Pure Cap-and-Trade

For all carbon pricing a consideration of primary producers is to reduce the risk of future very low product prices, bringing their resources on the market at higher volumes, and hence - collectively - lower prices now, see (Sinn 2012) for the reasoning involved in this paradox. (Edenhofer and Kalkuhl 2011) expand outcomes of the paradox to more precise assumptions and (Michaelowa 2012) gives a critical review indicating that the base assumption of equal cost for all production is highly unrealistic. Investment in most expensive new production capacity will be reduced by emission pricing, reducing the incentive to increase production now. (Harstad 2012) sees this leakage as undermining climate coalition efforts and proposes that the coalition buys all fossil resources to keep them in the ground, globally. This seems fully impossible, given the undefined nature of what constitutes resources and the extreme prices and related transfer payments that would be involved. More realistic leakage modelling methods have been developed by (Hoel 2011a, Hoel 2012) with more diverse outcomes. In their review on the subject (Fischer and Salant 2013) conclude that the paradox effect will be limited, if present at all. Other mechanisms induced by lower prices also due to climate policy may play a role however. The OPEC (Saudi) reaction to low oil prices, hardly induced by effective climate policy, has been to increase production to press US fracking production out of the market, accepting even lower prices. Such imperfect market mechanisms will remain.

Overall conclusion is that slowly but predictably rising prices reduce the cost of emission reduction, with a revision of the level after maybe two decades, to avoid underperformance or too substantial overperformance. Starting low and rising predictably has proven politically feasible in the UK price floor tax on CO2 emissions linked to ETS and the Canadian emission tax in British Columbia.

Table 5 Proceeds of emission pricing: emission price starting at 100€/ton in 2020 rising by 10€ per year

the pure-cap carbon pricing option is reduced by a fixed amount per year, legally established and binding for the total of the member states.

In the ETS (intended) pure cap-and-trade system the cap is fixed, with a fixed yearly percent point reduction relative to the base year 1990. It is now set at 1.74% per year and intends to rise to 2.2% per year in 2021. The long term reduction rate of 2.2% implies a rising reduction rate on a per annum percentage basis, extremely high towards 2050, rising to well over 10% reduction per year on a compound rate basis. As some emission domains are difficult to reduce very substantially, like iron and steel and aviation, most other activities would have to reduce even faster. A more modest reduction rate when the goal is near might be a better option. The current compound reduction rate of around 1.5% per year would have to rise to roughly 5% within the next decade to arrive the same 2050 target¹³. A reduction rate of 5% at a growth rate of 1.5%

would imply a reduced emission intensity of 6.5% per year or, formulated as the inverse, an eco-efficiency improvement by 6.5%.

For guiding long term decisions on decarbonization, the core question to answer is what the permit price will be in the near and far future, that is what price development might reasonably be expected. Answering that question must be based on estimating the factors determining demand for permits, as the supply is purely inelastic. The main factors determining demand are economic growth; the price development of fossil energy sources; cost development for low carbon production technologies; the system costs induced, substantially depending on market structure and the cost and market development in other sectors; the role demand spreading can take in the internet of things; the costs to be induced for decarbonizing the energy use chain and its speed of implementation; and the cost for energy storage systems, both for electricity and heat and other options; the volumes of iron and steel use and of aviation, and a few more. Most of these factors are highly uncertain, and can be brought under policy control to a very limited extent only, with the exception of the development of electricity markets. When explaining the price drop in the EU ETS, is a bit of everything (Feng, Zou et al. 2011) (Koch, Fuss et al. 2014), a difficult to disentangle mix of low economic growth; shift of material production to Asia, especially China; substantial CDM,; the upcoming of national climate policies like widely used feed-in tariffs; and technology development also induced by Chinese and US renewables policies.

Overall however, permit demand development is highly unpredictable. One thing is for sure: not only is supply fully inelastic in a pure cap system, also demand is highly inelastic in nature. On short and medium notice technologies are mostly given, so only market volumes can adapt to price changes of emission permits, with highly uncertain developments. This implies that also price development is highly unpredictable, not the one on short notice due to trading effects, but the longer one, over years and decades However, short term price fluctuations will also blur the long term picture somewhat, increasing at least subjective uncertainty.

Empirical analysis is available on substantial also longer term price volatility of the US SOx, and later also NOx cap on utilities, the major example for the EU ETS. There is some analysis on factors behind EU ETS price volatility, but hardly linked to time frames, see (Meng 2013) for the analysis of some factors, and on broader references. For energy, the long term elasticity of demand has been in the order of minus one, see (Lowe 2003), who analyses also long term expected

13 The cap reduction rate set by the Commission is now at 1.74% (of 2005 emissions), rising to 2.2% (of 2005 emissions) in 2020. This implies a rising compound emission reduction rate.

development of price elasticity of energy demand under effective climate policy. The emission permit price elasticity is different from energy price elasticity, more elastic long term as then shifts in technology use can be realized. Long term predictions of price elasticity of permit demand seem quite difficult however, and no predictions seem available. The more inelastic range (long term) might go to minus 0.2, let us assume as an option. This is not a prediction but an option a pure cap system would have to reckon with. Combining fully inelastic supply with this range in elasticities of demand indicates what the price effect of a disturbance could be. With a 10%

increase in demand, the unitary elasticity we mostly are used to in other domains would lead to a 10% increase in price. The same demand increase with a price elasticity of -0.2 would lead to a 50% price increase, see Figure 2. The figure conveys the mechanism but cannot be based on empirical data for the basic reasons mentioned. The converse would be true for a demand decrease by 10% leading to price decrease by 50%. Of course elasticities are not known or even fixed, depending on the time horizon and certainly are curves, not straight lines. But the message is clear, non-predictable factors can lead to very substantial price changes over a period of years.

Figure 2. Demand and supply for emission permits under pure cap-and-trade: graphical analysis.

The elasticities not only are hardly known, they will not be stable.

Substitution as a major factor depends strongly on the time horizon. Shifts of demand to other technologies and sectors are easier in the long term, as when investing in heat pump based heating systems to better use intermittent renewables. Such mechanisms in turn are also highly dependent on the future market structure in the electricity domain. In the long term elasticity then will be higher than on the short term. However, other mechanisms work towards reduced elasticity in the longer term. Picking low hanging fruit can only be done once; options for changing technologies tend towards increasing costs; and shifts in demand are limited as by given infrastructure and given tastes. The net effect on long term price elasticity of demand may therefore be in a broad range, with expected volatility in a broad range as well.

p

p₁

supply fixed q₁= q₂

p2u unitary demand

p2i inelastic demand permit supply:

fully inelastic

demand shift

q

This is the main long term weakness of the pure cap-and-trade system: there is no good basis to predict the long term price development of permits. This implies that there is no good price expectation possible for guiding public and private R&D and neither for long term investments towards a low emission society. ‘The permit price has collapsed before, so be careful.’

Substitution effects depend strongly on the time horizon. Shifts of demand to other technologies and sectors are easier in the long term, as when investing in heat pump based heating systems to better use intermittent renewables. Such mechanisms are also highly dependent on the future market structure in the electricity domain, see Section 4.4. In the long term elasticity then will be higher than on the short term. However, other mechanisms work towards reduced elasticity in the longer term. Picking low hanging fruit can only be done once; options for changing technologies tend towards increasing costs; and shifts in demand are limited as by given infrastructure and given tastes. The net effect on long term price elasticity of demand for emission permits may therefore be in a broad range, with expected volatility in a broad range as well. Price predictions can only be highly uncertain.

The pure cap-and-trade system is to be a major instrument in an instrument mix. Other instruments in the same domain, like national policies, local initiatives, and private contributions and innovations may seem interesting from their micro levels but don’t contribute to emission reductions, see (Sorrell and Sijm 2003, Sijm 2005) for an early convincing analysis of this relation with other instruments, and (Agnolucci and Drummond 2014) Chapter 5 for a recent literature survey with the same conclusions. The demand for permits is reduced, the demand curve shifting down along the vertical permit supply curve, see Figure 2. The lower price of permits just makes it cheaper for others to emit, these then emitting more. The relation also holds with other EU instrumentation. If a well-functioning electricity market is introduced to allow for easier introduction of renewables, that effect may occur, but the emission reduction not. Lower emission prices will make fossils so much more attractive that the cap is filled. This is an unavoidable characteristic of the cap-and-trade system being pure, and is not based on any empirical analysis or experience. This makes ingenuity and creativity in emission reduction superfluous from an emission reduction point of view. If, for climate reasons, you introduce an automat to lower your thermostat already before leaving your house, you just are silly. This demotivating characteristic is a main obstacle for the long term functioning of a pure cap-and-trade system. If all other instruments together are effective enough for to create the emission reduction of the cap, the permit price will collapse, making the pure cap scheme superfluous.