Pricing concepts

The long term quantitative climate policy target, above 80% emission reduction⁸, allows for different approaches to pricing, from emission tax and different versions of cap and trade, to using in reference price in using other instruments. Emission pricing by whatever means may relate to very different pricing concepts, though there is some correspondence with specific instruments. Behind all is a relation between emissions and some damage concept, fully quantified or not.

Putting a price on emissions may be done in different ways. One is to specify the social cost of carbon, indicating the value of damages a unit of emissions induces. There is a rich literature on this subject. In the domain of economics, see (Tol 2008) for a somewhat limited view, and (Stern 2008) taking a broader view, also on discounting options. There is not a clear outcome, for quite basic reasons: the damages are uncertain (Heal and Millner 2014) and spread out in time; the methods of discounting are varied and disputed and will remain so, see the survey by (Portney and Weyant 1999); and catastrophic effects are not amenable to the usual economic valuation methods (Weitzman 2009). Solutions are sought for but don’t give single outcomes, see for example (Dasgupta and Maskin 2005, Tol 2008, Hof, van Vuuren et al. 2010, van den Bergh and Botzen 2014). This pricing approach links to (Pigou 1932 (1920)), who advocated taxes as a policy instrument to internalize external costs into market considerations. The lack of consensus on the social cost of carbon is one reason why this version of emission tax has not clearly emerged.

Quantification is arbitrary, given the very substantial spread in outcomes. A second price concept involves a practical solution to the quantification problem, by setting a price required to arrive at an intended emission level, first proposed by (Baumol and Oates 1971, Baumol 1972). This takes the social cost issue to the political domain, when emission (reduction) targets are set. This approach is behind the caps set for tradable emission permits and also for setting emission tax levels, as discussed below. Both the Pigovian and the Baumol approach are in line with the OECD Polluter Pays Principle (OECD 1972), repeated in the Rio Declaration 1992. The Baumol line links with the major options, cap-and-trade and emission tax.

The analytics for different options for emission pricing are surveyed in Table 4. Pricing systems may be set up in covering fossil CO2 emissions only partly, in the sense of not covering all fossils, and for the fossils covered not covering all sectors and all firms, as now is the case with the EU ETS. Both cap-and-trade and tax can be expended to cover all fossils in all sector domains, including consumption and other final demand. In a long term strategy there are no good reasons for partial application of emission pricing so the full coverage is applied in both pricing systems.

8 The 2020 and 2030 targets of the EU are here for the short and medium term.

This position is reinforced by the administrative characteristic of CO2 emissions: They can hardly be measured at actual source and can be most easily measured fully at source, then covering all fossil carbon. The administrative level may vary. An early survey of options is in (OECD 1992) by (Victor 1992), p.244. Victor distinguishes four possible loci of administrative application, with points of carbon taxation at primary production; at key transport locations; at processing (refineries, power plants); at final sale to end-users (fuel sales and electricity delivery). The direct option, of pricing at source of emission, has been left out, as direct emission measurement was deemed too complex. Though lacking a full system perspective, main corresponding elements as specified here are visible, see Table 4. The option of taxing at end-use has been extensively discussed but falters at the lack of an accepted method of allocation of upstream emissions.

Allocating refinery emissions over all its products; power station emissions over heat and power;

and dairy emissions over meat, milk, butter and cheese is done in very different ways in different studies, with ISO 14040ff retreating from a choice. See (Guinée 2002) for a deep treatment of this allocation issue and (Guinee, Heijungs et al. 2004) for details and examples on the most generally applicable method, economic allocation. Lack of agreement on methods makes emission pricing at the final product level not feasible. Tools for carbon footprint specification as have been developed give a rough indication, not accepted facts to base pricing on.

Current ETS applies at the level of installations, based on the administration of the firm owning or operating them. It is the firm administration of inputs of fossil carbon, excluding biogenic carbon, not actual emissions. There is a double relation with sectoral domain. Smaller firms and households lack an administration of the fossil inputs into their actually emitting installations and it would require a disproportionate private and public administrative effort to cover them. Going up at least one level, to their suppliers, shifts the administrative application to different firms and sectors. At that level at least that of the suppliers, all downstream installations are covered implicitly. But why should it stay there, if a move higher up the supply chain would resolve the intricacies of overlapping flows, like coal used for electricity administered at a different place than coal used in industry and households? The move to upstream application has its logical end at primary production in the EU and imports of fossil energy products, reducing the number of regulatees very substantially, even as compared to the currently limited application of the ETS, covering only 45% of emissions by larger firms only, already in the tens of thousands. The move upwards in administrative level is applicable to both cap-and-trade and emission tax. The third dimension is how prices are set. The situation for CO2 is very different from that of SOx and later NOx, where the US EPA set up a price direct emission measurement system⁹ for larger emitters, who then had to purchase the corresponding emission allowances. The current ETS is a near pure cap-and-trade system, with the cap volume of emission allowances (see on the term (Schmalensee and Stavins 2013)¹⁰) adjusted only incidentally to keep the price up to more relevant levels. The Market Stability Reserve does not refer to price levels but only to oversupply.

However in the discussions in the EU Parliament the motive for accepting the MSR was “in order to support the price” (EP-PressOffice 2015). From a long term perspective it seems a curious position to not explicitly state the price goal when adjusting the system to that effect. A clear

9 CEM: Continuous Emission Measurement system.

10 The term allowance was introduced in the US SO² trading system because in US administrative-legal text the more common term permit had a different meaning already. The term permit is used here, as is common in economic and regulatory literature.

choice is due: stabilize the price, or not. If not, there is the pure cap-and-trade system, with advantages and disadvantages as discussed below. If the price is stabilized, with a floor and a ceiling, there is a clear cap-and-trade system as well, a hybrid one, also with its advantages and disadvantages. For the emission tax system the price is fully fixed, determined exogenously.

Table 4 Sectoral domain, administrative level and price stability of emission pricing instruments

Sectoral Domain Administrative Level Price Stability/Predictability

1. ETS now: electricity production, larger industries (+ some non-CO2)

1. ETS now: level of emitting firms and installations

1. ETS now: price dependent on volume, MSR to increase price coming 2019

2. Also natural gas in small firms and households heating

2. With electricity producers and energy distributors (for gas, possibly also some coal)

2. Price floor indicated (rising) floor, by reducing permit volumes in the market 3. Also all other natural gas 3. With primary producers and

wholesale traders, firm level

3. Bandwidth specified, indicative midlevel, permit volume following

4. Also all other coal 4. With primary producers coal and gas and importers, firm level

4. Bandwidth zero, long term specified permit price level, volume following

5. All coal, oil products and natural gas

 All fossils

5. With primary producers and importers, product flow basis

5. Bandwidth not an item, fully specified rising excise/tax level

The three dimensions distinguished in Table 4 are not fully independent. As indicated, moving upstream in the administrative level will by near necessity correspond to an upward shift the sectoral domains covered. The step to include all coal seems trivial, as most coal is covered under the ETS already. However, with a rising price of alternatives coal becomes attractive, with multi-fuel stoves also for coal coming in the woodstove market now already, next to wood pallets, etc.

The third dimension is more independent. One might try to implement a fixed price system at an intermediate administrative level to some but not all firms. There certainly is no good strategic reason for that, creating partial implementation, creating unnecessary cumbersome administrative implementation, and requiring other instruments to fill the gaps left. The pure cap at level 1 and the pure encompassing emission tax at level 5 are clear options. Price floors and bandwidth by adding ceilings may keep the price at deemed relevant levels. But if relevant price levels are the aim, not the emission level, then a move to a fixed (rising) price level, bandwidth zero, is the logical next step.

The current near pure cap-and-trade ETS system - fully applied to all fossil emissions on the road to 2050 - could first be transformed into a price stabilized system, then narrowing the collar between floor and ceiling in next steps. When floor and ceiling touch, the cap system comes close to an emission tax in pricing but not yet in administrative set up. In a last step the permit system can be replaced with an excise/tax system, functionally the same but with a much simplified administration. For all systems the administrative implementation can be by member states and the revenues of emission pricing may go to the member states. This is not fully the case now with the ETS, where auctioning is centralized, controls are decentral, and revenues remain partly at the EU level.

In a long term design of instrumentation a mechanism is preferably built in administratively, where the controlling body has the clear interest to get in what is due, and the paying bodies have

an interest in not paying more than is due. If an administrator has no disadvantage in overlooking emissions, and the firm involved deserves support, not seeing is a reasonable mode of operation, to be kept in check with forceful internal administrative procedures, backed up politically as they are unpleasant and costly. With opposed interests between regulatee and the tax-type receiving administrator they can come to terms on the best established carbon flow specification. Such a mechanism is not present in the current EU ETS, making that pricing system vulnerable in its administrative controls. See for such administrative issues for example (Künze 2012). The US experience in the cap-and-trade for SOx indicated that trust in others’ emission reporting was an important factor for success, see (Stone and Zaelke 2005) and (Kruger and Egenhofer 2005) on the essential role of reliable compliance in the SOx case. The development by the EPA of the CEM (Continuous Emission Measurement) system played an important role there, showing to all parties involved, a few hundred only, what the others emitted. In the US that cap system was halted due to political-administrative problems, see (Schmalensee and Stavins 2013).

Administrative implementation remains always open to limitations, fraud and misuse, requiring permanent controls and adaptations. Moving the administration upstream eases implementation. The ETS also has shown options for fraudulent tax evasion, for the years 2009-2010 estimated by Interpol to be in the order of five billion (5*10⁹) Euro see (Interpol 2013). Those options have been reduced, and perpetrators punished, also the banks that allowed the fraud to come up. However, with a non-mass product with different legal surroundings in different countries fraud remains possible see (Interpol 2013) for some insights, and will become more attractive with rising emission permit prices. Non-fraudulent tax evasion (like tax routing) will always remain an option. This is inherent in the trading system, including derivatives trading. Zero bandwidth cap and emission tax don’t share this characteristic.