It is time to stop quivering in our boots in pointless fear of the future and just roll up our sleeves and build it.
- Ray Pierrehumbert

Thursday, April 2, 2009

Open Thread 3

Are they gone? Are they coming back?

Suggested topic: somebody please explain to me what "cap and trade" is and why it is supposed to be a better idea than a simple carbon emission tax/sequestration subsidy scheme. It seems I am supposed to believe it but I don't even get it.


John Fleck said...

Cap and trade specifies level of carbon emissions and lets the price float. Carbon tax fixes the price and lets the amount of emissions float. But there is so much wiggle room in implementation details that either sentence above can be made wrong by how you structure the thing.

Zeke said...

I wrote a brief piece about the tax vs permit debate a few years back which might be useful:

The need for economic incentives for emission abatement

In the early 20th century, the economist Arthur Pigou argued that the existence of market externalities required government intervention in the form of a tax, in order to ensure that the private costs of an action are aligned with the social costs (1920). Pigou argued that cost internalization of environmental externalities would yield more effective solutions than command and control regulation, as it would provide an incentive, rather than a prescription, for reconciling private and social cost. Individuals would have an incentive to bring private information to bear in developing the most efficient possible abatement strategies, yielding a better outcome than top-down policies limited by the imagination of government bureaucrats.

As Ronald Coase famously pointed out in The Problem of Social Cost, government intervention to resolve market externalities is unnecessary in cases where transaction costs are low and property rights are well defined (1960). Coase argued that individuals will make mutually beneficial bargains to resolve any externalities that are not socially optimal, unless transaction costs are higher than the social cost of the externality. Likewise, the initial allocation of property rights is irrelevant to the equilibrium outcome, though it does have distributional implications. There are a number of important factors, however, which undermine a Coasian solution to climate change. First, there are information asymmetries and deficiencies. Most people in the world do not know the exact extent to which they will be effected by climate change, and know little about the costs of abatement. Future generations are unable to effectively bargain, and leaving the social value of future generations to the simple sum of individual posterity values is problematic from an ethical perspective. Second, collective actions costs impede action by those hurt by climate change. Each individual has an incentive to free ride on the efforts of others, as individuals benefit from collectively achieved reductions in GHG emissions whether or not they actually expend individual resources to bargain for reductions. Finally, there are simple transaction costs. For every person in the world to talk with every single emitter of greenhouse gases and bargain with them to reduce emissions to the optimal level would be effectively impossible, even if every person had perfect information. Such a solution would involve an enormous amount of time spent bargaining, and time is valuable.

Climate change has important temporal aspects that complicate a traditional Pigouvian taxation solution. In particular, the marginal social cost of carbon increases over time. This is because the costs of additional warming increase exponentially, such that 4°C warming imposes a disproportionately larger economic cost than 3°C (Stern 2006), though this result depends in part on the discount rate chosen. Thus, a ton of carbon emitted when there is a relatively low atmospheric concentration of carbon is less damaging than a ton of carbon emitted when there is a higher concentration. Additionally, the marginal costs of carbon abatement will likely decrease over time due to technological progress and the development economies of scale for new technologies. This means that the marginal cost of carbon emissions imposed by a Pigouvian externality price would have to increase over time to reflect the increasing social cost of carbon.

The efficiency of abatement mechanisms

Two competing approaches have emerged in the effort to internalize the cost of climate change in the price of carbon emissions. The first is a price-based approach, where emitters of carbon pay a fixed price for every ton of carbon they emit. Under such a system, firms will choose to reduce emissions whenever emission abatement is available at a cost lower than the level of the carbon price. The second approach fixes the total quantities of emissions, and is a tad more complicated. In a quantity-based approach, the quantity of permissible carbon emissions is capped and firms are required to possess permits for each unit of emissions. In most modern variants, these permits are tradable between firms, and may or may not be bankable for use in future years. At the start of the quantity-based system, permits are either auctioned off by the government, distributed to firms based on the emissions of a chosen baseline year, or distributed annually based on emissions in an updating prior period. Firms with relatively high abatement costs will choose to buy permits on the market if the price of permits is lower than the cost of abatement. Likewise, firms with low abatement costs will reduce emissions and sell extra permits until the point at which the cost of further abatement equals the market price of permits. In an ideal world of perfect information and no transaction costs, the equilibrium outcome of taxes and tradable permit systems would be effectively identical, as both would result in firms abating to the point at which price of emissions (either the price of a permit or the marginal tax) is equal to the cost of abatement. In a world characterized by uncertainties, however, taxes and tradable permits can result in quite different outcomes.

In his classic 1974 article, Prices vs. Quantities, Martin Weitzman pointed out that price-based and quantity-based approaches behave quite differently when the costs of abatement are uncertain. The relative effectiveness of each depends on the slope of the marginal abatement cost and marginal benefits curve. If marginal abatement costs are expected to be steep, e.g. each unit of emissions costs more than the prior unit to abate, a price approach will tend to limit potential deadweight loss that would occur under a quantity-based mechanism by capping costs should marginal abatement costs prove higher than expected, and allow for the optimal level of abatement should abatement costs prove lower than expected. If the marginal benefits curve is steep, however, a quantity-based mechanism can limit the deadweight loss that would occur due to inefficiently high or low abatement under a price-based mechanism. Thus steep marginal costs and flat benefits of abatement would tend to favor a price-based mechanisms, while flat marginal costs and steep benefits would favor a quantity-based mechanism.

In the context of climate change issue, it is likely that price-based mechanisms will outperform quantity-based mechanisms, at least in the short term (Pizer 2002). Carbon dioxide is a classic example of a stock pollutant, where the marginal damages of a unit of emissions is solely a function of the existing concentration of CO2 in the atmosphere and a single unit of emissions has a negligible effect on the total stock. Hence the marginal benefits of abatement should be relatively flat over the short term. Likewise, marginal abatement costs are expected to be steep, as each additional unit of abatement will be more expensive than the prior one, and there is little time for innovation in abatement methods. If a pure quantity-based mechanism is implemented and marginal abatement costs prove higher than anticipated, firms could be forced to pay permit prices well in excess of the social cost of carbon, leading to a loss of welfare. In the long term, however, the situation changes considerably. Here, marginal benefits of abatement are expected to be quite steep, as the stock changes dramatically over time. Likewise, marginal costs of abatement are expected to become more flat as both endogenous and exogenous technological development drive down abatement costs. This has led some to suggest that the ideal mechanism for greenhouse gas abatement is one that acts like a price mechanism at the margin with the flexibility to behave like a quantity-based mechanism in the long-term by changing the price in response to changes in the slopes of the marginal cost and benefit curves (Stern 2006).

Unfortunately, efficiency is not the only criterion that governs the policy-making process. In reality, efficiency is but one of many factors used in deciding which policies to adopt. Also important are concerns about effectiveness, in terms of how much emissions are actually reduced, concerns about equity, in terms of who has to bear the burden of emissions reductions, and concerns about what is political acceptable to the powerful actors. Quantity-based mechanisms (tradable permits) and price-based mechanisms (taxes) perform differently in their ability to satisfy these oft-competing criteria.

Tradable permits

Tradable permit systems have proven the most popular mechanism to date in both international climate negotiations and domestic policy proposals. They have the benefit of being more politically acceptable to firms, as virtually all tradable permit systems for greenhouse gas emissions implemented so far have grandfathered most permits to existing firms based on emissions from a particular baseline year. This reduces the compliance costs of firms vis-à-vis a tax and creates a barrier to entry for new firms in GHG-intensive sectors, giving a competitive advantage to those firms granted initial permit allocations. Trading in permits helps minimize abatement costs by allowing firms with high abatement costs to buy emission permits from those with low abatement costs. This is particularly important on the international level, where abatement costs differ widely between countries.

Tradable permit systems allow the explicit separation of effectiveness and equity. Effectiveness is solely governed by the level of the cap, as climate change occurs based only on the total quantity of emissions, irregardless of who actually emitted them. Equity, on the other hand, is solely determined by the initial allocation of permits. This separation of efficiency and equity can be a mixed blessing, however, as equity issues tend to be highly controversial and efficiency can suffer if no equitable distribution can be agreed upon under a given cap. These issues are particularly important in the context of developing an international tradable permit system.

Tradable permit systems also tend to lack a continuing incentive for abatement. Under a permit system, firms will reduce overall emissions to the point at which the total emissions of all firms is equal to the cap, but there is little systemic incentive incentive to make further reductions even if abatement costs prove lower than anticipated. The only way to achieve additional reductions is to reduce the total amount of permits in the system. This can be accomplished in two ways: either the government can reduce the emission allowance of all permits by a constant amount, or they can buy up permits from the market to remove them from circulation. The prior option could potentially wreck havoc on the market in the absence of a fixed schedule of reductions decided at the time of policy formation. As the recent experience of the European Trading System demonstrates, if people are uncertain what the value of permits will be in the future, changes in permit allocations can lead to dramatic spikes or drops in the market price of permits (Sterner and Muller 2006). However, creating certainty in the permit market limits the ability of the government to tighten permit allocations in the future in response to new scientific developments that clarify the social cost of carbon. The second option, buying up permits, may result in an unacceptable transfer of wealth from the government to firms, especially if initial permit distributions were grandfathered rather than auctioned. Additionally, the announcement of a large government buy up of permits could lead to a spike in permit prices and potential windfall profits to firms and traders.

Finally, international trade in permits poses potential compliance issues. Countries initially allocated excess permits will have an incentive to sell their permits at the onset and later withdraw from the treaty when emission restrictions start to impose an economic burden. In the absence of trade sanctions on countries that withdraw from the treaty, which would prove problematic under international trade law, non-compliance may pose a huge problem. One way around this may be what David Victor calls “buyer liability” (2001: 69-74). In a buyer liability system, a permit becomes void if the initial issuer defaults on their treaty obligations. Thus all permits in the market would be explicitly priced based on the perceived risk of default of the issuing nation, and nations would have a strong incentive to take steps to reassure potential buyers against the risk of default. A more subtle but no less problematic form of non-compliance has been seen in the empirical experience of the European Trading System (ETS). Countries have had an incentive to allocate emission permits in excess of their cap to help reduce the burden on their firms. The discovery of widespread overallocations recently led to the collapse of permit prices in the European market (from 30 euros to 10 euros per permit), and has led regulators in the EU to focus more on ensuring that permit distributions are compliant with overall country caps in the next phase of the trading system (Sterner and Muller 2006). Given the difficulty of uncovering permit overallocations in a trading system with a strong central regulatory oversight, one can imagine that preventing excessive permit allocations by a country in a truly global trading system would be quite problematic.

Emission taxes

Taxes tend to be favored over tradable permits by most economists, as they cap potential compliance costs while avoiding the problem of permit distribution. Revenue from taxes can be used to reduce existing distortionary taxes, such as income or payroll taxes. Many argue that shifting taxes from economic goods (productive labor) to economic bads (pollution externalities) can yield what is termed a “double dividend”: increased social welfare both through lower taxes on earnings and decreased harm from environmental externalities (Repetto 2001). A number of European countries have already implemented some tax-shifting policies, where revenue from taxes on landfilling, sulfur dioxide emissions, and increasingly carbon emissions is used to reduce other taxes (Brown 2001). Taxes also create a consistent price signal for firms considering abatement investments, providing guaranteed returns for abatement independent of actions taken by other firms (Hepburn 2006).

Taxes are arguably more flexible than tradable permits, as the level of the tax can be adjusted over time in response to the increasing social cost of carbon without the problem of having to remove existing emission rights from circulation. However, a system in which the level of the tax level increased at an unpredictable rate (given that the predicted social cost of carbon is an function of the inherently sporadic rate of scientific discovery) would come at an economic cost, as uncertainty about future tax levels would result in inefficient investment decisions. Additionally, future tax increases will likely prove politically controversial. Governments would have an incentive to resist increasing the tax over time, which would effectively undermine the system. An increasing, flexible carbon tax is essential to allow a tax-based system to behave like a quantitative constraint over time to avoid the potential large long-run inefficiencies discussed earlier. Carbon taxes are much less effective when they lack the flexibility to adjust the tax level based on changes in the expected social cost of carbon and the marginal cost of abatement.

Taxes may also be more difficult to implements on an international level than tradable permit systems. Creating an internationally harmonized carbon tax would pose a number of practical difficulties. Countries would have some incentive to neglect to enforce taxes as a result of rent seeking by powerful domestic interests. A simple look at the effectiveness of the current tax systems in countries like India, Brazil, or China suggests that a harmonized carbon tax would, in practice, not be uniformly enforced. Lax enforcement for export-focused industries could be potentially remedied by border tariffs on products based on their remaining externalized costs, though proposals to this end have so far proven extremely controversial (Bounds 2006). Countries may also disagree over tax levels based on equity grounds. Because taxes do not separate efficiency and equity concerns, there is no simple way to differentiate responsibilities without undermining the overall efficiency of the system.

A carbon tax would likely be opposed by many firms, as they may conclude that the increased costs of compliance (due to the taxation of all emissions, not just up to a certain point) vis-à-vis a tradable permit system would be larger than the potential benefits of a tax system should short-term compliance costs prove higher than expected. Likewise, taxes are extremely controversial in the United States for political reasons. While a pure tax system might be ideal from an efficiency standpoint, the political realities may require compromise. It may be necessary to use elements of both taxes and tradable permits to create a hybrid approach, combining the political acceptability of a tradable permit system with the efficiency of a tax.

Hybrid systems

The current political context in many countries, especially the United States, strongly favors tradable permit systems over tax-based systems (Hepburn 2006). With this constraint in mind, many economists have started looking for ways to make tradable permit systems behave more like taxes by limiting potential short-term deadweight loss when abatement costs are uncertain. They have built off the seminal work of Roberts and Spence (1976), who first proposed a “mixed system” of emission permits bounded by a penalty for excess emissions and a subsidy should the permit price fall below a certain threshold. This has the benefit of capping potential abatement costs at the penalty level while creating a continuing incentive for abatement via the subsidy should abatement costs prove lower than expected. Modern hybrids proposed for carbon abatement tend to include the price ceiling proposed by Roberts and Spence, but generally opt to not include a price floor. This may be in part explained by the potential for large monitary transfers from the government to the private sector to should abatement costs prove substantially lower than expected, or perhaps an implicit assumption that abatement costs will always be relatively high.

Under the most common hybrid varient, proposed by Victor (2001), Wilcoxin and McKibbin (2002), and others, potential compliance costs of a tradable permit system would be capped by an “escape valve” that would be triggered once the market price for permits exceeds a certain pre-determined level. Once the escape valve is triggered, the government would sell an unlimited number of annual permits for a ton of emissions at a fixed price. The escape valve effectively acts as a tax on the margin once permit prices reach a certain point. In practice, the escape valve would be set at a level where it was unlikely to be triggered unless abatement costs prove higher than expected.

Zeke said...

For more juicy economic details, is one of the better articles out there on the subject.

Matt said...

Mr. Fleck is dead on about the wiggle room. What I find interesting, too, is how LONG this idea has been around. A system has been out there since the early '00s ... an iteration, at least:

David B. Benson said...

Zeke --- Your llink is broken. YOu have to use the HTML "a" tags here.

Michael Tobis said...

I don't know what you are using to read this stuff, David, but copy and paste works for me.

I'd prefer if people would use or the like, though.

Hank Roberts said...

Although long links are tiresome, a thought --

One argument against tinyurl, or hiding links behind clickable colored words, or anything else but clear plain text:

I think it's safer in the long run to actually show people the real link you're meaning for them to open.

Someone could buy tinyurl and change the links.

Or hack somewhere and redirect them.

Someone could hack your website and change what's behind the colored clickable words.

The original destination could go bad.

If you can see the long html destination, you can search for it somewhere else. Even years from now.

Just a hobbyhorse, pardon.

Hank Roberts said...

Barry Brook does a good job in a recent thread and some earlier ones of documenting problems with cap'n'trade compared to a carbon tax.

Particularly the way cap'n'trade devalues individual changes.

Hank Roberts said...

more on cap and trade:
"New cap-and-trade climate legislation offers Wall Street a new opportunity to make money"

Hank Roberts said...

And (ick) the Republicans are off by an order of magnitude according to the author of the MIT study they claim as the basis for their cost numbers.

David B. Benson said...

MT --- This older Firefox won't follow those links on Google/Blogger sites.

ac said...

With a 'capntrade' system the government sets a cap on emissions, and auctions the permits.

The most compelling argument for it I have heard is that a hard cap is the only way a nation can commit to meeting emissions targets.

One popular argument against cap and trade is that individual contributions may be devalued. Example: I invent a dogshit powered methane generator to power my home. The reduced electricity demand that results benefits the operators of the coal fired power plant up the road, who can emit the carbon they would have emitted to power my house some other way.

Capntrade proponents respond that any reduction in emissions, and thus in demand for permits allows the government to reduce the cap the following year.

A strong argument against a simple tax is that it depends on the government being able to predict the response of the economy to a fixed carbon price.

So there you have it: who do you trust more, the invisble hand of the market or the ham fist of government. Me, I'm rooting for Venter's superbug.

Zeke - wtf? no-one is going to read an essay posted as a comment. Summary? Abstract?

Michael Tobis said...

I have yet to hear any argument in favor of "cap and trade"; it is like a half-baked tax policy with a cheap Halloween Zorro mask on. It won't fool anybody.

The tax idea still makes a lot more sense to me. If the economy had not yet been revealed as mostly ficitious I would have likes Hansen's revenue neutral idea too.

And the rate could easily be adjusted annually to approximate cumulative emission targets. The argument that it couldn't makes not a shred of sense to me.

Right now I have no clue about taxing or spending. It seems to me we are still collecting and spending money out of force of habit, even though nobody has any idea what the stuff really represents anymore. I guess inflation plus hastily scratched together socialism beats any other alternative on the table.

Obama needs to use all his mojo to rehypnotize people into thinking there is an economy. Nothing else will happen anytime soon.

If Obama succeeds he may actually amass some power and may get things done; say, health care, transportation, parks, education, good conventional liberal projects.

In other words, if the rehypnosis works out, he will come up with a much improved arrangement of the deck chairs. Bravo! But even that looks unlikely. As usual the Republicans make the messes and the Democrats pick up the pieces.

Anyway, the whole question looks moot for the time being.

It does not look like we will avoid crossing 450 ppmv. Time to fall back and start fighting for 500 yet?

ac said...

Funny joke but maybe you can explain exactly why you think there are no reasonable arguments in favour of capntrade? Sounds like you want to both not understand it and be strongly against it, which is not very convincing.

One of the other arguments is it is hard to coordinate taxes levied by multiple governments, far easier to link various national emissions trading schemes together.

Michael Tobis said...

There is a great deal of material and I suppose there are strong arguments. It's odd that I haven't seen any that are compelling. I'm tolerant of complexity, but not of obfuscation.

While I understand from reading climatology denial sites that one man's precision is another man's obfuscation, I don't think I am SO naive about economics that every explanation of a valid idea sails over my head. I guess all the oil men reading about climate change think the same about themselves.

Still, I just don't get it and if I don't I'll bet that most of the politicians themselves don't. On the whole they don't have that kind of intelligence. Which makes it a risky proposition indeed.

There will have to be national allocations, I suppose, but each nation would be wise to achieve them by simple and equitable mechanisms rather than endless and complex arrangements.

Mitchell said...

I believe the basic proposition is that with tradable emissions permits, the prescribed level of emissions reduction (set by the cap) will occur in the lowest-cost way.

ac said...

You don't trust that politicians are smart enough to understand an emissions trading scheme, but you trust that they're smart enough to determine the right value for a tax to have a prescribed impact on emissions?

John Fleck said...

Michael - Hard to know if you are being serious or just rhetorically glib when you say "I have yet to hear any argument in favor of "cap and trade";".

If you aspire to do useful communication, you've got to be able to be able to understand and communicate effectively on issues like this where smart and reasonable people have come to different conclusions. It's not about finding an argument that's compelling *to you*. It's about being able to understand and explain to me, as your reader, why smart people (in this case) find C&T compelling *to them*.

If you can't do that, your filtering-communicating will be of little use to me, because I'll have to either take your word that C&T has no merit (and I'm not going to do that) or hunt someone else to explain it to me.

The RFF people are a good example of how this is done right on this question. They're not "journalists," but their work has the same useful characteristics (just longer and more boring):

Michael Tobis said...

John, OK, it was too strong.

I am not trying to advocate, at least as yet.

I just feel foolish that I don't have much understanding of the C&T position. This, even though most people pulling for serious action in the US political world have lined up behind it.

I am trying to pry a simple argument out. Is it OK to conclude that there isn't one?

That doesn't make the c&t position wrong. The greenhouse gas mechanism is a bit complicated, too, but robust for all that.

This question doens;t much impact the scientific community. I asked Ray P what he thought once, and he basically shrugged and said, whatever they do, so long as emissions come down, is fine with me. It was a "them" question, not an "us" question to me.

So indeed, if I'm going to cover this turf, I need to make it an "us" question for me. It's surprisingly foreign and a bit embarrassing on that account.

Sorry again for the hasty grumble, and thanks for the link. I will slog through it.

Surely somebody by now has made the attempt, though. I wonder if you or anyone else out there has any pointers to something more accessible comparing the approaches.

John Fleck said...

The simple argument is in the first sentence I wrote in the first comment on this thread: you get a "cap". The starting point for the policy is a known quantity of emissions you're going to allow, based on your calculation of what PPM you're trying to avoid.

To accomplish the same thing with a tax requires information you don't have. Since we don't have a market yet, we don't know what kind of price (tax) we need to achieve a specified carbon reduction goal.

There are lots of pluses to a tax, and lots of reasons one might argue against cap and trade, but that's the simple foundational argument that leads lots of advocates to support C&T.

And a PS: I found the RFF thing I gave you the link to quite accessible. It's the reading my econ professor gave the class if undergraduates I'm in. But maybe I'll try my hand at C&T and carbon tax 101 on Inkstain.