The journal Energy Policy has recently published a paper by my colleague Frank Jotzo and myself:

Wood, P.J., Jotzo, F., Price floors for emissions trading. Energy Policy (2011), doi:10.1016/j.enpol.2011.01.004

The paper (as well as this blog) proposes that one way that a price floor could be implemented is for emitters to pay an additional fee or tax per tonne of emissions. The carbon price is then equal to the sum of the ETS permit price and the extra fee. The UK government has proposed to introduce a carbon price floor via this approach, and has been engaging in consultations. The proposal is the reform the Climate Change Levy so that it functions like a carbon tax. Because the UK is part of the EU ETS, firms would also pay for EU permits, and so the effective UK carbon price will be equal to the sum of the Climate Change Levy and the EU ETS permit price.

The discussion paper includes three different “illustrative carbon price scenarios” of £20/tCO2, £30/tCO2 and £40/tCO2 which is somewhat more ambitious than likely to be proposed for the carbon price in Australia, or the price floor that was proposed in the Waxman-Markey Bill.

Because most EU emissions are determined by the EU ETS, the direct effect on global emissions is likely to be minimal. Emissions in the EU are determined by the cap. If the whole EU ETS had a price floor, and the floor price was met, then that would reduce total EU emissions; but when a single country has a price floor, overall emissions are unchanged. For this reason Climate Strategies has made the important point that the UK should embed its policy in a strategy to strengthen EU emission reduction targets.

What the UK proposal does do is provide ‘learning-by-doing’ on carbon pricing, which provides valuable information to other jurisdictions that may consider a carbon price. A UK price floor proposal is consistent with a polycentric approach to climate change. It also provides much more certainty about the carbon price for investors in emission reductions. By eliminating the risk that the carbon price will go below a particular level, the cost of investing in emission reductions is significantly less.

The UK proposal has attracted a storm of controversy. It will mean that polluters will have to pay more, and steel-makers have already started to complain. This is to be expected – if firms can shape government policy to reduce their costs, then their investment in shaping policy could have a huge payoff. This is why rent-seeking is such a big issue in climate policy.

What was less expected was the opposition from two environment groups: the WWF and Greenpeace (presumably the UK branches of these organisations). They have claimed in a media release that because a price floor will raise electricity prices, and nuclear generators do not have significant emissions, their profits will increase, which will make a “mockery of the Coalition government’s stated opposition to any form of public subsidy for nuclear” and “this is yet another taxpayer handout to a failing nuclear industry.”

Any carbon price will increase the profitability of nuclear energy, just like it will increase the profitability of renewable energy or energy-efficiency. A carbon price is technology neutral and the claim that it is a subsidy or taxpayer handout for the nuclear industry in completely ridiculous. This proposal is good policy and the WWF and Greenpeace should be supporting it rather than attacking it.

For more on price floors, see


In Phase I of the European Union Emissions Trading Scheme (the EU ETS), between 2005 and 2007, the carbon price collapsed. In April-May 2006, the price for a permit to emit one tonne of carbon dioxide collapsed from over 30 euro to less than 10 euro. The carbon price then eventually declined to less than 0.10 euro by September 2007. The primary reason for this was that too many permits were allocated — the size of the cap was higher than the total amount of emissions. Has Australia learned from what happened? Is the carbon price likely to collapse during the initial phases of the Carbon Pollution Reduction Scheme (CPRS)?

This question is pertinent because the Australian House of Representatives economics committee has been instructed conduct an inquiry into the choice of emissions trading as the central policy to reduce Australia’s carbon pollution. A low carbon price is inefficient because there is a lost opportunity to make deeper reductions in greenhouse gas pollution. This means that deeper reductions will need to be made later (which will cost more), or impacts from climate change will be greater (which will cost significantly more). This issue applies, in a much worse way, when there is no carbon pricing at all (as is happening now).

Phase I of the EU ETS has not been the only time that cap and trade schemes have experienced low carbon prices. In the north-eastern states of the United States, an emissions trading scheme called the Regional Greenhouse Gas Initiative (RGGI) has commenced. The RGGI covers emissions from electricity generation, and allows for the purchase of offsets. It aims to stabilise emissions at 2002-2004 levels by 2015, and then reduce emissions by 10% by 2020. Permits are auctioned, and there have been two auctions so far, the emission permit prices were US$3.07 and US$3.38 per tonne of carbon dioxide.

One “design feature” of the CPRS is that Australia’s emissions trajectory is set 5 years in advance (see Chapter 4 of the White Paper). There is also a target range for emissions reductions to be achieved by 2020 to be 5-15% below 2000 levels (Policy position 4.2). The target range is perhaps the worst aspect of the CPRS White Paper, because an unwillingness by Australia to reduce emissions beyond 15% undermines a good comprehensive international agreement to reduce greenhouse gas emissions. The initial trajectory is given in Policy position 4.5 of the CPRS White Paper. I have converted these figures from percentages to megatonnes of greenhouse gases (carbon dioxide equivalent).

The first indicative national emissions trajectory will be:
• in 2010–11, 109 per cent of 2000 levels (602.5 Mt)
• in 2011–12, 108 per cent of 2000 levels (597 Mt)
• in 2012–13, 107 per cent of 2000 levels.(591.5 Mt)

How does this trajectory compare with what Australia’s emissions will be in the absence of of the CPRS? If the trajectory is higher, then the carbon price is likely to be very low (although this is affected by intertemporal flexibility measures, such as banking).

Projections for Australia’s emissions in the absence of of the CPRS are supplied have been estimated by the Department of Climate Change here. They project Australia’s emissions during the 2008-2012 Kyoto commitment period to be 599 Mt CO2-e. Note that this is below the CPRS trajectory for 2010-2011, about the same as for 2011-2012, and slightly higher than for 2012-2013. The projections include uncertainties in emissions for each sector, they range from 3% to 10%. They do not include uncertainties in emissions from land use change or from forestry. Aggregating these uncertainties is difficult, because emissions in different sectors are not independent. There can be covariance between different sectors. If we assumed the sectors were independent and ignore uncertainties in land use change and forestry, then the uncertainty in the projected emissions is 12-13 Mt CO2-e (about 2.2% of projected emissions). If the emissions in each sector are highly dependent, and we take into account the uncertainty from land use emissions, then the amount of uncertainty could be several times higher.

The Treasury modeling predicts that a 5% 2020 reduction scenario would be consistent with an initial nominal carbon price of $23 in 2010. Since the modeling was released, the global financial crisis has gotten worse and the global carbon prices have dropped significantly. The boom in Australia’s emissions intensive resource industries has also ended. The government has also unveiled a stimulus package that includes $3.9 billion in ceiling insulation for households and incentives for solar hot water.

There have been some very serious declines in the price of carbon as the global financial crisis has been unfolding. In the EU ETS, the carbon price has declined from around 30 euro in mid-2008 to 8-9 euro in February 2009. Credits for clean development mechanism projects (known as certified emission reductions, or CERs) have also declined in price, and are trading at less that 8 euro (around A$15). The price of CERs functions as a cap on the carbon price in the CPRS, because firms will be able to purchase an unlimited amount of CERs to account for their emissions (Policy position 11.5).

One thing that will provide stability for carbon prices is the banking of permits. The CPRS White Paper proposes that “Unlimited banking of permits will be allowed under the Scheme (except those accessed under the price cap arrangements)” (Policy position 8.2). Firms with extra permits can always hold them until a later year, and exercise them then. This means that even if there is an overallocation of emissions, there still will be some demand for permits if there is the expectation that there won’t be an overallocation of emissions in subsequent years. In Phase I of the EU ETS, there was banking of permits allowed, but not between phases. At the end of Phase I the permit price declined to less than 0.10 euro. When there is overallocation, banking spreads the impact in carbon prices across several years, instead of confining it to one or two years.

Because of unlimited banking of permits, the carbon price in the early years of the CPRS will also be affected by the medium term trajectory. The target range of 5-15% that has been set for Australia is not very steep, and is not consistent with what science and equity arguments suggest would be appropriate for Australia. Business as usual projections for 2020 emissions are available, but are even more uncertain than for 2010. Australia’s unconditional target is likely to be easier to meet than the EU’s unconditional target, because Australia’s high per capita emissions imply more opportunities for abatement.

At this stage it is too early to know for certain whether there will be an overallocation of emissions in the initial years of the CPRS, but is is very likely. In this sense, Australia appears not to have learned the lesson of Phase I of the EU ETS. International developments alone suggest that the carbon price is likely to be significantly lower than the price suggested by Treasury modeling. Banking of permits means that the carbon price probably will not be as low as the end of Phase I of the EU ETS. The US RGGI also has unlimited banking of permits, and that has not prevented the carbon price from being very low. It is therefore very likely that unless significant changes are made to the CPRS, it will start with very low carbon prices.

What changes need to be made to the CPRS to make it more effective and efficient? What changes need to be made to put in place long-term incentives for investment in clean energy and low-emission technology? Clearly the targets need to be tightened, the unconditional target could be significantly stronger. The conditional target only serves to undermine climate negotiations by ruling out Australia playing its part in a global effort to reduce greenhouse gas emissions. It is not appropriate for the government to rule out any level of emissions reductions.

There are also structural changes that should be made to the CPRS. The best way to rule out downward price volatility is by introducing a price floor. This could be implemented by having firms pay an extra fee when they surrender their permits, or could be implemented by having a reserve auction price. By having a price floor (and making it sufficiently high) we get many of the advantages that we get from a carbon tax. The CPRS also should not allow firms to purchase unlimited amounts of CERs, this is also because of credibility and additionality problems with the CDM, as well as the effect on prices. If the CPRS allows banking of permits, then for each permit banked, the regulator could reduce the cap by that amount of permits in the following year.

The overallocation issue suggests that permits should not be treated as property rights. This has lead to problems elsewhere, such as with the overallocation of the water in the Murray River. It is also inappropriate to create and allocate property rights that potentially infringe on the rights of future generations to a clean atmosphere. Emissions permits should instead be treated as limited compliance instruments. This suggests that Policy position 8.1 of the White Paper needs to be changed.

We argue that the best carbon price signal is a cap and trade system with a price floor. The floor can be maintained by having firms pay an extra fee when they exercise their permits, based on the amount of their emissions. The carbon price then becomes equal to the sum of the permit price and the extra fee. This is similar to having emissions trading scheme combined with a carbon tax.

Stern described climate change as the greatest of market failure we have ever seen. Part of the reason for this market failure is that firms have not had to pay for the damage that they are doing to the planet. Any solution to the climate change problem must therefore involve making the polluter pay, be it through prices, quantities, other forms of regulation, or a combination of these mechanisms.

Price based mechanisms are known as carbon taxes – the regulator sets the price, leading to a reduction in emissions. The main quantity based mechanism is known as emissions trading, or cap and trade. A cap and trade scheme involves the regulator setting the amount of emissions, with the market determining the permit price. One form of cap and trade is carbon rationing, which is a pure application of the polluter pays principle – emission permits are allocated to the public, and polluters are required to buy permits from the public. There is also a form of emissions trading known as “baseline and credit”, which includes schemes like the UN Clean Development Mechanism and the NSW Greenhouse Gas Abatement Scheme. Baseline and credit schemes have problems with establishing a baseline, which means that it is difficult to prove that emissions reductions are real. For the rest of the post, when we discuss emissions trading we will be discussing cap and trade schemes. We shall also discuss hybrid approaches, which combine elements of a carbon tax and a cap and trade scheme.

A cap and trade scheme is not more or less market based than a carbon tax. The key issue is uncertainty — nobody knows how much emission reductions you will get for a given carbon tax level, and nobody knows what the carbon price will be for a given level of the cap. There are two sources of uncertainty – uncertainty in the cost of mitigation and uncertainty in the cost of climate change. It is not too hard to estimate upper bounds for the cost of mitigation (John Quiggin does it for electricity generation here, and some advocates of a very low carbon tax such as William Nordhaus also assume the existence of a “backstop technology” that plays a similiar role). The uncertainties in the costs and impacts of climate change are much greater, involve a whole lot of “long tails”, and cannot be bounded – in the words of Weitzman ( 2008 ) On Modeling and Interpreting the Economics of Catastrophic Climate Change, they have “potentially unlimited downside exposure”. There is also the issue of the discount rate, and of environmental costs, which mean that aggregating costs is also an ethical question.

When choosing a price or quantity, it will either be too high or too low (because of uncertainty), and there will be a welfare loss. Weitzman (1974) Prices vs Quantities discusses the mathematics of this issue. Weitzman argues that the slope of the marginal cost and benefit curves are what matter and that if the cost of abatement goes up faster than the cost of climate change then a tax is better, if it doesn’t (such as when climate tipping point is a risk) then a cap-and-trade system is better. Hepburn (2006) Regulation by Prices, Quantities or both: a review of instrument choice has some good discussion of these issues.

The uncertainty in the science is a very strong argument for more mitigation at a faster rate, issues where the science has not been entirely resolved but have serious impacts on risk include melting ice sheets (Greenland and West Antarctica are worth about 13 meters if they both go), carbon cycle feedbacks, albedo feedbacks affecting carbon cycle feedbacks (a recent paper suggests that melting arctic ice will increase the permafrost melting), and very low probability high impact possibilities of methane hydrates being released. All of this points to getting emissions down to as low a level as possible as quickly as possible at costs that can be managed. It is important that costs are kept as low as possible in order to ensure that the mitigation task continues (this is a political economy issue), as well as reasons of utility and human welfare. The problem with a pure cap and trade scheme is that it is very difficult to choose an emissions reduction trajectory when much of the recent science suggests that it would be prudent to reduce emissions as quickly as possible to a level somewhere around zero.

This brings us to the issue of hybrid approaches. Most hybrid approaches can be thought of as a cap and trade scheme where there is a price floor, a price cap, or both. The Australian Government’s Carbon Pollution Reduction Scheme Green Paper proposes a price cap – firms can always buy more permits at the level of the price cap, this means that the emissions cap is no longer a strict cap. Having a price cap is problematic because is there is a risk that the social cost of carbon is greater than the cap. The issue of long tails in the damage function and of potentially unlimited downside exposure make this risk more significant.

The issue of potentially unlimited downside exposure, the latest science, and the politics of mitigation all suggest that the risk of climate change damages exceeding the costs of mitigation is far more serious than the risk of the costs of mitigation exceeding the damages from climate change. For this reason we should have a price floor and not a price cap. A price floor helps us to choose an emissions reduction trajectory which would reduce emissions as quickly as possible to a level somewhere around zero. The emissions cap should be chosen so that it is sufficient to avoid dangerous climate change and the price floor should be as close as possible to the expected social cost of carbon. Having emissions caps is also useful for international cooperation – it is more likely to be consistent with negotiations under the UNFCCC.

Additional benefits of a price floor are that it manages the risk that the carbon price collapses if the amount of emissions goes below the cap, which happened with the EU Emissions Trading Scheme. A price floor also provides an extra degree of certainty for investors in low emissions technologies.

According to Hepburn (2006) a price floor can be implemented by a commitment by the regulator to buy emissions permits at the floor price. In his Draft Report, Ross Garnaut has criticised this approach. He states (page 344):

A floor price for permits would require the scheme administrator to enter the market to purchase permits whenever the permit price fell below a specified value. A floor price is incompatible with international trade in permits as it would effectively create an unlimited liability for the Australian scheme administrator.

There is another approach to a price floor that does not have this problem. A traditional cap and trade scheme works by firms being required to have purchased enough permits that cover their emissions. When a firm reports its emissions, it surrenders an equivalent amount of permits. We can introduce a price floor by requiring firms to pay an “extra fee” for each tonne of emissions when they report their emissions. The carbon price is then equal to the sum of the permit price and the extra fee. This approach is only slightly more complex than a pure cap and trade scheme. This approach is also very similar to having emissions trading scheme combined with a carbon tax.

Note: An earlier version of this post was included as a comment on John Quiggin’s blog post Carbon Taxes vs Emissions Trading, which also included a comment by Warwick McKibbin where he describes his hybrid approach (where there is a price cap).