Colorado Journal of International Environmental Law and Policy, vol. 4, no. 2 (1993), pp. 344-368
By David Hurlbut
The diplomatic success of the 1987 Montreal Protocol to protect the stratospheric ozone layer proceeds from a cohesive new philosophy of international relations with which future environmental treaties will have to conform if they are to be more than hortatory declarations. Some of the features of this new world view: science is an increasingly important policy impetus, state sovereignty is less of an obstacle to multilateral action, the need to distribute the burden and benefits of environmental protection in an equitable way is implicit, and advanced industrial countries acknowledge the special needs of developing countries when environmental protection efforts affect economic productivity.
Balancing new rights and obligations enhances the legitimacy of both. When an environmental protection regime does not equitably distribute new rights and obligations-that is, when burden falls heavier on some states than on others-states who see themselves on the losing end of negotiations will have an incentive to block multilateral action by regressing to obstructionist claims of sovereignty. This has often been the pattern of tension between industrial and developing countries, and can destroy the consensus necessary for international environmental protection measures to work. When new rights and obligations are equitably distributed, however, there is greater sense of fairness among the parties and a greater likelihood of consensus.
This paper will argue that the cohesiveness of the principles underlying the Montreal Protocol ought to make it possible to extend the reach of the protocol's environmental protection regime to nations who have not yet acceded to the agreement. A corollary argument is that recent environmental treaties such as the 1992 Convention on Biological Diversity are weak because they ignore these same principles. (The effect the protocol has had on actually reducing the rate of ozone deterioration is outside the scope of what this paper is intended to accomplish. The purpose here is to evaluate how international law affects the design of strong environmental agreements.)
As of November 1992, ninety states had acceded to the Montreal Protocol. Of that number, more than half are developing countries. This paper explores the key elements of this remarkable consensus first by observing that as science became more integral to the negotiations, states were less inclined to use arguments of sovereignty to obstruct agreement on control measures. Next, the paper discusses the protocol's recognition of the special situation developing countries face. These two principles form the core of consensus from which three specific issues arise:
b. The extent of a country's obligation to cooperate in the gathering and sharing of data relevant to environmental protection despite national laws and policies that may restrict access to that data; and
c. The extent of a developing country's right to economic assistance when faced with international obligations that restrict its right to exploit its own resources.
Finally, all of the above issues will be examined in the context of (a) countries that are not parties to the Montreal Protocol and (b) other environmental protection treaties. What this paper aims to prove is that the Montreal Protocol is a strong statement of international consensus whose effect reaches beyond the protocol itself, and that certain important principles of a new international order are at work. Other environmental treaties are bound to fail if they attempt merely to ride the coattails of the Montreal Protocol's diplomatic success without emulating the principles behind it.
Science initiated the policy debate about ozone depletion in 1974 when atmospheric researchers at the University of California at Irvine called attention to possible damage to the ozone layer from chlorine- and bromine-bearing chemical compounds-chlorofluorocarbons (CFCs) and halons in particular. Diplomats met in Vienna in 1985 for the purpose of adopting an agreement to limit CFCs and other ozone-depleting substances. At the time of the Vienna meeting, however, there was still uncertainty about what the scientific evidence exactly meant. If the theories being weighed at that time were accurate, significant ozone depletion was already taking place, and failure to enact the strongest possible restrictions could result in deaths and serious damage to the human environment for most of the next century and beyond. The theoretical consequences of ozone depletion were never at issue, but there was no reliable estimate as to the probability of the worst-case scenario ever occurring and no reliable data to show that deterioration had even begun.
How much could international control measures demand in light of such uncertainty? Lacking clear guidance from the scientific community, customary law could say little. The evolution of the public trust doctrine in Europe and the United States established a state's continuing sovereign control over common resources (or resources whose disposition affects the well-being of society at-large) even if the state transfers rights of ownership or usufruct to private hands. Judicial concern historically has been over access to navigable waters and rights-of-way for public transportation, but more recent decisions have said that maintaining the ecological integrity of common natural resources are legitimate applications of the public trust doctrine. The test of whether a state is justified in altering rights under the doctrine is whether the change significantly improves the well-being of society (or, conversely, avoids social harm).
Consequently, the public trust doctrine would be applicable when the integrity of the ozone layer is affected. But what if the ozone layer might be affected? Some commentators contend that the public trust doctrine permits the state "to wield immense power to restrict and rearrange property rights and expectations," and that it is "appropriate for a court to err on the side of protection" when "the consequences of damage to public trust resources are so grave that contemporary society and future generations cannot permit such degradation 'accidentally or carelessly'." However, no recent court decision explicitly addresses the question of whether the doctrine applies when harm is possible but not certain. Under the most recent court interpretations, the public trust doctrine merely affirms the state's authority to act when necessary. None of the decisions affirms the state's authority to decide arbitrarily what constitutes "the public need" and to remain insulated from challenge. That question remains open.
At the time of the 1985 Vienna Convention, the operative precedent was still the cautious conservatism of the landmark Trail Smelter arbitration: collect all relevant scientific and technical data first, and then make a decision drawn directly from the data. This was logical for the Trail Smelter case, in which adherence to the rule of certainty affected nothing more than the size of the award to the injured party. Yet the problem of ozone depletion presented customary law with a dilemma: as the possible consequences of harm spread globally and increased geometrically, the logic of insisting on scientific certainty became increasingly tenuous. The social cost of error could be very high in terms of increased incidents of cancer and cataracts, and decreased agricultural productivity.
International "soft law" (treaties, international declarations, etc.) did not leave environmental protection in the straightjacket of certainty, however. By the time of the Vienna Convention, many international declarations on sea pollution and hazardous wastes had embraced the precautionary principle as a guide for environmental law. This principle means that "precautionary action must be taken to ensure that the loading capacity of the environment is not exhausted, and it also requires action if risks are not yet certain but only probable, or, even less, not excluded." The precautionary principle was already an accepted rule in a number of treaties and national statutes by the time of the Vienna Convention, thus opening the possibility of restrictive measures in the absence of scientific certainty.
In Vienna, the European Community and Japan argued that no restrictions could be established until there was positive scientific evidence of significant ozone loss due to anthropogenic chemicals. The United States, Canada, and the nations of Scandinavia advocated precautionary action, arguing that "the margin of error between complacency and catastrophe is too small for comfort." The final compromise struck a middle ground: the parties committed themselves to the general goal of reducing activities that harmed the ozone layer, and agreed to further research that would be used to decide specific control measures when the parties were to reconvene in Montreal two years later. Afterwards, as the scientific evidence accumulated and it became apparent that the mathematical models had in fact underestimated the loss of ozone, a strong consensus began to form within the framework set by the 1985 Vienna Convention to establish strong control measures. In Montreal, Germany eventually succeeded in bringing the rest of the European Community into the camp supporting cutbacks for six types of CFCs and two halons. Developing countries led by Mexico, Egypt, and Ghana also joined the effort, and at the same time they succeeded in obtaining a provision that allowed developing countries an extra ten years to meet the protocol's production and consumption limits.
Although advocates of the precautionary principle did not win the day in Vienna, they were vindicated by the time the parties met in Montreal to draft a specific protocol. As the parties became more comfortable with the precautionary approach, and as the technical assessment panels began providing a clearer understanding of the problem of ozone depletion, states were less inclined to assert their sovereign right to pursue independent courses of action.
One aspects of this consensus clearly indicates that something was subordinating sovereignty: the simple fact that each country would have been economically better off had it rejected the Montreal Protocol itself and relied on all other states to enact the control measures. Ozone protection presents the classic "free rider" dilemma of group action. That is, if it is generally believed that the benefits of collective sacrifice will accrue to everyone-including those who make no sacrifice-each member of the group will have an incentive to do nothing. The common interest would be harmed because, ultimately, nobody would do anything. Yet this did not happen in Montreal; a significant number of states agreed to take on the burden of phasing out ozone-depleting substances.
Altruism may have had a role in the decision of some states to address the problem of ozone depletion cooperatively, but utilitarian motives were also at work. Game theory can help explain both the pragmatic reasons behind the Montreal Protocol's consensus and the role of science in bringing the consensus about. In the game theory framework, each state is regarded as a "player" whose goal is to maximize its own gain (or minimize its loss) by means of a strategy that guides all choices it makes. The entire endeavor of reducing ozone depletion (the "game") is characterized by the information each state has prior to any given position it takes, by the propensity of states to form coalitions within the negotiations, by what states are and are not allowed to do under international law, by the relevant technological constraints, and by what each state expects to gain from the various alternatives available to it.
The course of play within the negotiations was affected at various times by the introduction of new information (on measured ozone depletion, availability of new technologies, and economic assessments, for example) available to all the parties. The accumulation of scientific evidence had a number of effects on the negotiations that were predictable in the game theory model. Generally, new information reduced the uncertainty associated with the effects of ozone deterioration, and as the probability of harm became more fixed, a country's expected value of the loss it could suffer as a result of ozone deterioration increased. Technological research also provided better estimates of the costs involved in restricting ozone-depleting substances. As new information changed the pattern of what states expected to gain as a result of various alternative positions, the utility of cooperation increased and that of non-cooperation decreased. Coalitions changed, but more importantly the compensation needed to draw and keep players in the main coalition (the one in favor of universal control measures on CFCs) also diminished. For a number of developing countries, this meant that cooperating and bargaining for special consideration carried more benefit than staying outside the agreement. For the coalition of states committed to the protocol's objectives, it was worth it to give developing countries the 10-year grace period in exchange for their cooperation. Trade restrictions in the protocol further served to discourage states from leaving the coalition once they joined.
The information provided by the technical assessment panels was thus the mortar that held consensus together. Science kept all the parties equally informed about the most recent findings on CFC accumulation in the atmosphere, the relative ozone-depleting potentials of various chemicals, how ozone loss could occur naturally, and estimated rates of ozone depletion under a variety of control regimes. Uncertainty did not vanish, but the scientific community nevertheless was able to define methodically the dimensions and limitations of what was known. This made some degree of rational action possible even in the face of uncertainty.
What this suggests for international environmental law is that the precautionary principle is appropriate when states have made an effort to find and share credible information. Science was not left out of the process in Montreal merely because it could not supply certainty. Even if one accepts the precautionary principle, other important questions still must be answered about the threshold of certainty at which the principle becomes a basis for law. Must there be some estimated probability of actual harm occurring? (And if so, how great must the probability be?) Or is it sufficient for harm to be possible regardless of how small its probability may be? The role of science is to answer these questions in a methodologically sound way. Thus the Montreal Protocol accepts the precautionary principle, but it also states that control measures "should be based on relevant scientific knowledge, taking into account technical and economic considerations." In other words, precautionary measures have little legal force when the environmental harm is purely speculative; there must be some scientific rationale for not ruling out the possibility of harm from a specified action.
As it has evolved, the Montreal Protocol is a multilateral policy process that steers between underinformed overreaction and risky inaction. Science is firmly at the front end of the decision-making process, and scientists have room to allow for the honest uncertainty that is always a part of an evolving body of knowledge. The control regime is adjusted as scientific uncertainty decreases, and so far this part of the process has proven to be relatively non-political. When uncertainty persists, science still has a role in informing the negotiations between parties who have varying degrees of risk aversion; politics is then left to arrive at the most rational consensus it can obtain.
The Montreal Protocol assumes that, barring exceptional circumstances, environmental protection measures place a greater burden on developing countries than on advanced industrial societies. Article 5 of the protocol gives developing countries who are parties to the agreement an extra 10 years to meet the established control measures. In addition, producing countries are permitted to increase their output of CFCs and other controlled substances beyond what the would otherwise be allowed as long as the excess is exported to Article 5 parties during the 10-year grace period.
Article 5 establishes a legal distinction between industrial and developing countries that is consistently reflected in the protocol's substantive provisions. This was an essential element of the strong consensus behind the Montreal Protocol; although science had done its part by providing an apolitical description of the problem of ozone depletion, there were still issues of North-South equity that needed to be reconciled politically. Developing countries saw little fairness in being denied the CFC-based technologies that had played an important part of other country's prosperity prior to the Montreal Protocol, and they argued hard for special consideration. India and China, the two largest developing countries, refused to join the agreement until it included an aid mechanism to help poor countries acquire the more expensive alternative technologies.
There are limits to the special consideration given to Article 5 parties, however, just as there are limits to the obligations and rights created by the Montreal Protocol overall. Because the special situation of developing countries will continue to be a dominant theme in multilateral environmental protection measures, and because the Montreal Protocol stands as a strong precedent, it is crucial to take careful measure of what Article 5 parties actually got in the way of new rights, and what they actually gave away in terms of new obligations. Three important features of the protocol will be examined here: the obligation to refrain from environmentally harmful activities even though they may be essential to a developing country's economy, the obligation to participate in information gathering, and the right to aid.
The obligation to refrain from harmful acts. Customary law requires states "to conform to generally accepted international rules and standards for the prevention, reduction, and control of injury to the environment of another state or of areas beyond the limits of national jurisdiction."22 The Montreal Protocol's "special consideration" principle is consistent with this element of customary law in that the 10-year grace period does not absolve Article 5 parties from their primary obligation to phase out the production and consumption of ozone-depleting substances.
Without the grace period, meeting the protocol's control measures can present serious problems for Article 5 parties, especially those that have already invested in capital equipment using CFCs. The normal life of refrigerators, cooling systems, and other such equipment is 30 to 40 years.23 This means that Article 5 parties who increased their domestic stock of equipment during the past two decades would lose as much as three-fourths of the useful life of some items. An equal application of the phase-out schedule would have placed a greater economic and social burden on developing countries because most of them have very low rates of domestic capital formation.24 Direct foreign investment also varies widely among developing countries according to the host government's trade regime and macroeconomic policies. The additional 10-year grace period allows Article 5 parties to recover more of their sunk investments, and gives them more time to switch to ozone-safe technologies.
The absorption of new technologies usually is faster in an industrial economy than in a developing one. Technology assimilation depends on the stock of human capital (such as trained personnel, sophisticated patterns of industrial organization, and good social infrastructure for health care) and the relative absence of market distortions (such as poor information systems, poor regulatory structures, and monopolies). Developing countries, most of whom have low stocks of human capital and extremely distorted markets, consequently cannot absorb new technologies as quickly as industrial countries.
Another issue is that the alternative technologies are being developed by chemical firms in the industrialized world. The private costs of research and development for new technological processes affect the average cost of the goods that are ultimately produced, and these costs are passed on to consumers. The lower purchasing power of developing countries would make it relatively more expensive for them to acquire the new ozone-safe technologies.
Thus there are logical reasons for believing it would take a developing country longer to make the necessary structural adjustments in its economy so that it could comply with a universal control regime for ozone-depleting substances. The principle that can be abstracted from the Montreal Protocol may be stated this way: "special consideration" refers to the means by which a developing country may reasonably be expected to comply with accepted international rules and standards; the developing country is still obliged ultimately to comply with those standards.
The obligation to participate in information gathering. Like all other parties to the protocol, developing countries are required to report their production, consumption, and trade of ozone-depleting substances. This not only enables the secretariat of the protocol to monitor compliance, it also provides UNEP, WMO, and other scientific agencies the data they need to evaluate the current state of ozone depletion. In addition, data collected by the secretariat provide information used by the executive committee of the multilateral fund to disburse aid to Article 5 parties.
However, Article 5 parties do not have an absolute obligation to cooperate with multilateral fact-finding efforts within its territory. Article 9 calls for such cooperation in order to determine the costs and benefits of control strategies, but only insofar as the demands placed on states are "consistent with their national laws, regulations and practices and taking into account the particular needs of developing countries ... ." This weakens the protocol as an instrument to promote exchange of information. Lacking the specificity of other parts of the protocol, this article adds little to what customary law already requires in the way of consultation with other states.25 As Article 9 is worded, a developing country may bar international fact-finding missions from its territory by claiming they would violate its domestic laws pertaining to national security. Developing countries have often relied on such claims to forestall inspections by donor nations of aid programs.26
The right to aid. The agreement also recognizes the right of Article 5 parties to receive financial assistance. Significantly, however, there are caveats to this aid. Any financial assistance under the aegis of the Montreal Protocol is through a multilateral fund established specifically to help Article 5 parties make the transition to ozone-safe technologies. Policies of this fund are guided by an executive committee comprising seven Article 5 parties and seven industrialized parties. The executive committee determines the criteria for awarding projects that explicitly "enable their [Article 5 parties'] compliance with the control measures of the protocol."27
The parties to the protocol attempted to structure the financial mechanism so that it would be fair both the developing countries receiving the aid and to the industrialized countries that would be giving it. In this instance, "fair" meant ensuring that neither bloc could control the executive committee. Thus when consensus on an issue is not possible, the committee can take action only if:
(b) the voting majority comprises at least four of the developing-country members and four of the industrialized-country members.
Three points about the Montreal Protocol's financial mechanism are especially pertinent to the question of developing country rights. First, the purpose of the aid is very specific to the objectives of the agreement. It cannot be said, therefore, that the Montreal Protocol recognizes a developing country's right to receive aid in general. Rather, it recognizes the right of a developing country to aid in the context of the particular multilateral objectives to which that country is a declared party.
Second, a developing country receiving aid through the mechanisms of the protocol does not have the sovereign right to use those funds as it wishes. Developing countries have input into the principles guiding the use of those funds, but they do not have the right to "call the shots" without regard to the interests of the countries that are providing the financing. The path between the conflicting interests of the South (which wants to get as much aid as possible) and the North (which wants to give as little as necessary) is to ensure that the aid is used in the most efficient manner. If corruption and waste exhaust the funds in an aid package before the related project objectives are met, the recipient country would not be entitled to more aid. It must use its own resources (or funds from other sources) to satisfy the control measures to which it has agreed. On the other hand, a recipient country would be entitled to additional aid if the multilateral fund's executive committee underestimated the need and if the original funds had been spent efficiently.
The third important implication the protocol has for a developing country's right to receive aid is that the assistance is not open-ended. There is a clearly identifiable point at which the industrialized parties' obligation to an Article 5 party may be considered fulfilled. An open-ended right to aid would be untenable both legally and morally: a rich country cannot be obliged to give money to a poor one forever, otherwise the rich country eventually would become poor and the poor country rich. The precedent set by the Montreal Protocol is that in order for aid to be a legal right recognized by treaty, there must be (a) a clear rationale for the entitlement that can be evaluated according to evidence, and (b) a measurable objective that will satisfy the agreed-upon goal once achieved. In the case of the Montreal Protocol, the objective is the elimination of technologies that use ozone-depleting substances. Once an Article 5 party has made the transition to ozone-safe technologies in such a way that it is on a "level playing field" economically with its trading partners, it has no right to demand further aid under the auspices of the protocol.
What remains ambiguous, however, is the issue of accountability in the use of aid. This is due to the weakness of Article 9 and its failure to articulate a firm obligation to cooperate in international fact-finding efforts. Developing countries have equal input into the guidelines for projects supported by the protocol's multilateral fund. But once a project is approved by the executive committee and the funds are disbursed by the World Bank, there is no clear obligation on the part of the recipient state to participate in outside audits. The executive committee is empowered "to review regularly the performance reports on the implementation of activities supported by the Multilateral Fund," but Article 5 parties are not explicitly obligated to open the books of their domestic agencies to independent auditors.
This is the set of rights and obligations established by the Montreal Protocol for developing countries that are parties to the agreement. But what-if anything-do they mean to developing countries that are not parties to the agreement? It is important to examine the impact on non-parties because the principles such an examination may uncover will be those which will best inform future negotiations on multilateral environmental protection regimes.
Let us consider a hypothetical case with an imaginary state: Garibistan, a developing country not party to the Montreal Protocol, plans to build a medium-scale production facility for CFC-11 and CFC-12 using technology it acquired in the mid-1980s before the protocol entered into force. The plant will produce enough of the chemicals to meet Garibistan's own aggressive domestic development program, with enough left over to export to other non-party countries and earn Garibistan a considerable amount of foreign exchange. Further, the plant will begin production in 2007, the first year production of CFC-11 and CFC-12 by developing countries is to be banned under the Montreal Protocol. Assuming that appropriate jurisdictional authority exists for a tribunal to hear a dispute,28 would the parties to the Montreal Protocol have the right to impose trade sanctions on Garibistan in order to force its compliance with the control measures of the protocol, thereby shutting down the plant?
The Vienna Convention on the Law of Treaties says that an agreement is binding on a non-party state only if (a) the parties to the treaty expressly intend the agreement to establish a principle of international law that would be binding on non-parties, and (b) the non-party state accepts the rights and obligations in writing.29 This establishes the presumption that a treaty cannot confer rights or obligations on a state who is not a party to it without that state's consent.
The Montreal Protocol was drafted with the intent of establishing unambiguous and universal norms of behavior. The sixth preambular paragraph states the protocol's ultimate objective of the elimination of ozone-depleting substances, which would only be possible under a universal regime. Article 18 prohibits any reservations to the protocol or its subsequent revisions, both underscoring the consensus behind the agreement and establishing that it is a unanimous statement of international norms with regard to a global common. The ban on trade in controlled substances with third-party states (set forth in Article 4) constitute an incentive for all states to join the protocol, and a punishment if they do not. Only once does the protocol state that a provision is intended to be "without prejudice to any future arrangements,"30 and that refers merely to the modality (but not the principle) of providing assistance to developing countries. In short, the Montreal protocol meets the first test of third-party applicability set forth in the Vienna Convention on Treaties.
Meeting the second test (acceptance by third-party states) depends on the actions of the non-party state. Rulings by the International Court of Justice have determined that if a treaty is intended for universal application, and if in any official communique a non-party state expresses support for the treaty, a state may be bound by the treaty even though it has not ratified the agreement.31 Moreover, an official objection carries less weight if it is entered only after a controversy erupts. To be valid before an international tribunal, the objection must have been raised soon after timely and objective consideration of the treaty. Many states will therefore register official objections to a new treaty soon after it is negotiated in order to ensure that the obligations of the agreement will not be applied to them.
Thus if Garibistan had issued an official statement or joined a multilateral communiquÚ that professed support of the Montreal Protocol's objectives, it could find itself bound by the protocol's obligations. The principle may even be taken further: since the Montreal Protocol was clearly intended to establish universal norms of behavior, and given that it was negotiated in an open world forum, one may argue that Garibistan's silence on the issue may be construed as a tacit acceptance of the protocol's control measures.
The mechanism of the Vienna Convention is predicated on consent. But what if Garibistan had in fact had registered its official objection to the treaty with the United Nations or the Montreal Protocol secretariat? In this case, the Vienna Convention would imply no obligation for Garibistan to comply with the Montreal Protocol unless the provisions of the protocol could be interpreted as statements of international customary law applicable erga omnes.32
A principle becomes part of the body of customary law when it is generally accepted through international tribunals, national tribunals, and treaties as a fair and necessary rule for relations between states. Once accepted as customary law, it may be extended to all states regardless of treaty accession or official statements. The International Court of Justice acknowledged this possibility in its Barcelona Traction judgment of 1970:
The court specifically cited human rights and prohibitions against aggression as examples, but the same reasoning may in certain instances be applied to environmental protection. As previously mentioned, international customary law recognizes a state's general obligation to abstain from actions that harm parts of the common global environment beyond national jurisdiction. Maintaining the integrity of the atmosphere so that it can sustain healthy human life is at least as much of a "concern of all States" as human rights or non-aggression. The legal question, then, would not be whether the Montreal Protocol would apply to a non-party state. Rather, a tribunal would have to determine whether the actions of that state posed a threat of significant harm to the ozone layer. If the tribunal found that the actions prohibited by the Montreal Protocol indeed constituted a bona fide threat to the global environment, then a state would be transitively obligated to the treaty's norms of behavior even though it was not a party.
A treaty that confers obligations on non-party states without their consent is referred to as an "objective regime." The most common applications of this concept have been with regard to the establishment of demilitarized zones, areas of neutrality, and freedom of navigation both on the open seas and on international rivers.34 But there is no consensus among legal scholars about the validity of the objective regime model. Some contend that no objective regime is possible without consent or custom, and that if either is present then it is the consent or custom (and not the treaty's supposed status as an objective regime) that creates obligations.35 The essence of their criticism is that a treaty cannot make law that is binding on non-parties, although it can articulate already-existing customary law.
Prof. Arrow says that any attempt to confer new international obligations upon a third-party state without its consent "would need to be affirmatively established as norms of customary international law, with the party seeking to establish the exception bearing the burden of proving that the four requirements necessary to the establishment of such a norm had been met."36 The first of these requirements is that enough states must follow the norm that it may be interpreted as a general practice. This may be established by the number of accessions to a treaty, widespread existence of domestic laws, or a consistent body of tribunal rulings. Second, it must be clear that states are refraining from prohibited activities out of a sense of legal obligation, and not because political or economic self-interest makes it convenient to do so. Third, the obligation must be taken up by those states for whom adherence is not inconsequential. Finally, the norms should have a history of compliance, or "where a new prohibitory custom is asserted to have developed quickly, the state practice must have been 'extensive and virtually uniform'."37 However, Prof. Arrow contends that no treaty has ever achieved objective regime status by meeting these four criteria, and that "any purported exceptions, to the extent of their validity, are ultimately based on consent." A similar conclusion was reached by the International Law Commission in 1966 when it decided against proposing a special article to the Vienna Convention on Treaties dealing with objective regimes.38
Using Prof. Arrow's own criteria, however, it may be argued that the Montreal Protocol is indeed an authentic exception. Once it successfully addressed demands by India and China for equity towards developing countries, the consensus behind the protocol was enough to satisfy all four tests. The ninety developing and industrialized states party to the Montreal Protocol include all of the major producers and consumers of ozone-depleting substances, and their compliance with the norms of the protocol will entail large investments and considerable industrial restructuring. It is clearly not convenient for these states to comply with the protocol; they are spending millions of dollars on research and development and on replacement equipment because they are legally bound to do so. Moreover, the parties to the protocol have responded to the problem of ozone depletion with an alacrity that is "extensive and virtually uniform" (to use Prof. Arrow's phrase) even though the scientific evidence is so recent and the history of compliance so brief. By meeting these criteria, the parties to the Montreal Protocol have articulated norms for the protection of the ozone layer that are indeed universally applicable.39
If the parties attempted to use trade sanctions to enforce the protocol's control measures on a non-party state, they would have to establish the nondiscriminatory nature of the enforcement. Article XX of the General Agreement on Tariffs and Trade provides that
The ban on the trade of controlled substances with non-party states as called for in Article 4 of the Montreal Protocol would clearly be consistent with Article XX of the GATT. The trade ban applies equally to all states who have not ratified the protocol, and there are no discriminatory obstacles preventing any state from acceding. The question here, however, is whether stronger trade sanctions would be legally feasible against a rogue non-party state who was producing and exporting significant quantities of CFCs. Could the parties to the Montreal Protocol legally impose a wider trade embargo (on all capital equipment, for example) to force the non-party state to stop its harmful activities?
The use of trade restrictions to enforce environmental norms is an undeveloped area of international law, as shown by a 1991 GATT arbitration concerning import restrictions based on tuna fishing practices. The United States, which has domestic laws requiring dolphin-safe practices by U.S. fishing fleets, had imposed import restrictions on tuna from Mexico where the fishing fleets had no such requirement. The panel ruled that the United States could not use its domestic laws as a basis for restricting tuna imports because to do so would allow the United States to "unilaterally determine the conservation policies from which other contracting parties could not deviate."40
The panel noted that environmental regimes and trade regimes are likely to conflict more often in the future because not all states will have the same environmental policies. Recognizing the fundamental importance of the issue and the lack of meaningful precedent to resolve disputes through arbitration, the panel called on the parties to the GATT themselves "to impose limits on the range of policy differences justifying such [trade-restricting] responses and to develop criteria so as to prevent abuse."41 The panel went on to suggest that international norms would have been less ambiguous than national laws in the context of the GATT's Article XX:
If parties to the Montreal Protocol wanted to impose an embargo on capital goods exports to a rogue non-party, it would have to prove (a) the harm of the action they wish to stop, (b) the violation of international norms (according to the four criteria discussed previously), and (c) the need for trade sanctions. To prove the latter, the parties would have to show how negotiations, credit restrictions, and other less-drastic means of diplomatic confrontation had failed to achieve the desired resolution.43
Returning to our hypothetical case, it would appear that the parties to the Montreal Protocol would indeed have a legal basis for using some measure of trade sanctions to prevent Garibistan from building its CFC production facility. The "soft" obligations-the duty to cooperate in gathering and exchanging information-are less clear, however. This is because those duties are less than definitive for the parties themselves due to the weakness of Article 9 of the protocol. Thus if the parties wanted an official fact-finding mission to ensure that Garibistan was not building its CFC facility illegally, Garibistan would be within its sovereign rights to not cooperate. Similarly, this would almost vitiate any implied requirement that Garibistan report to the protocol secretariat its production, consumption, and trade of CFCs.
Addressing the problem of stratospheric ozone depletion has been relatively easy; specific chemicals cause the problem, and the straightforward answer is to stop using those chemicals and find something else to use instead. But global warming and biodiversity are so fundamentally different from ozone deterioration that simply grafting the success of the Montreal Protocol onto treaties dealing with them would be nonsense. A more productive and realistic exercise would be a deeper comparison of the Montreal Protocol with subsequent treaties to see what might make the difference between success and failure. Here, the comparison will be with the 1992 Rio de Janeiro Convention on Biological Diversity.
Consensus was sine qua non to effective action on ozone protection. The first element contributing to consensus was the role of the scientific community; the strength of consensus grew as scientific evidence of ozone depletion became clearer. The United Nations and the UN Environmental Programme have attempted to replicate that aspect of the Montreal Protocol in its efforts to forestall global warming44 and to promote biological diversity.45
However, the relationship between science and diplomacy in the Biodiversity Convention is backwards from what it was for the Montreal Protocol. In the ozone protection agreements, science preceded diplomacy; it was the physicists, chemists, and atmospheric researchers who were on the cutting edge of discovering and defining the problem, its causes, and the probable damage due to inaction. But politics rather than science is defining the issue of biological diversity. Scientists are in essence being called upon to justify the political concern retrospectively, but at the same time are told that "lack of full scientific certainty should not be used as a reason for postponing measures to avoid or minimize [a threat of significant reduction or loss of biological diversity]."46
At face value, the above quotation from the preamble might suggest the precautionary principle. As discussed in an earlier section of this paper, however, the legal weight of the precautionary principle comes from the effort that has been put into defining the limits of what is known. What exactly is the worst-case scenario, what are its hypothetical causes, and what scientific evidence is there to suggest that the worst-case scenario is in fact possible? Absent any objective effort to answer these questions in a scientifically defensible way, the worst-case scenario would be nothing but ill-informed conjecture-which the precautionary principle clearly does not imply.
The misapplication of the precautionary principle in the Biodiversity Convention is seen in the fact that certain rights and obligations are rearranged prior to creation of the special scientific panel that is to inform the parties. Article 25 establishes the panel, but articles 8 through 14 call on parties to begin implementing a wide array of unspecified but potentially costly domestic policies before the panel begins its work. This in itself would not be a serious problem were it not for the fact that (a) articles 20 and 21 attempt to establish developing country parties' right to obtain aid and developed country parties' obligation to provide aid, and (b) paragraph 5 of article 16 attempts to rearrange international intellectual property rights. The creation and altering of rights and obligations prior to the accumulation of a commonly accepted body of scientific information (certain or not) is both inconsistent with the precedent of the Montreal Protocol and insupportable under the precautionary principle.
The two problems previously cited (aid and intellectual property rights) merit closer examination. The key features of the Montreal Protocol's aid mechanism, as mentioned previously, are that it is limited to the specific objectives of the protocol, it is not open-ended, and it allows neither developing parties nor industrialized parties to dictate the policies of the protocol's multilateral fund. The biodiversity convention falls short on all these counts; Article 20 amounts to an open pocketbook to pay developing countries in perpetuity to maintain unspecified objectives of biological diversity:
The extent to which developing country Parties will effectively implement their commitments under this Convention will depend on the effective implementation by developed country Parties of their commitments under this Convention related to financial resources and transfer of technology and will fully take into account the fact that economic and social development and eradication of poverty are the first and overriding priorities of the developing country Parties.47
The effect of this provision is to exempt a developing country party of its obligations under the treaty (a) if it does not receive enough aid and new technology from industrial countries or (b) if those obligations conflict with its domestic development policies. This contrasts with the Montreal Protocol which, while allowing developing country parties special consideration, does not exempt them from the universal norms of behavior established by the agreement.
Nor do developed and developing country parties have equal input into deciding criteria for the distribution of funds. The policies, strategies, and program priorities of the financial mechanism established in Article 21 are to be set by the conference of the parties48 (that is, by all the parties), which would increasingly favor developing countries as more countries acceded. This is significantly different from the approach used by the parties to the Montreal Protocol. There, the parties established a precise definition to distinguish between developed and developing countries (the protocol's Article 5), and then gave developed and developing countries equal representation on the executive committee that oversees the protocol's financial mechanism.
Regarding intellectual property rights, the Biodiversity Convention attempts to impose on the international trading system a simple decision rule for what is anything but a simple controversy. It has taken trade negotiators 12 years to come up with a draft agreement on the trade-related aspects of intellectual property.49 The difficulty has arisen out of the widely divergent ways various political cultures regard property rights in general and intellectual property rights in particular. Each state has the sovereign right to define its own domestic property rights according to its own values, but when states trade, the differences between their respective regimes can create considerable economic distortions. Especially with intellectual property, the differences are often philosophically irreconcilable and impose significant costs on either innovators or consumers. If this long-standing controversy somehow became simple on the road to Rio, the miracle escaped the notice of most trade economists.
Article 16 says that parties shall ensure that intellectual property rights "are supportive of and do not run counter to" the objectives of the Biodiversity Convention.50 In other words, if someone develops a valuable product using genetic resources from a protected ecosystem, it would be acceptable to pirate the technology (without paying royalties to the patent holder) in order to "equitably [share] the benefits arising out of the utilization of genetic resources," which is stated as an objective in Article 1 of the Biodiversity Convention.
It may indeed be a good idea to give environmental objectives precedence over trade objectives. But mere hortatory statements will not make it so. It will require the explicit delineation of decion-making rules that ensure environmental protection is not used as a disguised barrier to trade. Had the Biodiversity Convention been preceded by a careful consideration of the legal and economic implications of changing the international regime of intellectual property rights, it could have included an article outlining specific rules by which disputes could be decided.51 This would have enabled the agreement to achieve a better balance of new rights and obligations between developing and industrial countries. Absent such attention, the Biodiversity Convention is insufficient to effect a legitimate change in the international regime of intellectual property rights.
Diversity is an important quality of the global biota. Preserving it is a worthy international goal that needs the protection of a good treaty. But the Rio Convention on Biological Diversity is not a good environmental treaty. It is inconsistent with the principles reflected in the Montreal Protocol, and therefore cannot be heir to the latter's success. This is seen in the misapplication of the precautionary principle, the provisions for financial transfers to developing countries, and the inadequate consideration that went into the effort to change intellectual property rights.
Professor Aman observes that "[t]he Protocol's specific substantive response to the problems created by CFCs and related compounds is, perhaps, less significant than the global regulatory process that it establishes and the international precedent that it sets."52 This paper has shown how the regulatory process can extend to non-parties, and how the precedent can strengthen future environmental treaties founded on the same principles. Science needs to be an integral part of rule-making, while fairness demands special consideration for the situation of developing countries. Aid is a right only if it specifically furthers the objectives of the treaty, is not open-ended, and is subject ultimately to principles determined by a consensus of the parties to the treaty rather than the sovereign policies of the state receiving the assistance. Perhaps most significantly, the regulatory process is open-ended and evolutionary. The parties are required to meet at least every four years to review the Montreal Protocol's control measures and make adjustments as necessary-which they did in London in 1990 and again in Copenhagen in 1992.
Nature is no respecter of international boundaries. Consequently, ecological threats to the global commons require, among other things, a more enlightened way of understanding the relationship between sovereign states. Requiring a state to adhere to international norms without its consent is not to be done lightly, but as sustainable development becomes the defining North-South issue in the post-Cold War world, the common need to address global environmental problems will on occasion have to supersede interests of individual state sovereignty. Customary law allows for this, and the Montreal Protocol is perhaps the best example to date of how competing interests may be balanced under a new philosophy of the global public interest that relaxes the once-sacrosanct assumption of sovereignty.
The interplay of sovereignty issues and trade is especially tricky for developing countries, because as the failure of their import substitution policies proved in the 1960s and 1970s, development is not an entirely endogenous process. As long as a developing country needs foreign exchange to import goods and stimulate investment, trade will be a latent constraint on its otherwise "sovereign right" to exploit its own resources pursuant to its own environmental policies.53
Up to now, trade agreements-including the various negotiating rounds for the GATT-have disregarded environmental issues. The relationship between trade and environmental protection is, at this point in time, adequately understood neither by environmentalists nor by trade experts, yet trade could become an important means of encouraging and enforcing global environmental protection regimes in the near future. At least one trade expert has predicted that most major trade treaties concluded after 1992 will have to address environmental concerns in one way or another, and this may occur in the next negotiating round of the GATT.54 The fundamental question to be resolved is whether (and if so, how) environmental norms supersede trade norms in cases when the two conflict. Article XX of the GATT says that they can, but the decision-making rules are still embryonic.
Sovereignty is becoming less relevant in a global economy which increasingly relies on transnational firms and multilateral financial institutions. Individual states, in a practical sense, cannot do whatever they wish without regard to the interests of other states. The modern challenge of international law (especially with regard to protecting the global commons) is to abstract fair and functional rules that will guide good-faith cooperation in a new political environment.