Back to the Conference Page         Back to Conference Papers

 

A Commonsense Framework for Operationalizing the Precautionary Principle

Joel Tickner

Paper prepared for the Wingspread Conference on Strategies for Implementing the Precautionary Principle

Racine, WI

January 23-25, 1998

During the past 15 years, the Precautionary Principle has become a guiding principle of international environmental law. As a result, the Principle is customarily incorporated into international environmental treaties and agreements. Several nations around the world (and economic regions such as the European Union) have either implicitly or explicitly endorsed the Principle. However, while few would dispute the importance of precaution, the principle and its components have yet to be clearly defined and no framework for policy makers on precautionary decision-making currently exists. As such, there is a critical need to develop a framework to operationalize the Precautionary Principle in order to answer the fundamental question, "what does the principle mean in terms of decision-making under uncertainty?"

While other papers presented during this Wingspread Conference will offer specific recommendations on tools for implementing precaution, needs for precautionary decision-making, or a historical perspectives on the use of the principle, I will attempt to produce an overall guiding framework for implementing the Precautionary Principle in environmental, public and occupational health laws and policies. Following a description of the weaknesses in the current international use of the principle, I will describe some of the fundamental characteristics of a framework for operationalizing the Precautionary Principle and then outline in detail the key components and decision-making criteria of the framework. Finally, I will provide some examples of uses of such a framework. The framework I describe should not be considered a static document but rather one which is flexible and open to discussion, modification and adaptation to various situations, as well as continual improvement. It should also not be considered a direct replacement for current decision-making tools such as risk assessment and cost-benefit analysis which could play an important (albeit much different) role in conjunction with the principle.

Weaknesses of the current use of the Precautionary Principle

While the Precautionary Principle is growing in acceptance and use internationally, it still lacks a specific widely recognized definition. With few exceptions, the principle remains only a concept, provides few guidelines for policy makers and fails to constitute a rigorous analytical framework. While a wide range of international treaties call for precautionary action few indicate how to determine when to take precautionary action; what action(s) should be taken and to what extent; and who should be involved in decisions to take precautionary action. Although several frameworks for integrating the principle into environmental and occupational health decision-making for specific problems have been proposed (Stijkel and Reijnders, 1995, and Mee, 1995), as well as legal structures for the principle (Cameron and Abouchar, 1991), no comprehensive, systematic structure for precautionary decision-making has been applied on a national or international level.

The lack of a generally accepted formulation and criteria to guide the operationalization of the Precautionary Principle has limited its wide-spread use and in some instances has led to heated debate and controversy. Those against the principle see it as a calling for an immediate halt to all potentially damaging activities. Those calling for the principle, often note the need for a precautionary principle but do not know exactly what they are asking for in terms of action. The time is thus ripe for the operationalization of the Precautionary Principle in environmental and occupational health legislation as it offers a framework for the advancement of preventive and holistic policies. As such, there is a clear need to establish an institutional framework for the precautionary approach.

Basis of a Framework for Operationalizing the Precautionary Principle

The Precautionary Principle is a commonsense, prevention-oriented, principle which instructs policy makers and other decision-makers to think before acting and act based on suspected rather than proven effects (in other words, act in the face of uncertainty). The Principle admits the limitations in what we know and can know and bases decisions on available information rather than waiting for conclusive proof. The Precautionary Principle forces scientists and policy/decision makers to begin to ask a different set of questions about problems. Current decision-making approaches ask questions such as: "How safe is safe;" "What level of risk is acceptable;" and "How much contamination can a human (usually a healthy adult male) or ecosystem assimilate without showing any obvious adverse effects?" The Precautionary Principle asks a different set of questions such as: "How much contamination can be avoided while still maintaining necessary values?;" "What are the alternatives to this activity that achieve a desired goal (a service, product, etc.)?;" and "Do we need this activity in the first place?"

Changing the questions we ask about a problem necessarily engenders a different set of public policies and policy responses. Policies based on the Principle are preventive and solutions-oriented, whereas those based on current decision-making approaches focus on pollution control and remediation. Precautionary approaches are goal and alternatives oriented, lending themselves to technology innovation, pollution prevention, and impact assessment. Policy responses based current decision-making approaches focus on add-on, end-of-pipe technologies, personal protective equipment, and medical treatment for those negatively impacted.

The Precautionary Principle encourages those undertaking potentially hazardous activities to seek out the least hazardous alternative, and to think before acting. In this respect, the principle can be considered to put in place a type of "speed bump," which creates bottlenecks in the development process but which does not stop flows. The Principle calls for a permanent process of looking for the least hazardous alternative to achieve a specific purpose, continuously updating knowledge to avoid harm. Under the principle we would tend to focus on those options that are the most "error-friendly," those that would be least prone to environmental damage or for which harm would be most reversible (von Weizacker, 1996). Operationalizing the Principle also does not mean throwing out current decision-making tools, such as risk assessment and cost-benefit analysis. It does mean, however, using those for their intended purposes, to inform decision-making. In this respect, these tools are relegated to a send tier in the decision-making process. Instead of using these techniques to quantify an "acceptable" risk , they are used to compare alternatives to an activity (or to establish priorities) which is a much less complex and often more clear cut activity, requiring less rigorous quantitative analysis.

A framework for operationalizing the precautionary principle will focus on changing the types of questions asked and policy responses that are taken in response to a perceived or actual environmental or public health problem. It would also shift responsibility for identifying and preventing environmental harm from government agencies and the public, to those undertaking potentially harmful activities. This framework would need to introduce a common definition for the precautionary principle as well as a set of criteria for precautionary decision-making. Such a framework (and precautionary policies) would need to be flexible enough to adapt to changes in knowledge without intensive scientific analysis. While being open to information beyond the realm of science (public concern, social impacts, etc.), the decision-making criteria proposed in the framework should be repeatable under different decision-making conditions. Finally, such a framework would open lead to the identification of a variety of options and potential actions to address a particular problem.

Conceptually, the precautionary approach to decision-making would consist of the following elements: (1) A goal-setting, guiding principle of precaution in the face of uncertainty; (2) tools for decision-making in the face of uncertainty (e.g., decision-making criteria); and (3) methods to carry out precaution-based decisions (such as clean production) as well as means to continuously measure potential adverse effects. We can think of a framework for operationalizing the Precautionary Principle as having two separate applications: For decision-making regarding new activities and decision-making based on potential hazards which already exist. How the Principle can be operationalized for each of these two applications will differ slightly (this will be discussed below).

In general, we can think of five specific virtues of the precautionary principle:

1) Responsibility: The burden of proof on the initiator of an activity to demonstrate that there is no safer way of accomplishing what it needs to accomplish.
2) Respect: Action before full scientific certainty of cause-effect.

3) Prevention: A duty to seek out means prevent potential harm rather than to control and treat it.

4) Obligation to know and inform: A duty to understand, investigate, inform and act on potential impacts. Ignorance is no longer acceptable under the Precautionary Principle.

5) Obligation to share authority: Democratization of decision-making regarding science and technologies.

Components of a Proposed Framework for Operationalizing the Precautionary Principle

Below I present some of the components of a policy framework for operationalizing the Precautionary Principle. While their order does not reflect their importance in such a framework, it does reflect a logical order to their incorporation into an overall policy framework and potential legislation. There are likely other components that should be incorporated or ones within this proposed framework that should be removed, modified, or enjoined. This framework does not attempt to deal with the broader and fundamentally important questions of role of science (its role in decision-making, its primary focus on economic development and not the public interest, etc.) and redefining "good science;" the role of corporate power in determining public policy; and societal questions, such as changing material and consumerist behavior. The framework also does not deal with the intrinsic and very important question of economics and the costs/benefits of actions taken through the Precautionary Principle. Nonetheless, the elements which I propose should provide an outline and guidance for decision-makers and others promoting the Principle in preparing legislation to incorporate the Principle into environmental and public health decision-making. The elements of a policy framework for operationalizing the Precautionary Principle would consist of:

1. Definition and general duty to take precautionary action.

An important first step in developing a precautionary approach framework is the creation of a working definition for the precautionary principle and its specific components. This will ensure that the concept is clear, can lead to strict and enforceable policies and cannot be distorted by interest groups. The definition will establish a broad­based goal of precaution in the face of scientific uncertainty, which will set the stage for subsequent decision­making strategies. International non­governmental organizations (NGOs), have identified the following components of the precautionary principle: (1) action before damage, before firm scientific proof exists; (2) a shift in the burden of proof to the potential polluter to prove that an activity will not have any adverse environmental consequences; (3) prevention of contaminants from entering the environment; (4) implementation through the use of clean technologies (Greenpeace, 1991).

Several international treaties, as well as the United Nations Conference on Environment and Development (UNCED) and the European Union have developed definitions for the precautionary principle. Nonetheless, there is no specific definition of the Precautionary Principle in the literature, only descriptions of the principle.

Through a specific definition, the Principle can be incorporated as a general duty in environmental and other public health legislation, which is specifically how it has been used to date on a national and international level. General duty statements implement the Precautionary Principle virtues of respect and responsibility and provide important powers to government agencies to take action on perceived threats that may or may not be specifically regulated. The U.S. Occupational Safety and Health Act General Duty Clause (OSHA section 5) has been used in numerous instances to protect workers from chemical or ergonomic hazards. The U.S. Clean Air Act, Section 112r Risk Management Plan Rule contains a general duty clause to prevent chemical accidents, which regulators could use to require options analyses for safer processes. The ill-defined, uncertain outcome of this type of general duty created by the Precautionary Principle may be critical to its success by creating a certain air of uncertainty in regulation which may lead companies to take action which would normally not be taken if regulatory obligations were clear.

2. Aggressive goals for reductions in hazardous substances, processes, products and practices (backcasting).

The precautionary principle demands goal­oriented statements of where the jurisdiction would like to be with regards to reducing or eliminating a specific hazard. For example, Sweden has set ambitious goals for the reduction of some heavy metals and most recently for polyvinyl chloride (PVC) plastic (Ministry of Environment, 1997). These goals, with interim targets, are periodically measured and reexamined. They are established in cooperation with a wide cross­section of society so as to minimize economic displacement and over long periods of time so as to allow the development of alternatives. Clear, enforceable and achievable goals within a precautionary framework will thus drive the development of safer alternatives to hazardous activities. Many countries, such as the U.S., fail to establish far reaching goals for environmental health, which is clearly observable in the lack of vision in environmental policies.

Rather than trying to forecast potential impacts and uncertain futures as is often done with tools such as risk assessment, under a decision-making framework for the Precautionary Principle government agencies, industry, workers, and the public would develop a vision of where society (or a country, state, etc.) should be and then work backwards to determine steps towards that goal, a process called "backcasting". For example, the Dutch government establishes five year environmental plans with clear goals and then works with municipalities, industry associations and specific companies to establish "covenants". These covenants are voluntary agreements between the government and the industry which establish interim and final goals, but place responsibility on the company to achieve those goals in the most efficient way possible (without creating new risks). The covenants are backed by strong enforcement and regulation if goals are not met.

Such a backcasting exercise could be combined with some sort of scenario development to better understand the effects of different means to achieve goals or different levels of goals. This would also allow for flexibility in both goals and the methods undertaken to achieve them.

3. Shifting the burden of proof to those undertaking hazardous activities (reverse onus)

A key function of the precautionary principle is to require those companies/actors engaging in potentially hazardous activities to demonstrate some level of safety before engaging in that activity. As government authorities never have sufficient resources to study every chemical, factory, or ecosystem it becomes critical for those undertaking a potentially dangerous activity to have to prove that their activities will not adversely harm humans or the environment. Since wildlife, workers, and the public rarely have the power to control development, technologies, the production of products or production processes, those undertaking activities thus have a responsibility to prevent harm, which accompanies their control over choice of technology, development, and products.. Manufacturers currently have the burden of proof of safety for pharmaceutical products (and sometimes pesticides) in many countries throughout the world, though these burdens are often couched to some degree on economic benefits or effectiveness of the drug or pesticide. Nonetheless, these laws could serve as a model for future environmental and occupational health and safety regulations.

The burden of proof requirement has its most important role with regards to new activities. Before a potentially dangerous activity can start, the proponent would have to demonstrate that no harm would occur or that there were no safer alternatives for an activity. The government agency would then have the duty of permitting, restricting, or prohibiting such an activity or offering alternatives. In the case of existing activities, a threat of harm could shift the burden of proof from the government agency or the public to the company or actor which has undertaken an activity (for example when it is found that an emission could potentially cause harm, it would be up to the emitting company to prove that its activity is safe or there is no safer way to carry out that activity). Finally, as those undertaking a potentially dangerous activity have a vested interest in proving its safety, it has been suggested (Raffensperger, 1997) that any claim of safety go under an independent peer review, with the initiator providing all information (demonstrating harm or safety) used in that determination.

4. Criteria for decision-making under uncertainty

A framework for operationalizing the Precautionary Principle needs to provide clear instructions to inform decision­makers on how to weigh scientific and other evidence and what actions to take in the face of uncertainty (in other words a precautionary standard of inference). Rather than a quantitative approach potentially limited by a lack of data, uncertainties, and assumptions, the Precautionary Principle framework should provide methodological guidelines for weighing scientific evidence and qualitative/quantitative decision­making criteria that will instruct policy makers on how to proceed when dealing with limited or uncertain scientific evidence. The criteria for decision-making (in other words the precautionary decision-making structure) would have to be accompanied by an administrative structure to support this new, novel type of decision-making. . This administrative structure would detail: (1) how and where the precautionary principle can be integrated into agency decision-making processes and programs; (2) infrastructure for precautionary decision-making (e.g., expertise, stakeholders); and (3) specific policy language and guidance on how to use the decision-making criteria.

This is probably the most complex and difficult part of the framework for precautionary decision-making which I describe in this paper for several reasons. First, a legal regime for the Precautionary Principle would likely have to establish some standard of proof of harm at which level precautionary action would be taken. However, given our limited knowledge about ecosystems such a level might be hard to define (such as 95% proof, 51% proof, preponderance of the evidence, etc.) and might vary depending on who makes the analysis. Second, what information should be included in decision-making and how should boundaries to the types of information analyzed be established?. Third, how can criteria be designed so that decision-making is logical, clear and understandable, reasonable, repeatable, just, transparent, and is both quantitative and qualitative. In essence, we are looking for alternative criteria/methods to weigh scientific evidence that overcome the limitations of current decision-making tools such as risk assessment. If we can solve the dilemma of what criteria should be used in precautionary decision-making, we have solved a large part of the problem of how to operationalize the Precautionary Principle.

In the next few paragraphs, I will lay out some specific thoughts on how decision-making criteria for the precautionary principle could be designed. Decision-making will clearly differ depending on whether we are analyzing a current or proposed activity. For a proposed activity, a focus on the existence of alternatives can take some pressure off the need to established a standard for proof of harm. In general, the approaches I describe rely on what is termed the "weight of evidence," rather than some quantitative probability of harm (as is the case with risk assessment approaches). The weight of evidence approach to decision-making takes into account the cumulative weight of the many studies from numerous sources that address the question of injury or the likelihood of injury to living organisms.(IJC, 1995) Types of information that might be considered include: worker case histories, toxicological studies, exposure assessments, epidemiologic studies, monitoring results, etc.

In order to guide decision-making, criteria for considering information on potential harm would be used would need to be established. Numerous sets of criteria for analyzing cause-effect or potential harm exist, some of which are listed in Tables 1-5. As can be observed from these lists of criteria, some address criteria to guide causal inference (such as the Hill criteria and those of the Massachusetts Weight of Evidence Committee) while others address considerations when weighing evidence of potential harm. The Massachusetts criteria, which were developed within an ecological risk assessment framework provide some important criteria for assessing causality, but many of those criteria would require substantial quantitative evidence before causality can be established, potentially undermining precautionary action. The Dovers and Ludwig criteria address the fact that some decisions (or activities) certainly have greater potential impacts than others and that greater scrutiny should be afforded those activities that could potentially have irreversible, widely distributed, and multi-generational impacts. In other words, they address the "decision-stakes" of a particular decision under uncertainty (Funtowicz and Ravetz, 1991).

The weight of evidence determination (or the determination of whether to allow an activity to continue or restrict it) would necessarily vary in depending on the range and scale of impacts which might be caused and the availability of means to prevent the hazard. Table 6 presents some proposed criteria for precautionary decision-making, divided by causal inference and decision-stakes criteria. Taken together, these would help guide weight of evidence analyses and thus the amount of proof needed before precautionary action is taken.

Issue specific decision-making criteria could then be developed for decisions being made with respect to: chemicals, biodiversity, development, physical stressors (e.g., radiation), biotechnology, etc. These issue specific criteria would address specific causal inference and decision-stakes concerns related to each of these regulatory regimes. They would be considered in addition to the general criteria listed in Table 6.

To determine the extent of precautionary action to be taken, a sliding scale of causality could be established. This scale of actions to take would reflect both the level of proof/certainty of the adverse outcome from the weight of evidence determination as well as the potential impacts should that outcome occur (certainty of a specific adverse effect would also need to be considered). Different levels of proof would lead to different types of responses (e.g., study requirements or substantive requirements such as mitigation or alternatives development). For example, an activity for which we are 10% sure of harm and for which harm, if it were to occur is minimal, would require increased monitoring (if the potential harm were great and widespread we would possibly require remedial action); an activity for which we are 30% certain of harm would require some kind of preventive or remedial action and monitoring; and activity for which are 45% certain of harm would be prohibited depending on the extent of potential harm.

Finally, the decision-making criteria and weight of evidence determinations could be incorporated into a decision tree type format. The decision-tree analysis would consist of two parts: (1) One branch of the analysis is for dealing with existing hazards. In this case if the weight of the evidence demonstrated actual or possible harm, some type of action would be taken. If insufficient evidence was available to arrive at a weight of evidence determination, the institution undertaking the activity suspected of causing harm would have the burden of providing evidence of no harm. If this evidence was not provided, precautionary action would be taken. A second part of the decision-analysis would be oriented towards the introduction of new chemicals, products, or work activities with potential impacts. The initiator of an activity would need to conduct an initial impact statement identifying potential impacts of the activity, potential alternatives, and the proposed action. Precaution would serve as a default decision until the weight of evidence determination demonstrated that (a) there was no safer alternative for the activity that would fulfill the needs of the initiator and that there is a necessity for such activity; or (b) that the activity posed no real risk. The weight of evidence analysis would also identify potential adverse impacts of that activity and monitoring/investigation requirements for the initiator.

5. Tools for implementing precaution

The Principle of Precautionary Action calls for preventive actions before proof of harm has been established. These measures should be taken in advance of scientific evidence, and, when possible, at the design stage of a potentially hazardous activity. The Precautionary Principle does not fulfill its purpose unless preventive methods for carrying out precaution are implemented. There are numerous tools for carrying out precautionary policies that have been used throughout the world (many of which are discussed accompanying papers). These include:

6. Use of the "Polluter Pays" principle.

Under a regulatory framework based on the Precautionary Principle, those engaging in potentially hazardous activities would be responsible for the costs of the damage they caused. A defense of failing to act to prevent harm because of uncertain knowledge or undertaking an activity while failing to consider potential harm with would no longer be acceptable under this structure. The polluter pays principle is a fundamental principle of environmental law in several European countries and in some international treaties. The difficulties of such a principle lie in assessing and quantifying damage and attributing harm to those who have caused it.

Economist Robert Costanza has suggested the concept of the precautionary polluter pays principle (4P) as an incentive for firms to make preventive decisions (Costanza and Cornwell, 1992). The 4P principle encompasses precautionary assurance bonding to ensure that companies undertaking potentially hazardous activities have sufficient resources available to pay for any damages created. Companies would be required to pay a premium before commencing a potentially dangerous activity, that is based on the worst potential damage that the facility could possibly cause. This assurance bond could be returned if the potential damage does not occur. This type of system would move costs of potential damage to the present, where they will have the greatest impact on decision­making, thus leading the firm to invest in safer alternatives.

The Precautionary Principle could also impose strict and several liability on those undertaking a potentially dangerous activity. Strict and several liability means that a company or initiator of an activity does not have to be found negligent or even to have known that they were causing harm to be assessed blame for harm (e.g., ignorance or a poor understanding of potential impacts does not serve to eliminate responsibility for harm). Also, liability for harm caused by many actors can be attributed to a single actor. Liability for damage should not be covered under traditional insurance schemes which tend to distribute costs among policy holders. Strict and several liability, as well as performance bonding, would provide a strong economic incentive to avoid potentially dangerous activities.

Finally, the polluter pays principle provides an economic incentive for investing in cleaner safer technologies, products, and activities. However, for such a principle to work governments will need to remove subsidies from dangerous practices, technologies, and products, and redistribute them to cleaner and safer ones.

7. Scheme to systematically evaluate alternative activities, technologies, chemicals, etc.

The Precautionary Principle will not serve its role if it prevents one hazard while creating another. The potential creation of new, unintended risks, by application of the Principle is one of its most common critiques (Cross, 1996). Thus, a systematic and comprehensive scheme must exist to examine the impacts of alternatives to potentially hazardous activities. This scheme must look at the entire lifecycle of the product or activity, including raw material extraction, production, use of product, and disposal. For example, the decision to phase out a particular chemical should be inextricably linked to the development and analysis of alternatives to fulfill the services provided by that chemical. This type of scheme will ensure that one hazard is not replaced by an unknown, yet potentially greater one. This type of scheme will also ensure that those undertaking an activity systematically and comprehensively study (and thus understand) potential impacts before initiating an activity so as to avoid or make plans to monitor or mitigate such impacts.

8. A duty to monitor, understand, investigate, inform, and act.

In addition to providing convincing evidence of safety for new activities, the Precautionary Principle calls for a duty on those initiating an activity to investigate and understand the potential impacts of their actions and to act to prevent harm when a potential impact has been identified. Ignorance is no longer acceptable for those undertaking activities which may adversely impact human health or ecosystems. Lack of knowledge about cause-effect relationships between an activity and some adverse consequence, often caused by lack of adequate assessment, is frequently used as an argument for postponing actions to prevent potential harm. For example, recent research has found that for the vast majority of high volume/high concern chemicals we do not even have the most basic toxicological information (EDF, 1997). However, companies will often note that a chemical is safe or approved when what they really mean is that it has never been studied (innocence by association)! To combat this problem, the Danish government has proposed a chemicals categorization scheme which considers a chemical as toxic as the most toxic one in its class if there is no evidence to the contrary (Bro-Rasmussen, 1996).

The Precautionary Principle certainly calls for more, not less science to better understand the complexity of ecosystems and the impacts of different stressors on them. Harm cannot be prevented if it is unknown or poorly characterized. Poor characterization of a problem could lead to treatment of the wrong problem, an error of the third kind (Clemen, 1991). While potential harm should be studied (and avoided) before an activity commences, the promoter of an activity should be required to conduct on-going analysis of potential impacts, to inform regulators and the public of the results of this analysis (allowing the public to independently review these analyses and conduct its own analysis as is currently done under some "good neighbor" agreements), and to take action when potential impacts are identified. This duty could consist of periodic assessment/audit requirements, long term monitoring, and pre-manufacture/start-up impact statement any time a change is made to a product, process, or activity. This duty would apply to both new and existing activities. A suspicion of harm would require notification of the public and government authorities so that subsequent actions to restrict or further study the potential impact could be initiated.

9. Methods for participative/democratic decision-making.

Democratic decision-making can be considered an important virtue of the Precautionary Principle. Decision-making in the face of uncertainty is clearly a policy decision, though hard science (as well as the "softer sciences) plays a critical role in providing the information on which to analyze and base a decision. Thus, decisions surrounding to undertake or halt an activity are necessarily public decisions because of their potential to impact ecosystems and public health. Decision-making under the principle involves a greater emphasis on holistic analysis, qualitative measures, values, and the weighing of various types of evidence. This considerably differs from decision-making using risk assessment, which is based on closed, assumption-laden, quantitative models and the supremacy of objective science. It is almost impossible for citizens to participate in "expert" driven risk science due to the complexities of mathematical models and heavy reliance on assumptions. Indeed, most scientists would concur that there is no place for citizens in the risk assessment process. At any rate, research has demonstrated that secretaries of experts are often more accurate in predicting uncertain future events than the experts themselves, as a result of differences in the way they assimilate and analyze information (IJC, 1995).

Because environmental decisions are public decisions, structures need to exist for citizens (workers, communities, teachers, etc.) to participate both in the collection of information on which to base decisions and in the decisions themselves. This would apply to both scientific and technological decisions. A simple right to know is insufficient under the Precautionary Principle, as is the right to participate without adequate information. Some of the most interesting work on democratic decision-making structures is currently being undertaken in Denmark, Norway, Sweden, and the Netherlands. In the U.S., the Loka Institute is trying to develop, refine, and disseminate these models. Three particular models for democratic participation in decision-making have been discussed:

  1. Consensus conferences. Consensus or "lay person" conferences are processes to involve citizens in decision-making regarding complex issues. These are not multi-stakeholder type negotiations where often those with greater power have a greater say in the outcome (or have greater access to information and resources). In these conferences, a randomly selected group of citizens first receives background on a specific issue (e.g., genetic engineering). This citizen panel then participates in a hearing where they first day they hear testimony from experts supporting and against a particular activity. The second day the panel can cross-examine experts. The third day the citizens deliberate and then write a detailed final report in their own words. Approximately forty of these panels have taken place in Denmark. In Norway, a consensus conference on genetic engineering led that country's government to ban genetically modified crops, because of the potential risks involved and the lack of a need for such crops (Sclove, 1997).

    Clearly, these types of panels work best for big issues with large societal repercussions. Because of the time and financial resources required to establish such panels, they could not practically be established for each facility siting, chemical manufacture, or other activity. Nonetheless, a hybrid of the consensus conference, such as a citizen grand jury invoked for six month periods, or citizen advisory panels might be useful.
  2. Scenario workshops. Scenario workshops provide a participatory method for addressing broad societal questions such as sustainability. Critical to these workshops is the framing of questions to be answered. In these workshops, different stakeholders come together to analyze different scenarios to answer a particular question/problem. In doing this they address barriers to solving a problem and an action plan for solving the question/problem. This exercise would be similar to backcasting mentioned under point 2.
  3. Science shops. Key to citizen participation in decision-making is their ability to independently and critically examine evidence and conduct their own studies. This type of "citizen science" leads to new types of qualitative and quantitative input into decision-making processes. Science shops are independent university-based centers that provide sophisticated referral services for communities and community groups needing to undertake environmental research. The science shop responds directly to the citizen concerns and engages in participatory research. The citizens or citizen's group must be both willing and capable of undertaking social action as a result of the research (Sclove, 1997).

In general, these democratic decision-making methods allow those potentially impacted by a activity to participate in decisions affecting their lives and health. They also lead to a greater transparency in the science used to reach decisions and the decisions themselves. While they may be difficult to implement on a wide scale and for each individual decision, they provide a model of how to involve citizens in complex decisions under uncertainty. On a smaller scale, citizens can and should be involved on panels to make decisions regarding permitting, siting, impact assessments, and whether to take action on a potentially harmful activity.

10. Strong enforcement

Lastly, strong enforcement is necessary to ensure that precautionary, preventive decisions are being undertaken. Enforcement, combined with strong regulation plays a critical role in the development of safer, cleaner technologies and practices. Strong regulations alone will not achieve precautionary results. Financial and criminal repercussions for those failing to act in a precautionary manner need to exist. Even the threat of enforcement can create the necessary uncertain conditions for those undertaking an activity to act with precaution. Strong enforcement should be coupled with outreach to firms, farmers, developers, and others undertaking potentially harmful activities (as well as citizens) to assist them in understanding their impacts and in developing and implementing safer and cleaner technologies/activities which fulfill their needs and serve the public good.

Examples of the Precautionary Principle in Use

This section describes several examples where the Precautionary Principle has been either implicitly or explicitly incorporated into environmental decision-making. While these cases represent either proposed legislation or policy decisions which have yet to be fully implemented, they provide concrete examples of how the Principle can be used in decision making, resulting in preventive policies which could lead to the development of safer cleaner processes and products. Although these are examples of use of the principle with regards to toxic substances, there are numerous cases where the principle has been invoked to prevent harm to ecosystems from development, farming, fishing, deforestation, and genetic engineering.

1. Massachusetts Precautionary Principle Act. In early 1997, State Representative Pamela Resor introduced a bill to the Massachusetts legislature entitled "An Act to Establish the Precautionary Principle as the Guideline for Developing Environmental Policy and Quality Standards for the Commonwealth". This bill was introduced in response to industry supported bills which would require risk assessments and cost benefit analyses for any major environmental regulation and which would dismantle to successful Toxics Use Reduction Act, in favor of a voluntary pollution prevention scheme. The proposed legislation calls for state agencies to apply the precautionary principle where they "are reasonable grounds for concern that a procedure or development may contribute to the degredation of the air, land and water of the commonwealth." The act also calls on agencies to implement the precautionary principle through pollution prevention, evaluation of alternatives, and research into the effects of potentially environmentally harmful activities. While the bill was sent to study following a short hearing, it has provided the impetus for more comprehensive precaution based legislation (Commonwealth of Massachusetts, 1997).

2. London Dumping Convention. The London Dumping Convention (LDC) came into effect in 1972 to protect the marine environment from dumping of wastes and other matter. A need for the Precautionary Principle, similar to that incorporated into other international agreements was introduced into later revisions of the LDC. In 1996, parties to the Convention, noting the need for more stringent measures to protect the marine environment, passed a Protocol to the Convention. The Protocol first establishes a list of materials that can be dumped at sea (a reverse list). All other materials are prohibited from dumping. Second, those applying to dump allowed wastes must first undergo a waste assessment audit and demonstrate that there was no other way to reduce, reuse, or recycle those wastes. If a country determines that there are other alternatives for the waste, a permit can be denied. Permits are regularly reviewed to ensure that there are no alternatives to the dumping (IMO. 1996).

3. Swedish Chemicals Policy. In early 1997, the Swedish government published its chemicals policy for future sustainability. The concepts of substitution and precaution have been at the heart of the Swedish government's chemicals policy for years. For example, in 1989 the Swedish National Chemicals Inspectorate instituted the concept of "sunsetting" for those substances not compatible with sustainable development. Chemicals were identified for sunsetting based on different characteristics such as toxicity, persistence, and bioaccumulation. The scheme was developed because the government had realized that chemical by chemical analyses were too slow to avoid unwanted consequences.

The 1997 review looks at new issues confronting the Swedish government going into the next century, such as changes in chemicals in commerce, membership in the European Community, and endocrine disruption. The review sets ambitious goals for the future, including labeling of all toxic materials by the year 2002; products free from bioaccumulative and extremely toxic substances by 2008; production processes free of toxic substances by the year 2012; and a phase out of polyvinyl chloride plastic by the year 2007. According to the Swedish government, these medium and long range targets are necessary to create driving forces for change and are based on long enough time frames to allow sufficient transition planning for impacted communities.

The government justifies this ambitious program noting that chemical risks are more complex and difficult to assess than before; single substance assessments are progressing too slowly; and that the continued use of persistent and bioaccumulative substances poses a threat to sustainable development. The cornerstones of this new policy are: the precautionary principle (measures to restrict chemicals are based on suspected rather than proven effects); industry responsibility to prove that their products are safe; regulations and enforcement; and generic approaches to identifying chemicals and processes for restriction (Ministry of Environment, 1997).

4. International Joint Commission. In its Sixth Biennial Report on Great Lakes Water Quality, the International Joint Commission (IJC) addresses progress in achieving the goals of the Great Lakes Water Quality Agreement (including virtual elimination of the input of persistent toxic substances into the Great Lakes). The IJC noted the damage caused by persistent and bioaccumulative substances in the Great Lakes Basin and the critical need to address those. They also noted that attempts to manage such chemicals based on the notion of assimilative capacity in the environment, had failed miserably. As a result, the Commission proposed a strategy calling for the sunsetting of all persistent toxic substances in the Great Lakes Ecosystem, because they cannot be managed safely. "Such a strategy should recognize that all persistent toxic substances are dangerous to the environment, deleterious to the human condition, and can no longer be tolerated in the ecosystem, whether or not unassailable scientific proof of acute or chronic damage is universally accepted." In devising its policy, the IJC has specifically rejected risk assessment and cost-benefit analysis, noting the difficulties and complexities involved in such assessments and the fact that they are not relevant to the virtual elimination commitment.

To identify candidate chemicals for action, the IJC has developed a policy framework for addressing areas of scientific uncertainty, based on the weight of evidence approach, so that unequivocal evidence of cause-effect is not needed before taking action. Evidence contributing to the weight of evidence can come from various sources. The Commission notes

"If taken together, the amount and consistency of evidence across a wide range of circumstances and/or toxic substances are judged sufficient to indicate the reality or strong probability of a linkage between certain substances or class of substances and a conclusion of causal relationship can be made. This conclusion is made on the basis of common sense, logic and experience as well as formal science. Once this point is reached, and taking a precautionary approach, there can be no defensible alternative to recommending that the input of those substances to the Great Lakes be stopped….The burden of proof must shift to the proponent of the substance to show that is does not or will not cause the suspected harm, nor meet the definition of a persistent toxic substance."

Using this rationale, given the wide range of adverse human health and environmental impacts of organochlorine compounds, the IJC has called for a phase-out of the manufacture and use of chlorine and chlorine-containing compounds as industrial feedstocks in the Basin (IJC, 1992, 1994, 1994a, 1995).

Conclusions

In this paper I have attempted to develop a general framework for operationalizing the Precautionary Principle. The components of such a framework attempt to put into action the specific virtues of the Precautionary Approach which I outlined at the outset of this analysis. While it would be unrealistic to believe that all of these components could be implemented into law and policy in the short term there are parts of this framework which should possibly receive more immediate attention than others. For example, the first step towards bringing the Precautionary Principle to the forefront might be incorporation of the principle into local, state, and national laws, as a guiding principle of environmental and public health law and policy. This would lead to a clear definition of the principle and increase awareness of its role in decision-making under uncertainty. Simultaneously, efforts could be undertaken to refine existing laws to focus on prevention, alternatives assessment, and burden shifting. All the while, a different set of actors will need to be working to revamp scientific method so that it supports precautionary decision-making.

In the short term we need concrete examples of where the Precautionary Approach has been implemented (not just enunciated) on a local, state, national, and international law. These examples will help to provide momentum and greater acceptance of the principle as a guide to environmental and public health policy, For the principle to gain wider acceptance, we will also need to educate members of government agencies, legislatures, business, workers, and the public of its benefits and develop clear responses to its critics. It will not be an easy road as the Principle requires an entirely new way of addressing and solving environmental problems. As Einstein once said, "The problems we face today cannot be solved with the same level of thinking as when they were created."

References Cited

Ashford N et al. The Encouragement of Technological Change for Preventing Chemical Accidents: Moving Firms from Secondary Prevention and Mitigation to Primary Prevention, (Cambridge, MA: Center for Technology, Policy and Industrial Development, Massachusetts Institute of Technology, 1993)

Ashford N, et al., Worker Compensation for Radiation-Induced Illness: A Re-Examination of Past Practices and Options for Change, (Cambridge: Ashford Associates, 1996).

Barrett K, Raffensperger C. Precautionary Science, Draft paper prepared at the Wingspread Conference on Implementing the Precautionary Principle, January 23-25, 1998.

Bodansky D, The precautionary principle in US environmental law, In O'Riordan T and Cameron J Interpreting the Precautionary Principle (London: Earthscan, 1994)

Bro-Rasmussen F et. al. The Non-Assessed Chemicals in the EU: Report and Recommendations from and Interdisciplinary Group of Danish Experts, (Copenhagen: Danish Board of Technology 1996/5).

Cameron J, Abouchar J. "The precautionary principle: a fundamental principle of law and policy for the protection of the global environment," Boston College International and Comparative Law Review 14:1-27 (1990)

Clemen R, Making Hard Decisions: An Introduction to Decision-Analysis, (London: Duxbury Press 1991)

The Commonwealth of Massachusetts. An Act to Establish the Principle of Precautionary Action as the Guideline for Developing Environmental Policy and Quality Standards for the Commonwealth, H. 3140 (1997)

Contini, S. et al. Benchmark Exercise on Major Hazard Analysis. Luxemboug: Commission of the European Communities, EUR 13386 EN, 1991

Costanza R, Cornwell L. "The 4P approach to dealing with scientific uncertainty," Environment, 34:9 (1992)

Cross F, "Paradoxical perils of the precautionary principle," Washington & Lee Law Review 51: 851-925 (1996).

Department of Environment of the United Kingdom. Ministerial Declaration, (London: Second International Conference on Protection of the North Sea, 1992)

Dethlefsen V, et al. "The precautionary principle: Towards anticipatory environmental management." In Jackson T. Clean Production Strategies (New York: Lewis Publishers, 1993)

Dovers S. "A framework for scaling and framing policy problems in sustainability," Ecological Economics, 12:93-106 (1995)

Environmental Defense Fund (EDF), Toxic Ignorance: The Continuing Absence of Basic Health Testing for Top-Selling Chemicals in the United States, (New York: Environmental Defense Fund, 1997).

Funtowicz and Ravetz, In R. Costanza, Ed. Ecological Economics: Science and Management of Sustainability, (New York: Columbia University Press, 1991)

Gee D. Approaches to Scientific Uncertainty, Conference on Transport Policy and Urban Health, London School of Hygiene and Tropical Medicine, 4-7 April 1995

Gee D. Criteria for managing uncertainty and regulation in public policy. Draft manuscript (1997).

Greenpeace International. Avoiding Western mistakes: a guide to clean investment in Eastern and Central Europe. Conference on the Environment, Industry and Investment in Central and Eastern Europe, Budapest Hungary, Nov 20-22, 1991.

Hickey J, Walker W. "Refining the precautionary principle in international environmental law," Virginia Environmental Law Journal 14:423-454 (1995).

Hill AB. "The environment and disease: association or causation?," Proc. R. Soc. Med 58:295-300 (1965).

International Joint Commission (IJC). Sixth Biennial Report on Great Lakes Water Quality, (Windsor, Ontario: IJC, 1992)

International Joint Commission (IJC). Applying the Weight of Evidence: Issues and Practice, (Windsor, Ontario: IJC, 1994)

International Joint Commission (IJC). Seventh Biennial Report on Great Lakes Water Quality, (Windsor, Ontario: IJC, 1994a)

International Joint Commission (IJC). 1993-95 Priorities and Progress Under the Great Lakes Water Quality Agreement, (Windsor, Ontario: IJC, 1995).

International Maritime Organization (IMO). 1996 Protocol to the Convention on the Prevention of Marine Pollution by Dumping of Wastes and Other Matter, 1972 and Resolutions Adopted by the Special Meeting, LC/SM 1/6, 14 November 1996

Ludwig D, et. al. "Uncertainty, resource exploitation, and conservation: Lessons from history," Science 260: 17,36 (April 2, 1993)

Massachusetts Weight-of-Evidence Workgroup. Draft Report: A Weight-Of-Evidence Approach for Evaluating Ecological Risks, (Boston: Massachusetts Department of Environmental Protection, 1995)

M'Gonigle RM et al. "Taking uncertainty seriously: From permissive regulation to preventive design in environmental decision-making," Osgoode Hall Law Journal 32:99-169 (1994)

Mee L. "Scientific methods and the precautionary principle" In Freestone D, Hey E eds. The Precautionary Principle and International Law, (Boston: Klewer Law International, 1995)

Ministry of the Environment (Sweden). Towards a Sustainable Chemicals Policy, (Stockholm: Government Official Reports 1997:84)

National Academy of Sciences (NAS). Risk Assessment in the Federal Government: Managing the Process, (Washington DC: National Academy Press, 1983)

O'Riordan T, Jordan A. "The precautionary principle in contemporary environmental politics," Environmental Values 4:191-212 (1995)

Raffensperger C. Regulatory Agency Accountability Under the Precautionary Principle, Draft paper prepared at the Wingspread Conference on Implementing the Precautionary Principle, January 23-25, 1998

Sclove R. Democratizing Science and Technology, Lecture given as part of the University of Massachusetts Lowell Center for Competitive Enterprise Lecture Series, December 5, 1997.

Stijkel A, Reijnders L. "Implementation of the precautionary principle in standards for the workplace." Occupational and Environmental Medicine 52:304-312 (1995)

von Moltke K. "The vororgeprinzip in West German environmental policy," In Royal Commission on Environmental Pollution. Twelfth Report: Best Practical Environmental Option, (London: HMSO, 1988)

von Weizacker C, "Lacking scientific knowledge or lacking the wisdom and culture of not-knowing," in Van Dommelen A, ed., Coping with Deliberate Release: the Limits of Risk Assessment (Tilburg: International Centre for Human and Public Affairs, 1996).

Wynne B. "Uncertainty and environmental learning: Reconceiving science and policy in the preventive paradigm," In Jackson T. ed., Clean Production Strategies, (Boca Raton: Lewis Publishers, 1993)

Ziem G, et. al. Health-Based Exposure Limits and Lowest National Occupational Exposure Limits, Health-Based Exposure Limits Sub-Committee, American Public Health Association. Draft 9/1/1990.Table 1: Criteria for scaling and framing policy problems in sustainability

(Dovers, 1995)

Problem framing attributes:

1. Spatial scale of cause of effect: local-national-regional-international-global

2. Magnitude of possible impacts (on both humans and natural systems)

3. Temporal scale of potential impacts

timing (near, medium, long term

longevity (short, medium, long)

4. Reversibility (easily quickly reversed or expensive/irreversible)

5. Mensurability of factors and processes (well known, ignorance)

6. Degree of complexity and connectivity

Response-framing attributes:

7. Nature of cause(s) (simple, systemic)

8. Tractibility (availability of means, acceptability of means)

9. Public concern


Table 2: Attributes related to strength of association between assessment and measurement endpoints

MassachusettsWeight of Evidence Workgroup (MWEW, 1995)

1. Biological linkage between measurement endpoint and assessment endpoint

2. Correlation of stressor to response.

3. Utility of measure for judging environmental harm

4. Extent to which data quality objectives are met

5. Site specificity

6. Sensitivity of the measurement endpoint for detecting changes

7. Spatial representativeness

8. Temporal representativeness

9. Quantitativeness

10. Use of a standard method.

Table 3: Principles of Effective Environmental Management

(Ludwig, et. al, 1993)

1. Consider a variety of plausible hypotheses

2. Consider a variety of possible strategies

3. Favor actions that are robust to uncertainties

4. Hedge

5. Favor actions that are informative

6. Probe and experiment

7. Monitor results

8. Update assessments and modify policy accordingly

9. Favor actions that are reversible

10 Solicit all available information from all stakeholders

(taken from Raffensperger C. Incentives and Barriers to Public Interest Research, Switzer Fellows Speech, the Headlands, CA, September 1997.)

Table 4: Hill Criteria for distinguishing between association and causation in epidemiologic studies:

(Hill, 1965)

1. Strength of evidence

2. Consistency of evidence

3. Specificity of effect

4. Temporality of effect

5. Dose response of effect

6. Plausibility of effect

7. Coherence with existing knowledge

8. Experimental evidence

9. Analogy (structure activity)

Table 5: Criteria for managing uncertainty and regulation in public policy

(Gee, 1997)

1. Who/what gets the benefit of scientific doubt

2. Who takes the burden of proof

3. What level of proof is appropriate

4. Which are the toxicologically dominant and strategically dominant causes

(is it of public health significance but not of statistical significance)

5. What are the multiple benefits of risk reduction.

6. What is the optimal balance between time spend establishing causation and time spent reducing risk?

7. What is the likely size and distributions of false negatives and false positives?

8. What is the optimum mix of policy instruments, targets, and timetables that will maximize overall cost-effective public policy.

9. What mechanisms are needed for establishing need in public policy on new products/services so as so optimize the balance between innovation and risk.

Table 6: Proposed criteria to guide precautionary decision-making

Causal inference criteria

Strength of evidence (experimental and observational)

Amount and consistency of evidence across a wide range of circumstances

Temporality of effect

Coherence with existing knowledge

Plausibility of effect

Have all evidence and all plausible hypotheses been considered

Power of study(ies) to detect an effect

Have false negatives (type II errors) been minimized?

Is the evidence statistically significant or of public health significance?

Is there some presumption of causal relatedness based on previous experience which would lower the evidentiary standard? (i.e., is there evidence from any other similar case that would lead one to believe that the a similar impact could be considered in the present case).

What is the adverse effect being studied and is it the correct one?

Decision-stakes criteria

Spatial scale of cause of effect: local-national-regional-international-global

Magnitude of possible impacts (on both humans and natural systems)

Temporal scale of potential impacts

timing (near, medium, long term

longevity (short, medium, long)

Reversibility (easily quickly reversed or expensive/irreversible)

Mensurability of factors and processes (well known, ignorance)

Degree of complexity and connectivity

Is the action robust to uncertainties (error friendly)

Do alternatives or measures exist to reduce or eliminate potential harm (ease of prevention)?

What is the trade off between further study and potential impacts?

Appendix A: Historical development and rationale for the Precautionary Principle

The precautionary approach evolved as a response to the environmental and human health impacts caused by the rapid industrial growth following World War II and the weaknesses of early pollution control legislation. As environmental legislation developed in the late 1960s and 1970s, scientists and policy makers began to the realize the enormous difficulties in demonstrating cause-effect relationships between environmental contamination and ecosystem and human health harm. With increasing knowledge about the complexities of ecosystems, the human body, and the impacts of various stressors increases, scientists realized that we understand less than we actually thought we did about these complex systems (ie, the more we know, the less we realize we know) (ref Clark). During the 1970s decision-making tools, such as risk assessment and cost-benefit analysis, were developed to bridge the gap between uncertain science and the political need for decision-making to limit harm. However, in their development too much emphasis was placed on the role of science to model and predict harm in extremely complex ecological and human systems. In essence, risk assessment, which was originally developed for mechanical problems such as bridge construction where the technical process and parameters are well­defined and can be analysed, took on the role of predictor of extremely uncertain and highly variable events.

Current environmental decision­making processes, based primarily on level of risk, thus suffer from several limitations, which constrain their ability to identify, anticipate, and prevent potential harm to human health and the environment. Decisions to take action to limit the input of potentially hazardous substances (or hazardous activities) in the environment are often taken only in response to the scientific establishment of a causal association between a substance activity and a well-defined negative impact. Proving causal associations takes both extensive time and resources. During this research period irreversible human health or environmental effects often occur. Needing additional knowledge to model problems became a means for governments or those undertaking a potentially hazardous activity to stall action to prevent harm. Numerous difficulties also exist in proving such causal associations. Among these are: (1) Limitations in scientific knowledge. The capacity to identify and prevent potential adverse effects is limited by the present state of scientific knowledge (Gee, 1995). Though ecosystem harm is wrought with large scale uncertainties and indeterminacy (due to openendedness in systems), scientists continue to seek to threshold for negative effects, a so-called "assimilative capacity" for the environment (a predicable level of harm from which an ecosystem can recover) (Wynne, 1993); (2) problems in statistical power. Regulatory programs often fail to consider statistical power in decision-making and attempt to minimize type I errors (incorrectly concluding that there is an effect when one does not exist) in monitoring and health studies; (3) low level adverse effects (M'Gonigle, 1994). Evidence regarding the effects of several classes of chemicals and stressors alone or together at low levels of exposure is slowly evolving; (4) difficulties in addressing cumulative effects. Science has not even begun to address the wide range of physical and chemical stressors to which humans and ecosystems are exposed, tending to focus on single chemical, single medium effects; and (5) financial and resource limitations. Recent studies (EDF, 1997) have demonstrated that our understanding of the effects of toxic chemicals on human or ecosystem health is extremely limited despite the large quantity of funding that goes into studying adverse effects.

Given the limitations of science to "prove" cause effect relationships in the environment, there is a significant need for the development of new public policy approaches to anticipate and prevent harm. The question of what society should do in the face of uncertainty regarding cause and effect relationships is necessarily a question of public policy, not science. A decision not to act in the face of uncertainty, to await further scientific evidence, is a clear policy decision, not a scientific one. Though most scientists will not admit it, so-called "sound science" approaches to decision-making, such as risk assessment, are highly reliant on policy/scientific assumptions (which are frequently very unscientific or subjective in nature). Risk assessment, for example relies on at least 50 different assumptions about exposure, dose-response, and relationships between animals and humans which are certainly policy driven (NAS, 1983). As a result, two risk assessments on the same problem conducted by different researchers can vary widely in results (Contini, 1991). Nonetheless, science (and in some respects engineering) has a critical role in increasing our understanding of the complexities of environmental harm, monitoring for potential harm, and methods to prevent that harm.

The Precautionary Principle arose as a counterbalance to science's inability (and subsequently that of public policy) to predict, anticipate, and prevent ecosystem and human health harm due to environmental degradation. The principle is relatively new to environmental decision­making, having emerged during the early 1970s in West Germany ­­ "Vorsorgeprinzip" in German (Von Moltke, 1988). At the core of early conceptions of this principle in Germany was the belief that society should seek to avoid environmental damage by careful forward planning, blocking the flow of potentially harmful activities (O'Riordan and Jordan, 1995). The Vorsorgeprinzip developed into a fundamental principle of German environmental law and has been invoked to justify the implementation of vigorous policies to tackle acid rain, global warming and North Sea pollution.

The precautionary principle has since flourished in international statements of policy; conventions dealing with high­stakes, low scientific certainty environmental concerns; and national strategies for sustainable development (Dethlefsen, 1993, O'Riordan and Jordan, 1995). It has gained international acceptance as a guiding principle for environmental decision­making. The precautionary principle was first introduced in 1984 at the First International Conference on Protection of the North Sea. Following this conference, the principle was integrated into numerous international conventions and agreements including the Maastricht Treaty, the Barcelona Convention, and the Global Climate Change Convention, among others. Hickey and Walker (1995) 1995) provide a useful summary of the treaties in which the Precautionary Principle has been implemented. It has been implicitly incorporated into several environmental laws in the U.S., such as the Pollution Prevention Act of 1990. The International Joint Commission has called for a phase­out of persistent organic chemicals in the Great Lakes ecosystem based on "weight­of­evidence" criteria. The criteria state that action should be taken to prevent environmental damage when evidence from several studies taken together indicate actual or potential environmental harm.