Pricing Carbon: What Does It Actually Accomplish?

Pricing carbon may not work the way economists thought.

In theory, pricing carbon should incentivize emissions reductions.  In reality, it is unclear to what extent that takes place unless the carbon price is very high.  This is not to say that pricing carbon is useless, but the main benefits may take different forms.

Basically, there are two ways of putting a price on carbon.  One is a carbon tax. The other is to cap total emissions, requiring companies to pay for the right to a share of the limited amount of allowed emissions.  Some places like British Columbia have adopted carbon taxes, but emissions trading systems seem to be a more popular approach.  According to a BitQT test, in emissions trading systems, the price is unpredictable — basically, the cap determines the supply but the demand depends on market forces.  Price ceilings are a popular way of preventing the price from getting unacceptably high, while a price floor can prevent unacceptably low prices. Even if a trading system does not have explicit floor or ceiling prices, political actors are likely to intervene if prices get too low (making it clear that the system isn’t changing behavior) or to high (leading to vociferous complaints from firms and consumers).

A recent paper in the American Economic Review provides some illuminating insights into emission trading, using California’s emissions trading system as a model.  In theory, the cap on emissions should determine how many emissions take place.  But it turns out that what is important in California isn’t the emissions cap but the program’s price and floor ceilings.  That reason is that the amount of future emissions reductions is largely determined by other climate programs such as California’s renewable portfolio standard or by unrelated market forces.  Statistically, it turns out that most of the time the cap is actually satisfied without regard to the cap-and-trade program, so that the floor price goes into effect.  The second most likely outcome is that achieving the cap is much too expensive, so the price ceiling hits. Only in a minority of cases will the program actually cause emissions reductions.

One of the key assumptions is that cost of emissions reductions has a hockey-stick shape, so reductions are pretty cheap up to a certain point but then accelerate rapidly.  The upshot is that, once we hit the curve of the stick, carbon prices would need to be extremely high in order to incentivize additional reductions.  But even though we know the shape of the price graph, we don’t really know the exact numbers. So when we set a cap, we are unlikely to hit the sweet spot where price increases make a difference.  Essentially, the emissions trading program turns out to work like a carbon tax, except that we don’t know in advance whether the tax rate will turn out to be the floor price or the ceiling price.

Although the paper is about emissions trading, the implication seems to be similar in terms of the amount of emissions reductions that will be induced by a carbon tax. A carbon tax is functionally equivalent to a trading system where the ceiling and floor prices are the same.  It will be hard to set a price point where the tax will actually impact firm behavior, particularly in situations like California where there are also non-tax emission reduction measures that have preempted the most likely avenues toward reducing emissions. 

Perhaps the model exaggerates the limitations of emissions trading.  Uncertainty about carbon reductions in the absence of trading may be smaller, or the “hockey stick” might have a gentler curve.  Either one would increase the odds that the trading system’s price signal would impact emissions levels. Or maybe the model omits other features of the trading system that would weaken the result. But while these would weaken the conclusion, the model would still be a useful reminder of some important limitations on trading systems.

The model also should prompt us to think about the possible benefits of pricing carbon beyond any direct effect on emissions levels. As far as I can see, there are three fairly tangible benefits and another less tangible one.

First, as the authors point out, pricing carbon may provide an incentive for technological change, pushing the curve of the hockey stick further out.  The problem is not simply inventing new technologies. It’s providing a sufficient market advantage to fuel development and scaling up of those technologies.  A carbon price can help with that. In the long run, developing and deploying new technologies is probably at least as important for deep decarbonization as changing the utilization of existing technologies.

Second, we can use the money for other emission reduction projects such as mass transit, retrofitting buildings, and so forth.  That’s basically what California does.  California could also raise other taxes as an alternative, but politically it seems to be easier to tie these expenditures to a carbon price.  Something similar has been true of the states participating in RGGI, the carbon trading system established by the Northeast states.

Third, the price signal of a carbon tax or trading system  is most useful when it provides incentives to emitters who aren’t covered by other regulatory measures.  For example, California has recently decided to allow firms to satisfy their obligations by funding the preservation of tropical forests in Latin America.  Those rain forests obviously aren’t otherwise subject to California’s jurisdiction, and they may not be effectively regulated by their own legal systems. More broadly, pricing carbon may be more a effective strategy when there are no other emissions mandates, which is more true of the U.S. as a whole than for California specifically.  In states that have taken little or no action to address climate change, a national carbon price could help shut down the remaining coal-fired plants, reduce the attractiveness of natural gas, and increase renewables and energy efficiency. In California, however, those changes are already underway under other programs, muting the impact of carbon pricing.

There’s also a less tangible benefit of pricing carbon. Setting a carbon price or an emissions cap provides a clear, easily understandable signal of a jurisdiction’s commitment to fighting climate change.  California’s extensive suite of carbon reduction strategies is too complicated for a sound bite. It’s easier to say that California’s emissions cap will drop 40% by 2030.  That kind of clear signaling is important in the effort to gain international cooperation on climate policy.

In short, pricing carbon may be a useful strategy — but not necessarily for the reasons that economists have traditionally put forward.  The direct effect of the carbon price on emission levels may be less significant than its indirect effects: incentivizing new technologies, funding public projects related to climate change, encouraging  cuts in emissions that are otherwise unregulated,  and signaling seriousness of purpose to other jurisdictions.




, , , ,

Reader Comments

6 Replies to “Pricing Carbon: What Does It Actually Accomplish?”

  1. This assumes that there is a carbon budget. What do you do when you realize that, if there ever was such a thing as a carbon budget, there is no room in the biosphere for more CO2 emissions and you must stop emissions immediately?

  2. One of the challenges with California’s cap and trade is that the “complementary policies” do not add additional reductions beyond cap and trade, and instead allow others to continue polluting at a low price. When California mandates that all electricity must be clean energy that keeps emission allowance prices low and allows other emitters (transportation, industry, etc.) to keep emitting at a lower price.

    If we had a steadily rising carbon price this would not be true. Regulations and mandates on particular sectors would not affect the price that other sectors pay and they would continue having the same incentive to reduce emissions. Thus we could add complementary policies and gain additional emissions reductions, over and above those incentivized by the carbon price.

    In addition, California’s system allows for banking of allowances, carbon offsets, and free allowances to certain emitters. This makes it very very different than a simple carbon tax (even if the floor and ceiling price were identical).

    Plus with most cap and trade systems the price is allowed to fluctuate and it is thus harder for businesses to predict the future price of carbon emissions. A carbon tax with a predictable price would be more effective at inducing long term investment and innovation because businesses could plan for the long term.

    Not all carbon pricing is created equal, and the details of policy design make a big difference. Caution is needed when extending results based on one system or type of system to all forms of carbon pricing.

    1. Tony — I agree that it can be problematic to base too much on one case study. I’m not sure how much banking/borrowing or other design features may have made a difference. Complementary policies obviously were a big part of the reason for the findings. However, that may not be an exception, because those measures seem to be more politically appealing than the carbon price.

  3. Thanks for highlighting an important paper by Borenstein et al. I note the following quote in your blog:

    “Although the paper is about emissions trading, the implication seems to be similar in terms of the amount of emissions reductions that will be induced by a carbon tax. A carbon tax is functionally equivalent to a trading system where the ceiling and floor prices are the same.”

    We are at a point with over 30 years of combined CO2 cap and trade experience in Europe, New England and California. Cap and trade is very significantly different from a carbon fee policy. The conflation of the two is unfortunate and impedes substantive policy debate.

    As characterized by one of the members of the IEMAC cap and trade is basically a revenue generator at this point.

    The theory and narrative that cap and trade provides environmental certainty has not proven to be the case not only in CA but in RGGI and the EU ETS. The inherent design of complexity, opacity and multiple “features” has resulted in the weak systems we have today. The IEMAC annual reports of 2018 and 2019 point out the problems of cap and trade that are not problems for a carbon fee policy. They are: massive surplus of allowances impacting the carbon price; “complementary” regulatory programs that counteract cap and trade; resource shuffling, offsets that undermine emissions reductions (most significantly in the EU ETS); and the equity concerns of the Environmental Justice advocates. (Recall that the EJ Advisory Committee to the Scoping Plan process recommended dropping cap and trade and implementing a fee & dividend cum regulations.) While the EIMAC Annual Report highlighted the significant structural issues, it did not mention the greatest weakness that cap and trade cannot overcome–its vulnerability to the economy and its swings. The likely reasons CA achieved its goal “four years early” are (i) the nature of forecasting and cap setting means the cap will probably be inaccurate (Wara, U. Michigan); and (2) the recession was a predominant factor in reducing emissions and not cap and trade or the regulatory programs. None of these factor adversely affecting cap and trade affect a steadily rising carbon fee.

    Your highlighting the impact of a carbon price on innovation is welcome. Too little discussion on how this works takes place. To stimulate innovation, a predictable rising carbon fee is should be preferred by policy makers to the uncertain price of a cap and trade program even one bound by a floor and ceiling. The uncertainty of a government body sitting on a reserve fund to push out more allowances when the price gets too high politically or holding back on auctions or allowance distributions when it is too low undermines the predictability that is important for stimulating innovation and investment. Nobel Economist Paul Romer describes a steady predictable rising carbon tax as a powerful tool to bring on the needed innovation. That can only be brought on by a carbon fee and not a cap and trade system.

    Your basic doubts about the efficacy of a carbon fee were not clear. I’m persuaded by the work of the Stanford Modeling Forum and the reasoning and position of the 3500+ economists expressed in the Economists Statement at

    Thanks for your on-going contributions to the policy dialogues.

  4. Another, oft-overlooked problem of carbon pricing is the rebound effect that is most obviously present in the case of carbon taxes. As carbon taxes raise the cost of using carbon-based fuels, the rate of demand for those fuels would be expected to drop. Assuming a normally functioning market, as demand falls off, prices would fall. When prices fall, demand would rebound. For the carbon pricing mechanism to be effective in the long run, it would require a ratchet mechanism that would offset the rebound effect.

Comments are closed.

About Dan

Dan Farber has written and taught on environmental and constitutional law as well as about contracts, jurisprudence and legislation. Currently at Berkeley Law, he has al…

READ more

About Dan

Dan Farber has written and taught on environmental and constitutional law as well as about contracts, jurisprudence and legislation. Currently at Berkeley Law, he has al…

READ more