Negotiating Strategies for a Post-Kyoto Regime

Glacier lagoon - jökulsarlon
Negotiating Strategies for a Post-Kyoto Regime - Scott Barrett


What should we do about climate change? There is disagreement about what we should do in quantitative terms, but universal agreement about what we should do in qualitative terms. Our aim should be not only to avoid “dangerous interference” with the climate system, but to effect a technological revolution. How can these goals be achieved? The European Union’s favored approach is to tighten up on the Kyoto emission cuts, possibly supported by the application of trade restrictions. A better approach would be to break the problem up into its many parts, using the best means available to enforce each part.


Al Gore and the Intergovernmental Panel on Climate Change received the Nobel Peace Prize in 2007 “for their efforts to build up and disseminate greater knowledge about man-made climate change, and to lay the foundations for the measures that are needed to counteract such change.” Laying the foundations for taking action is an essential and important contribution, but now comes the much harder task: deciding what to do and, more importantly, figuring out a way in which the world can cooperate to address this colossal challenge.

In this paper, I address both of these matters. I explain that the question of what we should do about climate change is both more complicated and simpler than most discussions acknowledge. In brief, there is disagreement about what we should do in quantitative terms, but near universal agreement about what we should do in qualitative terms. Eventually, addressing climate change will require a technological revolution.

We have already tried one approach to addressing climate change—the targets and timetables approach embodied in the Kyoto Protocol. It is often suggested that we need to continue to rely on this approach but that Kyoto’s targets and timetables need to be tightened up (this is the European Union’s current negotiating position). This thinking misinterprets the reasons for the current failures. It is true that more needs to be done to address climate change, but the Kyoto approach is unable to enforce its own obligations. The strategy of using trade restrictions to enforce Kyoto is fraught with difficulties. A better approach is to break the problem up into parts, using the best means available to enforce each part.

What Should We Do?

You might think that the world cannot agree what to do about this challenge, and that this is why progress has been blocked; but, actually, the world agreed in 1992 what to do about climate change. The Framework Convention on Climate Change says that the world should avoid “dangerous anthropogenic interference with the climate system,” and do so “within a time frame sufficient to allow ecosystems to adapt naturally to climate change, to ensure that food production is not threatened and to enable economic development to proceed in a sustainable manner.” Nearly every country is a party to this treaty—the only non-parties today are Andorra, the Holy See, Iraq, and Somalia. So the world agrees.

There are, however, two problems. First, agreeing what to do is only a necessary condition for addressing a global challenge; it is not sufficient. In 1959, a proposal was made to the World Health Assembly that the world should eradicate smallpox. Nearly every country was represented on this assembly, and every member country voted in favor of this resolution. Even during the Cold War, there was global unanimity on this one issue. However, achieving the goal required resources, and these, countries agreed, should be supplied voluntarily. Analysis shows that the benefit-cost ratio for smallpox eradication was close to 500:1, making this perhaps the best public investment the world ever made. The budget for eradication was $100 million. So, how much money do you think was raised voluntarily from 1959-1966 to finance the supply of this global public good? The answer: just $27,345. If anyone tells you that free riding is not a problem, you might remember this story.1

So the first problem is that agreeing on what to do is a lot easier than getting countries to do what is required to achieve a collective goal. Indeed, this is why so many countries signed up to the Framework Convention. The agreement does not require that these countries do anything.

The second problem with the Framework Convention is this: Who could be against avoiding “dangerous anthropogenic interference with the climate system?” No one. And yet the agreement does not explain what “dangerous” means. Different countries might have different ideas about this.

For example, in a recent analysis of the effects of climate change on agriculture, William Cline estimates that, by around 2085, “agricultural output potential” can be expected to fall in many countries. In India, Cline predicts, it will fall twenty-nine percent; in Mexico, twenty-six percent. In China, however, agricultural output potential is expected to rise seven percent. It is also expected to rise in Canada, Europe, Russia, and the United States. Overall, agriculture is not affected much—by Cline’s calculations, global output potential declines only about three percent—but there will be winners as well as losers.2 To India and Mexico, Cline’s predictions will surely seem dangerous. Other countries may view his predictions with less alarm.

In 1996, the Council of the European Union declared that mean global temperature should not be allowed to increase by more than 2º C above the preindustrial level. It did so on the basis that, “once global warming exceeds 2ºC, climate impacts on food production, water supply and ecosystems are projected to increase significantly and irreversible catastrophic events may occur.”3 Is this the right level to aim for? It would be nice if the answer were a clear yes. But it isn’t obvious that this level is enough. For example, coral reefs are highly sensitive to temperature; a 1º C increase is commonly mentioned as a threshold.4 As well, a 2º C warming might trigger positive feedbacks, tipping the world toward “climate catastrophe.” If this increase were to cause permafrost to melt, releasing methane, the temperature increase would not stop at 2º C; it would climb higher.

Another problem is that we do not know the concentration level needed to limit temperature change to 2º C. According to the latest assessment by the Intergovernmental Panel on Climate Change, a concentration of 450 parts per million (ppm) of carbon dioxide (CO2) equivalent would probably limit temperature increase to 2.1º C, but the likely range is 1.4-3.1º C.5 Given this uncertainty in climate sensitivity, limiting concentrations to 450 ppm may mean overshooting the 2º C goal.

The 2º C target might also be too strong a target. If the consequences of meeting this target were inconsequential, I would say we should do more—after all, even a limit this low is much higher than experienced during the whole of human existence.

But reducing emissions this much, this soon will be consequential. It will be costly; the money spent reducing emissions could be spent on other worthwhile causes. Reducing emissions will also introduce additional risks. For example, the reduction in emissions needed to limit climate change to 2ºC would almost certainly require a radical expansion in nuclear power worldwide, and this would pose problems for waste storage—like climate change, a burden for future generations—and proliferation.

Economists, when they think about this problem, look at both sides of the climate ledger—the costs and the benefits. As is often the case, however, they draw different conclusions from their analyses. Many economists, led by the Yale economist, William Nordhaus, believe we should reduce emissions somewhat today, and “ramp up” emission reductions over time. Other economists, especially Nicholas Stern, now of the London School of Economics, believe we should do much more to reduce emissions now; according to Stern, “the benefits of strong, early action considerably outweigh the costs.”6 Stern also believes that atmospheric concentrations should be prevented from exceeding twice the pre-industrial level.7 Nordhaus, by contrast, calculates that the costs of limiting concentrations to such a level exceed the benefits.

The reasons for these different conclusions have been written about elsewhere.8 Here I want to emphasize their three points of agreement.

First, were we to know with certainty that a concentration level above some threshold would be truly catastrophic, then Nordhaus and Stern would disagree only about the path of emission reductions that would be best for avoiding catastrophe. Both economists would agree that catastrophe ought to be avoided.9 Second, both economists agree that virtually nothing is being done to reduce greenhouse gas emissions today, and that more should be done. Finally, they agree that concentrations should not be allowed to increase indefinitely. This last point of agreement is important because the physics of climate change tell us that if concentrations are to be stabilized at any level, net emissions to the atmosphere must equal zero. The long run goal must therefore be to reduce net emissions to zero—a challenge that will require a technological revolution.

Let me now turn to these two imperatives—the imperative to reduce (net) emissions now, and the imperative to reduce them by much, much more eventually.

Reducing Emissions

Because of the free rider problem noted previously, countries will not do enough to reduce their emissions unilaterally: the global benefit of reducing emissions exceeds the individual benefit. Making matters worse, should only some countries reduce emissions, comparative advantage in the carbon-intensive industries may move to the other countries, causing their emissions to increase— a phenomenon known as “trade leakage.” It is really for both reasons that so little action has been undertaken so far.

One proposal for getting around the “collective action problem” is for a country or group of countries (the European Union) to “take the lead.” The idea is that, should certain countries lead by reducing emissions, others will follow.

Leadership works for some international issues. For example, as more countries adopted the Greenwich meridian, yet more countries wanted to do so, because of the overriding need for countries to coordinate as regards time keeping. Unfortunately, reducing emissions does not have this same characteristic.10 Another approach is for each country to make its own climate policy contingent on those of other countries. This approach has also been tried several times, without success. In 1992, for example, the European Community proposed a mix of measures, including a carbon/energy tax, to stabilize EC emissions at the 1990 level. This tax was made contingent on other OECD countries adopting the same tax. The other OECD countries rejected the offer; the EC dropped the tax.

A third approach is to try to “bind” countries to act in a treaty. This, of course, was the reason so many negotiators traveled to Kyoto in December 1997.

Kyoto Protocol

An effective international agreement for climate change mitigation must do three things. Kyoto, unfortunately, does none of them.

First, a treaty must create incentives for broad participation. Kyoto failed to deter the world’s biggest emitter and only superpower, the United States, from not participating—reason enough, in my view, to call the agreement a failure.

Why do I blame the agreement rather than George W. Bush? One reason is that it is very unlikely that the Kyoto Protocol would have been ratified by the United States even if Al Gore had been selected as president in 2000. In July 1997, the Senate passed a resolution 95 to zero against the United States becoming a signatory to any international agreement on greenhouse gas emissions that (among other things) did not also require developing countries to reduce their emissions. Before this, the Clinton administration had agreed to negotiate on the basis that the industrialized countries must reduce their emissions first. The Senate resolution did not change the negotiating stance of the United States. Al Gore traveled to Kyoto to break a deadlock in the negotiations, but the treaty he negotiated failed to meet a pre-condition for ratification expressed previously by the Senate. Treaty ratification requires 66 affirmative votes in the Senate. It is almost inconceivable that a Gore administration could have orchestrated a turnaround of this magnitude.

The Senate, in my view, was justified in insisting on broad participation. It is easy to see this in hindsight. The countries that matter most are the countries whose emissions are growing the fastest—countries like China and India. They matter the most not only because of their size (today, China is the world’s largest emitter) but because they are currently investing in a capital base that will last a very long time. China is bringing on line at least one coal-fired power plant a week. (In a single year, China has added more coal-fired capacity than the entire installed capacity of the United Kingdom.) These plants last 40 years or more. The challenge was, and still is, to get the fast growing developing countries onto a new kind of development path as a priority, and for the rich countries to transition onto this path over time as their capital base depreciates. By design, Kyoto fails to do this.

Second, a treaty must create incentives for compliance. Canada, unlike the United States, is a party to the Kyoto Protocol. It is required to reduce its emissions 6 percent below the 1990 level through 2008-2012. In 2005, however, Canada’s emissions were 30 percent above this target, and Canada’s current government has given up on the idea of meeting the Kyoto target. It aims instead to reduce the rate of growth in emissions, hoping that Canada’s emissions will peak around 2010, the mid-point of Kyoto’s implementation period. A governmentfunded roundtable, however, has concluded that the government’s policies will not meet even this modest goal.11 Canada’s previous government—the one that ratified Kyoto but that did not adopt the policies needed to comply with it—predicted that Canada’s emissions would exceed its Kyoto target by 45 percent by 2010.12 This still remains distinctly possible.13 Canada’s situation is extreme, but it is not the only Kyoto party in jeopardy of non-compliance. The compliance gap is significant for many countries, including New Zealand and Japan. The caps for some members of the European Union are also challenging. Spain has the largest gap of any country. Denmark is well off its individual target. However, thanks to the European “bubble,” these countries are not bound by their individual limits so long as the original fifteen members of the European Union meet their collective limit (this is an advantage unique to the European Union). Australia recently ratified the Kyoto Protocol, but after taking land use, land-use change, and forestry activities into account, Australia is currently within its Kyoto limit; Australia will have to do very little to comply.14

Third, an effective agreement must require that countries take real action. Kyoto asks very little of its parties.  It requires that a relatively small number of countries reduce their emissions very slightly for a very short period of time. Even if Kyoto were implemented to the letter, global emissions will keep on rising. So will atmospheric concentrations of greenhouse gases.

I have heard seasoned negotiators say that the problem with Kyoto is that it did not ask countries to cut back on their emissions by enough. It is true that much more needs to be done. But this is not the only problem with Kyoto. Would the United States have been more inclined to participate if its emission cap were even tighter? Would Canada have been more likely to comply? I don’t think so. The real problem with Kyoto is that it does not create incentives for countries to participate and comply, while at the same time requiring that they cut back on their emissions substantially. All three of the above requirements have to be met simultaneously. Meeting only one or two will not do.

Montreal Protocol

While the world’s attention has been on Kyoto, another agreement has been working quietly behind the scenes to make a material difference. This is the Montreal Protocol—the agreement created to protect the ozone layer, not to limit climate change.

The connection between climate change and ozone depletion is complicated. Ozone-depleting substances are greenhouse gases, but so is stratospheric ozone. In addition, many of the substitutes for ozone-depleting substances are greenhouse gases, including the so-called HFCs (HFCs are controlled by the Kyoto Protocol). So the Montreal Protocol has both positive and negative consequences for the climate.

A recent study has shown that the net effect of the Montreal Protocol on greenhouse gases is helpful.15 Indeed, the study calculates that the Montreal Protocol has been, and will continue to be, much more helpful than the Kyoto Protocol in addressing global climate change, even assuming that Kyoto is implemented perfectly and with full participation. Already, this study estimates, the Montreal Protocol has achieved four times as much as the Kyoto Protocol could ever hope of achieving.

Indeed, in late 2007, the Montreal Protocol was revised again. This time, the agreement to phase out HCFCs was accelerated. HCFCs are a greenhouse gas. Moreover, their manufacture produces HFCs, another greenhouse gas, as a byproduct. Preliminary estimates suggest that the agreement negotiated in Montreal in September 2007 will do more to limit greenhouse gas concentrations than the Kyoto Protocol would have done, had it worked as intended. This is on top of the larger effect (noted above) Montreal has already had in limiting the concentration of greenhouse gases.

What is the Montreal Protocol’s secret for success? There are many differences between Montreal and Kyoto. Here I shall mention just four. First, Montreal required all countries to cut back on CFCs from the very beginning; developing countries were given a “grace period,” but they were expected to “catch up” to the limits borne by industrialized countries. Second, the Montreal limits apply to production and consumption whereas the Kyoto limits apply only to the emission of greenhouse gases in production. China’s rapid emissions growth partly reflects a shift in manufacturing from rich countries to China. Under Kyoto, China’s emissions are unconstrained. Under Montreal, not only was China subject to a production limit, but other countries were subject to limits on imports of CFCs. Consumption limits dampen the leakage problem. Third, the Montreal caps are permanent, whereas Kyoto’s, as mentioned before, last only five years. Montreal created an expectation that the market for CFCs would disappear. It therefore shifted the attention of producers to discovering and manufacturing substitutes. Kyoto creates no such expectation for any of the greenhouse gases, and so has stimulated very little innovation. Finally, Montreal created strong incentives for both participation and compliance. These include “carrots” in the form of financial payments from rich to poor countries, and “sticks” in the form of a trade restriction as between parties and non-parties.

The trade restrictions are particularly important, and they have been a great success despite never being used. The reason for their success is that the credible threat to use the restrictions has been enough to deter non-participation and non-compliance.16

Trade Restrictions

Why not use trade restrictions to enforce Kyoto? President Nicolas Sarkozy of France has suggested recently that trade restrictions be considered. So has the Nobel prize-winning economist, Joseph Stiglitz.17

Kyoto—particularly a strengthened version of Kyoto—can probably only be enforced using trade restrictions.  However, I have yet to be convinced that trade restrictions will make Kyoto a more effective agreement or make the world any better off. My objections are practical rather than ideological. International trade and global climate change are both collective action problems and they are interrelated. If trade restrictions could help supply the global public good of climate mitigation, their use should be considered, even if trade liberalization would be harmed. We need to consider the two global challenges together rather than separately.

However, there are problems with the proposal. Three seem especially important: First, Montreal restricted trade in CFCs and products containing CFCs. It did not restrict trade in products made using CFCs. The reason: specialists who looked into this decided it could not be done. Moreover, they decided, correctly it turns out, that the other restrictions would suffice to enforce the agreement. If trade restrictions were used to enforce the Kyoto Protocol, they would have to apply to products that, when produced, resulted in the release of greenhouse gases. That is, Kyoto would have to limit trade in all products. Calculating the emissions resulting from the production of every product would be a colossal task (two identical products, manufactured in the same country, may have very different “carbon footprints”).

Second, a choice has to be made as regards the level of the restrictions. A border tax adjustment could neutralize trade leakage. But cooperating countries will want to raise the tax to deter free riding. What, then, is the appropriate penalty? A very high tax (or even a trade ban) could ensure full participation. However, a high level may not be credible (it would hurt the countries imposing the penalty as well as those on the receiving end); larger penalties tend to be less credible.

Third, who should decide what it is that countries should be required to do? Suppose the agreement were perceived as being unfair by some countries. Suppose, for example, that the agreement required all countries cut back on their emissions by 25 percent. We know that such a uniform cut in emissions would have very different implications for different countries. Some would find the target much harder to meet than would others. If a country objects, and chooses not to be a party to the agreement, should the other countries be permitted to impose trade restrictions against this country? This is a question of legitimacy. It seems to me very likely that, should trade restrictions be used to enforce an agreement that lacks legitimacy, the outcome may very well be a trade war.

A New Approach

How to address the enforcement challenge? One way is for countries to declare their unilateral goals in a multilateral treaty that deliberately lacks enforcement. This approach had modest success in addressing acid rain.18 The idea would be to turn the problem of enforcement over to domestic institutions, which are designed for this purpose, and to let the agreement serve as a basis for determining “comparability.”

The Montreal Protocol points to another way to address climate change. Currently, the Kyoto Protocol combines all greenhouse gases in the same basket, as it were. This makes sense from the perspective of cost-effectiveness. However, the ability to enforce a comprehensive agreement depends on the weakest link. The three industrial gases in the Kyoto Protocol (including HFCs), bear a similarity to the industrial gases controlled by the Montreal Protocol. With hindsight, it would have been better to control these in a separate agreement, styled after Montreal. This agreement could impose obligations on all countries, but incorporate a mechanism for financing cuts in developing countries (in the Montreal Protocol, this is accomplished by an Ozone Fund). The offer of financing would provide a means for addressing the legitimacy problem mentioned previously. A limited agreement like this could be enforced using trade restrictions. The aim of these restrictions, like those in Montreal, would be to deter non-participation and non-compliance.

For the main greenhouse gas, carbon dioxide, reductions in the emissions of individual sectors may be easier to enforce than economy-wide limits. The aluminium sector is a prime candidate for a separate agreement.19 It is a concentrated industry: twelve countries account for 82 percent of global production; ten companies produce more than half of world output. The industry employs just two smelting technologies, and emissions can be reduced substantially by remelting aluminium scrap (the former is 95 percent less greenhouse-gas-intensive than primary aluminium production). Finally, twenty-six companies, making up 80 percent of world output, belong to the International Aluminium Institute, which has already adopted voluntary intensity targets. There exists a basis here for negotiating new global standards for the industry, backed by international enforcement. Other candidates for sectoral agreements include steel and cement.20

As noted earlier, addressing climate change fundamentally requires a technological revolution—that is, it requires basic research, development, and diffusion. Kyoto creates very modest “pull” incentives for innovation. Even if these incentives were stronger, however, pull incentives alone will not be able to stimulate a technological revolution. To do that will require basic knowledge—the kind that cannot be patented. Supplying basic knowledge will likely require government funded R&D—a “push” incentive. The returns to investing in basic knowledge, however, will depend in turn on the prospects of technologies embodying this new knowledge being diffused. This is why technology R&D efforts need to be strategic.21 They need to be directed at technologies that will be diffused. Some R&D can be undertaken unilaterally (when the technology offers national benefits in addition to those related to climate change mitigation). Some R&D will require international cooperation. An example of the latter kind, with an uncertain and much delayed payoff, is the International Thermonuclear Experimental Reactor being built now in France. One technology that needs to be developed and demonstrated as a priority is carbon capture and storage.

How to promote technology diffusion? Diffusion depends on certain features of technologies. Transportation exhibits “network externalities,” making the incentive for each country to adopt a new technology increase in the number of other countries that adopt the same technology. It is through network effects that the combination of a catalytic converter and unleaded gasoline became a global standard for the automobile.

Though emissions from international marine transport are excluded from Kyoto, this sector may be an attractive candidate for switching to hydrogen fuel.22 The economics of hydrogen for automobile transportation are less attractive, but this sector also exhibits strong network effects, particularly as regards refuelling. Moreover, substituting an alternative fuel for petrol may yield local environmental benefits—a further inducement for technology spread. Currently, the electric vehicle seems to be a particularly attractive option, with the plug-in hybrid possibly acting as a kind of bridge to a full-electric future. Network effects would include the availability of electrical outlets for recharging and of replacement batteries for long-haul travel. Of course, a switch to electric vehicles makes it even more imperative that emissions from electricity generation be cut very substantially. Unfortunately, electricity production does not exhibit network effects. This is partly why the economy wide emission limits may be useful, even though they would need to be enforced domestically.


Climate change is the most difficult collective action challenge the world has ever faced. It should not surprise us that we have failed to address it thus far. However, to ignore the lessons of our previous failures would be folly; to repeat them would be worse. The approach of setting targets and timetables for emission reductions without putting in place incentives for real changes in behavior has not worked. Setting even stronger targets and timetables in a post-Kyoto agreement will not work either. In this paper I have explained the reasons for this. I have also argued that it is unlikely that a Kyoto-style agreement can be enforced internationally. By design, a failure of any of its components brings down the entire structure. I have suggested a different strategy for a future climate regime, one that unbundles the problem, developing a multiple of linked agreements capable of addressing different dimensions of the problem. This will not sustain a perfect outcome, but it will improve upon the current arrangement and provide a platform upon which more improvements can be added as we learn more about how to address this critical and complex global challenge.

Notes & References

  1. Of course, the free rider problem was corrected in this case; smallpox was eradicated.I tell this story, and many others like it, in my book, Why Cooperate? The Incentive to Supply Global Public Goods (Oxford: Oxford University Press, 2007).
  2. See William R. Cline Global Warming and Agriculture: Impact Estimates by Country (Washington, DC: Peterson Institute, 2007). I am relying here on Table 5.9, on page 77 of this monograph. These estimates include the effect of higher CO2 concentrations in stimulating phyotsynthesis.
  3. “Questions and Answers on the Commission Communication, Limiting Global Climate Change to 2 ºC,” Memo/07/17.
  4. Note that a substantial percentage of corals are already severely damaged and thereare no more “pristine” coral reefs left, mainly as a result of over-fishing and land-based pollution. Limiting climate change is essential to protecting the coral reefs, but it is not sufficient.
  5. Intergovernmental Panel on Climate Change, Fourth Assessment Report, Working Group I: The Physical Science Basis of Climate Change, Technical Summary, 2007, Table TS-5; available at
  6. See Nicholas Stern, Stern Review: The Economics of Climate Change (Cambridge: Cambridge University Press, 2007, p. ii).
  7. In particular, Stern recommends that we stabilize atmospheric concentrations at alevel not greater than 550 ppm CO2 equivalent (a target expressed in terms of the concentration of all greenhouse gases and not only carbon dioxide. Today’s level is about 430 ppm CO2e; as mentioned before, CO2 concentrations are around 385 ppm today.) This level mirrors the European Union’s own recommendation for a target intended to limit temperature rise to 2°C.
  8. See, for example, Chapter 3 of my book, Why Cooperate?
  9. There is, of course, a risk of abrupt and catastrophic climate change, but we do notknow the level of concentrations that would trigger such change. Nor do we know how such change will be expressed. This uncertainty is the feature of climate change that most worries another economist, Martin Weitzman of Harvard University. His analysis of climate change urges extra caution. See weitzman/files/REStatFINAL.pdf.
  10. See Chapter 6 of my 2007 book, Why Cooperate? As explained in this chapter and later in this paper, coordination is relevant to the choice of some technologies for reducing emissions.
  11. For the relevant section of the roundtable’s report, see
  12. Government of Canada, Moving Forward on Climate Change: A Plan for Honouring Our Kyoto Commitment (Ottawa: Government of Canada, 2005). 
  13. The current government’s projections are that Canada will exceed Kyoto’s limits byaround 34 percent; again, see
  14. Land use, land use change and forestry is normally treated differently because ofvarious accounting and incentive problems. For example, carbon accumulated in forestry may later be released.
  15. Guus J.M. Velders, Stephen O. Anderson, John S. Daniel, David W. Fahey, and MackMcFarland, “The Importance of the Montreal Protocol in Protecting Climate.” Proceedings of the NationalAcademy of Sciences, 104 (2007): 4814-4819.
  16. See Scott Barrett, Environment and Statecraft: The Strategy of Environmental Treaty-Making  (Oxford: Oxford University Press, paperback edition, 2007); see especially Chapter 12.
  17. Joseph E. Stiglitz, “A New Agenda for Global Warming,”The Economists’ Voice 3 (2006): Art. 3; available at:
  18. See Marc A. Levy, “European Acid Rain: The Power of Tote-Board Diplomacy,” in Peter M. Hass, Robert O. Keohane, and Marc A. Levy (eds.), Institutions for the Earth: Sources of Effective International Environmental Protection, (Cambridge, MA: MIT Press, 1993): 75-132.
  19. I am drawing here from the excellent study by Rob Bradley, Kevin A. Baumert, BrittChilds, Tim Herzog, and Jonathan Pershing, Slicing the Pie: Sector-Based Approaches to International Climate Agreements (Washington, DC: World Resources Institute, 2007): 3738.
  20. Again, see the report by Bradley et al., cited in the previous footnote.
  21. Scott Barrett, “Climate Treaties and ‘Breakthrough’ Technologies, “American Economic Review (Papers and Proceedings), 96 (2006): 22-25.
  22. Alex E. Farrell, David W. Keith, and James J. Corbett, “A Strategy for Introducing Hydrogen into Transportation,” Energy Policy 31 (2003): 1357-1367.
Scott Barrett is Professor of Environmental Economics and International Political Economy at the Johns Hopkins University School of Advanced International Studies in Washington, DC. He has been a lead author for the Intergovernmental Panel on Climate Change and an advisor on climate change negotiations to several international organizations. He is the author of Environment and Statecraft: The Strategy of Treaty Making and, most recently, Why Cooperate? The Incentive to Supply Global Public Goods, both published by Oxford University Press.