Low Prices a Problem? Making Sense of Misleading Talk about Cap-and-Trade in Europe and the USA

Some press accounts and various advocates have labeled the Regional Greenhouse Gas Initiative (RGGI) as near “the brink of failure” because of the recent trend of very low auction prices.  Likewise, commentators have recently characterized the European Union Emission Trading Scheme (EU ETS) as possibly “sinking into oblivion” because of low allowance prices.  Since when are low prices (which in this case reflect low marginal abatement costs) considered to be a problem?  To understand what’s going on, we need to remind ourselves of the purpose (and promise) of a cap-and-trade regime, and then look at what’s been happening in the respective markets.

The Purpose and Promise of Cap-and-Trade

A cap-and-trade system– if well designed, implemented, and enforced – will limit total emissions of the regulated pollutant to the desired level (the cap), and will do this (if the cap is binding) in a cost-effective manner, by leading regulated sources to each make reductions until they are all experiencing the same marginal abatement cost (the allowance price).  Thus, the sources that initially face the highest abatement costs, reduce less, and those sources that face the lowest abatement costs, reduce more, achieving system-wide minimum costs, that is, cost effectiveness.  So, the purpose and promise, in a nutshell, is to achieve the targeted level of aggregate pollution control, and – if the cap is binding – do this at the lowest possible cost.

RGGI Allowance Prices

The Regional Greenhouse Gas Initiative (RGGI) – a downstream cap-and-trade system for CO2 emissions from the power sector in 10 northeast states (Connecticut, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Rhode Island, and Vermont, with New Jersey now in the process of withdrawing from the coalition), was launched with relatively unambitious targets, principally in order to keep prices down to prevent severe leakage of electricity demand and hence leakage of CO2 emissions from the RGGI region to states and provinces outside of the region (mainly from New York to Pennsylvania).

Emissions are capped from 2012 to 2014, and then, starting in 2015, the cap decreases 2.5% per year until it is down by 10% in 2019.  This would represent a level of emissions 13% below the 1990 level of emissions.  It was originally thought that this would be some 35% below the Business-as-Usual (BAU) level in 2019.  Sounds good.  What happened is not that the system performed other than designed, but that “business was not as usual.”  That is, what happened is that unregulated power-sector (BAU) emissions in the northeast fell significantly.  (See the graph below of the RGGI cap and historical emissions.)

For source, please click here.

So, Why Did Emissions Fall in the RGGI States?

This happened for three reasons.  First, because of increasing supplies in the United States of low-cost, unconventional sources of natural gas, prices for this fuel have fallen dramatically since 2008. (See the graph below of natural gas and coal prices.)  That has meant greater dispatch of electricity from gas-fueled power plants (relative to coal-fired plants), more investment in new gas-fired generating plants, less investment in coal-fired generating capacity, and retirement of existing coal-fired capacity, all of which has contributed to lower CO2 emissions.

For source, please click here.

Second, the worst economic recession since the Great Depression hit the United States in 2008, causing dramatic reductions in electricity demand in the industrial and commercial sectors, reducing emissions.  (See the graph below of quarterly percentage change in U.S. GDP, 2007-2009.)

For source, please click here.

Third and finally, moderate northeast temperatures have kept down CO2 emissions linked with both heating and cooling.

Low Emissions, Low Allowance Demand, Low Allowance Prices

So, for the three reasons above, BAU CO2 emissions from the power sector in the RGGI states are dramatically below what was originally (and quite reasonably) anticipated.  The supply of RGGI CO2 allowances made available at auction is – by law – unchanged, but demand for these allowances has fallen dramatically, hence the fall in RGGI allowance prices.  (See the graph below of RGGI allowance prices, 2008-2010.)

For source, please click here.

Given that emissions are below the RGGI cap and – due to expectations regarding future natural gas prices – are likely to remain below the cap, there is no scarcity of allowances.  Shouldn’t the price fall to zero?  In theory, yes, except that the system has an auction reservation price of $1.86 per ton built in, thereby creating a price floor of precisely this amount.

Is RGGI a Failure?

So, the cap put in place by the RGGI system is being achieved, but it is not binding.  RGGI may not be particularly relevant, but it is not thereby a flawed system; surely it is not a failure.  Rather, a great environmental success has been achieved by the “fortunate coincidence” of low natural gas prices, economic recession, and mild weather.  This is hardly something to be lamented.

True enough, the RGGI system does have flaws (such as its narrow scope limited to electricity generation, and its lack of a simple safety valve, as I have written about in the past).  But the low allowance prices are evidence of a success outside of the RGGI market, not evidence of failure within the RGGI market.

If the RGGI states have the desire and the political will to tighten the cap in the future, then the system can again become binding, environmentally relevant, and cost-effective.  That’s an ongoing political debate.

To be fair, I should note that the same outcome I have described here can be spun – perhaps for political purposes – quite differently.  Recently, a self-described “free-market energy blog” commentator claimed – not without some justification – that RGGI is irrelevant or worse:  “Bottom line, the program has raised electricity prices, created a slush fund for each of the member states, and has had virtually no impact on emissions or on global climate change.”

Phrased differently, due to exogenous circumstances (I’ve described above), the RGGI program is non-binding, and so has no direct effect on emissions, but its relatively low auction reservation price does lead to very small impacts on electricity prices, and produces revenues for participating states, revenues which those states would surely claim are of value for state-level energy-efficiency and other programs that indirectly do affect CO2 emissions.  So, the real bottom line is that low RGGI allowance prices are not a consequence of poor system design or a fatal flaw of cap-and-trade systems in general, but rather a consequence of what are in reality some exogenous coincidences that have turned out to be good news for the environment.

Now, let’s turn to the European Union Emissions Trading Scheme (EU ETS).

EU ETS Allowance Prices

Unlike RGGI, the EU ETS has not been irrelevant.  It has successfully capped European CO2 emissions, achieved significant emissions reductions, and it has done so — more or less — cost-effectively.  (More about this hedging on cost-effectiveness below).  Not surprisingly, like RGGI, the EU ETS has some design flaws (principally, its limited scope – electricity generation and large-scale manufacturing – and lack of a safety-valve), but as with RGGI, its low allowance prices should not be taken as bad news, but to some degree as good news, and certainly not as a sign of failure of the EU ETS.

Hand-wringing in Europe over Low Allowance Prices

There has been much hand-wringing in Europe over the “failure of the system” because of low allowance prices.  Indeed, Danish Energy Minister Martin Lidegaard said earlier this month that low carbon prices threaten the EU ETS.

Of course, he’s correct that EU ETS allowance prices are “low.”  They are down from their historic average of about $20 per ton of CO2 to about $9 per ton currently (having reached an all-time low of $7.88 in early April).  Here’s a graph of EU ETS allowance prices (EUAs) over the crucial period of change, January 2007 to January 2009.

For source, please click here.

At this point in this essay, I probably don’t need to say that this pattern is partly explained by the global recession, which has hit Europe particularly hard (and now threatens a double-dip recession in a number of European nations).  Lower European – and global – demand has meant decreased economic activity in Europe, hence lower energy demand, lower CO2 emissions, and therefore lower demand and lower prices for EU ETS allowances.

Even if we assume a growth rate of European CO2 emissions 1 percent less than the growth rate of GDP (represented by the dotted “counterfactual” BAU line in the graph below, which estimates what emissions would have been from 2005 to 2010 without the introduction of the EU’s Emissions Trading System), the evidence makes clear that the EU ETS has succeeded in reducing emissions significantly below what would be expected from the recession alone.

For source, please click here.

This is where an important caveat needs to be introduced.  Also feeding into this allowance price depression has been a set of national and regional energy policies, such as those promoting use of renewables, which have served to reduce emissions, demand for allowances, and hence allowance prices (while rendering the overall CO2 program less cost-effective by ensuring that marginal abatement costs remain heterogeneous).  So, to the degree that the low allowance prices are due to so-called complimentary policies, the low prices are bad news about public policy (in cost-effectiveness terms), not good news.  But this refers to misguided complimentary policies (which fail to bring about any incremental emissions reductions — under the cap-and-trade umbrella — and drive up aggregate cost), not to any design flaw in the EU ETS itself.

Multiple Goals Typically Require Multiple Policy Instruments

No doubt, Minister Lidegaard is aware of the allowance price impacts of the recession, and I hope he’s aware of the allowance price consequences of these other energy and environmental policies.  The problem arises, however, because he sees the fundamental purpose of the EU ETS as somewhat broader than what I described at the beginning of this essay (namely, achieving emissions consistent with some cap, and doing so cost-effectively – if the cap is binding).  For him – and many other European observers – “the purpose of the ETS was to cap CO2 emissions in the E.U. and ensure clear economic incentives for investment in renewables.”  So, the hand-wringing is not about a failure to achieve emissions reductions cost-effectively, but to have prices high enough to achieve other goals – in this case, greater use of renewable sources of energy.  For others, the “other goals” have involved allowance prices high enough to bring about some targeted amount of technology innovation.

As I have written at this blog in the past, having multiple policy goals typically necessitates multiple policy instruments.  For example, if the goal is a combination of reducing emissions cost-effectively and having prices maintained at some minimum (whether to bring about greater use of renewable energy sources or to inspire more technology innovation), then two policy instruments are needed to do the job:  a cap-and-trade system for the first goal in combination with a carbon tax in the form of a price floor (as in RGGI) for the second goal.

Don’t Throw Out the Baby with the Bath Water

In other words, the EU ETS has not failed, but the design was inadequate (that is, incomplete) for what politicians now seem to want.  If the Europeans want a price floor in their system (or better yet, a price collar, which would combine a price floor with a safety valve, i.e., price ceiling), then this is certainly feasible technically and economically.  Likewise, if the EU member states have the desire and the political will to tighten the cap in the future, there are a variety of ways in which they can accomplish this, rendering the program more stringent and increasing allowances prices.  But, in any event, the European Commission’s Energy division, Environment division, and Climate division should sort out the real effects of the “complimentary policies” that have contaminated the EU ETS, and which fail to bring about additional emissions reductions but drive up costs.  Whether any of this is feasible politically is a question that my European colleagues and friends can best address.


Author: Robert Stavins

Robert N. Stavins is the A.J. Meyer Professor of Energy & Economic Development, John F. Kennedy School of Government, Harvard University, Director of the Harvard Environmental Economics Program, Director of Graduate Studies for the Doctoral Program in Public Policy and the Doctoral Program in Political Economy and Government, Co-Chair of the Harvard Business School-Kennedy School Joint Degree Programs, and Director of the Harvard Project on Climate Agreements.

13 thoughts on “Low Prices a Problem? Making Sense of Misleading Talk about Cap-and-Trade in Europe and the USA”

  1. Rob,

    In general a useful comment. The main issue in Europe, however, seems to be that in addition to the ETS there is a separate goal of subsidizing, encouraging via feed tariffs, or simply requiring renewable electricity. These separate policies undercut the carbon price, leaving little or nothing for the ETS to do. The problem is that the renewables incentivized this way are probably not the least expensive way to reduce carbon. Thus, in some sense the low carbon price is a problem because its a sign that these other costly programs are doing the work of the cap and trade.

    While we have not examined the specifics of the European case, we looked at this interaction in proposed US legislation in

    Report 187.
    Combining a Renewable Portfolio Standard with a Cap-and-Trade Policy: A General Equilibrium Analysis

    Morris, J.F., J.M. Reilly and S. Paltsev, Joint Program Report Series (July 2010) (19 pages)
    [abstract] [PDF: 756 kB]

  2. John,

    I agree with your comment in part. As I stated in the essay above, these other EU policies aimed at renewables are part of the explanation for low allowances prices. But only if you believe that the EU renewable policies have had a greater impact that the recession on emissions should you argue that they have been “the main issue.” Of course, as I have written elsewhere, it’s true that the renewable policies, working under the umbrella of the EU ETS, have no incremental effect on CO2 emissions, but simply function to drive up aggregate costs by introducing heterogeneous marginal abatement costs.



  3. Dear Professor Stavins,

    My initial reaction to this very interesting post is that whoever is responsible for PR, promoting cap and trade, isn’t doing a very good job. If RGGI allowance regulations were, in fact, rationally set, so as to ensure that adequate supply would be available to meet expected demand with a price floor of $1.86, the implication would seem to be that the cost of offsetting greenhouse gas emissions, economy-wide, should be truly trivial.

    For simple context, a gallon of gasoline when combusted produces about 8.8 kg of CO2, what would be offset at the lowest realized RGGI allowance price ($2.04 per ton – what is 10% higher than the floor) by a per gallon “tax” of 1.8 cents, what is only about 0.5% of the current market price of gas! This notional 0.5% surcharge, of course, doesn’t include the offset of carbon embodied in the production of the gasoline (which is but a fraction of the direct contribution) but on the other hand, the combustion of gasoline is one of the more, er, carbon-intensive activities in our economy, so we could still call the 0.5% an upper-bound estimate to achieve carbon neutrality.

    But that cannot be right, can it?

    Regardless, it does point to the fact that the apparent “bottleneck” in the system is the provision of a sufficiently elastic supply of offsets. If the lowest-cost technology can produce offsets, with ordinary profit, at $1.86 per ton (again, assuming the RGGI regulations were rational) being able to sell them at a “discount” price of $7.88 per ton in Yurp would seem to be, um, an opportunity.

    Am I missing something here?


    Steve Schran

  4. Mr. Schran,

    Thanks for your comment, but perhaps you’ve missed something, or rather more likely, I failed to express it clearly. Due to exogenous factors that — subsequent to system planning — have led to a tremendous fall in RGGI emissions and hence abatement costs, the program is non-binding, period. Allowances have value only because of the “tax” that is charged at the auction, which is the reservation price.

    The RGGI system is NOT directly linked with the European system; and the European system has strict limits on CDM offsets, and RGGI does NOT currently permit CDM offsets at all; hence they are NOT indirectly linked. Therefore, different allowance prices can and will exist across these two un-linked systems. For more about direct and indirect linkage, see previous blog posts.

    Sorry, but the real world of public policy does not fit your definition of “rational.”



  5. Rob:

    I enjoyed your recent blog on RGGI and EU-ETS, but I think you missed one crucial element. In RGGI much of the revenue is used for incentivizing energy efficiency and that has been a very effective complement.

    RGGI auctions have raised $993.752 million to date. While the recession played some role in the decline of carbon emissions, economic analysis suggests that the main sources of the reduction were fuel substitution, including the development of renewable sources of power as well as conversions due to lower natural gas prices, the construction of combined heat and power plants, which are much more efficient, and energy efficiency.

    In Maine a decomposition analysis reveals that 52% of the decline in carbon emissions from 2005 (the year RGGI was announced) to 2010 (that latest year for which all data are available) was due to energy efficiency, 43% was due to fuel substitution and about 5% was due to the recession. (Maine does not have the same access to natural gas as other RGGI states.)

    In Maine megawatt-hour electricity consumption declined 13% during the 2005-2010 period. As a consequence of both lower natural gas prices and the drop in consumption, standard offer electricity prices also dropped. (For example for residential customers the decline in prices was 12.4% for Central Maine Power customers and 13.5% for Bangor Hydro customers.)

    Using the revenue in this fashion has mattered.



  6. Dear Tom,

    Thanks for your comments. Let me respond very briefly:

    1. I did note in the essay above that the states derive revenue from the reservation-price-driven auctions, and that they see this as valuable for funding renewable and energy efficiency programs.

    2. Whether those programs are themselves cost-effective or even effective is another question. The fact that a substantial share of the emissions reductions in Maine — 52% by your estimate — were due to increased energy efficiency does NOT indicate that share that were due to state energy-efficiency policies (as opposed to ordinary price effects).

    3. I stated in the blog essay that the most important factor as been low natural gas prices (fuel substitution in your comment).

    4. I’m sorry, but as I understand it, nothing in your comment indicates that the use of the auction revenue for state energy-efficiency programs was either particularly effective, let alone cost-effective.

    Thanks again for your comment.


  7. Dear Professor Stavins,

    You write: “EU ETS has not failed, but the design was inadequate (that is, incomplete) for what politicians now seem to want. If the Europeans want a price floor in their system (or better yet, a price collar, which would combine a price floor with a safety valve, i.e., price ceiling), then this is certainly feasible technically and economically.”

    It seems that you are essentially arguing for a carbon tax – indeed, the tighter the collar, the closer your proposed system gets to a tax, in most ways.

    Considering this, in your view, is the main benefit of Cap and Trade that political-economic forces lead to greater coverage than a carbon tax, because in Cap and Trade stakeholders fight for free allowances instead of fighting to be outside of the system entirely?

    Thank you,

  8. Dear Professor Stavins,
    1) In this non-binding situation, does it mean carbon tax is better than cap-and-trade?
    2) How can we better design CAT to achieve cost-effectiveness in this situation?
    3) In theory, how can we do research on policy design also considering exogenous shocks?

  9. Andrew,

    In a word, the answer to your question is “yes,” the major advantage of the cap-and-trade approach, in my view, is the political-economy difference between it and the tax approach. With the former, the politics is about alternative allocations of allowances, and this has no impact on environmental performance or aggregate cost; whereas with taxes, the politics tends to be about exemptions, which reduce environmental performance and drive up costs for what’s accomplished.


  10. Rob,

    RGGI and the EU ETS are not unique. Almost all trading systems have overachieved whether due to economic recession, technological change, complementary policies, changes in relative fuel prices, or other reasons. We have not yet built that possibility into the designs of trading systems.

    It is interesting that the proposed “remedy” (as you point out, it is not clear that a failure has occurred) often is a minimum price. A minimum price higher than the market price in effect turns the trading program into an emissions tax.

    There are reasons why a trading program, rather than an emissions tax, was implemented. I think alternatives to a minimum price should be evaluated before it is adopted.

    A trading program is intended to provide quantity certainty but cost uncertainty. In practice every GHG trading program includes features to reduce cost uncertainty — banking, offsets, limited borrowing, etc. The cap for a GHG program is a political choice rather than an economically optimal level (the cap covers only a small fraction of global emissions for a few years when global emissions need to be reduced significantly over a long period).

    Responding to overachievement with a minimum price converts the trading program into a politically determined (not economically optimal) carbon tax (equal to the minimum price) and means it will now have cost certainty and quantity uncertainty.

    The trading program structure can be maintained by automatically “locking in” environmental benefits when overachievement occurs. An automatic quantity adjustment could take the following form — when banked allowances (plus offsets) exceed X% of the annual cap, the cap for the next year is reduced by the amount of the surplus. This lowers the cap when it is economic to do so (prices are low) but not if it is costly and it maintains a trading system rather than converting it to a minimum price (tax) regime.

    Economists have devoted a lot of attention to options to deal with cost uncertainty in a trading program and such measures are part of virtually every trading program. We have not devoted much attention to options for dealing with overachievement, although it is a common occurrence (or the rare cases, such as NOx emissions by electricity generators in RECLAIM during the electricity crisis, where achievement is virtually impossible). Reasonable elaborations of a trading program design are possible to enable it to better achieve its environmental goal.

    Adjustments that maintain the trading structure should be considered before converting successful trading programs into emissions taxes based on political considerations that create economic and environmental uncertainty.


Leave a Reply