Economic Impacts of Alternative Emission Reduction Scenarios

By Alan S. Manne and Richard G. Richels

Executive Summary

This paper has three purposes: To identify the near-term U.S. and global economic costs of ratifying the Kyoto Protocol; to assess the significance of the Protocol’s “flexibility provisions”; and, perhaps most important, to evaluate Kyoto’s policy approach (sharp near-term greenhouse emission reductions) in the context of the Framework Convention on Climate Change’s long-term goal of stabilizing atmospheric concentrations. With respect to economic costs, we estimate U.S. GDP losses from implementing the Protocol at between $20 billion and $90 billion annually in 2010 and between $45 billion and $105 billion annually in 2020. Discounted global consumption losses through 2100 could total between $1 trillion and $2.5 trillion (1990 dollars). International emissions trading is essential to reducing costs, but options are limited in the Kyoto Protocol and nations disagree on their application. Lastly, the Kyoto Protocol does not appear consistent with the Framework Convention’s call for cost-effective,1 long-term policies to stabilize CO2concentrations, unless the concentration target for CO2 is well below 550 ppmv. In contrast to the Kyoto Protocol, a least-cost strategy, in which CO2emissions gradually depart from the baseline, would stabilize concentrations at a fraction of the cost.


The Kyoto Protocol2 to the Framework Convention on Climate Change represents the first time that negotiators have adopted binding emission reduction targets and timetables for Annex I countries (developed nations plus economies in transition). The Protocol’s stated goal is for these countries to reduce their aggregate anthropogenic carbon dioxide (CO2) equivalent emissions by approximately 5 percent below 1990 levels in the first commitment period, 2008-2012. However, the Protocol will not enter into force unless 55 countries, including a sufficient number of Annex I countries representing 55 percent of total Annex I CO2 emissions in 1990, ratify it. The United States and other governments clearly have an interest in understanding the economic implications of ratification.

“Kyoto Forever”

We begin with an examination of a “Kyoto Forever” scenario. This analysis is based on MERGE, a model for evaluating the regional and global effects of greenhouse gas reduction policies.3 MERGE is an intertemporal market equilibrium model. It divides the globe into nine geopolitical regions: 1) the USA; 2) OECDE (Western Europe); 3) Japan; 4) CANZ (Canada, Australia, and New Zealand); 5) EEFSU (Eastern Europe and the former Soviet Union); 6) China; 7) India; 8) MOPEC (Mexico and OPEC); and 9) ROW (the rest of world). Note that the OECD (regions 1 through 4) together with EEFSU constitute Annex I of the Framework Convention. In the Kyoto Forever scenario, the Kyoto emissions constraints are maintained throughout the twenty-first century. With regard to non-Annex I emissions, we assume they will be bounded by their business-as-usual baseline (see Figure 1). The latter constraint is imposed in order to prevent carbon leakage. Later on, we will explore the impact of relaxing this constraint.

[Figure 1: Regional Carbon Emissions: Reference Case]

Numerous studies have shown that global mitigation costs can be reduced substantially by allowing emission reductions to take place wherever it is cheapest to do so-regardless of geographical location.4 The Kyoto Protocol includes several provisions allowing for a limited amount of “where” flexibility. These include emission trading and joint implementation among Annex I countries. They also include provisions for a Clean Development Mechanism (CDM) that is intended to facilitate joint implementation between Annex I and non-Annex I countries.

The Protocol, however, leaves many critical details unresolved. For example, it remains unclear whether there will be limits on the extent to which a country can rely upon the purchase of emission rights to satisfy its obligations. The Protocol states that “the Conference of the Parties shall define the relevant principles, modalities, rules and guidelines”. Similar ambiguity surrounds the CDM. Again, the elaboration of “modalities and procedures” is left to a future meeting of the Conference of the Parties.5

In this section, we explore three options that are representative of alternative implementations of the Kyoto Protocol: 1) no trading; 2) Annex I trading plus CDM; and 3) full global trading. Each option has its own advocates and opponents, but we do not consider them equally likely. In our opinion, there is little likelihood of enticing all major countries to participate in a global market in emission rights during the initial commitment period (2008-2012).

The full global trading scenario places an upper bound on the CDM’s potential to reduce GDP losses. Because of the difficulties in implementation of the CDM, we assume that only 15 percent of the potential would be available for purchase through this mechanism. Figure 2 reports the incremental value of carbon emission rights to the United States in 2010 and 2020. We focus first on 2010. In the most constrained scenario, no trading, the United States must satisfy its emission reduction requirements within its own geographical boundaries. In this case, the value of emission rights approaches $240 per ton. With Annex I trading plus CDM, the value drops to slightly less than $100 per ton. As might be expected, the value of emission rights is lowest with full global trading. Here, it falls below $70 per ton.

[Figure 2: Incremental Value of Carbon Emission Rights in the United States Under “Kyoto Forever”]

For the two scenarios in which trading is permitted, the value of emission rights increases in 2020. This is because projected emissions for Eastern Europe and the former Soviet Union (EEFSU) lie below its negotiated constraint for 2010. It has been allocated more emission rights than it needs to satisfy its internal obligations. By 2020, however, EEFSU’s economic growth is expected to be such that it no longer enjoys an excess of emission rights. As a result, there is more competition for emission rights in the international marketplace, and their price increases.

Another way to view the costs of abatement is to show the GDP losses. Figure 3 contains those for the United States. Losses are highest in the absence of trade. Here, they approach $90 billion dollars in 2010. This is approximately one percent of U.S. GDP. To the extent that trade is introduced, losses decline. Under the most optimistic option (full global trading), losses are approximately $20 billion or one-quarter of one percent of GDP in 2010.

[Figure 3: Annual U.S. GDP Losses Under Kyoto Forever Scenario]

Of the three options, “Annex I trading plus the CDM” is most consistent with the Kyoto Protocol as it currently stands. However, the U.S. Senate has stated that the United States should not be a signatory to the Protocol if it does not mandate specific commitments for developing countries. If this were to result in full global emissions trading, we move in the direction of the right-most group of Figure 3.

There is, however, strong sentiment among many parties to the Framework Convention to substantially limit the extent to which Annex I countries can meet their obligation through the purchase of emission rights. Several influential developing countries have expressed strong opposition to the concept altogether. Figure 3 shows the costs of the no-trading scenario. We now turn to the case where trading is permitted, but with limitations on the purchase of emission rights.

Limits on the Purchase of Carbon Emission Rights 

With full global trading (the least-cost of our three options under Kyoto Forever), trading is used to satisfy more than 50 percent of the U.S. obligation. But suppose that limits are placed on the purchase of emission rights? For example, suppose international negotiators agree that Annex I buyers can satisfy only one-third of their obligation through this means. What would be the impact on GDP losses?

Figure 4 compares the three options. All assume full global participation in an international market for carbon emission rights, but only the first assumes no limits on the amount a country can buy. The second and third cases are based upon the one-third limitation. We further make the distinction between a buyers’ market and a sellers’ market. With the former, sellers of emission rights are price takers. Buyers exert sufficient market power to hold the international price to the marginal cost of abatement in the selling countries. However, since a country is only able to satisfy one-third of its obligation through the purchase of emission rights, it must eventually rely upon its own domestic marginal abatement capabilities to meet its obligations. Hence, there is an important distinction between the international price and the domestic price. Conversely, with a sellers’ market, buyers face but one price. Here, the rents accrue to the sellers.

[Figure 4: Incremental Value of Carbon Emission Rights With and Without Limits on the Purchase of Emission Rights]

Figure 5 shows the GDP losses associated with the three options. Note that losses in 2010 are two and one-half to three times higher with the constraint on the purchase of carbon emission rights. That is, the benefits from “where” flexibility are greatly diminished. The message is clear. Developing country participation in the market for carbon emission rights is a necessary, but by no means sufficient, condition for reaping the full benefits of “where” flexibility. To achieve a cost-effective solution, buyers must also be unconstrained in the manner in which they fulfill their obligation.

[Figure 5: Annual U.S. GDP Losses With Full Trading and When Annex I May Satisfy Only One-Third of Obligation Through the Purchase of Emission Rights]

Also note that the distribution of the rents makes a difference to GDP losses. U.S. losses are 25 percent higher in 2020 when market power resides with the sellers. The analysis provides an additional message for Annex I buyers. If at a given point in time, low-cost sellers are concentrated among a few countries (e.g., EEFSU), they may have considerable potential for extracting monopoly rents.

The Issue of Carbon Leakage 

The Kyoto Protocol refers specifically to the period centered about 2010. During this period, the onus for emission reductions falls on Annex I countries. We assumed that non-Annex I countries would be constrained to their business-as-usual baseline. In this section we relax this constraint. That is, no specific obligations are imposed on countries outside Annex I. There is now the possibility of “leakage.” That is, the reductions in Annex I might be partially offset by increased emissions from China, India, Brazil, and other countries that do not belong to Annex I. Energy-intensive industries would tend to migrate out of Annex I.

The model results suggest that the Protocol could lead to serious competitive problems for energy-intensive sector (EIS) producers in the United States, Japan, and OECD Europe. The Protocol would lead to significant reductions in their output and employment, and there would be offsetting increases in regions with low energy costs. U.S. output of energy-intensive products such as steel, paper, and chemicals could be 15 percent less than under the reference case by 2020 (see Figure 6). In contrast, countries such as China, India, and Mexico would increase their output of energy-intensive products. One can easily anticipate calls for protection against “unfair competition.” In its present form, the Protocol could lead to acrimonious conflicts between those who advocate free international trade and those who advocate a low-carbon global environment.

[Figure 6: Ratios of Domestic Energy-Intensive Sector Supplies to Demands: Kyoto With Leakage]

Evaluating the Kyoto Protocol in the Context of the Longer-Term Goal

The objective of the Framework Convention is “the stabilization of greenhouse gas concentrations at a level that would prevent dangerous anthropogenic interference with the climate system.”6 The drafters of the Kyoto Protocol focused exclusively on the initial steps to be taken by Annex I countries. Little attention was paid to the ultimate goal.

We now examine the Protocol in the context of a long-term stabilization objective. From Figure 1 it is clear that the Kyoto Forever scenario will fail to stabilize global emissions much less concentrations; that is, global emissions will continue to grow without developing country participation. A particular concentration target can be achieved through a variety of emission pathways. In this section, we explore three pathways for stabilizing concentrations at 550 ppmv (twice pre-industrial levels) by 2100. We stress, however, that the issue of what constitutes “dangerous interference” has yet to be determined. Indeed, it is likely to be the subject of intense scientific and political debate for decades to come. Hence, our choice of a target is meant to be purely illustrative.

The three pathway scenarios are intended to illustrate the benefits of “when” flexibility. They are titled: 1) “Kyoto followed by arbitrary reductions”; 2) “Kyoto followed by least-cost”; and 3) “least-cost.” As their names imply, the first two are designed to be consistent with the Protocol during the first commitment period. The third assumes a clean slate in the choice of emissions pathway throughout the twenty-first century.

For the first scenario, we assume that Annex I countries reduce emissions through 2030 at the same rate as the OECD during the first decade of the twenty-first century (2 percent per year). During this period, non-Annex I countries are permitted to emit up to their reference case levels. By 2020, emissions in the developing nations are larger than those in Annex I. We then choose a pathway to stabilization which represents a relatively smooth transition to the target. As for the post-2030 burden-sharing scheme, we assume that between 2030 and 2050 all regions move to equal per capita emission rights (based on their 1990 population). Equal per capita emission rights have been proposed as one approach to international fairness, but there are others that might also serve to separate the issue of equity from that of economic efficiency.

With “Kyoto followed by least-cost,” the Protocol is adopted for the initial commitment period. Thereafter, the most cost-effective pathway is followed for stabilizing concentrations at 550 ppmv. With “least-cost,” the most cost-effective pathway for stabilizing concentrations at 550 ppmv is followed from the outset. The latter two scenarios adopt the same proportionate burden-sharing scheme as the first.

All three scenarios assume Annex I trading plus the CDM. However, they differ as to the timing of developing country involvement in the international market for carbon emission rights. By definition, least-cost assumes that emission reductions will be made where it is cheapest to do so, regardless of the geographical location. Hence, in the least-cost scenario, we assume global emission trading from the outset. In the case of “Kyoto followed by least-cost,” we assume that global emission trading is delayed until after the first commitment period. With “Kyoto followed by arbitrary reductions,” global emission trading does not begin until 2030, the year that developing countries agree to lower their emissions below business-as-usual.

Global Carbon Emissions

Figure 7 shows global carbon emissions for the reference case and the three stabilization scenarios. Following a least-cost strategy from the outset results in an emissions pathway that tracks the reference path through 2010 and then departs at an increasing rate thereafter. There are several reasons why a gradual transition to a less carbon-intensive economy is preferable to one involving sharper near-term reductions.

[Figure 7: Global Carbon Emissions-Reference Case and Three Alternative Emission Pathways for Stabilizing Concentrations at 550 ppmv]

Concentrations at a given point in time are determined more by cumulative, rather than year-by-year, emissions. Indeed, a concentration target defines an approximate carbon budget, i.e., an amount of carbon that can be emitted between now and the date at which the target is to be reached. At issue is the optimal allocation of the budget. Reasons for relying more heavily on the budget in the early years include: 1) providing more time for the economic turnover of existing plant and equipment; 2) providing more time to develop low-cost substitutes to carbon-intensive technologies; 3) providing more time to remove carbon from the atmosphere via the carbon cycle; and 4) the effect of time discounting on mitigation costs.7

We next turn to the two scenarios where we adopt the Protocol for the first commitment period. Notice that the two emission pathways behave quite differently post-2010. “Kyoto followed by least-cost” follows the least-cost pathway once the Protocol’s constraints are relaxed. “Kyoto followed by arbitrary reductions,” on the other hand, bears no resemblance to the least-cost pathway. What is striking about Figure 7 is that with a 550 ppmv target, the Protocol is inconsistent with the cost-effective mitigation pathway, i.e., “least-cost.” Indeed, it appears that the ultimate target would have to be considerably lower than 550 ppmv for the Protocol to be justified in terms of cost-effectiveness.

Near-Term Losses

The incremental value of emission rights varies widely over time in our three stabilization pathways. Following the least-cost path, we estimate the value of an emission right at $11 per ton of carbon in 2010, rising gradually to $113 per ton in 2050. With “Kyoto followed by least-cost” actions, the value is $133 per ton in 2010, then falls to the least-cost path around 2020 and tracks it thereafter. In the case labeled “Kyoto followed by arbitrary reductions,” the incremental value of emission rights starts at $164 per ton and climbs to $296 per ton in 2050.

U.S. GDP losses in 2010 and 2020 also vary widely under the three stabilization scenarios (see Figure 8). Notice that GDP losses in 2010 differ for the two scenarios involving the initial adoption of the Protocol. Because of the long-lived nature of energy investments, investors are concerned both with what happens in the initial commitment period and what happens thereafter. In the case of the more rapid transition away from the baseline (“Kyoto followed by arbitrary reductions”), investors will be forced to invest more heavily in high-cost substitutes in the early years. With “Kyoto followed by least-cost,” they will have more flexibility.

[Figure 8: U.S. GDP Losses Under Three Alternative 550 ppmv Stabilization Pathways]

It is striking by how much GDP losses can be lowered under the “least-cost” scenario. This strategy involves a more gradual transition away from the baseline in the early years. It relieves much of the pressure for premature retirement of existing plant and equipment and for dependence on high-cost substitutes (both on the supply and demand sides of the energy sector). Relative to the reference case, the United States also receives some benefits as an oil importer. A carbon constraint decreases the overall demand for oil and lowers its price on the international market.

Global Losses

Finally, it is instructive to compare the global consumption losses through 2100 (discounted to 1990 at 5 percent) of “Kyoto Forever” and the three scenarios that stabilize concentrations at 550 ppmv. We estimate the global loss of “Kyoto followed by arbitrary reductions” at $2.4 trillion by 2100, “Kyoto followed by least-cost” actions at $890 billion, and the “least-cost” strategy at $640 billion (see Figure 9).

[Figure 9: Global Consumption Losses Through 2100 Discounted to 1990 at 5 Percent-Kyoto Forever vs. Three Scenarios for Stabilizing Concentrations at 550 ppmv]

What is surprising is that “Kyoto Forever” turns out to cost the global economy around $1 trillion by 2100, or more than a policy of “Kyoto followed by least cost” or “least cost” actions from the outset. Furthermore, “Kyoto Forever” may produce sharp global emission reductions during the early decades of the twenty-first century, but this strategy does not stabilize emissions, much less concentrations, by 2100. By contrast, the other scenarios all lead to stabilization at 550 ppmv. In short, “Kyoto Forever” costs more and buys less long-term protection. Non-Annex I countries must participate to achieve a least-cost result.


Some suggest that models such as MERGE tend to overestimate the costs of mitigation. They argue that, when prospects for technical progress are incorporated, the costs of a carbon constraint, even a sharp near-term constraint, will be minimal. We, too, are optimistic about the outlook for technical innovation. Indeed, such innovation is embedded both in our reference case and in the policy scenarios. The disagreement is over the rate at which such progress will occur. We do not believe that economically competitive substitutes will become available at such a rate as to trivialize the costs of a Kyoto-like Protocol.

A more valid concern may be that we are underestimating the costs of a carbon constraint. There are several reasons why this may be the case. To begin with, optimization models assume that decision makers have perfect foresight. That is, they assume that investors are fully informed about the nature of future constraints, and act accordingly. Given the present state of uncertainty, this is highly unlikely. Models such as MERGE also tend to ignore short-term macro shocks. For example, the higher energy prices brought about by a carbon constraint are likely to be inflationary. If this leads to higher interest rates, investment may be dampened. The result would be a slowdown in economic growth.

In addition, we assume that policies will be efficient. That is, market mechanisms will be chosen over command-and-control approaches to accomplishing environmental objectives. Whereas, in recent years, there has been an increasing trend toward market mechanisms, the approach to be taken with climate policy is by no means assured. Moreover, even if such a commitment were made, we have no assurances that the requisite international institutions will be available when needed.

Although it is easy to quibble over the numbers, the real value of analysis lies more in insights than in numerical values. We believe that the current exercise has yielded several important insights for policymakers:

  • First, it is extremely unlikely that a “Kyoto Forever” scenario will stabilize emissions-much less concentrations. Non-Annex I emissions are quickly overtaking those of the OECD and the economies in transition. Hence, meeting the stabilization goal of the Framework Convention will eventually require the participation of developing countries.
  • Second, international trade in emission rights is essential to reducing mitigation costs. The magnitude of the savings will depend upon the number of countries participating, the marginal abatement costs of participants, and the extent to which buyers can satisfy their obligation through the purchase of emission rights. For example, limiting a country’s emission purchases to one-third of its obligation could triple GDP losses in 2010.
  • Third, the Protocol’s near-term costs will depend heavily on expectations regarding future energy investments. Analyses that focus solely on 2010 may underestimate Kyoto’s costs.
  • Fourth, and most important, the Kyoto Protocol appears inconsistent with a cost-effective long- term strategy for stabilizing CO2 concentrations, unless the concentration target for CO2 is well below 550 ppmv.

Rather than requiring sharp near-term reductions, it appears that a more sensible strategy would be to focus on emission reductions at the point of capital stock turnover. This would eliminate the need for premature retirement of existing plant and equipment and would provide the time needed to develop low-cost, low-carbon substitutes.


  1. Intergovernmental Negotiating Committee for a Framework Convention on Climate Change, Fifth Session, Second Part, New York, 30 April-9 May 1992, states that it is essential that “policies and measures to deal with climate change should be cost-effective so as to ensure global benefits at the lowest possible costs.”
  2. Conference of the Parties, “Kyoto Protocol to the United Nations Framework Convention on Climate Change,” Third Session, Kyoto, 1-10 December 1997.
  3. MERGE covers nine geopolitical regions and combines a bottom-up representation of the energy supply sector with a top-down perspective on the remainder of the economy.
  4. R. Richels, J. Edmonds, H. Gruenspecht, and T. Wigley, “The Berlin Mandate: The Design of Cost-Effective Mitigation Strategies,” EMF-14, Working Paper 3, Stanford University, Stanford, Calif., 1996.
  5. Conference of the Parties, “Kyoto Protocol to the United Nations Framework Convention on Climate Change,” Third Session, Kyoto, 1-10 December 1997.
  6. Intergovernmental Negotiating Committee for a Framework Convention on Climate Change, Fifth Session, Second Part, New York, 30 April-9 May 1992.
  7. T. Wigley, R. Richels, and J. Edmonds, “Economic and Environmental Choices in the Stabilization of Atmospheric CO2 Concentrations,” Nature 379 (18 January 1996).