Climate Conference that took place in December 2015. Although the Commis- sion proposal addresses all these issues, there is emerging evidence that only addi- tional reform steps will bring the EU ETS back to its intended role of becoming the cornerstone of EU climate policy.
In addition to garnering a better understanding of authoritative knowledge making, there is more to be gained from STS’s elaboration ofthe performativity of epistemic representation. One concept that is particularly relevant for concerns with governance and politics is a generic conception of power as an effect of representation: The entity (or person) who “does” a representation assumes the power to enact the represented in a particular way; it is actualized and brought into being in a particular situation as the thing that it is represented to be. Yet, there is no representation that does not entail a transformation but any representation involves the selection and subsuming of some aspects of what the entity is or maybe of how it is said to be. The power to perform this transformation is at the heart of a concern with “mediators” in actor-network theory (ANT) and it is central to the analytical project to make the mediation, the transformation and thus the power that is connected with the practicing of representations visible. The basic concept is contained in various ANT terms like blackboxing, spokespersonship, point representations (Callon and Latour, 1981), obligatory passage point (Callon, 1986), centre of calculation (Latour, 1987, p. 215-48), or oligopticon (Latour, 2005, p. 174). Such a generic understanding of representational power can be specified for areas of social life other than science and its claim to represent nature.
This paper focuses on the review of existing economic models regarding their potential application for a quantitative assessment ofthe effects of linking initiatives. As a matter of fact, most ofthe economic assessment criteria to evaluate the effects of a bilateral linking of ETS cannot be measured empirically, but need to be modelled by aid of economic modelling. The various models that are available in the market have been developed over the past years and decades, with partly very different economic foci and level of detail. This paper examines different modelling approaches to quantify the impact of link- ing Emission Trading Schemes (ETS) on those selected economic assessment criteria that cannot be measured with empirical data. The aim is to give a comprehensive overview on selected economic models suitable for assessing the economic effects of linking. To this end, eleven economic models that were principally deemed suitable for analysing the economic effects of linking have been assessed: six CGE models (Aim-CGE, EPPA(-EU), GEM-E3, G-cubed, IMACLIM-R and PACE), one macro-econometric model (E3ME) and four PE models (POLES, PRIMES, REMIND-R, TIMES-M.). Five models will be de- scribed in more detail in this report (E3ME, GEM-E3, PACE, POLES, TIMES).
As part of its unilateral climate change policy, the EU has established a series of policy targets it aims to meet over the coming decades. What is more, the EU is determined to keep its emissionstrading scheme for CO 2 as its flagship climate policy instrument. Over the past two
decades, tradable permit systems have become a well-established policy instrument for regulat- ing environmental externalities. Emissionstrading has been credited with substantially reducing the costs of environmental regulation (Ellerman et al., 2010). However, practical experiences with trading schemes for conventional pollutants have revealed some unintended consequences that arise from the spatial distribution of pollution. With a uniform permit price, the single cri- terion for the allocation of pollution in space is the marginal abatement cost. If the market shifts pollution from low-damage regions to high-damage regions, this may create large inefficien- cies that interfere with the goal of efficient environmental regulation (Muller and Mendelsohn, 2009). Moreover, the redistribution of pollution to places with low-income populations may exacerbate inequality (Fowlie et al., 2012). At a glance, these issues may seem irrelevant for the EU ETS because CO 2 is a harmless gas with no known local impacts. However, if ancillary
none of them doubted the effectiveness ofthe fundamental design ofthe EU ETS and did not aim to abolish and replace it with another instrument.
The policy process ofthe EU ETS revision was shaped by the then forthcoming UN climate conference in Copenhagen and the oil and gas concerns ofthe new EU member states of Eastern and Central Europe. As a consequence, the EU ETS Directive became part of a newly implemented climate and energy package. Four other directives were also part of this package and were negotiated in parallel. Moreover, the then forthcoming climate summit in Copenhagen caused a sped up policy process. This meant that the Commission‟s proposal did not have to pass all stages ofthe co-decision legislative procedure. Furthermore, the European Council participated in the process, something that is very uncommon. On the one hand, this had the positive effect that the revision was highly supported by the heads ofthe member states. However, on the other hand, the Commission‟s proposal was in a way watered down and the Parliament‟s opportunities to influence the policy making process were weakened. The economic and financial crisis considerably changed the attitudes ofthe actors. During the time ofthe negotiations about the climate and energy package, the support for climate policy was very high, something that significantly changed with the crisis. Economic issues became the highest priority on the political agenda and other political issues were delayed. This was also the case for the EU ETS, whose ineffectiveness became clear. The economic crisis had led to an automatic reduction ofemissions and thereby an oversupply of allowances and a decline in their prices. Moreover, the crisis resulted in a greater split between the member states of Northern and Western Europe and the member states of Central, Eastern and Southern Europe. Furthermore, the crisis created a feeling of disappointment around environmental politics as the ineffectiveness ofthe revised EU ETS became visible. Although the economic and financial crisis in the EU can be considered as being caused by neoliberal policies, this did not lead to a turning away. The EU tried to counteract the crisis with further neoliberal instruments. This was also the case when dealing with the problems ofthe EU ETS, as further neoliberal measures will be added to it.
assumptions rarely hold in practice. Indeed, emission permits markets can suffer from several impediments such as uncertainties, transaction costs 5 , market power 6 and cheating behaviors 7 .
In this paper, we focus our attention on uncertainty. Large scale experiences have shown that well designed markets minimise transaction costs, cheating behaviors, and the risk ofthe exercise of market power, but do not succeed in reducing the various sorts of uncertainty that firms may face in such markets: permit price uncertainty, demand uncertainty which means production and emissions uncertainty, abatement costs uncertainty and regulatory uncertainty among others. A number of researches have already analyzed the role of uncertainty in emission permits markets. The first conclusions come from experimental economics. In different experimental settings , and  show that uncertainty faced by regulated firms regarding their total emissions creates price instability, which is higher when banking is not allowed. Moreover, price peaks are higher in high rate emission periods. In a theoretical and numerical paper about marketable permits,  analyzes the effects of trade approval and transaction cost uncertainties on market performance and aggregate control costs. Although uncertainty and transaction costs suppress exchanges that