Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Pollution control, administrative discretion and science: a journey through the maze of environmental… Findlay, Caroline K. H. 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_fall_findlay_caroline.pdf [ 5.57MB ]
JSON: 831-1.0077528.json
JSON-LD: 831-1.0077528-ld.json
RDF/XML (Pretty): 831-1.0077528-rdf.xml
RDF/JSON: 831-1.0077528-rdf.json
Turtle: 831-1.0077528-turtle.txt
N-Triples: 831-1.0077528-rdf-ntriples.txt
Original Record: 831-1.0077528-source.json
Full Text

Full Text

POLLUTION CONTROL, ADMINISTRATIVE DISCRETION, AND SCIENCE: A JOURNEY THROUGH THE MAZE OF ENVIRONMENTAL LAW  by CAROLINE K. H. FINDLAY B.A. (Hons.), University of Western Ontario, 1982 LL.B., Queen's University, 1985  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF LAWS in THE FACULTY OF GRADUATE STUDIES (FACULTY OF LAW) We accept this thesis as conforming to the required standard.  THE UNIVERSITY OF BRITISH COLUMBIA September 1993 © Caroline K.H. Findlay, 1993  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  DeDakftlent of  vi  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  11q3  ii ABSTRACT  This thesis explores the interface between environmental regulatory discretion and scientific complexities. Two key observations about the nature of scientific information are made which lead to the argument in this thesis that science must be made more explicit in environmental decision making processes. First, scientific analysis is crucial to understanding the impact of pollution on the natural environment. Thus, it is fundamental to the design and implementation of environmental laws. Secondly, scientific information has certain methodological limitations and inherent uncertainties which often make it subject to interpretation and value judgments. These judgments involve important policy choices about environmental risks. In Canada, the use of science to shape environmental laws is a matter for bureaucratic discretion that is rarely subject to external scrutiny. This thesis argues for an express statutory obligation on environmental administrators to disclose and to explain the scientific analysis used to support the exercise of their discretion in making regulatory decisions. Discretionary powers are a necessary and permanent part of the Canadian legal landscape of environmental protection. Within this regulatory context, science is segregated as the rational "factual" basis for decisions when, in fact, science cannot be disentangled from the economic, political and social dynamics that influence regulatory discretion. The implications of this are illustrated by the experience of the United States which is considered for purposes of comparison in this thesis. Under U.S. environmental laws, regulators are required to disclose and to explain the  iii scientific analysis used to support their regulatory decisions. While it is clear that a procedural "fix" is not a panacea, it does offer some distinct advantages that enhance the democratic legitimacy of environmental decision making. In short, a legal duty of this kind will improve the process of environmental decision making in Canada: (1) by requiring a more thoughtful analysis of the administrative task and the relevant information and thus increasing the accountability of regulators; and (2) by helping to harness and make accessible a valuable pool of knowledge. As a result, the integrity of the decision making process in Canada will be improved by making analysis more transparent and subject to challenge.  iv TABLE OF CONTENTS Abstract^  ii  Table of Contents^  iv  Excerpt from the Heidelberg Appeal^  vi  Chapter One^Introductory Overview^ A. B. C. D.  1  2 Lessons From Our Past^ Making Decisions under Environmental Laws^9 The Context of Science^ 16 The Methodology and Scope of this Thesis ^20  Chapter Two^The Administrative State and Environmental Laws^25 25 A. Introduction^ B. The Age of Bureaucracy and the Nature of Discretion^27 C. The Theoretical Framework --The Political, Administrative and Legal Contexts^39 D. Discretion and Environmental Laws^51 57 E. Conclusions^ Chapter Three^The Role of Science in Environmental Regulation^59 A. B. C. D. Chapter Four  Introduction -- The Need for Science^59 The Nature and Use of Scientific Knowledge^66 The Debate: The Separateness of Science^83 Conclusions^ 85  The Case for Scientific Explication -- An Examination of the U.S. Experience^ 88 A. Introduction^ 88 B. The Clean Air Act - An Overview^92 C. Standard Setting and Science -A Closer Look^103 D. Conclusions^ 123  Chapter Five^Concluding Remarks^  126  Bibiliography^  135  vi  We want to make our full contribution to the preservation of our common heritage, the Earth. We are however worried, at the dawn of the twenty-first century, at the emergence of an irrational ideology which is opposed to scientific and industrial progress and impedes economic and social development. We contend that a Natural State, sometimes idealized by movements with a tendency to look toward the past, does not exist and probably never existed since man's first appearance in the biosphere, insofar as humanity has always progressed by increasingly harnessing Nature to its needs and not the reverse. We fully subscribe to the objectives of a scientific ecology for a universe whose resources must be taken stock of, monitored and preserved. But we herewith demand that this stock-taking, monitoring and preservation be founded on scientific criteria and not on irrational preconceptions.... We intend to assert science's responsibility and duties toward society as a whole. We do however forewarn the authorities in charge of our planet's destiny against decisions which are supported by pseudo-scientific arguments or false and non-relevant data.*  *Excerpt from the Heidelberg Appeal, signed by 218 scientists worldwide and addressed to the heads of state attending the Earth Summit, Rio, 1992 (Source: "Beware the False Gods in Rio", Wall Street Journal, June 1, 1992)  1 CHAPTER ONE INTRODUCTORY OVERVIEW  The face of pollution control in Canada has changed over the last two decades: it lacks the innocence of the earlier years, but now possesses the first wrinkles of wisdom gained through experience.'  The primary concern motivating this project has been to explore, from a legal perspective, the interface between science and law in the area of environmental protection. On one hand, it appears obvious that scientific understanding is fundamental to both addressing and preventing the impact of pollution on the natural environment. It is key knowledge. On the other hand, scientific uncertainty in this area often means that this information is subject to interpretation and value judgments. There is a conspicuous absence on the face of many Canadian environmental laws as to how much reliance should be given to scientific analysis. In Canada, this is a matter for bureaucratic discretion that is rarely subject to external scrutiny. The combined effect of these two observations, that science is both necessary to, yet subject to the vagaries of, the administrative process, leads to the somewhat obvious legal argument presented in this thesis, namely, that science is so crucial to the shape of environmental policies and laws that the role of science must be made more explicit in decision making processes. It will be argued that there should be a 1 K. Webb, Pollution Control in Canada: The Regulator y Approach in the 1980's. Administrative Law Series, A Study Paper prepared for the Law Reform Commission of Canada, (Ottawa: Minister of Supply and Services, 1988), at pg. 9.  2 statutory duty on environmental regulators to disclose and explain the science behind their decisions. In order to give this argument context, this introductory chapter begins by examining the general Canadian experience to date with regulatory environmental protection. It then provides an overview of the general nature of environmental laws and how science interfaces with this regulatory process. This discussion will also explain the key terms used in this thesis. Finally, the methodology and limitations of this thesis will be set out.  A. Lessons From Our Past Canadian environmental policies and laws have, but for a few exceptional events, developed in an evolutionary manner, gradually adjusting to the understandings brought by experience. Over the last twenty years, this experience has taught us a number of valuable lessons about the dynamics and realities of the design and implementation of environmental protection laws. We have learned that the solutions to environmental concerns can be costly and that they are extremely complex, both technically and politically. There is increasing judicial recognition of this situation. For example, in a recent Ontario case dealing with PCB contamination, Fraser J. made the following observation about the the "arbitrary" limit of 50 parts per million of PCBs set as the level in regulations triggering legal action, in this case, clean-up: "Such regulatory decisions reflect a complex weighing of choices in terms of economic activity, costs, technological options and social  3 goals."2 The costs and the scientific/technological and political complexities associated with environmental protection are important factors in the history of Canadian environmental regulation that will inevitably continue to shape the future. Although these themes are not necessarily comprehensive of all that has been learned, they are considered central to the Canadian experience to date.3  The Costs. Evidence of the often enormous financial costs associated with efforts to  control or eliminate pollution are plentiful4, but consider just one example. Recently, the Canadian federal government has put in place three sets of regulations applicable to the pulp and paper industry.5 It is estimated that the pulp and paper industry will incur costs of about $4.85 billion in order to comply with these  2  R. v. Consolidated Mayburn Mines Ltd. et. aL (1992), 73 C.C.C. (3d) 268 (Ontario Prov. Div.),  at pg. 277.  3 These themes, and others (such as the ideological dimensions of the environmental movement), are elaborated on by many commentators in the literature. For some interesting Canadian perspectives, see the following (references in full in bibliography): Boardman, Canadian Environmental Policy; Doern, Getting It Green; MacDonald, The Politics of Pollution; Schrecker, Political Economy of Environmental Hazards; Tingley, Into the Future: Environmental Law and Policy for the 1990's; and Webb, Pollution Control in Canada. For example, it has been estimated that the costs of environmental regulation in the United States average about $90-100 billion/year, about half of the total cost of government regulation. It is estimated that by the year 2000 this will increase to $155 billion or 2.5% of the nation's GNP. These figures are cited by E. Donald Elliott, "Environmental Law at a Crossroad", 20 Northern Kentucky Law Review 1 (1992), at pg. 1. It is difficult to readily obtain any comparable Canadian statistics. These three regulations are: (1) Pulp and Paper Mill Effluent Chlorinated Dioxins and Furans Regulations, promulgated under the Canadian Environmental Protection Act ("CEPA"), SOR/92 267; (2) Pulp and Paper Mill Defoamer and Wood Chip Regulations, promulgated under CEPA, SOR/92 268; and (3) Pulp and Paper Effluent Regulations, promulgated under the Fisheries Act, S0R192 269. -  -  -  4 regulations.6 One analyst has commented that one of the most disturbing aspects of these regulations is that it is unclear what overall benefits will be gained by this expense.7 Furthermore, the allocation of these costs has presented a constant stumbling block for achieving effective environmental control: "Who pays for pollution prevention - governments, through spending, grants, and tax incentives? the polluting industry? the worker who loses his or her job? the province or country that sees economic growth transferred to its trade competitors because of its environmental regulation?"8. These are common questions in the environmental debate. The Science. Very little scientific assurance can be given about the real risks  or impacts that many human activities will have on the environment. Indeed, scientific uncertainties abound in this area. Again, the experience of the pulp and paper industry is exemplary of the nature of the scientific information that typically 6 This figure is based on cost estimates set out in the Regulatory Impact Analysis Statements accompanying the publication of the final regulations, supra, fn 5. The costs are computed over 20 years and discounted at 10%. This figure is also cited by William T. Stanbury, "Reforming the Federal Regulatoty Process in Canada, 1971 1992", Report to the Sub-Committee on Regulations and Competitiveness of the Standing Committee on Finance, published as Appendix SREC-2 in the Minutes of Proceedings and Evidence of the Sub-Committee, House of Commons, Third Session of the Thirty-Fourth Parliament 1991-1992, at pg. 10 and Table 6-3. -  7 Stanbury, supra, fn 6, at pg. 10. Professor Stanbury notes: "The Regulatory Impact Analysis Statement failed to provide any estimate of the benefits of the measure in either physical or economic terms." However, the Statement does attempt to set out qualitative benefits based on improved watercourse quality and the value of passing on to future generations improved environmental conditions. Professor Stanbury's comment is compounded by the fact that monetary values may not begin to capture the benefit or values inherent in the various perspectives that underlie environmental laws. While the advantages and disadvantages of "cost-benefit" analysis may be debated, it is at least arguable that without some systematic analysis, it will be difficult to justify such a huge cost. Query: could this money be better spent? 8 Doug MacDonald, The Politics of Pollution: Why Canadians Are Failing Their Environment, (Toronto: McClelland & Stewart Inc., 1991), at pg. 12.  5 exists in environmental regulation. Dioxins and furans, two substances known as organo-chlorines that have been linked to human health concerns and are associated with pulp mill effluent, have been a pollution issue for regulators over the last five years. Scientists have different views about the potential health risks these substances pose. A world-renowned expert on cancer causing substances, Dr. Bruce Ames, referring to dioxins (TCDD) states: "there is no convincing evidence the TCDD is carcinogenic or teratogenic [causing fetal abnormalities] in man, although it is in rodents."9 Compare this opinion with that of Dr. David Schindler, a scientist with the Canadian federal government for 22 years: "organo-chlorines are an extreme human carcinogen."10 Who is to be believed in such a case, and what are the  ultimate implications of this to a legal decision to regulate? With respect to dioxins and furans, despite these differing scientific opinions, the Canadian government has assessed the toxicity of these substances and has concluded "that these compounds are highly toxic and that they may enter the environment in quantities which have immediate and long term harmful effects on the environment and they constitute a danger to human health in Canada."11 This raises the question of what kind of scientific evidence was relied upon and why. In many ways, the answers can be cast as political ones.  9 "Dioxins: time for tough limits." The Vancouver Sun, Letters in the OP/ED page, pg. A13. [emphasis added]. 10  Ibid. [emphasis added].  11  Pulp and Paper Mill Effluent Chlorinated Dioxins and Furans Regulations, supra, fn 5, at pg.  1946.  6 The Politics. The last lesson learned is indeed a political one. For  environmental decisions to have integrity and credibility with Canadian citizens, the decision making processes need to be democratized. For too long, regulation of the environment has been seen as a closed-door negotiation process between regulators and industry alone, backed by a legal system allowing for little room to participate in or challenge decisions.12 Concern about the alienation that this has caused and cries for greater participation by other interested parties have resulted in reform efforts by policymakers to improve the processes of decision making. Indeed, this concern has prompted one observer to state that process is all that really matters in this area: "In democratic policymaking, process always matters. But in environmental policymaking, process seems to be everything."13 The dissatisfaction with how environmental policy and law have been managed in the past is a continuing theme in present reform efforts. One does not have to look far for evidence, or at least the rhetoric of it, regarding the recent preoccupation of Canadian governments with improving environmental decision making processes. In British Columbia, a discussion paper outlining a five-year action plan for a new legislative framework (tentatively entitled the British Columbia Environmental Protection Act or the "BCEPA") expressly addresses the need to make  Bi-partite bargaining has been endemic to the process of Canadian environmental regulation. For further discussion, see, for example, A.R. Thompson's classic study, Environmental Regulation in Canada: An Assessment of the Regulator), Process, (Vancouver: Westwater Research Institute, 1980). 12  13 G. Bruce Doern, Canadian Environmental Policy: Why Process is Almost Everything Commentary #19 for the C.D. Howe Institute, (Ottawa: Renouf Publishing Company Limited, July 1990), at pg. 2.  7  procedural reforms: The provisions of the new act should promote public involvement in setting standards...Few things are more suspect to an increasingly wellinformed public than to learn that proposed regulatory initiatives have been shared with industry alone... The BCEPA should entrench the principle of stakeholder consultation in policy development initiatives, through "notice and comment" procedures.14 Political accountability is also at the heart of Ontario's proposed Environmental Bill of Rights:  In examining present environmental laws, it became apparent that the public does not have a consistent, clear right to participate in significant environmental decisions by government. Government makes such decisions in the form of policy, regulations or when it issues licences, permits, approvals or orders controlling activity which may result in environmental harm...A state of law which does not expressly recognize the public as a partner working with government and industry to promote a better environment is not likely to inspire public confidence or participation. The public requires an accessible and transparent decision-making process in order to measure government's accomplishments and to hold government accountable where it fails to meet public expectations respecting environmental protection.15 The foregoing discussion has distilled the major Canadian environmental regulatory "lessons" from the past in three key areas: economic impact, scientific complexities and political dynamics. These lessons highlight the fact that much of the environmental dialogue centres around competing value systems, whether that be in the form of how best to allocate costs, how to deal with risk uncertainty or how to  Ministry of Environment, Land and Parks, New Approaches to Environmental Protection in British Columbia: A Legislation Discussion Paper (April 27, 1992), at pg. 31. 14  15 Ontario Ministry of the Environment, Report of the Task Force on the Ontario Environmental  Bill of Rights (July, 1992), at pg. 9.  8 accomodate the interests of various groups. They pose great challenges for the future of environmental protection on the economic, scientific, political and legal fronts. Several of the underlying themes suggested by these lessons are of central interest to this thesis. These themes are: (1) the importance of democratic legitimacy to the process of environmental decision making; and (2) the inevitability and implications of the involvement of science in these decisions. As already stated, this thesis will argue that the scientific or technical reasoning upon which environmental regulators base their decisions should be made explicit in decision making processes. In legal terms, this means that there should be an express statutory obligation on environmental administrators to disclose and to explain, both before and after a regulatory decision, the scientific analysis used to support the exercise of their discretion. The focus of this argument, as noted below, will be on the standard setting processes of regulation making and issuing permits as the decision points requiring scientific explication. In short, this legal duty will result in several significant improvements to the process of environmental decision making designed to meet the challenges ahead: (1) it will increase the accountability of decision makers by forcing a more thoughtful analysis of the regulatory task and the relevant information; (2) it will help to harness and make accessible a valuable pool of knowledge; and (3) these first two consequences, in turn, will improve the integrity of the process by making analysis more transparent and subject to challenge.  9 B.^Making Decisions under Environmental Laws  Over the last two decades, a great deal of literature has been generated describing and analyzing the plethora of environmental protection laws16 and regulations that have emerged in both Canada and the United States. It is widely recognized by these commentators that one of the hallmarks of environmental legislation is that it confers broad discretionary powers° on delegated administrative authorities: They can be viewed as faceless purveyors of state policy, as benevolent servants of society or as pawns manipulated by the capitalist power elite. No matter how they are perceived, there can be no doubt that government bureaucrats occupy a pivotal role in the pollution control process. Behind every piece of environmental legislation, behind every Minister of Environment lies a bureaucrat, and it is largely in his or her hands that the instruments of government policy are wielded.18 Nonetheless, there is a noticeable dearth, particularly in Canada, of either general or legal literature providing any enlightening examination of the administrative process in the context of environmental regulatory decision making 16 "Environmental protection law" covers broad legal terrain and may include common law doctrines and statutes ranging from local zoning and public health laws to comprehensive schemes for resource management. However, for the purposes of this thesis, the following definition is adopted: Environmental laws means only those measures that are aimed at protecting the natural environment and, more specifically, means those laws or regulations that regulate potentially harmful conduct. These latter laws are pollution licensing statutes and are typically labelled as "command and control" laws. 17 The concept of discretionary powers will be discussed in greater detail in Chapter Two. In short, the notion of discretion as used herein means "an express grant of power conferred on administrative officials where the determination of the standards according to which power is to be exercised is left largely to them". (Denis J. Galligan, Discretionaty Powers: A Legal Study of Official Discretion, (Oxford: Clarendon Press, 1986), at pg. 1.)  Kernaghan Webb, "Between Rocks and Hard Places: Bureaucrats, Law and Pollution Control" in Robert Paehlke and Douglas Torgerson, eds., Managing Leviathan: Environmental Politics and the Administrative State, (Peterborough: Broadview Press, 1990), at pg. 201. 18  10 A current researcher had this to say about Canadian environmental policy: University and other researchers who are now studying the field are struck by the paucity of literature and data. After more than 20 years of attempted action - and some notable progress - on the environment, we still know very little about the policy and implementation record. Data on monitoring, testing, enforcement, costs and benefits, natural stock, and so on, remains thin or, where available, extremely unreliable and user unfriendly. 19 Even within the realm of general administrative law, writers provide little analytical treatment of the inner realities of the administration despite its clear connection with many of the preoccupying concepts of administrative law. 2° For example, one leading scholar in this area observes that administrative law is "concentrated on questions of jurisdiction, errors of law, natural justice, and remedies, each endlessly fascinating in its own way, but at the same time seeming to miss the most important question, namely just how administrative powers do get exercised and what contraints they are subject to". 21 This conception of administrative law reflects the fact that lawyers tend to view the administrative process from the perspective of dysfunction, that of judicial review. It is the premise of this thesis that a more systemic notion of the administrative process must be adopted in order to suggest any improvements. In environmental law, this gap in the understanding of "how" the law "in action" works 19 Tom Conway, "Taking Stock of the Traditional Regulatory Approach" in G. Bruce Doern, Getting It Green: Case Studies in Canadian Environmental Regulation, (Ottawa: Renouf Publishing Company Limited for the C.D. Howe Institute, 1990), at pg. 26. 20  The lack of proper treatment in administrative law with respect to the "realities" of the process are acknowledged by two administrative law scholars in their excellent treatises. See Rene Dussault and Louis Borgeat, Administrative Law: A Treatise, 2nd Edition, (Toronto: Carswell, 1985; Translated by Murray Rankin) and Galligan, supra, fn 17. 21  Galligan, supra, fn 17, at pg. vii. [emphasis added].  11 is especially troublesome given that the vast majority of the content, and thus the real impact, of the law is dictated by environmental regulators. A minor sampling of some of the discretionary powers exercised by environmental regulators in British Columbia, as illustrated by the following example, gives one a sense of the vast scope and diversity of these powers. Consider the situation of an imaginary plastics manufacturing company wishing to establish a plant outside of Vancouver on the Fraser River. Assume that the company's operations will cause some "waste" in the form of toxic air emissions and effluent discharge into the river. First, the company may be required to conduct an environmental impact assessment "if the Minister [of Environment] considers" that the operations would have a "detrimental environmental impact."22 Very little elaboration is given in the Environment Management Act as to the content of the term "detrimental environmental impact". Subsection (1)(2) states that, for the purposes of the Act, a "detrimental environmental impact" occurs "when a change in the quality of air, land or water substantially reduces the usefulness of the environment or its capacity to support life". What does substantial reduction mean? Does it mean a more than 50% risk of harm? If so, consider the regulatory consequences of this in the event that the decision maker has the burden of proving this. In certain cases, it has been observed that such statutory requirements "are seen by agency officials as reasons for caution, and may generally predispose an agency to an approach which errs on the side of industry, at least unless the agency has made a clear commitment 22 Environment Management Act, S.B.C. 1981, c. 14, s. 3. This section is permissive (ie. the Minister may require an environmental impact assessment).  12 to (and has the financial and personnel resources to support) a strategy of clarifying the limits of regulatory authority through test cases. In Canada, such commitments are rare, if they exist at all."23  In addition, assuming that the air and water emissions of the company are "wastes" (as defined in the relevant statute), the company must obtain the necessary waste discharge permits under the B.C. Waste Management Act (the "WMA").24 These permits will be issued by the regional waste manager "subject to requirements for the protection of the environment that he considers advisable"25. The WMA does not set out any criteria that the manager must meet in this regard although subsections 8(1)(a)-(f) do enumerate some items that the manager may include in the permit such as monitoring, the posting of security, conducting studies, etc. Once the company is ready to operate, the WMA requires that spills of any "polluting substances" be reported. A "polluting substance" is defined in the WMA as any substance that could "in the opinion of the minister, substantially impair the usefulness of land, water or air" if it were to escape or be spilled. 26 What set of criteria does the minister use to determine if the environment is "substantially impaired"? Again, does this mean a greater than 50% risk of harm? How is this  23 Ted F. Schrecker, Political Economy of Environmental Hazards. Protection of Life Series, A Study Paper prepared for the Law Reform Commission of Canada, (Ottawa: Minister of Supply and Services, 1984), at pg. 11. [emphasis added]. 24  Waste Management Act, S.B.C. 1981, c. 41, s. 3(3)(a).  25  Ibid, s. 8. [emphasis added]  26  Ibid, ss. 10 (1). [emphasis added]  13 information made available to the company so that it is certain of its spill reporting obligations? Lastly, where a manager is satisfied "on reasonable grounds" that a substance is causing "pollution", meaning a substance "that substantially alters or impairs the usefulness of the environment", a pollution abatement order may be issued.27 What are reasonable grounds in this case? Is scientific data showing cancer in rats from the substance in the effluent a "reasonable" basis to issue an abatement order that may cost millions of dollars to comply with? The above example is not a comprehensive picture of all of the environmental laws that the company would be subject to. It is merely intended to illustrate that each legal requirement is subject to administrative discretion with virtually no statutory criteria as to how this discretion is to be exercised. 28 The implications of this are that companies are very much in the hands of the administrators, and within this system, there is little advance certainty as to the rules of the game. How does a regulator determine whether an activity has a detrimental environmental impact, or whether it substantially impairs the environment? If it does, what measures are necessary or appropriate to protect the environment? Should he quantify the potential risks and weigh these against the costs of the advisable protective measures?  27  Ibid. s. 22.  28 To further exacerbate this, one of the peculiarities of environmental regulation is that it is very difficult to obtain information about how these administrative judgments are made. This may be contrasted with the practise of other regulatory intensive areas of the law, such as tax or securities, where administrative officials frequently publish "Information Circulars" or "Policy Statements" advising the regulated community as to how the administrator will exercise his discretion.  14 The above example is also intended to illustrate the types of environmental decisions that a regulator makes. In general, there are three types of environmental decision making processes to consider: policies, projects and standards. 29 Policies, in the ordinary course of events, are at the top of the hierarchy and are considered the bailiwick of politicians. They are "statements of general intent and direction that set out a government's major purposes in pursuing particular goals and some of the means to be used to achieve them." 3° In the above example, it may be said that it was a policy decision to design the WMA using a regulatory scheme of pollution permits rather than, say, to completely prohibit pollution. Project decisions, usually relating to large facilities such as airports, dams or pipelines, are more specific and are typically governed by impact review assessment legislation. Assessment legislation is outside of the scope of the focus of this thesis. Lastly, the standard setting process refers to the process for developing rules for "specific effluents, pollutants, or production technologies pursuant to parent legislation and/or regulations." 31 In the example above, the requirement to obtain a waste permit under the WMA is an example of the standard setting process. The standard setting process is a fundamental part of environmental protection laws, typically called "command and control" laws, because they focus on what particular polluters must do to meet the law and control this behaviour through penalties. These laws and the process of  29  Doern, supra, fn 13, at pg. 3.  30  31  kid  kid, at pg. 5.  15 setting regulatory standards will be the focus of this thesis for two reasons. First, there is more history and thus more experience and information available about these processes. Secondly, and more significantly, scientific issues are particularly salient in this context. Although, conceptually, there are generally three types of decisions in environmental laws - policies, projects and standards - they are rarely so discrete in practise. It should be clear from the above example that the very nature of the discretionary powers granted under environmental laws require that policy decisions be made in order to implement the laws. Each time that a regulator issues a permit and determines what is "advisable" to protect the environment or issues an order to rectify the "substantial alteration" of the environment, he is making a policy choice about what constitutes an acceptable environmental risk. Furthermore, the substance of these decisions are, by and large, very difficult to challenge under the narrow principles of Canadian administrative law. The fact that there is a limited formal institutional check on this discretion adds to the true power of discretionary decision making under Canadian environmental laws. This is not to suggest that regulators act with unbridled powers. There are, of course, constraining forces such as political pressures, economics, and institutional factors which place very real limits on discretionary powers. One of the key factors influencing environmental decision making is scientific knowledge and the understanding it brings to any analysis of the environmental impacts of commercial/industrial activities. Until recently, the role of science in environmental decision making has received limited attention in the legal  16 literature despite its clear relationship to the substance of administrative policies and regulatory schemes.  C.^The Context of Science  The natural or physical sciences32 are the central fountain of knowledge from which human understanding about environmental impacts and hazards is obtained. Without this understanding, we would have little knowledge of, for example, ozone depletion33 or the potential risks associated with many toxic substances. The type of scientific knowledge available to and used by regulators in environmental problem solving is as vast and varied as the types of environmental issues encountered in today's world. Yet, it is fair to say that one of the fundamental aspects of environmental standard setting decisions is that they are about understanding environmental risks. In the jargon of risk analysts, "risk" in this context means the probability of harm multiplied by the magnitude of an adverse event for humans or the environment34. These risks result from human and commercial/industrial  32 "Natural science" refers to that branch of knowledge that is concerned with the physical world and its phenomena. More generally, "science" means accumulated and accepted knowledge that has been systematized and formulated with respect to the discovery of general truths or the operation of general laws. See the Webster's Third New International Dictionary. 33 See Douglas A. Smith, "The Implementation of Canadian Policies to Protect the Ozone Layer", in Doern, supra, fn 19, at pgs. 111-128, for a discussion of the role that science played in achieving  international agreement on ozone depletion (arguing that there was significant convergence of scientific opinion to provide the basis for useful assessments of costs and benefits). 34 This definition of "risk" is adopted from the general literature on risk assessment. See Anne V. Whyte and Ian Burton, eds., Environmental Risk Assessment, SCOPE 15, (New York: John Wiley & Sons, 1980), at pgs. 2-3; and Donald T. Hornstein, "Reclaiming Environmental Law: A Normative Critique of Comparative Risk Analysis", 92 Columbia Law Review 562 (1992), at pg. 571.  17 activities, particularly those associated with the use or release of synthetic chemicals. Accordingly, the discussion in this thesis will draw upon the science related to toxic chemical risk anaylsis.35 Despite this focus on standard setting and toxic chemicals, this analysis and its conclusions provide general insights into the environmental law/science interface.36 In the context of environmental standard setting, it is generally thought that science brings certainty and rationality to the discretionary framework of environmental decisions. However, as will be demonstrated in Chapter Three of this thesis, scientific understanding of environmental risks is limited. These limitations require scientists and/or regulators to use their discretion to fill in the knowledge gaps. What are the implications of this to the regulator? The regulator may not be equipped to evaluate the scientific information he is using and thus to understand the methodological and knowledge limitations. However, even assuming that the regulator is versed in the scientific uncertainty that he deals with, it is clear that he must exercise discretion to interpret and assess the uncertainties presented by the scientific data. As one begins to explore the nature of discretion in environmental decision making and its intersection with scientific knowledge, it becomes apparent that the  35 A large part of the literature dealing with the debate about science in environmental decision making that I have surveyed examines the use of science in the context of identifying and quantifying the risks of toxic chemicals.  36 This observation has been made by several writers who have studied science and environmental regulation. See, for example, Ted Greenwood, Knowledge and Discretion in Government Regulation, (New York: Praeger Publishers, 1984) and John D. Graham, Harnessing Science for Environmental Regulation, (New York: Praeger Publishers, 1991).  18 core of this intersection revolves around the theme of democratic integrity. In the final analysis, the role of science in the context of environmental decision making must be viewed against the larger backdrop of the role of regulatory power in democratic societies. The scholarly analyses measuring how good a job our environmental regulators do when dealing with science is a debate inspired by the question: "How should we structure our institutions so as to take account of scientific complexity without sacrificing " - and here you can fill in a list of whatever democratic political traditions are considered valuable such as accountability, fairness, or rationality.37 Typically, it is concluded that our regulators could indeed do a better job. Often the suggested solution for reform is to implement better rules better legislative rules and better procedural rules, both of which are oriented towards controlling the exercise of discretion. If we had better rules, so the argument goes, we would have socially acceptable and workable decisions. This is the rationale behind the American rulemaking framework38 imposed on U.S. administrative 37 The observation that this question drives the scholarly discussions about the role of science in environmental decision making is made by Professor Stephen Carter in his astute article, "Separatism and Skepticism", 92 The Yale Law Journal 1334 (1983). Carter states that perhaps this is the wrong question. He posits that the better question might be to ask "Why decisions involving science should be treated any differently from other complex decisions in our society?". Other decisions which we accept often defy rational analysis. Although I think this is a very perceptive remark and undoubtedly of some philosophical significance, the answer is that science is indeed viewed differently. There are probably many explanations for this some of which will be canvassed in this thesis. However, let me say at the outset that I accept that the proper analytical starting point is to consider the question of how we can best use science to make environmental decisions so that they have legitimacy and respect in our democracy. 38 The American system is examined in greater detail in Chapter Four. "Rulemaking" generally refers to the administrative actions which result in the promulgation of "general rules" (as opposed to case-by-case adjudication) which, for all intents and purposes, are equivalent to the making of regulations and issuing permits under Canadian law. Rulemaking is subject to procedural requirements under the U.S. Administrative Procedure Act, as interpreted by the courts: "The courts have said, consider all the "relevant" evidence, respond to all "significant" comments, and weigh all  19 agencies and for this reason, it is often pointed to as an improvement on the "unruly" Canadian approach. But the rulemaking approach also has its critics. Too many rules can result in a defensive preoccupation with information accumulation at the cost of efficiency39, or they can create rigidity which can sacrifice scientific legitimacy. 40 In addition, rather ironically, the result of these rules is that they often create more discretion. One American scholar observes that the uncertainty created by judicial review of the rulemaking process in the United States results in agencies looking for ways to "avoid the rulemaking quagmire...So instead of more rules, we have more discretion."41 So we return to the original question. How can we make the best decisions and accomodate scientific complexities? One author has observed that the challenge in this area is to "design a regulatory process that captures the knowledge of science while safeguarding the proper domain of political choice."42 To meet this challenge we need to better understand the regulatory process and the use of scientific knowledge within it. This thesis seeks to elucidate this understanding and to propose  "reasonable" alternatives" (as noted by R. Shep Melnick, "Administrative Law and Bureaucratic Reality", 44 Administrative Law Review 245 (1992), at pg. 247). 39  ibid.  4° See, for example, Sheila Jasanoffs critique of the U.S. administrative process (in the context of analyzing the Occupational Health and Safety Agency's program for assessing carcinogens), in "Science and the Limits of Administrative Rule-making: Lessons from the OSHA Cancer Policy", 20 Osgoode Hall Law Journal 536 (1982), at pg. 554 ("Values cherished by American administrative agencies - uniformity, continuity, certainty - seem frequently at odds with the demands of science, particularly in areas of rapidly developing knowledge"). 41  Melnick, supra, fn 38, at pg. 247.  42  Graham, supra, fn 36, at pg. 1.  20 a viable legal solution for addressing the concerns raised by the interface between environmental regulation and science.  D.^The Methodology and Scope of this Thesis  This thesis seeks to examine the dynamics of "how environmental regulators regulate" with a particular focus on the interface between scientific information and the administrative process. The purpose of this examination is to gain a better understanding, from a legal perspective, of the role of science in environmental decision making and to provide some insights into the tough job faced by our regulators. Although it is often recognized that scientific uncertainties plague the environmental administrator's task, it is rare to find solutions as to how to address the problems resulting from these uncertainties. The argument in this thesis is simple. Regulators need to explain the assumptions, methodologies and gaps relating to the science behind what they are doing and why. Chapter Two of this thesis examines the notion of environmental regulatory discretion in the context of the modern administrative state. Discretionary powers are a necessary and permanent part of the legal landscape in environmental protection and the exercise of this discretion is influenced by many competing interests, including science. Within this context, in order for regulatory decisions to be legitimate, they must adhere to the underlying principles of Canadian democracy, such as accountability, openness, rational decision making and fair procedures. In general, the regulatory process of decision making under Canadian command and  21 control pollution laws falls short of meeting these objectives. This regulatory process,  as it currently operates, segregates science in a way that allies it with expertise and rationality; it does not subject this information to open scrutiny through participatory procedures. This exploration of the regulatory context of administering Canadian environmental laws in Chapter Two sets the stage for a consideration of the role that science plays, the subject of Chapter Three. In Chapter Three a fuller examination of the nature and use of scientific knowledge in environmental standard setting is considered. This is done through case examples and a more detailed discussion of the risk assessment tools used in standard setting, epidemiological and animal studies. This discussion supports the argument that science is not separate from the economic and political dynamics of regulatory discretion. This argument lays the foundation for the legal argument in this thesis, which is further explored in Chapter Four through an examination of the U.S. experience. In the United States, the basic tenets of administrative law and procedure impose general duties on regulators to explain their regulatory result through various mechanisms, including structured nilemaking and liberal judicial review. With respect  to environmental decision making and standard setting, U.S. regulators are required to disclose and to explain the scientific analysis used to support their regulations. Accordingly, the U.S. experience affords an illustration of how a procedural rule requiring scientific explication functions. In Chapter Four two key standard setting provisions of the U.S. Clean Air Act are discussed. From this discussion, it is clear  22 that a procedural fix is not a panacea; there are both advantages and disadvantages of such a procedural approach. Based on the U.S. experience, one of the benefits of a legal procedure of this kind is that it forces a link between the regulatory decision and the scientific analysis. This can reveal genuine deficiencies in the scientific data, and through public scrutiny, it can also foster a more fullsome participatory debate about environmental risks. On the down side, this link between the regulatory rationale and science may be construed as requiring a reliance on science which can produce unwanted results such as the promotion of delay (through the need for more and better science) or bureaucratic fence-sitting, and the disguising of policymaking behind the veneer of science. Finally, Chapter Five summarizes the substance of the argument in this thesis and examines what a procedural rule dealing with the disclosure and explication of science might look like in Canada and why it would improve environmental decision making. It also presents some concluding reflections on the interface between environmental regulatory demands and scientific complexities. Some preliminary comments about the scope and limitations of this thesis are necessary. Treatises have been written on each of the separate topics of administrative process and scientific knowledge and about their interface. This is a technically complex and analytically challenging area. My goal in this thesis is to synthesize some of this learning so as to give lawyers a critical context in which to examine the role that science plays in the framework of environmental law, its procedures and norms. Lawyers not only need to understand what the law is; it is  23 equally important for lawyers to appreciate how the law is made. In environmental law, the "how" of lawmaking goes beyond understanding the due process aspects of decisions. It is also vital to understand the substance of the scientific content behind the formulation and implementation of environmental laws and policies. The authors of a recent Canadian study dissecting the risk analysis controversy surrounding the cancellation of the herbicide alachlor had the following to say about the valuable contribution of the public interest lawyers in that case: ...they raised many important questions about the uncertainties that plagued the [alachlor] studies presented to the Review Board by others, and identified the assumptions underlying many of the interpretations of the data.43 Given my own tendency to think of science as a "black box", a tendency which I know is shared by many in the legal profession, I think it is valuable for environmental lawyers to confront this intellectual barrier and to assess the implications of the science/law relationship in the sphere of environmental protection. In addition, the scientific component of this discussion, attempts to distill a lot of highly technical and complex information. This distillation is a layman's understanding of this information in order to meet the overall goal herein of highlightling the methodological parameters and limitations of science as it relates to regulatory discretion. Some comments about the nature of the sources relied on are also needed. Although there is some very thoughtful and useful Canadian literature on environmental decision making and science, it is limited. The vast majority of the  43 Conrad Brunk et al., Value Assumptions in Risk Assessment: A Case Study of the Alachlor Controversy, (Waterloo: Wilfrid Laurier University Press, 1991), at pg. 21.  24 relevant literature on, for example, environmental regulatory discretion and risk analysis is American. Accordingly, this thesis has relied heavily on U.S. material. Indeed, many Canadian sources draw on U.S. writings. In addition, much of the science used by American regulatory agencies is relied upon in Canada.' Nothing in the Canadian literature leads me to believe that this reliance on the U.S. literature and experience is unwarranted, outside of some caution that is needed as a result of the obvious political and legal institutional differences between Canada and the U.S. Finally, the scientific component of this thesis discussion dealt with in Chapter Three attempts to distill a lot of highly technical and complex information.  ' Schrecker, supra, fn 23, at pg. 3 (there is "widespread reliance" by Canada on American test data with respect to environmental hazards).  25 CHAPTER TWO THE ADMINISTRATIVE STATE AND ENVIRONMENTAL LAWS Administrative agencies of the regulatory kind are established to cany out the terms of the treaties that the legislators have negotiated and ratified. They are like armies of occupation left in the field to police the rule won by the victorious coalition.'  A.^Introduction One of the discriminating and persistent features of Canadian environmental protection laws is that they are pregnant with discretionary powers. In general, this means that administrative officials are given broad power and "the determination of the standards according to which power is to be exercised is left largely to them".2 Consequently, much of the form, substance and enforcement of environmental laws is determined by the occupying armies, the environmental regulators. As was observed in Chapter One, British Columbia's Waste Management Act (the "WMA") is a classic example of this situation. Section 8 of the WMA provides for the issuance of waste permits by an official, the regional manager, "subject to requirements for the protection of the environment that he considers advisable".3 1 George Hoberg, Pluralism By Design: Environmental Policy and the American Regulatoiy State, (New York: Praeger Publishers, 1992), at pg. 17, fn 2 therein (citing Earl Latham in "The Group Basis of Politics: Notes for a Theory" 46 American Political Science Review 376, at pg. 391). 2 Galligan, supra, Chapter One fn 17, at pg. 1. The concept of discretionary powers will be discussed in greater detail below. 3 Supra, Chapter One fn 25. As previously noted, the WMA does not set out any criteria that must be met by the manager in this regard although subsections 8(1)(a)-(f) do enumerate some items that the manager may include in the permit such as monitoring, the posting of security, conducting studies, etc. This is particularly important to the scheme of pollution control laws in British Columbia since the vast majority of legally binding environmental protection standards are determined through  26 Given the extensive nature of administrative activity and power in environmental law, any consideration of environmental law cannot be divorced from the larger context of administrative law. The term "administrative law" is used here in its broadest sense to mean that area of law that is concerned with the use of power by public authorities.4 Although it is fair to say that environmental law has evolved into its own discrete field, it is helpful for the purposes of the analysis in this thesis  to consider environmental law as a subset of administrative law. This is the task of this chapter -- to consider the phenomenon of discretionary powers in environmental law against the backdrop of administrative law and administrative realities. The purpose of this discussion is to develop an appreciation of the regulatory context in which environmental laws are administered. This analysis will set the stage for a consideration of the role that science plays in this context, the subject of the next chapter. This discussion begins by examining the nature of bureaucracy and discretionary powers in the machinery of modern government, which machinery will be labelled "the administration". The term " the administration" will be used throughout this thesis to refer to the plurality of government entities comprising the permit conditions. The other method of creating legally binding standards is through the use of regulations, a tool used rarely under B.C.'s environmental statutes. 4 See generally Dussault and Borgeat, supra, Chapter One fn 20, at pgs. 12-28 for a discussion of the definition and scope of this area of law. I adopt their view that administrative law extends beyond the notion of judicial review: "...this approach underestimates other parts of administrative law dealing with structures and acts of the Administration as well as the major management systems upon which all administrative decision-making is based In its full dimension, administrative law consists essentially of two major functions: providing rules for the organization and internal operations of the Administration and establishing relationships between the Administration and the public." (Mid, at pg. 28).  27 State's "administrative organs", ranging from Cabinet to governmental departments (or ministries) to consultative or adjudicative independent agencies or tribunals.5 Bearing in mind that Canada is a federal system, this notion of the administration is meant to encompass the structures of both the federal and provincial governments based on the constitutional division of powers. This examination will be followed by an outline of the theoretical political, administrative and legal frameworks within which discretion is situated and consider why it is viewed as anathema to the traditional conception of parliamentary democracy. This theoretical framework is important because it reveals the underlying values that continue to shape our institutions. Lastly, the nature of discretion as it is structured in environmental laws will be examined together with the political dynamics that impact the admininistrative "realities" of environmental laws. This last section will focus on an examination of traditional Canadian pollution control statutes as a model for understanding the implications of discretion.  B. The Age of Bureaucracy and the Nature of Discretion  No matter what disciplinary perspective one brings to the study of governments in North America today, whether it be as a political theorist, sociologist, public administrator or legal scholar, it is uncontroversially accepted that we live in the era of the administrative state. An era in which the organization and operation of government is highly dependent on bureaucratic structures and the exercise of 5^This understanding of "the administration" is developed by Dussault and Borgeat, supra, In 4, at pgs. 43-47.  28 discretionary decision making powers by technocratic expertise. The administration plays a predominant role in modern political, economic and social life; indeed, it has been observed that "it constitutes the main reality of State power for the majority of [Canadian] citizens."6 This reality appears to fulfill the prediction made in 1918 by Max Weber, one of the earliest and most influential students of administrative theory, that "the future belongs to bureaucratization°. Indeed, as early as 1888 academics in England were commenting on the increasing bureaucracy of the time: We are becoming a much governed nation, governed by all manner of councils and boards and officers, central and local, high and low, exercising the powers which have been committed to them by modern statutes.8 Bureaucratization, or the delegation of decision making power to the administration, has become the norm in Canada: The Administration is now situated at the centre of collective life; it translates the choices expressed by those who govern into concrete acts. Through its constant and multifaceted activity, it regulates and transforms the relationships existing between the authority of the State and society, and also amongst citizens themselves. In fact, very few individual or collective projects may now be advanced without the endorsement of one or another of its agencies.9  6  ^  lbid, at pg. 9.  7  Paehlke, supra, Chapter One fn 18, at pg. 24 (citing Max Weber in Economy and Society). For additional reading on Weber's examinations of bureaucratic organizations and a useful summary of his theories and their place in modern ideology, see Galligan, supra, fn 2, at pgs. 117-128. 8 Denys C. Holland and John P. McGowan, Delegated Legislation in Canada, (Toronto: The Carswell Co. Ltd., 1989), at pg. 6 (citing Maitland's text, Constitutional History). 9  ^  Dussault and Borgeat, supra, fn 4, at pg. 3.  29 Despite the many complaints and criticisms against sprawling bureaucracy and overregulation, it is reluctantly acknowledged that this modus operandi of government is here to stay. For example, in its study on the powers of independent administrative agencies, the Law Reform Commission of Canada (the "LRCC") makes the following introductory comment: "And while we do not seek either to legitimize or question any role they presently play, we are confident that they will remain for some time an important aspect of modern government".1° Another commentator observes: We have a modern administrative state that cannot be dismantled or denied. And it is not merely a minimalist state, with proportions irreducible in light of modern life's countless opportunities for savagery. It is, rather, the ambitious state constructed as an instrument for common purposes.11 Why is it that bureaucracy with its accompanying delegation of discretionary powers has come to play such a crucial role in how we govern ourselves? One of the main reasons is the ever increasing complexity of the issues that legislatures must deal with. These complexities have increased with the expanded role of the modern welfare State. Government today has become the fixer of social dysfunction: in this role, the State limits abuses (legislation on labour, environment and consumers), attempts to rectify inequalities among citizens (various kinds of insurance, social assistance and allowances), and proceeds to create physical infrastructures (transportation, energy and communication) or social infrastructures (education, health and  10 Law Reform Commission of Canada, Report on Independent Administrative Agencies: A Framework for Decision Making (Report 26), (Ottawa: Law Reform Commission of Canada, 1985), at pg. 5. ii Christopher F. Edley, Jr., Administrative Law: Rethinking Judicial Control of the Bureaucracy, (New Haven: Yale University Press, 1990), at pg. ix.  30 recreation).12 The complexities that accompany these concerns, the inherent resource limitations of legislatures and the flexibility needed to deal with individual situations and changing circumstances, all encourage delegation to subordinate bodies to allow the development of specialized policies and strategies. These subordinate bodies can be departments within the government or they can take the form of independent administrative agencies, such as labour relation boards or utilities commissions. Independent administrative agencies, in particular, have a special attraction for governments trying to cope with their growing involvement in social and economic regulation. The LRCC attributes the creation of these agencies to: the desire to divert the responsibility for the resolution of politically sensitive issues to discrete, non-partisan governmental bodies; the need for specialization and expertise to manage progressively more complex governmental tasks; the perceived inability of a then partisan, nonprofessional civil service to perform such tasks; and a reluctance to bog down courts in matters that, because of their nature or their volume, were not suited to the judicial process.13 A good example of the use of an independent agency in British Columbia to address the problematics of environmental conflicts can be seen in the government's establishment in 1992 of "CORE", the Commissioner on Resources and the Environment. The legislation creating the office of the Commissioner states that his role is to "advise the Executive Council in an independent manner on land use and Dussault and Borgeat, supra, fn 4, at pg. 7 fn 12 (citing L. Borgeat, R. Dussault and L. Ouellet, l'administration quebecoise: organization et fonctionnement, 1982). 12  13  LRCC, supra, fn 10, at pg. 5.  31 related resource and environmental issues in British Columbia and on the need for legislation, policies and practices respecting these issues."14 Although policy complexities and concerns for allocative efficiencies are real and valid reasons for delegation in modern government (whether it be to governmental departments or to independent agencies), these reasons only explain part of the story. As alluded to by the LRCC in the above quotation, politics also plays a factor. Laws are often the product of political compromise. With respect to environmental laws, it has been said that this results in laws that contain "sweeping, politically attractive but administratively impractical statements about environmental protection, and then leaves the tough decisions to the bureaucrats and courts."15 The point that environmental laws are designed to deflect the "tough decisions" to the bureaucracy is a key one and should be kept in mind throughout this discussion. Whether the delegation to specialized bureaucracies is considered a selfevident need arising from the complexities of modern government or, perhaps, more cynically, is seen as political "buck passing", the granting of discretion to the administration is more often than not viewed with skepticism. It is considered a "necessary evil" in democratic societies. However, prior to discussing the theoretical dilemmas posed by administrative discretion, it is necessary to expand on what is meant by the concept of discretion, with a particular understanding of its parameters in the environmental regulatory context.  14  Commissioner on Resources and Environment Act, S.B.C. 1992, c. 34, s. 3.  15  Webb, supra, Chapter One In 1, at pgs. 49-50.  32 Discretion, as one scholar has aptly explained, "may be best defined as the power to make a decision that cannot be determined to be right or wrong in any objective way."16 More particularly, discretion in the administrative law sense means a delegation of authority from the legislature to the administration in order to make decisions or regulations to achieve general statutory objectives.17 By its nature then, this delegation requires the transfer of powers of a legislative nature to the administration; that is, the authority to make laws. This can be done by granting powers to make regulations pursuant to a statutory enabling clause or by granting powers to adjudicate disputes akin to a court, often referred to as "quasi-judicial" powers. As an aside, it should be noted that this description of legislative, or rulemaking 18 powers is somewhat oversimplified for the purposes of this discussion. ,  In administrative law, certain consequences will arise based on the classification of delegated powers as either "legislative", "administrative" or "judicial". In general, if a power is considered "administrative" in the sense that it "is really nothing more than a legislative acknowledgement of the right to make....managerial rules necessary for  16  J. H. Grey, "Discretion in Administrative Law", 17 Osgoode Hall Law Journal 107 (1979), at  pg. 107. 17  For this reason, as Galligan points out, each grant of discretion must be examined separately in order to ascertain its own peculiar characteristics. Nonetheless, on a more general level, there are recurring features of discretionary authority that can be identified. The intent of the following analysis is simply to highlight these recurring features. A more detailed analysis of discretion in the context of environmental laws follows in Part D of this chapter. 18  In this discussion, the term "rule-making" will be used to denote the power to make law under delegated legislation; this is contrasted with the non-hyphenated version of the term "rulemaking" used elsewhere in this thesis to describe the American administrative process of regulation making.  33 the efficient operation of an organization"19, it will not be legally binding nor subject to judicial review whereas a "legislative" power will be. Similarly, if a power is considered to be 'judicial" versus "legislative", then the common law rules of natural justice, such as the duty to hold a hearing, will apply. These classifications have caused much confusion in the case law and are as malleable as the imagination of a creative lawyer: "the terms are inherently ambiguous, capable of ostensibly describing either the nature of the power, or the manner in which it is to be exercised, or perhaps both."2° It is beyond the scope of this discussion to do more than alert the reader to the fact that these labels import a certain legal significance. For the present purposes, the key concept to the analysis is that delegation turns on the nature of the product authorized by statute: "if the making of rules has been authorized, then the power to make delegated legislation has been conferred".21 Given that the focus of this thesis is on standard setting powers in pollution control laws, this complication is not a major concern. These powers, either through the issuance of permits or regulations, are clearly powers to make binding law through delegation. As noted above, discretionary delegation is a function of the modern welfare State in which concern for the collective good is an explicit theme. This model of  19  Holland and McGowan, supra, In 8, at pg. 108.  20  Ibid, at pg. 102. For an analysis of the confusion in the jurisprudence as a result of the terminology, See Holland and McGowan, supra, fn 8, at pgs. 102-114 and Grey, supra, In 16. 21  16.  lbid, at pg. 114. This functional approach is also recommended by Professor Grey, supra, fn  34 legal authority has been called the "public law model": ...the predominant concern is with the achievement of policy goals, whether of public order or social welfare, and thus private rights and interests are subordinate to notions of public interest. But what constitutes public interest is likely to depend partly on the way political activity at the broadest level is translated into legislative standards, and partly on the more precise working-out of those standards in the course of discretionary assessment by administrative officials.22 These "public interest" activities are broad ranging and varied and most involve some form of policy making.23 Protection of the "public interest" lies at the heart of many environmental laws. One of the most analytically penetrating examinations of administrative discretion from a legal perspective is D. J. Galligan's book, Discretionag Powers.24 At the outset of Galligan's discussion he notes that according to its etymological origins, the idea of discretion is good judgment. This is ironic given the reflexive tendency of the law to treat discretion as imbued with arbitrariness or other not so "good" qualities. In describing the nature of discretion, Galligan points out that it involves two principal variables: (1) the scope for personal assessments in the course of a decision; and (2) the nature of the institutional arrangements or the "internal  Galligan, supra, fn 2, at pg. 88. Galligan contrasts this model with the "private law model" which is "based around a firm separation of powers, both in terms of functions and institutions, with the definition and protection of private rights at the centre of the stage." (lbid, at pgs. 87-88). 22  This is certainly true with respect to the functions of independent administrative agencies. for example, LRCC, supra, at pg. 6. However, the same can be said of governmental departments. Enviromental laws are fraught with policy making at the administrative level. 23  See,  24  Supra, fn 2. The following discussion draws heavily on Galligan's perceptive analysis regarding the nature of administrative discretion, at pgs. 8-14. This is a cursory examination of the ideas that Galligan develops more fully in chapters 3 and 4 of his book.  35 point of view" 25 . Since discretion by definition emanates from statute, the statute is of key importance to the first variable, the scope for personal assessments. An official's personal assessment can be viewed as involving three decision elements: finding facts, settling standards and applying the standards to the facts. There can be personal assessments, or discretion, within each of these elements depending on the statutory mandate; that is, the extent of discretion will depend on the absence or relative absence of binding standards. The extent of discretion, or its functional context, can be characterized along a continuum depending on the relevant enabling statute. At one end of the spectrum, a statutory framework can exist for the decision maker by outlining broad objectives and little guidance as to how to exercise discretion. In such a framework, the administrative task may require the creation of standards where none exist or the interpretation of given standards in order to apply them. 26 Again, section 8 of B.C.'s WMA, mentioned at the outset of this chapter, is a good illustration of this bare-bones statutory approach. Recall that the regional manager's task is to issue waste permits for the protection of the environment "as he considers advisable". 27 How does he do this? Essentially, this section creates personal assessment in the finding of facts, the settling of standards and interpreting them on an individual case by case basis. This type of statutory framework can be 25 Ibid, at pg. 12. This "internal point of view" refers specifically to the attitudes and strategies developed by officials to approach their tasks. 26  Ibid, at pg. 11.  27  Supra, fn 3.  36 contrasted with statutory delegations of "mandatory powers" which are very precise and detailed and leave the decision maker no latitude. 28 One writer observes: If regulations extend only to details of mechanical procedures, no real discretionary powers are delegated. However, where the statutory provisions are only a skeleton and it is left to regulations to say 'what, where, when, why, how and who', then we have created meaningful discretionary powers...29 It is this latter kind of discretion that is most prevalent in Canadian environmental laws. In the 1980's, a Law Reform Commission of Canada study commented as follows on the legislative framework of Canadian environmental laws: "Rarely, if ever, does legislation specify in any detail the criteria which are to govern the development of regulations, nor does it set out timetables for achieving particular sets of objectives."30 Reference has already been made to section 8 of the WMA which adopts this approach. With respect to Quebec's Environment Quality Act, it has been observed that the powers it grants to the administration and the restrictions it imposes upon business are of "a scope without precedent in Quebec."31 Another commentator observes that Alberta's environmental assessment law contains no fewer  28 Dussault and Borgeat, supra, fn 4, at pg. 240. These powers may also be labelled "ministerial powers" in administrative law, but as Professor Grey points out this term is not consistently applied and thus the label "leads to confusion and should be eschewed." (Grey, supra, fn 16, at pg. 113). 29 William T. Stanbury, Business-Government Relations in Canada: Influencing Public Policy, (Scarborough: Nelson Canada, A Division of Thomson Canada Limited, 1993), at pg. 73 (citing Eric Hehner in "Growth of Discretion - Decline in Accountability" in Kenneth Kernaghan, ed., Public Administration in Canada: Selected Readings, 5th edition, at pg. 342). " Schrecker, supra, Chapter One fn 23, at pg. 10. This comment is made with respect to environmental "hazard" laws which generally means pollution control laws, occupational health and safety laws, and hazardous consumer products laws. (lbid, at pg. 5). 31  Dussault and Borgeat, supra, fn 4, at pg. 34.  37 than three layers of discretion and thus as enforceable legislation it is "virtually meaningless".32 Although it is arguable that recent provincial and federal environmental legislative initiatives evidence a greater trend to specifying applicable environmental goals and criteria33, it remains true that the vast majority of Canadian environmental laws are startingly vague as to the substantive criteria against which final administrative decisions can be judged. In general, the Canadian approach sharply contrasts with U.S. environmental laws which impose substantive environmental limits on an agency's decision -- such as enshrining protection of health, protecting species from extinction or ensuring sustained yield of timber.34 Institutional arrangements, and in particular the attitudes that officials have towards the exercise of their power, are also important in shaping an official's sense of discretion. Influences such as individual or institutional attitudes about discretionary powers and the constitutional and political order within which an official operates can create patterns as to how powers are exercised and the justifications for  32 Stewart A. G. Elgie, "Environmental Groups and the Courts: 1970-1992", in G. Thompson, eds., Environmental Law and Business in Canada, (Toronto: Canada Law Book, 1993), at pg. 193. The relevant section of the Alberta Land Surface Conservation and Reclamation Act cited by Elgie reads as follows: "When any person proposes to undertake any operation or activity and, in the opinion of the Minister the operation or activity will result or is likely to result in surface disturbance, the Minister may order the person to prepare and submit to the Minister in the time prescribed in the order, a report containing an assessment of the environmental impact of the proposed operation or activity if the Minister considers it in the public interest to do so". (mid). ' See, for example, the Canadian Environmental Assessment Act, S.C. 1992, c. 37; the Commissioner on Resources and Environment Act, S.B.C. 1992, c. 34; and Alberta's Environmental Protection and Enhancement Act, S.A. 23, 1992, c. E-13.3. 34 Elgie, supra, fn 32, at pg. 193. The more detailed expression of environmental objectives, and the implications thereof, will be evident in the discussion of the U.S. Clean Air Act in Chapter Four.  38 decisions. The political dynamics in the environmental area are particularly important and will be considered in greater detail below. Institutional arrangements may also dictate the administrative practise of a government body. For example, to administer their statutory mandate, a department may decide to formulate comprehensive standards for the guidance of administrators and officials may well regard these standards as binding. This can occur despite a highly discretionary grant of power in the enabling legislation. As Galligan notes: "What may be discretionary from an external, legal point of view, may be anything but discretionary from the internal point of view of officials within the system."35 This comment is particularly relevant in the Canadian context because so many administrative actions by environmental regulators flow from internally generated, yet inaccessible, policies. Although these policies are generally not legally enforceable, they do have very real implications for those who must deal with an administrator. Galligan concludes: "It is only when there is a significant freedom of choice in the officials, and when that choice is recognized and respected by the courts or other authoritative body, that we may talk of a reasonably discrete notion of discretionary power." 36 It is this two-pronged sense of discretion that is truly reflective of the nature of discretion granted to Canadian environmental regulators. In summary, this discussion has noted the following recurring features of discretionary powers as they pertain to environmental laws:  35  Galligan, supra, fn 2, at pg. 13.  36  Ibid, at pg. 14.  39 • •  •  discretion exists when there is a delegation of rulemaking authority from the legislature to the administration; the legal foundation for this authority is based in statute and thus the statute determines the functional context of discretion - in environmental legislation, the typical pattern is one of very limited statutory guidance as to how to achieve the legislated mandate; and discretion often applies to the pursuit of that nebulous concept, the "public interest", and as a result it is inherently subjective in the sense that there is no "right answer";^institutional arrangements and political dynamics become particularly important to shaping discretion in this milieu.  These features pose some recurring problems for the operation of modern democracies. Why? To understand this, the theoretical framework of our political, administrative and legal structures must be considered. This theoretical framework reveals the underlying values that shape our political culture. These values in turn help to assess the role of the bureaucracy and the role of the law in modern governance.  C. The Theoretical Framework -- The Political, Administrative and Legal Contexts Although discretionary powers are accepted as a functional necessity in the modern state, they are, as noted above, typically characterized as a "necessary evil"37: "necessary" in order to accomodate legislative inabilities to deal with the 37 For example, Galligan makes the following observation about discretionary powers: "....the attitude so often is encountered that, while its presence may be inevitable, it is at the same time slightly deviant." (Ibid, at pg. 1) [emphasis added]. Other writers also discuss the "dilemma" that discretion poses for constitutional or parliamentary democracies. See Paehlke, supra, fn 7; James R. Bowers, Regulating the Regulators: An Introduction to the Legislative Oversight of Administrative  40 complexities and variabilities inherent in modern societies and "evil" because of the threat posed to democratic ideals. This characterization of discretion is part of the gospel in administrative law. In order to understand why this is so, the foundations and evolution of our political, administrative and legal structures must be explored. However, it should be noted that there is an increasing trend in both public administration and law to accept discretion as a good thing and to reject the view that it is as an inherently bad aspect of an otherwise sound system. 38 I am persuaded by this more open understanding of discretion and view the law as needing to adapt to this reality. The broad delegation of discretionary powers often associated with the public law model is perceived to conflict with two fundamental theoretical principles of parliamentary democracy: (1) accountability via representative government; and (2) legality or the "rule of law". Canadian parliamentary democracy is a system of responsible government where executive functions are performed by ministers who must answer to Parliament (or the provincial legislature, as the case may be) for their actions. Accountability is the essence of this model of responsible or representative government -- that those  Rulemaking, (New York: Praeger Publishers, 1990); and Dussault and Borgeat, supra, fn 4. 38 For example, Galligan does not see discretionary powers as an undesirable deviation from an ideal of government through rules: "It will be argued that discretionary powers are important in any system of authority, that there are good reasons for having discretion, and that discretionary powers are neither necessarily nor typically in some way arbitrary and beyond the law far from legal rules and standards being its antithesis, they are always present in constituting, defining and constraining discretion." (Tbid, at pgs. 2-3). See also, Edley, supra, fn 11, preface to text.  41 who are to be governed choose those who will do the actual governing.39 This accountability gives the government legitimacy. If there is dissatisfaction with what the government is doing then, so the argument goes, the answer is at the polls. However, the delegation of broad discretionary authority to unelected administrative officials or agencies violates this accountability link between the public and the politicians. For this reason, particularly with respect to the delegation of powers to independent agencies, "some have perceived it as the last blow to parliamentary democracy and as foreshadowing Government by decree".40 On a theoretical level, this concern emanates from the classical notion of the separation of powers between the legislative, executive and judicial branches. Unlike the American constitutional system, this tripartite division of powers was not strictly adhered to in the establishment of the Canadian constitition. Our constitution does not exclusively allocate the three functions of government to three distinct organs: "There is no general 'separation of powers' in the Constitution Act, 1867. The Act does not separate the legislative, executive and judicial functions and insist that each branch of government exercise only "its own" function."41 In particular, our parliamentary structure permits an overlap of authority between the executive and the legislature. Nonetheless, there is a strong notional separation of powers with  39  Bowers, supra, fn 37, at pg. 10.  LRCC, supra, fn 10, at pg. 17. This is, as the LRCC admits, a rather antiquated view of the implications of delegation given the complexities of modern government. 40  Peter W. Hogg, Constitutional Law of Canada, Second Edition, (Toronto: The Carswell Company Limited, 1985), at pg. 150. 41  42 respect to the functions of the executive branch and the administration. Theoretically, it is the domain of the executive to have "political" rule-making power and it is the responsibility of the administration to apply and enforce these rules, under the watchful eye of the judiciary. However, in practise, it has been observed that this theoretical dichotomy is unrealistic since the activities of political/legislative rule-making and administrative rule-application "overlap to such a degree that the distinction between the two becomes purely formal, from the point of view of administrative law."42 To further understand how accountability has been attenuated to this degree in Canada today, it is necessary to consider the basic organization and dynamics of regulatory bureaucracies. The next section of this discussion will focus on some of the fundamentals of the organization of governmental departments and the complexities of policy formulation in order to elaborate on the general "realities" of administrative decision making. The focus is on governmental departments because the vast majority of Canadian environmental laws are administered by specialized departments or ministries rather than independent agencies. In the overall structure of government, departments or ministries43 play a prominent role. "They constitute an essential source of technical competence in the modern administration" and are the "principal delivery arm of government". 44 The  42  Dussault and Borgeat, supra, fn 4, at pg. 52.  43  The terms "department" and "ministry" are used interchangeably in this discussion.  44  Dussault and Borgeat, supra, fn 4, at pg. 83.  43 legislative delegation of complex schemes for administering policy is typically done using the vehicle of specialty departments. For example, some obvious departments in the environmental area are Environment Canada at the federal level and the B.C. Ministry of Environmnent, Land and Parks (the "BC MOE") at the provincial level. Each of these governmental institutions, along with other specialized and complementary departments or ministries (such as the Department of Fisheries and Oceans, federally, or the Ministry of Forests, provincially) play a crucial role in developing and implementing environmental laws and policies in Canada. Under our system of responsible government, ministries are subject to the direction and management of the Minister. The Minister is given his power pursuant to enabling legislation.45 Accordingly, the first place to look in order to determine the degree of autonomy of any given Ministry and the controls it is subject to, is the relevant legislation. For example, section 2 of B.C.'s Environment Management Act states that the "duties, powers and functions" of the provincial Minister of Environment "extend to all matters relating to the management, protection and enhancement of the environment" and then includes a specific list of the types of matters this encompasses; for example, the "preparation and publication of policies, strategies, objectives and standards for the protection and management of the environment".46 In theory, at least, the departmental structure is consistent with the democratic interest of accountability because the Minister is answerable to both ' The enabling legislation always confers powers on the Minister. See Dussault and Borgeat,  supra, fn 4, at pg. 85. 46  Environment Managemnent Act, supra, Chapter One fn 22.  44 Parliament and public opinion for the management of his Ministry. Although power is vested directly in the Minister, in practise, he relies heavily on his staff to address the formulation, drafting and implementation of Ministry policy. This should not be surprising given the nature of delegated tasks and the limited experience of Ministers in their portfolios. Accordingly, although the legislative framework outlines the powers of a given Minister and his ministry, the administrative organization and hierarchical structure of the ministry must also be considered to understand the operational complexities and division of labour. Before examining the hierarchy, one key point must be made. The Minister is a member of Cabinet and Cabinet is, of course, one of the defining features of the Westminster model of parliamentary government. In this system, there are few bodies within the executive (or the legislature) able to act independently of the Cabinet: "Thus there is no body able either to constrain the behaviour of Cabinet, nor to provide information to the public about either the quality of Cabinet decision making or the performance of its policies."47 Usually, a government ministry or department has three levels.48 The first level has already been mentioned, it is the political one occupied by the Minister. By its nature this position is subject to the winds of political change. Secondly, there is the administrative head, the Deputy Minister (the "DM"). This is typically a permanent position and his role is to provide a link between political change and 47  Stanbury, supra, Chapter One fn 6, at pg. 92.  48 The following discussion with respect to the three levels of governmental departments is adopted from Dussault and Borgeat, supra, fn 4, at pg. 89.  45 public service continuity. The description of the DM as an "administrative head" is not meant to underestimate the power and influence of this position. The DM is the principal policy advisor to the Minister and thus must be carefully attuned to changing political considerations: "the role of Deputy Minister demands awareness of both public acceptability in the general sense and consideration of the philosophic and political direction of the government he serves."49 Lastly, there is the administrative layer subdivided into various divisions and services to reflect the duties undertaken within the administrative apparatus. For example, simply glancing at the listings in the B.C. provincial phonebook gives one a sense of the wide division of tasks that exist at the B.C. MOE. It is this dispersion of responsibility that raises the concern for a loss of accountability in the system and this concern is one of the main forces stimulating political and legal reform in modern bureaucracies. For example, many of the reform efforts over the last two decades to improve regulation making in Canada at the federal level have been driven by the concern to tighten executive accountability through improved processes. These are processes that focus on making the analytical basis of decisions more transparent and on centralizing control and information. Professor Stanbury, who has completed a comprehensive examination and evaluation of the federal government's regulatory reform efforts, notes: "Most of the process reforms in Canada have consisted of efforts to improve the quality of new regulations, largely by means of requiring more/better analysis to assist decision makers in  " Ibid, at pg. 90 (quoting a comment made by A. Blakeney, former Premier of Saskatchewan).  46 departments and the Cabinet."50 For example, these efforts have included more centralized oversight of the regulation-making process aimed at improving effectiveness through the use of Regulatory Impact Analysis Statements, adherence to the policy document entitled "The Citizen's Code of Regulatory Fairness", and the re-evaluation of regulatory programs every seven years. These efforts have also included increased public participation through modified notice and comment procedures (eg. the advance publication of proposed regulations in the Canada Gazette, the use of a regulatory calendar or agenda and more stakeholder  consultations).51 Reform efforts at the provincial level have not been as comprehensive. For example, in British Columbia, provision for public notice and consultation is at the discretion of the originating minister.52 Discretion not only challenges notions of political accountability, it is also contradictory to principles of legality or the "rule of law". The rule of law is a concept which has special importance in the western ideology of law. 53 It is the "mortar that holds the edifice of the State together".54 The rule of law is a rather elusive concept. The Supreme Court of Canada has expressed the rule of law as follows: "The 'rule of law' is a highly textured expression...conveying, for example, 50  Stanbury, supra, fn 47, at pg. 14.  51  See Stanbury, supra, fn 47, for a more comprehensive history and discussion of these matters.  52  Treasury Board Secretariat, Administrative Policy Branch, Regulatory Affairs, "How Regulators  Regulate: A Guide to Regulatory Processes in Canada", at pg. 7. 53  Galligan, supra, fn 2, at pg. 61.  54 Gregory Tardi, The Legal Framework of Government: A Canadian Guide, (Aurora: Canada Law Book Inc., 1992), at pg. 19.  47 a sense of orderliness, of subjection to known legal rules and of executive accountability to legal authority".55 It essentially means that State power is to be exercised "according to binding general rules made and known in advance, and of sufficient specificity to allow individuals to know with tolerable certainty their rights, obligations, and liabilities".56 With respect to the administration, this means that the administration is subject to both the legislative acts of parliament and to the courts which are responsible for enforcing and interpreting the law. Again, the theory of the separation of powers is a central tenet of this principle: "The powers of officials are to be closely circumscribed by rules, which are to be adjudicated by courts, which in turn are themselves in the main separate from and independent of the legislative and executive branches."57 Implicit in this concept is the idea of certainty as stated in legal rules which establish the authority of both parliamentarians and public administrators. The hybrid nature of discretion, involving both legislative policymaking functions and adjudication, threatens this model of functional separation and rule certainty upon which the rule of law is based. Again, the underlying notions are of key importance. In order to maintain political legitimacy, those who exercise power must do so within the constraints of pre-determined fixed rules, the law. This theoretical discussion has highlighted two themes of importance in 55 Dussault and Borgeat, supra, fn 4, at pg. 231 (citing the Court's Resolution to Amend the Constitution, [1981] 1 S.C.R. 753 at 805-806). 56  Galligan, supra, fn 2, at pg. 61.  57  Mid, at  pg. 6.  decision in Reference re  48 Canadian democracy: (1) political accountability; and (2) legality/certainty through the legal prescription of authority. A consideration of the underlying political values represented by these themes is central to a legal perspective and understanding of discretionary powers. Our legal principles and institutions are part of the political and social composition of society and are instrumental to upholding these underlying values. The themes of accountability and legality can be further divided into additional concepts that reflect the values inherent in Canadian society: stability in legal relations, rational decision making, fair procedures and political morality.58 In other words, a decision should be arrived at for known and good reasons, it should meet the underlying policy objectives of the governing statute and it should do all of this in a fair manner that opens it to external public scrutiny. Ultimately, these ideas prevent decision making by "whim, caprice, chance, or ritual."59 The catalogue of values identified by the LRCC in its study on Independent Administrative Agencies is consistent with these notions. The LRCC's list is reproduced in full in Exhibit 1 below.  58 These four categories are developed by Galligan, 'bid, at pg. 90 wherein he states that a "reflection on the political theories that underlie modern, democratic, liberal societies would tend to suggest" that these principles "have a position of importance". 59  Ibid, at pg. 6.  49  EXHIBIT I -- LIST OF CANADIAN DEMOCRATIC VALUES°  Accountability: Having to answer for the exercise of what  is essentially governmental authority to affect public and private interests. The rule of law demands that governmental authority not be exercised arbitrarily, and that agencies account, sometimes within a political framework, sometimes within a legal one, and often within both, for the decisions they make and for the policies they pursue as decision makers.  Authoritativeness: Making decisions that are accorded full  recognition. Where authority is ostensibly given to an agency to decide a matter, those who deal with the agency are entitled to have the decision made by the agency, not by a politician, judge or other decision maker. Comprehensibility: Making the administrative process as  understandable as possible to those whom it affects. Interested persons must know whom to address, about what matters, and how to address them. Effectiveness, Economy and Efficiency: "Getting the job done"  without wasting human and material resources. Administration can be a drain on both public and private resources. It demands constant attentiveness to new and better ways of doing things.  Fairness: According appropriate recognition to the interests that  may be affected by agency decisions. Without fairness there will be neither the trust and credibility that lend integrity to a process, nor the co-operation that is essential if it is to be efficient.  Integrity: Operating in a manner that is true to the objectives laid  down for the agency; having a full commitment to its purposes. An agency must be sufficiently free of background pressures to project competence and confidence in carrying out its duties. The administrative process needs both the self-respect of the agencies and the respect of those with whom they deal.  60 These values are taken from the LRCC, supra, In 10, at pgs. 8 9. -  50  Openness: Making the administrative process accessible to those it  affects, and providing a window through which it can be seen. There must be openness if other values are to be adequately realized. For instance, those who are affected by decisions look to openness as a further guarantee of fairness and accountability. Principled Decision Making: Rationally correlating the information the  agency has, the interests of which it is aware, and the objectives, policies and criteria that are to guide its decisions.  Although the foregoing list was developed to provide a framework for appropriate procedures to guide independent agencies, the values identified can be taken to represent Canadian democratic ideals. The LRCC developed the list "based on a mixture of socio-legal values drawn from Canadian political traditions and aspirations, contemporary approaches to public administration, existing doctrines of administrative law, and constitutional tenets." 61 These values, at least in part, animate the structures and operation of government and the role that law plays in the exercise of administrative discretion. The political and legal frameworks of the Canadian and American democratic systems attempt to create structures and processes that give these ideals some substance while attempting to balance the demands for certainty with the needs for regulatory flexibility. From an administrative law perspective, one of the common legal strategies that is advanced in order to keep discretion to a minimum is to  61  ^Ibid, at pg. 8.  51 institute pre-determined fixed and certain rules.62 These rules may be in the form of clearer guidance from the legislature. In addition, procedural rules (such as the Canadian principles of natural justice or procedural fairness or, in America, noticeand-comment rulemaking) have developed as a way of checking the exercise of discretion. Despite these efforts, criticisms prevail that these rules miss the mark of obtaining rationality, purposiveness or morality. Common criticisms directed against environmental regulators from all perspectives include accusations about the arbitrariness of decisions (for example, that they lack foundation in science); the politicization and thus unfairness of the process; and a lack of concern for economic or environmental consequences, depending on one's perspective.  D. Discretion and Environmental Laws  The purpose of this section is to provide a more thorough understanding of discretionary powers as they exist in environmental laws. As previously noted, the label "environmental law" is a broad descriptive umbrella that covers a wide variety of common law doctrines and statutes which impact on the conservation and protection of the natural environment. For the purposes of this analysis, the focus will be on traditional "command and control" legislation; that is, those statutes that establish fairly comprehensive pollution control regimes and which will undoubtedly remain an important aspect of the Canadian regulatory approach for some time. In 1988, the LRCC conducted an examination of the command and control regulatory  ' Galligan, supra, fn 2, at pg. 1.  52 framework and concluded: "significant components of the legal framework for effective pollution control are now in place."63 In British Columbia, the Waste Management Act is the centrepiece of regulatory activity and it appears that its  control approach will continue to be the foundation for environmental law in B.C. The proposed British Columbia Environmental Protection Act, a more comprehensive statutory model, is being designed around the WMA64. In addition, in order to understand the nature of discretion in these regulatory schemes one must consider the variety of forces, including important extra-legal forces, that dictate the behaviour of government departments. In this discussion, these laws will be discussed from the following interrelated thematic perspectives: the legislative mandate and the political dynamics. By the mid-seventies, virtually all federal and provincial jurisdictions had passed command and control regulatory legislation dealing with air, water and land pollution. These environmental statutes represent the first true efforts at comprehensive environmental control in Canada.65 In general, the design of these laws was to "command" those subject to the 63  Webb, supra, In 15, at pg. 8.  64  See the B.C. MOE discussion paper, supra, Chapter One fn 14.  Professor Alastair R. Lucas describes these laws as the "first generation" of environmental statutes and notes that they arose from a recognition by government that common law doctrines for the resolution of private disputes and the existing public health laws and miscellaneous provisions in natural resource development statutes "were not equal to the task of comprehensive environmental control". See A. Lucas, "The New Environmental Law" in E. Hughes, A. Lucas and W. Tilleman, eds., Environmental Law and Policy, Volume I, Preliminary Edition, (Toronto: Emond Montgomery Publications Limited, 1992), at pg. 4:1. For a more thorough discussion of the historical roots of pollution control laws in Canada, see Webb, supra, In 15, at pgs. 11-15 and MacDonald, supra, Chapter One In 8, at pgs. 135-142. 65  53 law, by threat of penalty, to "control" the amount of pollution escaping into the environment as dictated by the statute. Although there is no uniform legislative model adopted by all Canadian jurisdictions, there are representative characteristics found in most of these laws. The essential means of control is through administrative tools, such as: (a) (b) (c) (d)  standards which define how much pollution is permitted; administrative systems of permits or licences intended to control pollution before it occurs; administrative powers to require clean-up after pollution has occurred and/or clean-up obligations; and requirements to provide information, eg. through supply monitoring data or by submitting to inspections and spill reporting.66  This overall statutory framework, with its heavy reliance on administrative tools, has meant that these schemes are dependent on the bureaucracy for implementation. For the most part, this has meant administration by specialized environment departments, that are largely technical agencies, "staffed by the scientific and engineering experts necessary to implement the permit schemes and develop "safe" standards for waste discharge."67 In effect, these laws require the "routine use of administrative judgment in determining the significance of pollution impacts."68 Consequently, this reliance on the administration has driven decision-making  66 Dianne Saxe, Environmental Offences: Corporate Responsibility and Executive Liability, (Aurora: Canada Law Book, 1990), at pg. 55. 67  Lucas, supra, fn 65, at pg. 4:2.  68  Schrecker, supra, Chapter One fn 23, at pg. 11.  54 processes  underground to the less visible world of regulation and license-  negotiating.69 As previously noted, this state of affairs is no longer considered tolerable and certain changes have been made, and continue to be proposed, to address the lack of participation and integrity that this underground administration creates. Another defining aspect of the discretionary scope of Canadian environmental laws is the limited role that the courts have played in overseeing the exercise of administrative discretion. Although this reality may be attributed to many factors79, it seems clear that one of the basic reasons is the nature of the legislation itself: Administrative suits, by definition, challenge government authorities who fail to follow their statutory duties. It is axiomatic to add that if statutes impose very few requirements on administrative decisionmakers, there will be few opportunities for administrative actions. Such has been the case in Canada.71 This is not to say that administrative law principles exert no influence on the exercise of discretion by environmental regulators. The guiding tenets of Canadian administrative law are applicable to administrative decisions under environmental  69  Webb, supra, fn 15, at pg. 15.  " One obvious factor might be the prohibitive costs of litigation especially in light of judicial deference to administrative decision-makers with expertise. A comprehensive canvass of judicial review of environmental administrative actions, from a public interest advocacy perspective, can be found in Elgie, supra, fn 32. Elgie attributes the limited role of the Canadian courts (at least prior to 1987) to: the standing problem; the lack of judicial sympathy to environmental issues; unfamiliarity with the public interest component of environmental suits in the context of administrative law which traditionally concerns itself with protecting the individual from the overeaching arm of government; and the limiting nature of the statutory language (as noted in the text above). After 1987, Elgie notes an improvement in most areas with the exception of the substance of the statutory powers. (mid, pg. 205).  71^Elgie, supra, fn 32, at pg. 11.  55 laws. Accordingly, judicial oversight of decisions may be available on the basis of the traditional concepts of: (1) procedural fairness; and (2) substantive challenges concerning the scope and content of decisions or the jurisdiction of environmental decision makers. Nonetheless, based on the vagueness and non-mandatory nature of command and control laws, these traditional avenues for judicial review pose significant hurdles.72 Given the lack of case law in this area, it is difficult to gage exactly what the judicial influence is. Arguably, it is minimal. It is fair to conclude that in Canada the courts provide a limited institutional check on the exercise of environmental administrative discretion and that this lack of judicial presence clearly enhances the power of the regulators. Despite the promise of these environmental laws, in practise, their implementation has proven difficult. What accounts for this "implementation gap"?73 Again, the legislative mandate is partly to blame. The sweeping and vague pronouncements in environmental laws give little credence to the importance of operational realities: Pollution control legislation is typically drafted in language which suggests that implementation is a straightforward, almost mechanical process, when in fact government officials are attempting to cope with unstated unresolved scientific, political, technical and economic fa ct ors.74 72 For a good example of the difficulties of showing that a Minister's discretion is not "reasonable", see Re Reese et. al. v. The Queen in right of Alberta (1992), 87 D.L.R. (4th) 1, a case seeking judicial review of a forest management agreement between the Alberta Minister of Forestry, Lands and Wildlife, and Daishowa Canada Co. Ltd. 73 This is a term coined by Webb, supra, fn 15, at pg. 7, to describe the disparity between the promise of statutory language and the bureaucratic experience with implementation. 74  Ibid, at pg. 24.  56 Canada's experience with regulating the environmental emissions from pulp and paper mills attests to these complexities. Writing in 1991, one observer comments that: "After twenty years of provincial and federal regulation and after receiving literally billions of public money in grants and tax exemptions intended to assist with pollution control, the majority of mills still do not meet federal toxicity standards and the industry remains one of Canada's largest polluters."75 This historical failure is attributed to political, economic and technical issues involved in mill regulation.76 Another example of the influence that these operational realities can have on the shaping of regulations is manifested in the case of regulating PCBs. In the view of some commentators the public fear of "micro-chemo phobia"77 created political pressure resulting in regulatory action that was unwarranted given the scientific evidence. As was observed by Justice Fraser of the Ontario Court of Justice (General Division), the political imperative may lock the law into an untenable scientific position:  ' MacDonald, supra, fn 65, at pg. 227. Note that this comment was made prior to the enactment of the three new regulations passed by the federal government in 1992, supra, Chapter One fn 5. 76 See, for example, the federal government's own assessment of this situation in which it openly acknowledges these complexities, Environment Canada, Conservation and Protection Branch, William F. Sinclair, "Controlling Pollution from Canadian Pulp and Paper Manufacturers: A Federal Perspective", (Ottawa: Minister of Supply and Services, 1990). See also MacDonald, supra, fn 65, at pgs. 225-240 (noting that Canada's attempts to regulate have been impeded by "major political factors").  77 R. v. Consolidated Maybum Mines Limited, supra, Chapter One fn 2, at pg 276. This term was used by one of the expert witnessess in this case to refer to the "public anxiety about minuscule concentrations of chemicals now detectable in the environment thanks to technology particularly when the chemical has a 'nasty two or three syllable name to it'". (Ibid).  57 I find as a fact based on the evidence called that there is a broad scientific consensus that PCBs are not direct carcinogens nor are they carcinogen promoters in humans. There is no evidence presently available that would justify a finding that PCBs have an adverse effect on humans based on the realistic exposure levels that could be obtained from environmental exposure... The Crown submits that the Court should not be drawn into pronouncing on the wisdom of regulating PCBs at a level of 50 ppm...Considering the clear evidence that the level of public concern about PCBs is not in step with scientific consensus, this proposition is undoubtedly correct. Such regulatory decisions reflect a complex weighing of choices in terms of economic activity, costs, technological options and social goals. Mr. Wong testified that the Ministry's goal is the complete elimination of toxic materials from the environment. This probably has no scientific foundation in terms of necessity but may well be an accurate statement of the public expectation. 78 It is appreciated that the two foregoing examples of regulatory situations are complicated and cannot be analyzed superficially. The only point in referring to them is to illustrate that the design of environmental laws necessarily sets up the resolution of difficult economic, political and social issues at the regulatory level. Consequently,  the influence of "extra-legal" factors becomes crucially important to understanding the landscape of environmental laws and their importance cannot be denied.  E. Conclusions  Broad discretionary powers are a necessary and permanent part of present day environmental laws, which powers are also part of the larger phenomenon of the increased role of bureaucracies in modern government. Within the context of environmental decisions, this discretion requires the resolution of important policy  78  R. v. Consolidated Mayburn Mines Limited, supra, fn 77, at pgs. 275-77.  58 choices amidst a panoply of competing forces. Discretion is acceptable within the framework of the Canadian political-legal culture so long as it meets underlying democratic values, such as accountability, openness, rational decision making and fair procedures. Currently, the regulatory process of decision making under command and control pollution laws falls short of meeting the values cherished in the theoretical framework. One of these values is the notion that regulatory decisions must be rational, a concept that is particularly interesting to this discussion. In the context of environmental decisions, rationality often translates into reliance upon scientific evidence and expertise. However, the characterization of science as rational and thus as a rule-certain constraint on discretion portrays an overly simplistic picture of the interface between scientific information and regulatory discretion. The focus of the next chapter is to elucidate the nature of the type of scientific information used by regulators and to examine the implications of this.  59 CHAPTER THREE THE ROLE OF SCIENCE IN ENVIRONMENTAL REGULATION  Scientific information has been the single most important force in shaping the environmental agenda.1  A.^Introduction - The Need for Science There is little doubt that science plays an integral part in environmental decision making processes and, particularly, in the setting of environmental standards. This chapter will consider the nature and use of scientific information by regulators to set environmental standards. This discussion requires a caveat. It is difficult to make generalizations about the role of science in environmental decision making without engaging in a case by case study. Nonetheless, some general observations can be drawn from the case study literature and this is the approach taken here. These observations also highlight the parameters of the philosophical debate in this area which is considered in Part C below. Science is crucial to environmental standard setting due to a combination of factors including the knowledge requirements dictated by statutory mandates and the need to found administrative decisions on a rational basis. Recall the broad statutory language of B.C.'s key environmental statutes that were discussed in Chapter One. In order to meet such general statutory mandates as, for example, preventing  of  Government of Canada, Canada's Green Plan for a Healthy Environment, (Ottawa: Minister Supply and Services, 1990), at pg. 147.  60 "detrimental environmental impacts" or "protecting the environment as is considered advisable"2, it is necessary to have knowledge about the nature of polluting activities or substances and their effects on human health and/or the natural environment. The knowledge that helps regulators to understand what may harm the environment or what is toxic is the domain of the natural sciences; disciplines such as biology, toxicology and soil chemistry. For example, our understanding of the impacts of CFCs on atmospheric ozone depletion, of lead pollution from auto emissions or of the harm caused by the pesticide DDT have all been informed by scientific inquiry and analysis. Scientific information is essential to the development of appropriate regulatory responses under environmental laws. Scientific information about environmental problems helps regulators to prioritize harmful pollution activities and to devise and defend protective control measures relating to such activities. For example, the Annual Report 1991-1992 to Parliament produced under the Canadian Environmental Protection Act states that regulations under the Act "typically begin  with a [scientific] assessment report, establishing a scientific basis for control."3 Furthermore, legislatures often confirm the need to obtain and rely upon scientific information in environmental decision making by providing express statutory authority  2^These two references are to sections of certain B.C. environmental statutes that were considered in Chapter One, supra, fns 22 and 25.  3 Canadian Environmental Protection Act, Report for the Period April 1991 to March 1992,  (Ottawa: Minister of Supply and Services Canada, 1992), at pg. 23.  61 to conduct research or to form technical advisory committees.4 In addition to the statutory context, administrative law principles require that regulatory decisions be "reasonable". Science, which is synonymous with rational thinking in our culture5, provides a sound basis for the rational exercise of administrative discretion. Given this, science does more than simply inform discretion. In an effort to grapple in a reasonable way with the complexities and uncertainties surrounding environmental issues, regulators often turn to science for answers. In this way, science satisfies the law's need for certainty and acts as a powerful legitimizing force.6 In short, it can be readily accepted that science is needed to inform regulatory decisions about environmental protection. Furthermore, using science, with its perceived objective facts and norms, to substantiate these decisions conforms to the prerequisite that these decisions be rational. Accordingly, it seems clear that a regulatory decision made without any consideration whatsoever of available and relevant scientific evidence would be considered patently unreasonable. However, the complaint that no science was used by an environmental regulator is seldom made. More often than not, the complaint is one of scientific incompetence or, perhaps more accurately, a disagreement with the way discretion was exercised in 4 See, for example, the B.C. Environment Management Act, supra, Chapter One fn 22, s. 2 which elaborates on the duty of the Minister to protect the environment by giving him the power to do "planning, research and investigation with respect to the environment". 5^Science connotes the opposite of ignorance or misunderstanding. See the Webster's Dictionary. " Mary L. Lyndon, "Risk Assessment, Risk Communication and Legitimacy: An Introduction to the Symposium", 14 Columbia Journal of Environmental Law 289 (1989), at pg. 296.  62 assessing and interpreting scientific evidence.7 The Canadian case involving the cancellation of the herbicide alachlor exemplifies how scientific information must be assessed by regulators in order to make a regulatory decision.8 In February 1985, the Minister of Agriculture cancelled the registration for the herbicide alachlor in Canada. Monsanto Canada, one of the main manufacturers of alachlor, appealed the Minister's cancellation to a review board. Alachlor had been used for almost 16 years by Canadian corn and soybean farmers. The cancellation action was taken as a result of new scientific studies indicating that alachlor induced cancerous tumours in lab rats and mice during longterm feeding trials. Health Protection Branch ("HPB") officials at Health and Welfare Canada had thus concluded that: "Alachlor is one of the most potent carcinogenic pesticides presently in use and should be removed from the market as soon as possible."9 Under the provisions of the Canadian Pest Control Products Act of 1969, the Minister of Agriculture can cancel the registration of a product when, "based on current information available to him, the safety of the control product or its merit or value for its intended purposes is no longer acceptable to him".1° Neither the Act nor the regulations give any guidance to the Minister on the meaning of "acceptable"  Greenwood, supra, Chapter One fn 36, at pg. 2.  7  ^  This following discussion of the alachlor case is based on the analysis of this case set out in the excellent book by Brunk et. al., supra, Chapter One fn 43. 8  9^Ibid, 10  Ibid.  at pg. 10.  63 nor how the notions of safety, merit and value interact. This provision was at the centre of the controversy over the cancellation of Monsanto Canada Inc.'s registration of alachlor. The legal dispute in this case revolved around the government's determination of acceptable risk. The government interpreted this provision as an absolute safety standard whereas Monsanto argued that as a matter of regulatory fairness, a formal risk-benefit analysis was required. How did the science influence these determinations? The details of the scientific data need not be repeated here in detail other than to note that the lab data and other studies revealed a variety of different cancer responses in rats and mice depending on the amount of alachlor fed to the animals. Health and Welfare Canada interpreted the data using the lowest dose level information, which was not statistically significant, but which it viewed as biologically significant because of the rare type of tumours induced: The appearance of tumours at the low dose rate was especially significant to HPB officials because this rate was within the range of expected dose that HPB had calculated for the applicators of alachlor. Thus, from their point of view, there was no "margin of safety" at all between potential dose rates experienced by human users of the chemical and those dosages that produced tumours in rats.11 Alternatively, Monsanto argued that the rat studies did not provide any basis for the HPB's conclusions that alachlor posed a risk of cancer to humans. Monsanto agreed that at high dose levels, alachlor posed a risk of cancer to humans, but held: (a) that there was no reason to extrapolate from the carcinogenic response in the rats to the assumption of  11^  Ibid, at pg. 12.  64 such a response in humans, and (b) that even on the assumption that such extrapolations were reasonable, the rat studies suggest that alachlor is not carcinogenic at the levels of exposure to be expected among human applicators and others. The company believed that the fact that there were no statistically significant occurrences of cancers in rats at the low dose rates indicated a "threshold" level below which alachlor could be assumed to be noncarcinogenic.12 On appeal, the review board found that it was not appropriate for the government to consider safety alone: The Board admitted that the risk-benefit approach was not required by the [Act] but, nevertheless, it was the approach the Board thought most reasonable for regulatory bodies to follow. All other "absolute standards" it considered to be arbitrary and unclear in their application.13 The board's interpretation is illustrative of how risk analysis intrudes on the regulatory process despite, arguably, a legislative intent to adopt a safety only approach. Although it is indisputable that scientific information is essential to environmental regulatory decision making, it is far from clear how much and what kind of scientific analysis is enough or appropriate to warrant regulatory action. Is it  appropriate to rely on animal studies as evidence of carcinogenic risks to humans especially if they contradict negative human studies? Or, vice-versa, what are the implications of this to the human data? If animal studies are going to be used, what  12 13  Ibid, at pg. 14. Ibid, at pg. 20. [emphasis added].  65 sort of statistical model should be used to determine the "right" dose level applicable to human exposure if a very high dose, as is common, was used in the animal experiment? Is effect A in fact caused by variables x, y and z? These decisions intermingle scientific asssumptions with administrative discretion. The result of this state of affairs has been noted by one of the leading U.S. writers on environmental law: A recurring issue in environmental law is scientific and technological uncertainty...Lawyers like to think that scientists have clear and definitive answers to certain factual questions - is this level of pollution harmful, and precisely what damage will it produce? The answer is likely to reflect a judgment, rather than a statement of fact...The resolution of such mixed questions - questions of fact which carry the law with them - is one of the earmarks of environmental law. It is also one of the reasons for the problems of government agencies that seem to lag in the promulgation of standards. The impatience with agencies may be justified - but the delays in regulation and in adjudication are understandable because the field is very complex...There do not seem to be any answers as yet to the many good questions lawyers and administrators ask the scientists.14 The exercise of discretion to interpret and apply science lies at the heart of the controversy surrounding the use of science in environmental regulation. This matter has fuelled some intensive litigation battles in the U.S., and to a lesser degree, in Canada. In order to understand the nature of the debate as to the role of science, it is necessary to further consider the nature of scientific knowledge and how it is used by environmental regulators.  14 Frank P. Grad, Treatise on Environmental Law, Volume I, (New York: Matthew Bender & Co., Inc., 1973; 1992 Supplement), at pgs. 1-25 - 1-26.  66  B. The Nature and Use of Scientific Knowledge  As was observed in Chapter One, the type of scientific knowledge available to and used by regulators in environmental problem solving is as vast and varied as the types of environmental issues encountered in today's world. However, it is fair to say that one common approach that is typically used in standard setting is risk analysis. Risk analysis is an analytical methodology used to identify, quantify and manage the environmental risks associated with toxic chemicals. In its classical expression, the first two activities of identifying and quantifying risk, the risk assessment stages, are said to be purely scientific factual inquiries relating to the  characteristics of risk, in as quantitative a way as possible.15 The final result of a risk assessment is generally expressed in a numerical estimate of the maximum individual risk or as the number of cancers expected to result from the substance at issue. Cancer is used as an indicator for other health effects because there is more available data on it. In addition, regulators have reasoned that controlling for the cancer effects of a pollutant establishes a level of exposure that would be likely to control for other health effects.16 Thus, a commonly accepted calculation of cancer risk is "one in one million". This expresses the acceptable de minimus risk standard  15 See Howard Latin, "Good Science, Bad Regulation, and Toxic Risk Assessment", 5 Yale Journal on Regulation 89 (1988), at pg. 89; Whyte and Burton, supra, Chapter One fn 34, at pg. 11; and Brunk, supra, Chapter One fn 43, at pg. 4.  Lyndon, supra, fn 6, at pg. 290, fn 4 therein. See also Richard Wilson and E.A.C. Crouch, "Risk Assessment and Comparisons: An Introduction", 236 Science 267 (1987), at pg. 269 (cancer risk 16  assessments act as surrogate for other chronic risks such as heart problems).  67 based on an exposure level calculated to reflect the probability that one person in one million will die of cancer. For example, in the B.C. Ministry of Environment's document on the applicable criteria for the risk management of contaminated soil, it is stated, without further elaboration, that the one in one million risk criterion is commonly regarded as a 'de minimus' risk level, below which agencies do not take regulatory action to control risks." 17 Environmental regulators use risk assessment calculations to help them to: (1) set priorities; (2) adjust regulations to the degree and distribution of the risk to be controlled; and (3) to make site specific decisions by considering the nature of the pollutant, the sensitivity of the environmental setting and the availability of the control techniques. 18 The supposedly objective assessment carried out by scientists is used to evaluate risk so that it can be managed. Risk management involves judgments about the significance and acceptability of risk and is openly value based. It is accepted that environmental regulators act as risk managers by using the scientific expertise provided by risk assessments. 19 There is some dispute as to whether risk assessment is a science or an art. Although often labelled as a science, it appears that, at best, it can be described as  17  British Columbia Ministry of Environment, Land and Parks, Waste Management Branch,  "Criteria for Managing Contaminated Sites in B.C.", Draft #6 dated November 21, 1989, at pg. 4.  is Milton Russell and Michael Gruber, "Risk Assessment in Environmental Policy-Making", 236 Science 286 (1987), at pg. 286. 19 For example, the U.S. EPA's guidelines for estimating carcinogenic hazards, provide that risk assessments should be carried out "independently from considerations of the consequences of regulatory action", as cited by Latin, supra, fn 15, at pg. 89.  68 a "developing science".20 In any event, risk assessment does endeavour to provide a systematic approach for solving complex problems relating to environmental hazards. Risk assessment and risk analysis arose in the late 1960s and early 1970s as a direct consequence of the first wave of green consciousness and environmental legislation marking that time. In the United States, the passage of the federal environmental impact assessment statute, the U.S. National Environmental Policy Act, and the mandates of several federal agencies, such as OSHA and EPA, to protect the public from environmental risks were catalysts in the development of the formal practise of risk analysis.21 In Canada, similar federal and provincial policies, laws and agencies emerged during the same period. In response to criticisms about the inconsistencies in government policies, the U.S. government has made several concerted attempts over the last decade to coordinate and standardize risk assessments both within the federal bureaucracy and, particularly, within EPA. For example, in 1982 the U.S. Congress passed the Risk Analysis and Research Demonstration Act to "help perfect the use of hazard  assessment by federal agencies concerned with regulatory decisions related to the  2" K. S. Shrader-Frechette, Risk Analysis and Scientific Method: Methodological and Ethical Problems with Evaluating Societal Hazards, (Dordrecht: D. Reidel Publishing Company, 1985), at pg. 5. However, note Lydon's comment that "in a real sense, risk assessment is an art." (Lyndon, supra, fn 6, at pg. 291. 21 K S. Shrader-Frechette, Risk and Rationality: Philosophical Foundations for Populist Reforms, (Berkeley: University of California Press, 1991), at pg. 6.  69 protection of human life, health and the environment."22 Another example of the government's coordinating efforts involved the publication of the "principles for cancer risk assessment" by the White House Office of Science and Technology Policy (1985).23 In 1986, EPA issued its own guidelines for carcinogen risk assessment. These guidelines were followed by a widely referenced policy report in 1990 produced by the EPA's Scientific Advisory Board. This report, Reducing Risk: Setting Priorities and Strategies for Environmental Protection, is aimed at shifting national environmental  policy to a comparative risk approach. This is considered to be a more rational strategy based on the principle of achieving the greatest risk reduction for the least money and for anchoring EPA decisions more to the "scientific understanding of risk" than to the public's risk "perceptions".24 Reducing Risk provides an excellent example of the institutional entrenchment of the notion of rationality made synonymous with scientific risk assessment. Senate Committee hearings were held to discuss the Reducing Risk Report. One writer has observed that these hearings elicited fundamental disagreements between governmental, scientific and citizens groups about the scientific issues in evaluating risk and the underlying questions of human values.25  22  Ibid.  23  Graham, supra, Chapter One, fn 36, at pg. 6.  ' These comments were made by EPA Administrator William Reilly at the announcement of the Reducing Risk Report, as cited by Hornstein, supra, Chapter One fn 34, at pg. 563. 25 David Clarke, "Looking at Risk", 8 The Environmental Forum 12 (March/April 1991), at pg. 16.  70 The notion that risk assessment is a purely factual and objective scientific exercise reflects our cultural world view that science is a fact finder/truth seeker.26 It is said that science stands beyond politics.27 Science is merely a way of gaining knowledge through inquiry and experimentation. In particular, science "seeks to find the best among available answers to questions, to demonstrate that one hypothesis is so consistent with observation and experiment that we are justified in calling it "the truth"."28 This observation raises two important points about the nature of scientific methodology and knowledge. First, science is an experimental endeavour focussed on finding the best possible answer and by its very nature it inherently encourages doubt. Uncertainty is part of the investigative process and thus scientific knowledge can legitimately be subject to honest differences of opinion between equally credible experts. Secondly, before scientific consensus or acceptability emerges, a very high standard of proof is needed, usually 95%.29 These scientific norms can hinder the regulatory process. As many commentators have observed, the implication of this to the administrator who bears the burden of proving that his decision meets the relevant regulatory mandate is that any imperfections in scientific knowledge about human risk can A great deal has been written about the emergence of scientific thought and its impact on the North American world view and our cultural myths. For a provocative look at how this scientific world view is integral to modernity and is, arguably, connected to present day socio-economic ills, see Morris Berman, The Reenchantment of the World, (Ithaca: Cornell University Press, Bantam Edition, 1984; original publication, 1981). 26  27  Carter, supra, Chapter One fn 37, at pg. 1337.  28  Ibid.  29 Schrecker, supra, Chapter One fn 23, at pg. 26.  71  operate to permit continued exposures to toxic chemicals: It is now well recognized that the inherent conservatism of science, that a good scientist does not allow commitment to a hypothesis until the case is proven, can work against the prudence embedded in most laws and regulations drafted to protect public health.3° The observation that the disciplinary norms of science can hamper the regulator's mission was made by a former administrator of the pesticide division of the U.S. Department of Agriculture (the "USDA"), Ned Bayley. During his tenure  as administrator during the 1960's, Bayley noted that the scientific viewpoint dominated the pesticide division. This caused reluctance to respond to the types of issues raised by writer Rachel Carson in her seminal book, Silent Spring. This USDA orientation led: bureaucrats to emphasize the weakness of some of Carson's evidence, to dismiss harmful effects not proven with reasonable scientific certainty, to request more funding for long-term pesticide research than regulation, to adopt the extended timeframe typical of scientific research and to underestimate the political clout of the nascent environmental movement.31 This led Bayley to conclude: "Although scientific data are essential to decision making on pesticide regulations, I believe (as a scientist who has had to modify his thinking drastically as an administrator) that having a regulatory function operate under a science-dominated administration is a mistake."32  30  Graham, supra, fn 23, at pg. 1.  Howard Latin, "Regulatory Failure, Administrative Incentives, and the New Clean Air Act", 21 Environmental Law 1647 (1991), at pg. 1671 (quoting Bayley, "Memoirs of a Fox", 2 Environmental Affairs 332 (1972-73), at pg. 335). 31  32^  Ibid.  72 Implicit in these observations is the recognition that "uncertainty" is part of the state of affairs for the scientist. One writer who has comprehensively studied the interaction between science and the exercise of policy discretion by environmental regulators in the United States has characterized the types of scientific knowledge used by regulators into four conceptually useful categories: • • • •  Information may be available, but its interpretation is uncertain or in dispute; Information may exist or could be obtained with limited research, testing, measurement, or analysis, but is not available to agencies when needed; Information may not exist and could be obtained only with substantial effort; and Information may be unobtainable in principle.33  Lastly, what is the source of scientific information? The scientific community is unlike some of the other players in the environmental regulatory arena in that its institutionalized expression is less precisely identifiable.34 It is a diverse collection of knowledge whose members can be found in many institutions such as universities, governments and private organizations. For example, one of the premier science organizations in Canada is the National Research Council of Canada, a government body. It is in "charge of such matters affecting science and industrial research in Canada as may be assigned by the Governor in Council"35 but it does not appear  33 Greenwood, supra, fn 7, at pg. 73. Greenwood's book analyzes OSHA's health-based standards program and the EPA's hazardous air pollutants program for the years 1970-1981.  34 See G.B. Doern, The Peripheral Nature of Scientific Controversy in Federal Policy Formulation, (Ottawa: Science Council of Canada, 1981). National Research Council Act, R.S.C. 1985, s. 4. In 1992-93, the NRC had an annual budget of about $429 million and employed about 3,000 people (as cited in Canadian Government Programs and Services, (Toronto: CCH Canadian Limited, 1993), at pg. 5030.) 35  73 to have a specific responsibility for environmental risk assessment. In Canada, there is no single scientific lobby or body known for its environmental research and pressure: "Instead, key Canadian scientific concerns have been advanced by scientific 'notables' or through the ongoing work and activity of scientists in government, business, ENG0s, and the universities."36 What are the implications of all of the above observations to the task of the environmental regulator? These implications can best be understood through a case illustration. A fascinating case dealing with the regulator's task and how courts respond to scientific uncertainty is the decision of the United States Supreme Court in Industrial Union Department, AFL-CIO v. American Petroleum Institute et. al.37 . This litigation represents one of the first attacks on a regulator's use of science and is the only U.S. Supreme Court decision on the question of how much scientific analysis is necessary to support a regulation in the U.S.38 It is also indicative of the problems that repeatedly plague the interface between science and law in the area of environmental regulation. As a result of new scientific evidence linking the substance benzene to leukemia, the U.S. Occupational Safety and Health Administration ("OSHA") issued regulations requiring employers to reduce benzene exposure levels in the workplace  ' G. Bruce Doern, Green Diplomacy: How Environmental Policy Decisions Are Made, (Ottawa: Renouf Publishing Company Limited for the C.D. Howe Institute, 1993), at pg. 19. 37  448 US 607 1980 (USSC), 65 L. Ed. 2d 1010. This decision is hereinafter referred to as the  Benzene case and all subsequent citations are to the L.Ed. report of the case. 38  Graham, supra, fn 23, at pg. 4.  74 from the pre-existing standard of 10 parts per million ("ppm") to 1 ppm. Although OSHA expressly recognized that it could not quantitatively measure the difference in risk between the 10 ppm standard and the 1 ppm standard, it reasoned that the reduced standard would qualitatively reduce the risk to workers and thus was in keeping with their statutory mandate. OSHA's regulation was made the subject of pre-enforcement judicial review pursuant to the provisions of the Occupational Safety and Health Act (the "OSH Act"). In a majority decision39, the Supreme Court invalidated OSHA's new standard of 1 ppm. Under the OSH Act the Secretary of Labor has broad authority to promulgate occupational safety and health standards. In the benzene case, two key sections were at issue. First, the general standard setting authority in section 3(8) providing that an "occupational safety and health standard" means a standard "which requires conditions.. .reasonably necessary or appropriate to provide safe or healthful conditions of employment."40 In addition, with respect to toxic materials, section 6(b)(5) required, in relevant part, that the Secretary set a standard "which most adequately assures, to the extent feasible, on the basis of the best available evidence, that no employee will suffer material impairment of health..."41 The majority of the Supreme Court invalidated the OSHA regulation primarily on the basis that OSHA exceeded its statutory authority. The court found that the 39 The decision in this case was split 5 to 4, with a very strong dissent being written by Justice Marshall. 40  Benzene case, at pg. 1017.  41^Ibid.  75 administrative record did not support the view that 1 ppm was "reasonably necessary or appropriate to provide safe and healthful employment" as required by section 3(8) (which as a matter of statutory interpretation was read to limit the toxics standard section). Furthermore, and this is the crucial point, the majority determined that section 3(8) confined the Secretary's authority to the regulation of "significant risks".42 The dissenting judges expressed the view that the OSHA standard was fully in accord with their statutory mandate with respect to toxic substances and could be  supported by "substantial evidence in the record as a whole" as required by the relevant judicial review section of the OSH Act. One scholar has argued that the critical flaw made by the majority in the benzene case was that it equated scientific uncertainty with minimal risk; in other words, the court interpreted the notion of "significant risk" to mean a risk that could be demonstrated with scientific certainty.43 This observation was also made in Justice Marshall's strongly worded dissent in which he notes that in their legal quest for certainty, the court subverted the legislative preference to act preventatively when dealing with toxics in the workplace." The influence that science played in the interpretations given to these statutory provisions by each of OSHA and the Court is one of the most noteworthy aspects of the Benzene decision. What scientific information was before OSHA and how was Howard Latin, "The Significance of Toxic Health Risks: An Essay on Legal Decisionmaking under Uncertainty", 10 Ecology Law Quarterly 339 (1982), at pg. 343. 43  Ibid.  76 it dealt with? There were primarily two types of scientific evidence linking benzene and cancer or other health effects: epidemiological45 studies and animal studies, both of which are common risk assessment tools. Based on this evidence, OSHA and industry experts agreed on two points: (1) that there was valid and acceptable scientific evidence to show that at high exposure levels, benzene would cause nonmalignant blood diseases, leukemia and chromosal damage; and (2) that there was very little direct evidence of harm at the 10 ppm leve1.46 As a backdrop to this evidence, it should be noted that the vast majority of carcinogens are what are called "non-threshold" pollutants which means that scientists cannot determine a dose level below which no adverse health effects can be observed. This is further compounded by the fact that the causal mechanisms of cancer are not understood by scientists. What does this information mean to a regulator responsible for determining a "safe" exposure level for humans? It was OSHA's view that this lack of scientific understanding meant that no "safe" exposure level could be determined and that section 6(b)(5) of the OSH Act therefore required that an exposure limit be set at "the lowest technologically feasible level that will not impair the viability of the industries regulated."47 This was an explicitly risk-conservative policy decision made by OSHA which attempted to err on the side of the worker and shift the implications of scientific uncertainty to industry. The Benzene litigation arose ' Epidemiology is "a branch of medical science that deals with the incidence, distribution, and control of disease in a population." See Webster's Dictionary. 46  Latin, supra, fn 42, at pg. 360.  47  Benzene case, at pg. 1017.  77 because industry was strongly opposed to the costs that it would bear for a minor improvement in risk reduction.48 Thus the familiar battle lines are drawn. The centre of the dispute in the Benzene case was the issue of the distributive costs and benefits in connection with unquantifiable environmental risks. This discussion now turns to the matter of the scientific information in the case, which is typical of the type of information generated by risk assessments, epidemiological and animal studies. Epidemiological Studies. Epidemiology uses direct observation of human  disease and employs biostatistical techniques to control for confounding factors.49 The aim of these techniques is to show a positive association between an agent and a disease. For example, dietary factors such as meat consumption and coffee intake have been linked to various cancers based on epidemiology. This association is generally accepted as the most convincing evidence on human risk.50 Despite this, epidemiological studies pose significant evidentiary problems. For example, it is difficult to obtain accurate case histories and to distinguish the effect of mixed and multiple exposures in a population. This can be especially troublesome given the long latency periods of many cancers, often 40 years or more. It was precisely these kinds of concerns that were mentioned by OSHA in the Benzene case as deficiencies  Ibid, at pg. 1027. The court noted that OSHA estimated that the new standard would require approximately $266 million in capital investments for engineering controls, first year operating costs of $187-205 million and annual costs thereafter of about $34 million in order to benefit about 35,000 workers. (kid). 49  Shrader-Frechette, supra, fn 20, at pg. 20. Ibid, at pg. 23.  78 that made the data unreliable.51 One writer has stated that these limitations mean that epidemiology is at best a "crude science" which is "often useless".52 Given these drawbacks and the need to rely on human exposures, animal and other tests are an essential part of risk identification methodology. Toxicological Experiments and the Dose-Response Relationship. Animal  experiments, typically using rats and mice, are concerned with the effects of different exposure levels of a suspect substance. Their primary purpose is to test hypotheses about carcinogenicity. These experiments permit greater precision in the control of exposure levels and other factors that do not compromise the results. However, they are expensive and time consuming.53 Several hundred chemicals have been shown to be carcinogenic based on animal studies whereas epidemiological investigations have been able to classify only 36 agents as strongly suspected human carcinogens.54 Some scientists have interpreted such statistics as indicative of the inability of animal tests to predict the potency of substances in humans. The argument being that protective biological mechanisms in humans will prevent reaction to such low doses. There are a couple of major methodological problems with animal tests.  51  Latin, supra, fn 42, at pg. 361 (Latin recites in detail 9 deficiencies identified by OSHA).  James P. Leape, "Quantitative Risk Assessment in Regulation of Environmental Carcinogens", 4 Harvard Environmental Law Review 87 (1980), at pgs. 92-93. 52  53 These factors necessarily limit the size of the animal populations that can be tested. Leape writing in 1980 stated that an experiment involving only 500 animals costs over $250,000. Yet in order to detect a risk level of even one in 100 persons would require over 10,000 animals. (kid, at pg. 94, fn 51 therein). Ian Burton et. al., eds., Living with Risk: Environmental Risk Management in Canada, (Toronto: The Institute for Environmental Studies, 1982), Chapter 2, at pg. 14.  79 First, it is somewhat speculative to base human risks on results from rodent experiments. This is particularly true when, as noted above, science just does not understand what causes cancer. Secondly, very high dose rates must be used in order to induce cancer in a limited number of animals. This then requires extrapolation to human dose levels. This quantification aspect of animal testing is one of the most problematic. By way of dose-response mathematical modelling the human response to contaminants is predicted based on the animal tests. These calculations build in a safety factor of one-hundred to one-thousand-fold below the lowest animal dose. It is beyond the scope of this paper to give a careful treatment to many of the difficulties associated with the dose-response analysis but several points are worth noting.55 There are at least six different modelling approaches available and nearly all can yield plausible results. The results of these various models can vary greatly, often by a factor of 100,000. For example, in the risk assessments done for saccharin, three different models predicted risks ranging from 5 to 1,200 cancers per one million people exposed.56 A fourth model, offered by an industry group, predicted one death per billion people exposed.57 In another case relating to TCE, a drinking water contaminant, it was observed that the "estimates provide a range of uncertainty  55 The most thorough and understandable presentation on this topic that I have found is by Dennis J. Paustenbach in his article, "Health Risk Assessments: Opportunities and Pitfalls", 14 Columbia Journal of Environmental Law 379 (1989). 56  57  ^  Leape, supra, fn 52, at pg. 103. Ibid.  80 equivalent to not knowing whether one has enough money to buy a cup of coffee or pay off the national debt." 58 A colourful explanation of the range involved in these modelling approaches. After a dose-response relationship is developed, the populations at risk and the dose they are likely to receive from a given agent are considered. There are a wide variety of factors that scientists must consider in estimating the concentration of chemicals in the environment and what pathways will cause resulting ingestion by exposed humans. Conservative assumptions relating to the "maximum exposed individual" rather than the typical person are used at this stage. This approach of using the maximum exposed individual answers the regulator's need to achieve a high degree of protection. One writer is sympathetic to this concern but demonstrates how extreme it can be in one analysis relied upon by the U.S. Environmental Protection Agency to determine the dioxin hazard posed by municipal waste incinerators: At first review, the analysis seemed reasonable until one noted that the child ate about two teaspoons of dirt each day, that his house was down-wind of the stack, that he ate fish from a pond near the incinerator, his fish consumption was at the ninety-fifth percentile level, he drank contaminated water from the pond, he ate food grown primarily from the family garden, and he drank milk from a cow which grazed on forage at the farm. This is not quite the description of a typical person living near a municipal incinerator. 59 In the Benzene case, in light of science's lack of comprehensive theoretical  58  Latin, supra, fn 15, at pg. 92.  59  Paustenbach, supra, fn 55, at pgs. 403-404.  81 understanding of cancer causation, which essentially provides an inadequate base for which to achieve scientific consensus on the proper method of extrapolation, OSHA ackowledged these difficulties and adopted a qualitative approach to the risk." The purpose of reviewing the science behind the Benzene case was not to be persuasive one way or the other as to whether the best scientific/regulatory approach was used by OSHA.61 The point of this discussion has been to illustrate the impact  that scientific uncertainty has on environmental standard setting. Clearly, environmental regulators must exercise discretion to interpret and assess the uncertainties presented by scientific data. This is the fundamental point - the way regulators deal with science is an exercise in discretion. Yet the debate is not framed this way. The debate is framed in terms of "good science" versus "bad science", in terms of scientific incompetency or as contests between "objective" scientific opinion versus "subjective" scientifically uninformed, but well-intentioned, public opinion.62 The important value-driven policy choices in this area are masked by these dichotomies of "actual" against "perceived" risks and by aligning the "actual" risks with science and claims of political and moral neutrality.63 This amounts to playing the politics-of-science game: "That is, the need for 'more research' is advocated and, as  60  Latin, supra, fn 42, at pgs. 364-371.  In addition, it is worth noting that there have been alterations in the regulatory posture of OSHA and EPA as a result of the Benzene case and of additional regulatory experience gained over the last decade. For example, the prevalent risk theory at EPA these days is a "comparative" risk idea as explained above in Part B. 61  62  63  ^  Brunk, supra, fn 8, at pgs. 1-2. Ibid, at pg. 7.  82 a consequence, substantive action is continuously postponed until greater certainty of knowledge is acquired. In other cases, the tactical politics of science is a none-toothinly-veiled shield for self-interest."64 The essence of these criticisms is not about science, it is about the exercise of discretion. As one articulate writer observes: Knowledge and discretion are closely intermingled in a decision-making environment; the boundary between them is rarely precisely definable. As a result, debates ostensibly about knowledge and how to improve the knowledge base of decision-making are often actually about the exercise of discretion and its implications for the distribution of costs and benefits in society. Decisions based largely or even totally on discretion are easily mistaken for - and are frequently claimed to be decisions based on knowledge. In public policy debates, the confounding of knowledge and discretion provides many opportunities for both government officials and nongovernmental advocates to obscure arguments actually based on ideology or interest behind the more respectable veil of science.65 The next section of this chapter will examine the parameters of the philosophical debate regarding the role of science in environmental law and its influence on administrative discretion.  64 Doern, supra, Chapter One fn 13, at pg. 8. 65  Greenwood, supra, fn 7, at pg. 2. That there are implicit value assumptions built into certain scientific methods and also into their interpretations is not a unique observation. For a fascinating example of a Canadian case (dealing with the cancellation of alachlor and briefly described above), see Brunk, supra, fn 8, at pgs. 1-7 (issues surrounding environmental risks are cast as battles about "objective" scientific opinion versus "subjective" scientifically uninformed, but well intentioned, public opinion).  83  C. The Debate: The Separateness of Science  The discussion in Chapter Two of this thesis highlighted the fact that the raison d'etre of the administrative state is the need to create and rely upon expertise. In theory, the delegation of discretion to administrators relies on the separateness of expertise from the legal, political and ethical considerations traditionally reserved for the legislative/executive policymakers. The ideology of science as a separate fact finding discipline in search of the "truth" about the physical world fits particularly well into this conceptualization. However, as was also noted in Chapter Two, the separation of powers is an illusion in today's world. Environmental regulators routinely perform policymaking activities balancing many considerations. Nonetheless, the idea that science is separate from policy and that it should be the exclusive domain of scientists continues to prevail as the dominant paradigm in the 66 • context of environmental standard settmg. Evidence of this can be found in the  segregation of risk assessment and risk management, previously noted. The school of thought which espouses the view that science is "separate" from politics, law and ethics has been labelled, not surprisingly, as the "separatist view".67 Professor Yellin in an enlightening article on the separatist/non-separatist debate, sets  66 For a thought provoking examination of the paradigm of "scientific" expertise and its separateness from the paradigms of "politics" and "fairness" in the context of U.S. judicial review doctrines, see Edley, supra, Chapter Two fn 11. 67 Joel Yellin, "Science, Technology and Administrative Government: Institutional Designs for Environmental Decisionmaking", 92 The Yale Law Journal 1300 (1983), at pgs. 1305-1316.  84 out three main components of the separatist view.68 First, separatists assert that it is possible to distill "scientific" and "technical" issues of fact from other matters in environmental controversies and to assign them to appropriate non-political institutions. In this way, environmental decisions could be made in a simple two-step process: "Scientists, engineers or statisticians would first gather, compile, and analyze data, and legislators or administrative policymakers would then consider the firststage results and come to ultimate findings."69 Second, separatists argue that this approach is consistent with democratic principles since it reserves the policy questions for those who are accountable. Third, separatists claim that adversarial processes such as administrative hearings and judicial review are the best method for dealing with the open uncertainty of many technological issues. Yellin observes that these arguments have "undeniable force" and are powerful justifications for maintaining the existing institutional balance. However, as Yellin and other proponents of the non-separatist view explain, if the separatist view was true, the scientific and other dimensions of environmental issues could be readily distinguished and these distinctions should be manifest in the actual performance of our institutions. Yet there is ample analytical and empirical data to support the argument that this is not the case. Recall the Benzene case discussed above! Non-separatists argue that it is not possible to segregate the scientific and political components of environmental decisions. These decisions by their very nature  68  The following discussion is based on Professor Yellin's analysis at pgs. 1305-1309.  69  Yellin, supra, fn 67, at pg. 1306.  85 are complex policy decisions. They argue that the value aspects of a decision inevitably impact the scientific understanding of an issue and that they are so intertwined any effort to separate them is misguided. Accordingly, non-separatists claim that the best approach to improving how the administrator can best deal with science is to improve the process of how scientific evidence is used so as to make the value laden issues and uncertainties more explicit. It is precisely because the most difficult environmental issues that confront regulators turn on uncertainty that these controversies cannot be split into scientific and other components. This acknowledgement inevitably leads the non-separtists to conclude that in environmental decision making, it is not the substance of scientific questions that is important but rather the procedures by which the significance of scientific evidence is evaluated.70  D. Conclusions  The modest aim of this discussion has been to provide a better understanding of how scientific knowledge is used by environmental regulators in determining environmental standards. When environmental law meets science, the regulator is faced with a perversely circular set of problems: policy decisions must be made in the face of scientific complexity and uncertainty; these judgments are complicated by economic, political and ethical complexities and uncertainties which, in turn, are obsured by the lack of acknowledgement as to the breakdown of facts and values in  7° Ibid, at pg. 1310.  86 the scientific information. In the most problematic cases of toxic risk analysis, where knowledge uncertainty prevails, it is apparent that assumptions arise at every point where uncertainty arises: in calculating "safe" exposure levels; in relating animal tumour data to human response; and, finally, in the very mechanisms by which disease is believed to work. To the extent that uncertainty exists, there is always the possibility to err on one side or the other, either of over-estimation or under-estimation of risk. The assumptions that are used to fill in the knowledge gaps created by uncertainty implicitly require values to be used in the sense that the assessor must chose between a risk-aversive or a risk-taking approach.71 For these reasons, the non-separatist school of thought is more persuasive. Their proposed procedural fix as a way of meeting the task of harnessing science while safeguarding legitimacy is also persuasive. What is particularly notable, from a Canadian perspective, about the Benzene case is the fact that OSHA's interpretation of available science and its analysis in connection with the statutory mandate was openly disclosed in the administrative record. This disclosure allowed the testing of the regulatory rationale through the courts and thereafter induced a lively scholarly debate on the topic. Putting aside the question of whether one agrees with the court's reasoning, it is clear the the U.S. regulatory process permits more careful scrutiny of the implications of scientific uncertainty. The next chapter examines the experience in the United States with  71^Brunk, supra, fn 8, at pg. 36.  87 efforts to impose procedural rules about the disclosure and explanation of the scientific information used in environmental standard setting.  88 CHAPTER FOUR THE CASE FOR SCIENTIFIC EXPLICATION -- AN EXAMINATION OF THE U.S. EXPERIENCE  Let me propose the following generalization about public bureaucracies in the United States: They are almost always given huge, even utopian, goals and are then saddled with a large number of constraints that prevent them from achieving these goals efficiently - or even at all. We tell EPA, for example, to protect the public health with an adequate margin of safety, but advise it not to spend too much money or put anyone out of work. We tell them to use the best scientific evidence, but refuse to let them pay enough to recruit top-flight scientists, and then we tell them, "By the way, do it within 90 days." We expect bureaucrats to account for every penny of public money, to record every conversation with a member of an interest group, to show that they have treated everyone equally, and to consider all the relevant information and alternatives - but to stop producing all that red tape and being so damn slow.1  A.^Introduction The purpose of this chapter is to consider how a legal rule requiring environmental regulators to disclose and to explain the science behind their decisions might operate. The experience in the United States is useful in this regard. Over the last two decades, the collision of environmental regulatory demands and scientific complexities has been given considerable legislative, judicial, administrative and academic attention in the U.S. This has resulted in two noteworthy developments. First, the rulemaking2 procedures of the Environmental Protection Agency (the  1  ^  Melnick, supra, Chapter One fn 38, at pgs. 257-58.  =^"Rulemaking" generally refers to those administrative activities that result in the promulgation of "general rules" (as oppposed to case-by-case adjudication) pursuant to which procedural requirements are imposed under the U.S. Administrative Procedure Act as interpreted by courts. For all intents and purposes in this thesis, rulemaking is equivalent to the Canadian administrative activity  89 "EPA" or the "Agency") have shifted over time to provide for more rigorous scientific analysis and explanation. Secondly, there have been adjustments to environmental policy making and statutory mandates to increasingly accomodate the use of risk analysis, including cost-benefit analysis, in environmental decision making. Although these two developments are interconnected in the sense that the disclosure of the treatment of scientific data by EPA has enhanced the understanding of risk analysis methodology, and vice-versa, it is the first development, with respect to rulemaking, that is of particular interest to this analysis.3 One of the pertinent aspects of the U.S. experience, which makes it a compelling case for the purposes of this thesis, is the fact that the development of environmental law has been synonymous with the development of administrative law. As observed by a leading U.S. authority on environmental law: Environmental law, which involves standard setting and rulemaking of a sophisticated technical nature by government agencies, as well as administrative determination of controversies and the issuance of administrative orders and their enforcement is clearly a field of administrative law. Because of the high level of activity in the field, and because of the complex nature of the area, decisions with environmental implications now make up the majority of administrative law decisions in the federal courts, and, in consequence, the leading coursebooks on administrative law could probably be adapted quite easily to teach a course in environmental law.4  of regulation making. 3^This second development is generally not the subject of examination in this chapter or this thesis, however, some aspects of the risk analysis experience in the U.S. were previously noted in Chapter Three.  4^Grad, supra, Chapter Three fn 14, at pg. 1-7. This sharply contrasts to the Canadian experience where there has been very little correlation between the development of administrative law and environmental law.  90 With respect to rulemaking, the courts in the United States have required EPA to establish a "reasoned" basis, supported by the administrative record, in scientific evidence when setting regulatory standards. For example, in one of the first cases dealing with the impact of scientific information on standard setting, Kennecott Copper Corp. v. EPA5 , the EPA record provided no explanation for the standard set  in a regulation. The court remanded the standard back to EPA and directed the "Administrator to supply an implementing statement that will enlighten the court" as to the numerical standard set in the regulation. The rationale for this requirement was to facilitate the court in judicial review on the basis that more information about the regulatory analysis would help the court assess whether the regulation was reasonable or not. The Kennecott Copper case will be examined in greater detail later in this chapter but, for the moment, suffice it to say that such judicial pronouncements have affected the rulemaking procedures of EPA. In fact, the procedures imposed on EPA as a result of the judicial review of many environmental decisions have been codified in most major federal U.S. environmental statutes.6 Consequently, in the United States there are administrative procedural rules requiring the disclosure and explication of the scientific evidence used to support a  5^  462 F. 2d 846 (D.C. Cir. 1972). This decision is hereinafter referred to as the Kennecott  Copper case.  6 For example, see Ernest Gellhorn and Ronald M. Levin, Administrative Law and Process in a Nutshell, Third Edition, (St. Paul: West Publishing Co., 1990), at pg. 335 (noting that in several regulatory statutes adopted in the 1970s Congress experimented with variations of the Administrative Procedure Act rulemaking models, often borrowing devices from case law). In particular, under the Clean Air Act, which is considered in detail later in this chapter, the judicially created "paper hearing" requirements were adopted in the rulemaking provisions of that statute. See discussion infra.  91 regulatory decision. For this reason, the U.S. experience with environmental standard setting is a useful reference point for this thesis. It provides some valuable insights for the Canadian standard setting experience despite the obvious political-legal institutional and cultural differences between Canada and the U.S. This is so because the common denominators of administrative discretion and scientific analysis exist in both countries and raise identical concerns with respect to the legitimacy and rationality of the decision making process. In addition, the hybrid nature of the Canadian system, consisting of American administrative institutions and English judicial review7, permits the translation of the U.S. experience to the Canadian context, especially given the common issues in environmental regulatory protection. In order to bring this analysis down to a meaningful level of specificity, the discussion in this chapter will examine certain aspects of standard setting by the EPA under just one program, out of the vast number of regulatory schemes, administered by EPA. This examination will focus on air quality standards as mandated by the U.S. Clean Air Acts. The Clean Air Act (the "CAA" or the "Act") is a seminal command and control statute with a rich twenty-two year history. It is also one of the key environmental laws under which administrative rulemaking procedural  The notion of the Canadian experience as a "hybrid" between American and English concepts is part of the conventional wisdom in Canadian administrative law. However, as pointed out by Professor Bryden, American judicial review ideas have increasingly supplanted their English competitors over the last 25 years: "We have gone from a situation in which we understood English legal developments and accepted them as our own, to one in which we do not fully understand American law but are constantly dealing with its influence, sometimes consciously but as often as not unconsciously". (Philip Bryden, "Canadian Administrative Law: Where We've Been", 16 Queen's Law Journal 7 (1992), at pg. 8). 42 U.S.C. s.7410 et. seq.  92 requirements and scientific uncertainties have dovetailed. This chapter begins with an overview of the Act's history, the rulemaking context and a brief introduction to the two key standard setting provisions that will form the analytical substance in the last half of this chapter. In this discussion, the "Administrator" refers to the head of EPA. The Administrator is "independent" in the sense that he is not elected, but this is not to say that he is immune to political pressures. For the purposes of this discussion, the Administrator can be considered the functional equivalent of a Minister of Environment in Canada.  B. The Clean Air Act - An Overview  The Clean Air Act, originally enacted in 1970, is one of the earliest and most far-reaching examples of the command and control approach to environmental regulation that blossomed in the United States during the 1970s.9 The essence of the CAA as originally conceived was the establishment of a framework for the attainment and maintenance of air quality standards. In its present form, it also provides for the setting of emission standards for motor vehicles and fuels, the regulation of hazardous air pollutants, the protection of stratospheric ozone, and control over acid rain. This descriptive overview and the balance of this Part B are  9 The CAA served as a model for other federal environmental legislation in the U.S. such as the Clean Water Act. For a discussion of the history of the CAA and a concise overview of its main provisions, see the three part series of articles by Theodore L. Garrett and Sonya Winner, "A Clean Air Act Primer: Part P, 22 Environmental Law Reporter 10159 (March 1992) (hereinafter "Part I"); "A Clean Air Act Primer: Part IF, 22 Env. L. Rep. 10235 (April 1992) (hereinafter "Part II"): and "A Clean Air Act Primer: Part III", 22 Env. L. Rep. 10301 (May 1992) (hereinafter "Part III"). A more comprehensive discussion of the Act can be found in Grad, supra, fn 4.  93  a decidedly simple explanation of a very complex statute. However, a schematic overview of the CAA is useful "to prevent confusion and disorientation resulting from statutory overload."10 The statutory history and regulatory experience under the CAA provide a wealth of information and some fascinating insights into the benefits and shortcomings of pollution control in modern North American society. In 1983, Professor Shep Melnick wrote a book analyzing the first decade of the CAA's life and the influence of the judiciary on environmental policymaking.11 Professor Melnick explains why the CAA is a useful case study for understanding environmental law, which he describes as part of the "new social regulation": ...not only has the Clean Air Act had a major effect on the economy and environment, but it presents a leading example of the new regulation. One finds here the key issues of how to define acceptable social risk, how to enforce costly control requirements, and how to reconcile protection of public health with economic growth and energy development. Second, the judicial decisions issued under the act have themselves become models provoking emulation. Many doctrines developed under the Clean Air Act cases rank as the guiding principles of the new administrative law. Since these decisions have received praise from the academic legal community, Congress, and even the EPA, they provide a test case, not a straw man or a sideshow.12 As has been noted elsewhere in this thesis, the themes that Professor Melnick  10  Grad, supra, fn 4, at pg. 2-76.  ii Melnick's insightful book is Regulation and the Courts: The Case of the Clean Air Act, (Washington: The Brookings Institution, 1983). 12  Ibid, at pg. 19. [emphasis added]  94 identifies as central to the experience under the CAA are representative of many environmental standard setting and regulatory policy decisions in both Canada and the United States. For this reason, the CAA provides a useful microcosm in which to explore the central argument of this thesis.13 Over its more that two decades of life, the CAA has been the focus of voluminous amounts of litigation and has undergone two periods of Congressional scrutiny and amendments, one in 1977 and more recently, in 1990. One commentator writes: "Congress had considered amendments to the Act for most of the 1980s, but the debate was protracted and frequently acrimonious".14 In 1990, these troublesome issues were dealt with in the The Clean Air Act Amendments of 1990 (the "1990 Amendments"). The 1990 Amendments are exceptionally long, extremely complex and include a great deal of material that previously would have been committed to the regulatory process.15 As noted at the outset of this discussion, the CAA provides for the attainment and maintenance of air quality standards. Two key provisions of the CAA aim to do  13 Professor Melnick's book is not just about the influence of the courts. As he explains: "Viewed in one way, this book is about how a regulatory bureaucracy reacts to stimuli from an outside force. Viewed in another, it is a description of how various institutions and private to resolve troublesome regulatory issues. It is my hope that even those with no interest in the courts can learn something about bureaucratic decisionmaking and the nature of regulatory issues from this study." (lbid, at pg. 23) 14  Garrett and Winner, Part I, supra, fn 9, at pg. 10162.  Grad, supra, fn 4, at pg. 2-81. Professor Grad notes that the 1990 Amendments extend the CAA by "314 pages,...making it one of the longest environmental laws on the books." (Ibid). One key section, section 112, which will be discussed in greater detail in this chapter, was completely rewritten in 1990 to expressly include material previously left to the administrative process. This approach, according to Professor Grad, "indicates the Congressional distrust of the executive, and particularly EPA." (kid). 15  95 this and these sections will be the focus of discussion in this chapter. These two provisions, section 109 and section 11216, are examined in an overview manner initially and are then given more careful consideration in Part C below. However, it is first necessary to take a minor excursion through the governing principles of U.S. administrative law that are relevant to how the EPA conducts its business under the CAA. Needless to say, the political and legal institutional structures in the United States are different than those in Canada thus creating some important consequences to the administrative process and the exercise of discretionary powers. One of the key differences between the two countries is that the United States has a federal statute, the Administrative Procedure Act17 (the "APA"), circumscribing how administrative discretion and rulemaking, in particular, are to be exercised. The basic rulemaking procedure prescribed by section 553 of the APA is of interest to this analysis as it is the foundation for the procedural developments under the CAA. The APA provides for what is known as "informal"18 or "notice and comment" rulemaking. The APA's informal rulemaking process is simple and flexible consisting of only three procedural requirements: (i) notice; (ii) comment (ie. the written  submissions of interested parties); and (iii) the issuance of the final rule with a  Throughout this discussion, references to the relevant sections of the CAA will be made to the Public Law citations rather than to the U.S.C. citations. 16  17  5 U.S.C. s. 500 et seq. The APA was first enacted in 1946.  18 "Informal" rulemaking is contrasted with "formal" rulemaking or "rulemaking on the record". The latter type of rulemaking requires more formal adjudicative procedures (such as oral hearings and cross-examinations) and is usually triggered by the express provision of the governing statute.  96 "concise statement of basis and purpose".19 As was noted in Chapter Two, the Canadian federal government and some provinces have, as a matter of policy rather than statute, adopted modified versions of notice and comment procedures for regulation making. The basics of the APA procedural model were supplemented by amendments in 1977 to the CAA. These amendments adopted judicial interpretations of the APA's informal rulemaking provisions: When the EPA set its first national standards in 1971, it followed the minimal requirements of the Clean Air Act and the Administrative Procedure Act...Shortly thereafter the Court of Appeals for the D.C. Circuit announced that this procedure did not provide the courts with enough information to evaluate the reasonableness of the EPA's standards. Without requiring the agency to conduct trial-like hearings, the courts ordered it to comply with a set of procedures designed to increase public participation and to provide the courts with an administrative record to review. Congress included these so-called "paper hearing" requirements in the 1977 amendments."2° The administrative procedure to be followed under the CAA is expressly dealt with in section 307(d) of the Act. Rules must be promulgated by notice and comment procedures, beginning with the publication of a notice of proposed rulemaking in the Federal Register, the American equivalent of the Canada Gazette. This notice must  contain a statement of basis and purpose of the proposed rule. This statement must include a summary of the factual data on which the proposed rule is based, the methodology used in obtaining and analyzing the data, and an explanation of major  19  Gellhorn and Levin, supra, fn 6, at pgs. 323-324.  20  Melnick, supra, fn 11, at pg. 263.  97 legal interpretations and policy considerations underlying the proposa1.21 The CAA also provides that if the proposed rule relates to any subject on which the Scientific Review Committee and/or the National Academy of Sciences has made findings, recommendations, or comments, this information must also be provided in the notice.22 All supporting documentation for the rule must be placed in the official EPA "docket" (or administrative record) and made available for public review. After notice has been published, the EPA is required to accept written comments, data, or other documents from "any person" and to give interested persons an opportunity for "the oral presentation of data".23 All written comments and other documents submitted during this time must also be available in the public docket. The CAA does not set out a minimum time for public comment. The final rule promulgated by EPA can be based only on the information and data that are in the docket.24 The final rule must be accompanied by a response to all "significant" comments, criticisms, or data provided during the comment period, as well as an explanation for any major change in the final rule from the proposed rule.25 The CAA also provides for pre-enforcement review in the U.S. circuit courts 21 CAA, s. 307(d)(3). Ibid. The Scientific Review Committee is an independent body established by s. 109(d)(2) of the Act to review national ambient air quality standards at five-year intervals, commencing in 1980. 22  23 CAA, S. 307(d)(5). The statute does not require formal hearing procedures to be used, other than the preparation of a transcript of the proceeding.  24  CAA, S. 307(d)(6)(c).  25  CAA, s. 307(d)(6).  98 of the air quality standards set by EPA under sections 109 and 112 of the Act.26 These standards generally mimic the APA judicial review standards. A court may invalidate a regulation if the regulation is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law", it is unconstitutional, in excess of statutory jurisdiction or without observance of procedure required by law.27 Accordingly, the CAA establishes a procedure which, at least ostensibly, links the EPA's final regulatory decision to the underlying facts through a participatory process that tests the EPA's explanation of its analysis. The statute also provides an institutional check on the ultimate regulatory product through judicial review. What is striking about this system, from a Canadian perspective, is the fact that any interested party to a rulemaking can obtain access to the relevant facts, methodologies and explanations for the regulations, scrutinize them, and challenge the decision in a much more open manner than is permitted in the Canadian approach to environmental regulation. Section 109 and "NAAQSs". Section 109 creates the principal regulatory  program of the CAA. It has two elements: (1) nationwide air quality goals set by  26  Judicial review is governed primarily by sections 304 and 307 of the Act.  CAA, s. 307(d)(9). S. 307(d)(8) also states that in reviewing procedural errors, the court may only invalidate a rule if the errors were so serious and of such central relevance to the rule that there is a substantial likelihood that the rule would have been significantly changed if such errors had not been made. Also, compare the judicial review provisions of the CAA to those found in section 706 of the APA: a court may set aside an agency action found to be "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law"; it violates the constitution; it is in excess of statutory jurisdiction; it does not observe procedure required by law; it is unsupported by substantial evidence; and it is unwarranted by the facts. (APA, ss. 706(2)(A)-(F)). 27  99 the EPA; and (2) individual state implementation plans, known as "SIPs", approved by EPA in order to implement the national goals. Implementation is obviously a key component of any regulatory program and although SIPs are a fundamental aspect of the CAA, it is beyond the scope of this analysis to deal with SIPs. This discussion deals only with the EPA's experience in establishing national air quality standards. Under section 109, EPA is directed to promulgate health-based primary and secondary national ambient air quality standards, or "NAAQSs", for what are known as "criteria" pollutants. Criteria pollutants are defined in section 108 of the CAA as air pollutants that may reasonably be anticipated to endanger public health or welfare and that result from numerous mobile or stationary sources. These pollutants are called "criteria" pollutants because they are pollutants for which the EPA has issued "air quality criteria" -- scientific information about the extent and nature of the effect of the pollutant on public health or welfare.28 As of 1991, EPA had listed seven criteria pollutants and promulgated NAAQSs for six of these pollutants: sulfur oxides (sulfur dioxide), particulate matter, nitrogen dioxide, carbon monoxide, ozone, and lead.29 Section 112 and "NESHAPs". In addition to section 109, the CAA seeks  to control hazardous air pollutants or "air toxics" under section 112. Section 112 has a very checkered 20 year history under the CAA and it was significantly amended in  28 Janet L. McQuaid, "Risk Assessment of Hazardous Air Pollutants under the EPA's Final Benzene Rules and the Clean Air Act Amendments of 1990", 70 Texas Law Review 427 (1991), at pg. 429. 29 40 CFR pt. 50 (1991). Volatile organic hydrocarbons are listed but no NAAQS has been promulgated.  100 1990 to respond to the many criticisms about its failure. For the purposes of this discussion, the historical experience under section 112 is of most interest because of the understanding it sheds on the implications of dealing with scientific uncertainty and the regulation of toxic substances. Accordingly, the pre-1990 section 112, referred to as the "old section 112" or "section 112", will be the subject of this analysis. The 1990 revision of section 112 will be referred to as "the new section 112". By way of introductory overview, some salient points about the old section 112 and the new section 112 are worth making to give the history of section 112 some context. The old section 112 authorized EPA to establish health-based national emission standards for hazardous air pollutants, or "NESHAPs". Hazardous air pollutants were defined under the old section 112 as air pollutants to which no ambient air quality standard was applicable and that, in EPA's judgment, might reasonably have been anticipated to increase mortality, serious irreversible illness, or incapacitating reversible illness.30 Section 112 pollutants are typically found at high concentrations close to the source and are thought to have more serious health impacts than criteria pollutants.31 The old section 112 envisaged that EPA would follow a two-step process: (1) develop a list of air toxics; and (2) then establish NESHAPs at a level adequate to protect the public health. Between 1970 and 1990, only eight substances or  30  31  ^  McQuaid, supra, fn 28, at pg. 429 (citing old s. 112). Ibid.  101 categories of substances were listed for regulation: asbestos, benzene, beryllium, mercury, vinyl chloride, radionuclides, inorganic arsenic, and coke oven emissions.32 Standards for some, but not all, sources of seven of these air toxics had been promulgated and four were remanded to EPA by the courts for reconsideration.33 The regulation for vinyl chloride was one such remanded decision and it will be discussed in greater detail below. By the time Congress came to consider section 112 in 1990, it was perceived as highly defective; there was a "widespread, virtually universal recognition that [EPA's] current efforts have been a failure".34 As a result, section 112 was completely rewritten in 1990 to establish a new regime for hazardous air pollutants. In particular, the new section 112 expressly listed 189 pollutants as air toxics for regulation; a telling contrast with the eight substances listed by EPA to that point. This step dramatically reduced EPA's discretion to determine what substances were to be listed presumably to avoid the regulatory gymnastics which had characterized the first 20 years of section 112. As noted above, the details of the new section 112 are not of major analytical interest other than to note the basic result - the statutory listing of 189 pollutants. In short, in addition to the express list of the 189 air toxics, the principal elements of the new section 112 are as follows:  32  40 CFR pt 60, s.61.01 (1991).  33  McQuaid, supra, fn 28, at pg. 431.  Leslie F. Chard III, "The 1990 Clean Air Act Amendments: Section 112 Comes of Age", 59 University of Cincinnati Law Review 1253 (1991), at pg. 1253. See also McQuaid, supra, fn 28, at pg. 428, fn 13 therein citing additional support for this perception. 34  102 • •  •  it creates a strict timetable for the promulgation of regulations governing emissions of those pollutants; it sets a new technology-based methodology for regulating hazardous emissions, with the principal focus on maximum control of total source emissions rather than on strict pollutant-by-pollutant emission standards; and it establishes a revised strategy for regulating any "residual risk" remaining after the imposition of technology-based standards.35  One crucial to this analysis should be made at this juncture. As noted, both sections 109 and 112 were drafted as health-based standard setting provisions meaning that NAAQSs and NESHAPs standards were to be set at a level adequate to protect the public health. This language, like most environmental laws, raises perplexing questions about what is "adequate"? One commentator, writing about the CAA, has observed that the upshot of this vague language, which is more symbolic than functional, was to "fob off on the EPA and the courts the difficult problem of defining acceptable risk".36 Lastly, as was observed earlier in this thesis, the understanding of adverse health impacts is inextricably linked to scientific information and analysis. For the purposes of the following discussion, it might therefore assist the reader to think of health-based standards and scientific analysis (or risk assessment) as one and the same thing; that is, defining what is appropriate to  35 Garrett and Winner, Part II, supra, fn 9, at pg. 10246. For further discussion and analysis of the details of these provisions, See Garrett and Winner, Part II, lbid, and Chard, supra, fn 34. One economist notes that the cost of the new scheme in section 112 when fully implemented may be between $6 - 10 billion annually (as cited in Daniel A. Farber, "Risk Regulation in Perspective: Reserve Mining Revisited", 21 Environmental Law 1321 (1991), at pg. 1322). 36  McQuaid, supra, fn 28, at pg. 433.  103 protect health is primarily seen as a scientific pursuit.  C.^Standard Setting and Science - A Closer Look  As noted in Part A above, section 109 of the CAA forms the cornerstone of the Act's air pollution control program through EPA's promulgation of NAAQSs. The NAAQS prescribes the maximum permissible concentration of a harmful pollutant in the ambient air. For example, the NAAQS for airborne lead is "1.5 micrograms per cubic meter, maximum arithmetic mean averaged over a calendar quarter".37 How is such a determination made? Under the relevant sections of the CAA, the promulgation of a NAAQS involves a three-step process, all taken within the context of the Act's administrative procedural requirements, previously noted. First, the Administrator is required to publish a list of air pollutants.38 Second, within one year after listing, the Administrator must issue an air quality criteria document describing the kind and extent of the effects on public health and welfare that may be expected from the listed pollutant.39 At the same time that the criteria document is issued, a proposed NAAQS is also published. Lastly, after a reasonable time for public comments on the proposed standards, the Administrator must promulgate the final NAAQS.4° One essential point to note about this statutory  37  40 CFR pt.50, s.50.12.  38  CAA, s. 108(a)(1).  39  CAA, s. 108(a)(2).  40  CAA, s. 109(a)(1)(B).  104 scheme is the express link it makes between scientific information, as set out in the criteria document, and the regulatory standard ultimately promulgated by EPA. To list a substance as an air pollutant, the Administrator must determine that (i) the substance is an "air pollutant" as defined in the CAA 41 (ii) it is emitted from numerous or diverse mobile or stationary sources; and (iii) the emissions will, in his judgment, cause or contribute to air pollution which may reasonably be  anticipated to endanger public health or welfare. This language reveals that the initial determination to list, which ultimately results in regulation, is highly discretionary. In particular, the last prong of the listing requirement expressly recognizes the discretionary exercise of the Administrator's judgment with respect to anticipated health risks.42 It has been stated that: "Although authority granted to the Administrator is broad...Congress intended to give the Administrator authority to exercise his judgment while requiring that the judgment be rational and subject to adequate judicial review of its reasonableness".43 Since listing automatically generates a scientifically-based criteria document and then a NAAQS, this reasonableness notion appears to get tested at these later stages of the procedure.  41 "Air pollutant" is very broadly defined in S. 302(g) of the CAA and, in part, provides that it means "any air pollution agent or combination of such agents, including physical, chemical, biological, radioactive...substance or matter which is emitted into or otherwise enters the ambient air..." 42  Compare this approach with the definition of "air contaminant" as provided in section 1(1) of the B.C. WMA: an "air contaminant" means a substance that is emitted into the air and that injures or is capable of injuring the health or safety of any person, etc. It is arguable that the essence of the administrative discretion in these two cases is the same despite the differences in the statutory schemes, namely, to determine what air emissions will endanger public health. 43 Garrett and Winner, Part I, supra, fn 9, at pg. 10166.  105 The statutory requirement of an air quality criteria document is especially interesting to this thesis because it creates an express recognition and statutory foundation for scientific input into standard setting. Thus, it can be argued that the CAA creates a regime for rational administrative action based on scientific analysis. Under section 108(a)(2) of the CAA, criteria documents must "accurately reflect the latest scientific knowledge useful in indicating the kind and extent" of a pollutant's effects on public health or welfare. In addition, the section goes on to provide that, "to the extent practicable", a criteria document must include factors that alter the effect of a pollutant, the interaction of one pollutant with another, and any known or anticipated adverse effect on public welfare. The CAA also provides that the issuance of air quality criteria documents must be announced in the Federal Register, and made available to the general public.44 The production of the criteria document creates a background factual basis and process leading to the issuance of the primary and secondary air quality standards for a pollutant. It became clear in the early stages of EPA's standard setting efforts, that disputes over the scientific analysis would necessitate procedures for greater scrutiny of the data. As noted above, the 1977 amendments to the CAA imposed a more detailed administrative process for the promulgation of NAAQSs. These procedures include permitting interested parties to submit written comments on the proposed rules; providing an opportunity for oral presentation; the creation of an administrative record; and providing for EPA responses to "significant"  '4  CAA, s. 108(d).  106 comments, criticisms, and new data submitted in oral presentations.45 The listing of a substance and the production of the criteria document ultimately determined the basis for the establishment of a NAAQS. The statutory language defining what a NAAQS means, as set out in section 109, is particularly crucial to understanding the interaction between administrative discretion and scientific analysis. It states that national primary ambient air quality standards shall be: ambient air quality standards the attainment and maintenance of which in the judgment of the Administrator, based on such criteria and allowing an adequate margin of safety, are requisite to protect the public health.46 The breadth of the Administrator's discretion is readily apparent in this language. What standard represents an "adequate margin of safety" to "protect public health"? It is evident that many of the same questions facing Canadian environmental regulators are relevant to the U.S. administrator's task. Notably, the Act presumes that a "right level" of health impact can be determined. Are the economic costs of the health risk to be incorporated into the decision? Does the technological feasibility of obtaining this emission level bear on this assessment? On the basis of the description of section 109 given thus far, it appears that the CAA presents standard setting as a purely scientific investigation founded on the  45  46  See CAA, s. 307(d) and discussion above in Part B.  CAA, s. 109(b)(I). This is the standard for primary air quality. The Act also requires the Administrator to promulgate secondary standards which "in the judgment of the Administrator, based on such criteria, is required to protect the public welfare". (CAA, s. 109(b)(2)). In this discussion, secondary standards are not examined since for all practical purposes, they are the same. See Garrett and Winner, Part I, supra, fn 9, at pg. 10167.  107 data set out in the criteria document. To the uninitiated this process sounds rather mechanical. However, this seemingly straightforward process belies the complexity of the nature of the scientific analysis. As was argued in Chapter Three of this thesis, the segregation of science denies that the scientific analytical process is a political as well as a technical one. The fallacy of this separation myth came to light in the initial efforts by EPA to establish both NAAQSs and NESHAPs. The difficulties posed by this fallacy were also evident in the first judicial examination of a NAAQS, in the case of Kennecott Copper Corp v. EPA47 , which will now be considered. In April 1971, in accordance with the EPA's mandate under the CAA, the EPA promulgated its first NAAQSs for six air pollutants, including sulfur oxides (sulfur dioxide).48 The promulgation of the sulfur dioxide standard was challenged in the Kennecott Copper case. In this case, Kennecott Copper Corp. brought a petition for judicial review of the EPA's national secondary ambient air quality standard for sulfur oxides of "60 micrograms per cubic meter - annual arithmetic mean". The essence of Kennecott's attack on the EPA standard was that there was no adequate basis for it in the criteria document. Three objections, in particular, were raised in the petition for judicial review regarding the EPA standard: (i) it was not based on the underlying air quality criteria document; (ii) it was not accompanied by a "concise statement of general basis and purpose" as required by  47  Supra, fn 6.  36 Fed. Reg. 1502 (proposed) and 8186 (final) (1971). In the proposed rulemaking it states that "sulfur dioxide is an indicator of the presence of sulfur oxides in the air" and thus it is used as an index for sulfur oxides. (36 Fed.Reg. 1502). 48  108 the Administrative Procedure Act49; and (iii) it was not adequately supported by a statement of the regulatory basis of action necessary for judicial review. The air quality criteria document on which the sulfur dioxide NAAQS was purportedly based mentioned no adverse effect from the pollutant at a level below 85 micrograms per cubic meter. The EPA's statement of basis and purpose accompanying the regulation simply repeated the statutory language: National secondary ambient air quality standards are those which, in the judgment of the Administrator, based on the air quality criteria, are requisite to protect the public welfare from any known or anticipated adverse effects associated with the presence of air pollutants in the ambient air." The Natural Resources Defense Council (the "NRDC"), as intervenor, argued that the EPA's standard was supportable on the basis of the Administrator's expressly permitted 'judgment"; in short, to protect the public health and welfare the Administrator correctly built in a margin of safety to avoid the adverse effects noted at the 85 level. The court declined to rule on the particular issue as to whether the Administrator exercised his judgment reasonably in picking the 60 or 85 level. Instead, the court cast the issue as a procedural matter by holding that the Administrator must disclose the basis of his decision for the standard he selects thus affording the court enough information to assess his actions. In examining the EPA's  49 Note that this case pre-dated the 1977 procedural amendments to the CAA which incorporated specific judicial developments interpreting the APA. The APA was applicable to rulemaking activities prior to 1977. 50  Kennecott Copper, supra, fn 47, at pg. 848.  109 procedure in this case, the court made several key observations revealing its sensitivities to the need to balance speedy disposition of the Agency's mandate with the court's special responsibility to review air quality standards: Inherent in the responsibility entrusted to this court is a requirement that we be given sufficient indication of the basis on which the Administrator reached the 60 figure so that we may consider whether it embodies an abuse of discretion or error of law. The provision for statutory judicial review contemplates some disclosure of the basis of the agency's action.. .We are keenly aware of the need to avoid procedural strait jackets that would seriously hinder this new agency in the discharge of the novel, sensitive and formidable, tasks entrusted to it by Congress... Particularly as applied to environmental regulations, produced under the tension of need for reasonable expedition and need for resolution of a host of nagging problems, we are loath to stretch the requirement of a "general statement" into a mandate for reference to all the specific issues raised in comments.51 Accordingly, the court remanded the matter to the Administrator ordering that he "supply an implementing statement that will enlighten the court as to the basis on which he reached the 60 standard from the material in the Criteria".52 This remand had significant consequences in that it ultimately resulted in the EPA's revocation of the standard in 1973.53 In re-examining the scientific evidence in its criteria document in order to provide the requisite rationale to the court, EPA found that it had misinterpreted the one study that might have supported its standard:  51  Ibid, at pgs. 849-50.  52  Ibid, at pg. 850.  It is interesting to note that a secondary standard for SOx now exists, 40 CFR pt.50 (1991). In fact, it is the only criteria pollutant that has been regulated as a secondary standard (Garrett and Winner, Part I, supra, fn 9, at pg.10167.) 53  110 The Office of General Counsel attorney charged with constructing an explanation for the court gave the following report to his superiors: "We were at first struck by the lack of written data supporting the levels set forth in the standards. As we went deeper into the matter it became more and more apparent that to some extent, the lack of documentation was due to lack of supporting data. ..There appears to be no doubt that the secondary standard for SOx was incorrectly established."54 What were the implications of the Kennecott Copper case? One of the key lessons from this case undoubtedly must be that a procedural requirement which forces the regulator to provide an explanatory connection between the scientific evidence and the legal standard ultimately promulgated, is a useful rule. In this case, the need to explain the regulated result revealed an error in analysis that had clear implications to the administrative product - it could not be supported on the relevant facts. To the extent that a requirement to disclose information and to give an explanation regarding it creates a rational connection between the science and the regulation, it does help avoid arbitrary decision making. In addition, the Kennecott Copper case is often cited by commentators, from varying perspectives (such as EPA,  Congress, academics, the National Academy of Sciences, industry and environmental litigants), as a watershed case forcing the upgrading of the scientific competence of federal regulators.55  54 Melnick, supra, fn 11, at pg. 241, fn 3 therein (citing an internal EPA Memorandum of 1972). ' lbid, at pg. 240. Professor Melnick is more skeptical about the court's pronouncement as a definitive catalyst for improved science at EPA_ He explains that this matter resulted from many forces over time. See, Ibid, pgs. 241-43 for a further elaboration. However, Melnick's explanation does not detract from the argument that a procedural rule requiring scientific explication is beneficial and logically it may also have this effect.  111 However, one of the difficulties that the Kennecott Copper case spawned was the implicit notion that EPA must base its regulatory actions on scientific evidence, a concept that arguably was established in the statutory scheme. This has had the  unfortunate effect of creating many of the perceived problems with environmental regulatory action such as extensive delay and even subterfuge. As Professor Melnick has explained, the Kennecott Copper case (and other similar judicial developments) resulted in the following regulatory approach at EPA: To guard against further embarrassing reversals the EPA has adopted an elaborate record-building process, replete with multiple opportunities for public comment and review of scientific evidence by panels of technical specialists. The agency now spends years developing the criteria documents on which air quality standards are based.56 The need to take agency action only after the accumulation of enormous amounts of technical information is one of the hallmarks of the U.S. administrative regime in environmental decision making. This result is perhaps the opposite extreme of what the court was hoping to achieve in the Kennecott Copper case. In that case, the court's rationale for disclosure was sound in that it posited that an exposition of the utilized information would be useful in validating the exercise of regulatory discretion. Presumably, the result of this would be that in some circumstances this exercise may reveal deficiencies prohibiting regulation. However, it does not necessarily follow that such disclosure should impose a regulatory burden  of epic proportions. This has been the tendency in the United States in this regard  ' Ibid,  at pg. 241.  112 and, in my submission, is a case of the pendulum swinging too far. The Kennecott Copper case represents the genesis of the notion in the U.S. that EPA is required to  base its action on science, rather than to explain the science. This distinction is important because, as was seen in Chapter Three of this thesis, it is rare that the science does not exist, rather it is the interpretation of uncertain and imperfect scientific evidence that is at issue. This point is also illustrated by EPA's experience under section 112 of the CAA, to which this analysis now turns. Hazardous air pollutants were segregated out for separate regulatory control under the provisions of section 112. As originally drafted in 1970, section 112 was enacted to deal with what was believed to be a narrow and manageable problem: setting stringent emission standards for a limited number of dangerous air pollutants, including carcinogens. 57 However, over time and certainly by the time of the 1990 Amendments, increased scientific attention and public awareness of the cancer problem led to increased public pressure for the regulation of air toxics. 58 Section 112 established a simple two-step rulemaking process for EPA to follow: (1) list a specific pollutant as hazardous; and (2) set a NESHAP for each category of sources of that pollutant. Thus, the initial listing decision imposed a nondiscretionary duty on EPA to promulgate a NESHAP in much the same way that listing under section 109 mandated the promulgation of a NAAQS with one 57  McQuaid, supra, fn 28, at pg. 428.  One commentator explains that the Toxic Release Inventory published in 1989, based on data gathered pursuant to the federal Emergency Planning and Right-to-Know Act of 1986, revealed for the first time the "actual extent of hazardous air emissions nationwide" thus adding to the momentum for section 112 reform in 1990. (Garrett and Winner, Part II, supra, fn 9, at pg. 10245). 58  113 significant difference - no criteria document was necessary under section 112.59 As most commentators of section 112 have noted, it was the statutory healthbased test for setting these standards that prevented effective implementation. Why was this so? The health-based concept for standard setting used in section 112 was quite similar to the approach set out in section 109 of the CAA. Section 112 directed the EPA to set emission standards for air toxics at a level low enough to "protect the public health" with an "ample margin of safety"60, as contrasted with the phrase an "adequate margin of safety" used in section 109. The requirement to provide control to protect public health with an "ample margin of safety" was interpreted by some environmental groups, and for that matter by EPA, as requiring zero emission standards. This argument was based on the fact that carcinogenic air pollutants do not have identifiable threshold levels below which there is no risk of an adverse health effect.61 If no threshold level can be identified, then it followed that an "ample margin of safety" was required to be set at zero. EPA's policy document on this point stated the following: The method used to establish a margin of safety for a threshold pollutant - setting the standard somewhere below the demonstrated effects level as a point at which the absence of adverse health effects 59 However, this did not negate the importance of scientific information. As one writer has noted, the determination of whether a substance was hazardous or not was based on relevant scientific data (John 0. Graham, "The Failure of Agency-Forcing: The Regulation of Airborne Carcinogens under Section 112 of the Clean Air Act", 1985:1 Duke Law Journal 100 (1985), at pg. 108).  McQuaid, supra, fn 28, at pg. 430 (citing old section 112, 42 U.S.C. s.7412 (b)(1)(B) (1988)). 61  As was noted in Chapter Three, most carcinogens are "non-threshold" pollutants meaning that there is no identifiable "threshold" at which scientists can determine a level below which no health effects can be observed. Under section 112, every pollutant the EPA listed was a non-threshold carcinogen (as observed in the Vinyl Chloride case, infra, fn 64, at pg. 1153, fn 1 therein.)  114 is predicted - therefore cannot be used to set standards (other than zero) for carcinogens under section 112, since risk of cancer is believed to exist at any exposure level greater than zero.62 The perception that section 112 might require zero emission limits put the EPA in an untenable position because such standards would cause many major U.S. industries to go out of business.63 In an EPA policy document, EPA determined that a zero-emissions standard for non-threshold pollutants would result in the elimination of such activities as: the generation of electricity from either coal-burning or nuclear energy; the manufacturing of steel; the mining, smelting, or refining of virtually any mineral (eg., copper, iron, lead, zinc, and limestone); the manufacture of synthetic organic chemicals; and the refining, storage, or dispensing of any petroleum product." The reluctance of the EPA to do this was compounded by the fact that industries would be shut down without "assurance that [the] regulations would result in health benefits that are not grossly disproportionate to the costs of control."65 The result of this dilemma, as was pointed out above, was the listing of a mere eight pollutants in 20 years, most of which occurred before 1977, while the 1990 amendments  identified 189 air toxics.  62  McQuaid, supra, fn 28, at pg. 430, fn 23 therein(citing EPA policy document).  63  Garrett and Winner, Part II, supra, fn 9, at pg. 10245.  64 Natural Resources Defense Council v. EPA 824 F.2d 1146, at pg. 1154 (citing NESHAPs; Policy and Procedures for Identifying Assessing and Regulating Airborne Substances Posing a Risk of Cancer, 44 Fed.Reg. 58,642 (1979)). This decision is hereinafter referred to as the Vinyl Chloride case.  65 McQuaid, supra, fn 28, at pg. 431 (citing the Comptroller General of the U.S., "Delays in EPA's Regulation of Hazardous Air Pollutants" (1983)).  115 What role did science play in EPA's approach to regulation under the old section 112? The gap in scientific knowledge appears to have had two significant consequences. First, it permitted the playing of the "politics-of-science" game66: the promotion of delay and self-interest. EPA delayed the passing of NESHAPs as long as it could by creating a cumbersome process of "sound science" for listing air toxics in the first place.67 This is a manifestation of the administrative concern that action must be based on solid scientific facts. It has also been observed that the lack of available data on health effects and "other scientific data" meant that the "agency had to rely on studies of lab animals such as rats and mice, which had been exposed to very high doses of a pollutant, and extrapolate from those studies the risk to human health from exposure to the relatively low doses of the pollutant in the environment".68 As was seen in Chapter Three, this type of information is subject to interpretation and is not free from uncertainty. For those chemicals that survived the listing process, the EPA rewrote the statute "either openly or, if the political costs or risks of judicial reversal were too high, silently."69 These consequences are evident in the EPA's experience with attempting to establish NESHAPs for the eight listed pollutants. As noted above, of the seven chemicals for which emission standards had been set before 1990, 4 were remanded ' Doern, supra, Chapter Three fn 64. 67 Graham, supra, fn 59, at pgs. 116-123. See also, Margaret L. Claiborne, "The New Air Toxics Program", 7 Natural Resources and Environment 21 ( Fall 1992), at pg. 21 (attributing listing delays  and standard setting delays to the lack of data on health effects and "other scientific data"). 68  Claiborne, supra, fn 67, at pg. 22.  69  McQuaid, supra, fn 28, at pg. 424.  116 by the courts." The radionuclide proceeding was particularly acrimonious and is worth examining briefly. In 1979, the EPA determined that radionuclides were carcinogenic and it listed them as hazardous air pollutants. However, it failed to further regulate any NESHAPs. This lack of action prompted the Sierra Club to sue EPA to take action and the court ordered the Administrator to issue NESHAPs. The EPA Administrator initially ignored the court's order to issue final standards. As a result, the court found the Administrator, William Ruckelshaus, in civil contempt71. Professor Melnick's summary of this situation is worth repeating in full: At first Ruckelshaus surprised everyone by refusing to obey the court. Not only was the scientific evidence unclear - it always is - but EPA estimated that the health risks were tiny and the potential cost large. Estimated risk from some facilities was one cancer death every fifty years, from other facilities, one cancer death every thirteen to seventeen years. Ruckelshaus was in the middle of a major effort to educate the American public about the nature of environmental risks to let them know that there is no such thing as a completely safe environment, and to force the government and the public to be honest about the level of safety they are willing to pay for. He told the court, "Given the inevitable burdens that regulation imposes just by its existence, and the shortage of resources to deal with the real health risk both in EPA and the society at large, these risks did not appear to me to be large enough to warrant regulation." The judge responded by calling Ruckelshaus a "scofflaw", describing his actions as "outrageous", and expressing shock that such a "responsible person" (and a lawyer to boot) would so cavalierly disregard an order of the U.S. District Court. After being found in contempt of court, Ruckelshaus published regulations that would have virtually no effect on radionuclide levels. EPA estimated that it spent $7.6 million in 70 Supra, fn 34. The four standards subject to remand were: vinyl chloride, radionuclides, benzene (equipment leaks) and asbestos (demolition procedures) (as noted by McQuaid, supra, fn 28, at pg. 431). 71  See Sierra Club v. Ruckelshaus 602 F.Supp. 892 (N. D. Cal. 1984).  117 contract funds and 150 staff-work years on this regulation.72 Another case that exemplifies the interface of science and regulatory discretion under section 112 is the EPA's twelve year saga of rulemaking in connection with the substance vinyl chloride. This rulemaking was ultimately tested in NRDC v. EPA73, known as the Vinyl Chloride case, in which the D.C. Circuit confirmed that the EPA was not required to adopt a zero exposure standard in promulgating NESHAPs.74 Vinyl chloride is a gaseous synthetic chemical used in the manufacture of plastics and is a strong carcinogen. In 1975, EPA issued a notice of a proposed rulemaking which on its face gives a fairly explicit explanation of the EPA's policy analysis of the scientific data. The scientific data included both rodent experiments and human exposure studies. In the notice, the EPA asserted that the scientific data linked vinyl chloride to liver cancers and other disorders and that "reasonable extrapolations" from this data suggested "that present ambient levels of vinyl chloride may cause or contribute to such disorders".75 The EPA noted that vinyl chloride is a non-threshold pollutant and thus concluded that in such cases EPA assumes, in  72 Melnick, supra, Chapter One fn 38, at pgs. 250-51. This battle has not ended. In 1989, EPA issued final rules for radionuclides which were the subject of a pending appeal in the D.C. Circuit Court and "EPA is already reconsidering major elements of the radionuclide regulations". (Garrett and Winner, Part II, supra, fn 9, at pg. 10246, fn 14 therein.) 73  Supra, fn 64.  74 The chronology and substance of this rulemaking as described in the next few paragraphs is based on the court's description of EPA's action as set out in the Vinyl Chloride case and a review of the Federal Register notices of the relevant rulemaking activities: 40 Fed.Reg. 59,532 (1975); 41 Fed.Reg. 46,560 (1976); 42 Fed.Reg. 28,154 (1977); and 50 Fed.Reg. 1182 (1985). 75  Vinyl Chloride, supra, fn 64, at pg. 1148.  118 the absence of strong evidence to the contrary, that "there is no atmospheric concentration that poses absolutely no public health risk".76 In the face of this scientific analysis, EPA concluded that it was faced with two alternative interpretations of section 112. Either it could set the standard at zero emissions on the basis that a "zero emission limitation will be the only emission standard which would offer absolute safety"77 or it could use a technology-based standard based on the "best available control technology", or "BAT". In its final rule issued in 1976, EPA opted for the BAT standard, setting the general emission level at 10 ppm, because it believed that it would "produce the most stringent regulation of hazardous air pollutants short of requiring a complete prohibition in all cases."78 As noted previously, EPA simply found the regulatory consequences of a zero emission standard untenable." The BAT approach was challenged by the Environmental Defense Fund which threatened to sue EPA on the ground that section 112 prohibited the Administrator from considering costs and technology in his standard setting. This suit was settled upon EPA's agreement to propose new and more stringent vinyl chloride emissions and to establish an ultimate goal of zero emissions. Pursuant to this settlement  76  Ibid.  77 Ibid. 78  lbid, at pg. 1149.  79 In particular, in the Vinyl Chloride rulemaking, the EPA's proposed notice indicated that the EPA found this alternative "neither desirable nor necessary" because "complete prohibition of all emissions could require closure of an entire industry," at a cost the EPA found "extremely high for elimination of a risk to health that is of unknown dimensions." (Ibid, at pg. 1148)  119 agreement, EPA again issued proposed regulations in 1977 which sought to impose a stricter standard of 5 ppm (as opposed to the 10 ppm originally proposed in 1975) and by establishing an aspirational goal of zero emissions. The result of these proposed stricter standards was that EPA took no further action for seven years. Ten years after its first proposed rule, in 1985, EPA withdrew the proposal and returned to a proposal similar, but for minor revisions, to the 1976 standards. EPA explained that it was abandonning the 5 ppm standard because it imposed "unreasonable" costs on industry and that no control technology "has been demonstrated to significantly and consistently reduce emissions to a level below that required by the current standardum of 10 ppm. As a result of this action, the NRDC petitioned for judicial review arguing that the Administrator must regulate under section 112 exclusively on health-related factors and that any other considerations amounted to actions that were arbitrary and capricious. Alternatively, the EPA argued that in the face of scientific uncertainty, it is permitted to set emission levels at the standard which is the lowest level attainable by BAT (whenever that level is below that at which harm to humans has been demonstrated). The court rejected both positions and instead posited its own two-step analysis for EPA to follow based on its statutory interpretation of "protecting public health with an ample margin of safety". First, said the court, EPA was to determine, on health considerations alone, the minimum level of control necessary to protect public health. In the court's words: "[Section 112] requires the  80 Ibid, at pg. 1149.  120 Administrator to make an initial determination as to what is 'safe'".81 Interestingly enough, the court expressly observed that "safe" did not mean "risk free"82 but it did mean setting a standard that resulted in an "acceptable risk to health" without considering costs or technological feasibility.83 With respect to determining acceptable risks, the court advised the EPA to "determine what inferences should be drawn from available scientific data and decide what risks are acceptable in the world in which we live."84 Presumably, on the court's reasoning, costs and technology and any other non-health factors, are divorced from real world considerations of acceptable risk. After this determination, EPA was to identify whatever level of control was appropriate to provide "an ample margin of safety". At this stage, EPA was free to consider health effects, economic factors, technological feasibility, scientific uncertainty, and other relevant factors. The "ample margin" language, as the court explained, requires the exercise of discretion: "This language permits the Administrator to take into account scientific uncertainty and to use expert discretion to determine what action should be taken in light of uncertainty."85 What has the court's analysis of the NESHAPs mandate meant for EPA? Given the 1990 Amendments and their establishment of a specific list of 189 air toxics 81  Ibkl, at pg. 1164.  82  Ibid (citing the Benzene case which was discussed in Chapter Three of this thesis).  83  Ibid, at pg. 1165.  84  Ibid.  85  lbid, at pg. 1165 [emphasis added].  121 to be regulated on a technology-based approach, this question is, for the most part86, moot today. However, after the Vinyl Chloride decision and prior to 1990, two final NESHAPs were adopted for various benzene and radionuclide sources. It is instructive to briefly examine an analysis of these rules because of what it reveals about the continuing struggle to set standards and what it might say about environmental standard setting in Canada. One analyst has stated that the benzene and radionuclide rules indicate that the EPA "has used this discretion silently to consider non-health factors in its acceptable risk determinations."87 In other words, EPA has brought cost and feasibility factors in through the back door. This is because, as was pointed out in Chapter Three of this thesis, scientific-based risk assessment determinations simply cannot be made on a value-free basis absent considerations of costs and other factors. For example, in the post-Vinyl Chloride benzene standards, EPA chose a "presumptively acceptable level of risk" to the maximum exposed individual, over a lifetime, of 1 in 10,000 (meaning that their regulatory standard would be premised on the basis that 1 out of 10,000 people could be expected to die of cancer). As the regulation explains this number was selected from a survey that produced individual 86 This is a qualified comment because some of the old issues under section 112 may continue to haunt the application of the new section 112. As one commentator observes: "By shifting from health-based to primarily technology-based standards, the agency may avoid many of the problems encountered under the old scheme. Even under the new scheme, however, old issues may be the subject of renewed debate, including how to calculate the risk to public health without the necessary scientific data and how to factor in costs when establishing emission standards." (Clairborne, supra, fn 67, at pg. 56).  McQuaid, supra, fn 28, at pg. 438. See McQuaid's analysis at pgs. 438-43 demonstrating that, based on the EPA's rationale in its final benzene and radionuclide rules, the EPA continues to weigh non-health factors in establishing "acceptable risks" in spite of the Vinyl Chloride case. 87  122 risks of between 1 in 10 and 1 in 10,000,000. The EPA concluded that no single factor by itself could define acceptabilility in all cases and chose the level of risk presumed to be common "in the world in which we live"; that is, 1 in 10,000. As one writer has noted: "EPA could have chosen a[] [maximum individual risk] anywhere in this range, thus assuring the flexibility to weigh costs and technological factors when it adds an ample margin of safety to the acceptable risk level."89 The benzene regulations also explain that the presumptive level is not a bright line; it is simply a presumption that can be adjusted on the basis of "other health measures". The regulations expressly state that other health measures include such intangibles as science policy assumptions, estimation uncertainties in the risk estimates, and the weight of the evidence of human health effects." These three intangible factors allow the EPA to justify an acceptable risk of more than 1 in 10,000 in the first step, again making "room for consideration of cost and technological feasibility in the second, the 'ample margin of safety' step".9' The point illustrated by these observations is simple - acceptable risk is inherently connected to these other factors and the continued segregation of healthbased factors perpetuates the myth of the separateness of science. This segregation, in turn, appears to bolster the overall rationality of the rulemaking, when ironically  88  Ibid, at pg. 439, fn 81 therein (citing the 1989 Benzene Standards).  89 Ibid. 99 Ibid, 91  Ibid.  at pg. 440.  123 in fact, it disguises the important policy choices in the decision.  D. Conclusions  The EPA's regulatory experience with standard setting under the Clean Air Act echoes the basic point of Professor Melnick's remarks at the outset of this chapter; that is, EPA is given a herculean task and then is blamed for the failure due to its bureaucratic paralysis or incompetency. The CAA is a particularly intriguing example of the regulator's task in environmental law because it captures the tough issues the lack of scientific understanding with respect to carcinogens and the nonsegregation of costs and technology from the realities of environmental standard setting. It also exemplifies the results - extensive periods of delay or inactivity (even resulting in a contempt order against the Administrator in one case!) and endless debates and posturing over acceptable risk, leading to conclusions of administrative failure resulting in legislative "cure" (eg. the listing of the 189 air toxics in section 112). Understanding the nature of the underlying scientific anaylsis helps to shed light on the administrative dilemmas and can suggest ways to more appropriately cope with environmental regulations. One answer may be to improve legislative mandates by increased recognition of the policy choices that must be made but this seems somewhat unrealistic in the era of the modern administrative state. In this thesis it has been argued that the disclosure of scientific information and a corresponding duty to explicate the use and reliance thereon will help to distill the regulator's role thereby fostering a more transparent societal debate about the  124 values that are at stake in environmental regulation. In this chapter, the U.S. experience with scientific disclosure and explanation was examined in order to make the case for such a procedural rule in Canada. What can be learned from the U.S. through the lens of the CAA experience? In general, it was seen that the advantages of scientific disclosure and explanation are: 1.  The exercise of forcing a regulator to explain the correlation between the scientific evidence and the regulated result may reveal genuine deficiencies in analysis. Witness the Kennecott Copper case.  2.  This focus may also help to upgrade the regulator's scientific/technical proficiency. This ability will surely help a regulator to utilize scientific information more intelligently and given the complexities of this information, this is a good result.  3. Disclosure helps interested parties assess what the regulator has done with the information; how he has filled in information gaps and exercised his discretion. This will help expose the factual premises of a decision to public scrutiny and for those wanting to test the administrator's role, it will more readily facilitate judicial review by providing a record of what has been done.  125 On the down side, the CAA experience indicates that the need to understand scientific complexities can, in the administrative context with its need for rationality, turn into reliance on scientific evidence that simply does not (and may never) exist. This has the following consequences: 1.  It may promote delay and/or bureaucratic fence-sitting on troublesome political issues by insisting on "further study" and scientific information to back up decisions. To some, this approach is one that simply waits for the "body count".  2.  It may also shift policymaking further underground by disguising it behind the veneer of science. This is illustrated in the EPA's rulemakings that followed the Vinyl Chloride case. Such results achieve the complete  opposite of what is aimed at by a "better process". What does all this mean for the design of a Canadian procedural rule about scientific disclosure and explication? That is the subject of the next and closing chapter of this thesis.  126 CHAPTER FIVE CONCLUDING REMARKS  Clarity and fullness of explanation are central to the legitimacy of decisions affecting individual health and safety.'  At the outset of this thesis, it was stated that one of the challenges facing environmental law is to "design a regulatory process that captures the knowledge of science while safeguarding the proper domain of political choice."2 In order to answer this challenge, this discussion has journeyed through the maze of environmental regulatory discretion and considered its interface with scientific knowledge. Pollution control laws are characterized by the delegation of ample discretionary powers to regulators. As a result, it has been argued that the design and implementation of the law primarily occurs at the administrative level, a level of government at which there is limited means of ensuring that decisions are made for good reasons and that they meet the underlying policy objectives of the statute, pursuant to a process that is fair and open to external public scrutiny. For political and practical reasons, it can be expected that discretionary delegation will continue as a feature of Canadian environmental law. The exercise of this discretion is shaped by a confluence of complex, and often competing, economic, political and social forces that merge in this regulatory arena. It is within this context that scientific  1  ^  Yellin, supra, Chapter Three fn 67, at pg. 1324.  Supra, Chapter One fn 42.  127 information enters the fold. With respect to science, this thesis has argued that scientific information and analysis is a key influence on the exercise of regulatory discretion: it informs it, it constrains it, but it also confounds it. From a legal perspective, what makes science so singularly fascinating is the perception that it is grounded in "fact". It brings with it the norms and imperatives of objectivity and rational evidence. In the context of an administrative process seeking to ground its decisions on "rationality", science can be deceptively seductive. Consequently, the process of environmental standard setting, as it currently operates, encourages the use of science as a sword or a shield; it becomes the instrument, not the source, of conflict, as astutely observed by one writer: To the extent that decisions derive from the exercise of discretion, they are neither predetermined nor immutable. They may be influenced by persuasion or the exercise of political power. Frequently, they can be reversed...But scientific and engineering knowledge is compelling on its own. It is not subject to change through the political process. Actions based on knowledge are, or seem to be, immutable. Both decision-makers and advocates therefore have strong incentives to claim that their actions and arguments are grounded in knowledge, not interest or values, and can often thereby muster support...The result is a special brand of politics in which the pretense of knowledge yields power.3 This acknowledgement is not meant in any way to denigrate scientific information. Rather, what is being said is that science is simply not a neutral factor in the environmental dialogue. If it was, surely a consensus on scientific information would be much more common than is typically the case. Yet one of the constant  3 Greenwood, supra, Chapter One fn 36, at pg. 252. [emphasis added].  128 themes evident in the various regulatory examples examined in this thesis, from alachlor to benzene to vinyl chloride, is that the battle lines are drawn around the science. In each case, the scientific evidence was at issue precisely because of how it was interpreted to achieve the objectives of vaguely worded statutes. What does this mean for the shape of environmental laws? Many commentators who have considered the relationship between environmental regulatory demands and science have observed that it raises implications about the role of various institutions: the legislature, the administration and the judiciary. A wide variety of solutions, all of which at some level address institutional roles, have been suggested to reform environmental decision making processes in order to account for scientific complexities and uncertainties. For example, some recommendations seek to increase the regulator's reliance on scientific expertise by, for example, the creation of expert panels4 or "science courts" to adjudicate scientific disputes.5 These reformers can be categorized as members of  Expert panels or committees have infiltrated the decision making processes in many cases of environmental regulation. For example, in Canada the main federal law dealing with the control of toxic substances contemplates the establishment of advisory committees, which presumably includes technical/scientific committees given the nature of the statutory scheme (Canadian Environmental Protection Act, R.S.C. 1985, c. 16 (4th Supp.), s. 5). In the United States this notion has been institutionalized at EPA through the creation of the Scientific Advisory Board (the "SAB"). The SAB was authorized by Congress in 1978 to provide the EPA administrator with independent, expert advice on a wide range of issues. For a fuller discussion of the SAB's role see, Graham, supra, Chapter One fn 36, Chapter Two therein. 4  "Science courts", proposed specifically to deal with factual technical issues, have not caught on but a lively academic debate ensued over them in the late 70's and early 80's in the U.S. For a comprehensive survey of the literature on science courts, see, for example, Yellin, supra, fn 1, at pg. 1307 fn 35 therein.  129 the separatist school of thought.6 Other reform suggestions which have been made can be classified as more integrative perspectives, or "non-separatist".7 These include reforms to increase education about environmental risks and to promote "scientific literacy".8 In his last term as EPA Administrator, William Reilly was determined to educate the American public about environmental risks using science as a foundation of understanding: As we gear up to deal with the environmental problems of the 1990s and beyond, I think the time has come to start taking aim before we open fire. In short, we have to find a better way of setting environmental priorities. And this is where sound science comes in. Sound science can help us establish priorities and allocate resources based on risk, to the extent that statutory mandates allow. Obviously there are a number of other important factors that go into shaping our priorities - public values and perceptions, economic constraints - but sound science is our most reliable compass in a turbulent sea of siren songs. Science can lend much-needed coherence, order and integrity to the often costly and controversial decisions that must be made.9 Still others have suggested adopting the goal of "zero pollution" as a way of avoiding the prickly risk determination questions entirely. In this thesis, a middle ground is adopted; it is one that acknowledges the "separateness" of science as an expert discipline but seeks to integrate it into the 6^Recall that the "separatist" school of thought refers to those who view science as "separate" from politics, law and ethics. See discussion, infra, at Chapter Three, Part C.  7^The "non-separatists" do not believe in separating science, supra, fn 6. 8^One writer, Dorothy Howell, has suggested that scientific illiteracy has disenfranchised society in the U.S. For a proposal on the reenfranchisement through the promotion of scientific literacy, see Dorothy J. Howell, Scientific Literacy and Environmental Policy: The Missing Prerequisite for Sound Decision Making, (New York: Quorum Books, 1992). 9 William K Reilly, EPA Administrator, "Aiming Before We Shoot: The Quiet Revolution in Environmental Policy", (Global Environment Change Report; Policy, Science and Industry News Worldwide, Cutter Information Corp., (Document #306), September 1990), at pg. 4.  130 participatory processes of environmental decision making. It has been argued that regulators need to openly address the assumptions, methodologies and gaps relating to the science behind what they are doing and why. In particular, there should be an express statutory obligation on environmental administrators to disclose and to explain, both before and after a regulatory decision, the scientific analysis used to support the exercise of their discretion. What might such a duty look like and what would it achieve? The discussion of the U.S. experience under the Clean Air Act helped to illustrate how a legal rule about the disclosure and explanation of science might operate. It was seen that when the EPA proposes a regulation under the CAA, it must provide a statement of basis and purpose, including a summary of the factual data on which the rule is based, and the methodology used in obtaining and analyzing the data.1° Also, pursuant to the statutorily entrenched notice and comment procedures of the CAA, this regulatory explanation is subject to the scrutiny of interested parties and the courts. This administrative framework has had some successes and some failures.11 The good aspects of this U.S. experience can be 10 For a slightly more comprehensive version of this requirement, see the proposed U.S. Regulatoty Reform Act of 1981, Senate bill 1080. This Act was never passed but it is a good example of statutory language to address scientific/technical disclosure. Section 3(c)(1)(E) provides that a notice of proposed rulemaking shall include: "a description of the data, methodology, reports, studies, or other information upon which the agency plans to substantially rely in the rule making, and of the purposes for which the agency plans to rely on such information, including an identification of the authors or sources of such information." (Ibid). ii For an amusing parody on the regulatory hurdles and the process excesses facing EPA when it regulates a potentially hazardous substance, See Douglas M. Costle (ex-Administrator of the EPA), "Brave New Chemical: The Future Regulator), History of Phlogiston", 33 Administrative Law Review 195 (1981).  131 tailored to the Canadian regulatory system, while the bad aspects can be left behind. In Canada, the "due process" parts of the U.S. approach, the disclosure and explanation of scientific data through notice and comment procedures, are amenable to the trends in increased public participation in environmental decision making generally.12 Similarly, the institutional excesses of the U.S. system do not exist in Canada and are not likely to be duplicated simply by virtue of making improvements to the regulatory process. The generation of an evidentiary record on which to test regulatory discretion does not by corollary mean an extensive accumulation of information to fend off judicial activism. In fact, there is a good case to be made for the argument that by improving the process up front there will less dissatisfaction with the process and thus less need to involve the courts in dispute settlement. A legal duty imposed on environmental regulators to disclose and to explain their scientific analysis will significantly improve the process of environmental standard setting by making it more open and legitimate. A statutory duty, rather than a discretionary policy dealing with process reform which really amounts to selfregulation, is needed. It will impose greater accountability on decision makers at the administrative level, where it should be. Any requirement to disclose information and to give the reasons for a decision will force the decision maker to be more thoughtful in his analysis. This will result in several benefits: a more careful reflection of the administrative task; more diligence in identifying and specifying both statutory  12  Evidence of informal efforts to involve stakeholders in environmental regulation and of reform proposals to open up environmental decision through notice and comment procedures were noted above in the discussions in Chapters One and Two, infra.  132 objectives and relevant information; greater care in applying the rationale to individual circumstances; and the direction of arguments by interested parties more efficiently. Disclosure and explanation will also improve the integrity of the process by avoiding the temptation to use science to veil politics or economics. This will thus remove some of the ambiguity or uncertainty as to why a decision was made. In addition, the requirement to explain the scientific analysis, will help to establish a common and accessible information base about the science used in environmental decisions. This will help harness and centralize a valuable pool of knowledge. A spin off effect of this may also be to provide a better focus for the tailoring of scientific research and development. This may also lead to a consideration of institutional support systems, such as the Science Advisory Board. All of these benefits are considered important contributions to improving the democratic legitimacy of the environmental regulatory system in Canada. The concept of a specific legal duty to give the supporting reasons for a regulatory decision is neither revolutionary nor surprising coming from a lawyer. In fact, the idea is not entirely foreign to the operation of some aspects of British Columbia's environmental laws. 13 The argument in this thesis is essentially an effort to improve the process of environmental decision making in a way that is palatable within the current framework of Canadian environmental law. It is also limited for the purposes of this thesis to a singular component, science, of a multi-faceted  13 For example, section 4 of the B.C. Environment Management Act, supra, Chapter One fn 22, empowers the Minister of the Environment to issue environmental protection orders which must "state the reasons for making the order".  133 decision process. It is a mere tinkering at the edges which emanates from the view that incrementalism14 is the most realistic approach to achieving change in Canadian environmental policy and law. Furthermore, it must be appreciated that improvements in the legal system are only part of the formula for overall "sound governance"15 in environmental protection -- political willpower, better understanding of risks and increased resources are also of key importance. Of course, any argument seeking process reform is in danger of the usual criticisms about "processitis". It is said to add to an already overburdened administrative process thus increasing public costs and furthering delays. The view presented here is that although matters of efficiency and cost are important, the benefits from the duty to disclose and explain the science supporting the regulatory decision outweigh these concerns. In the final analysis, the essence of environmental regulatory decisions requires a clear understanding of what balance, if any, is being struck between competing values. This recognition is important because it has enormous implications to determinations about how society's limited resources will be used to achieve environmental protection. A legal rule about the disclosure and explication of scientific information seeks to squarely address the government's role in dealing with  14  "Incrementalism" is used here to mean the idea that change is best effected by marginal adjustments to existing processes. 15  This is a phrase used by Christopher F. Edley in his book, Administrative Law: Rethinking Judicial Control of the Bureaucracy, supra, Chapter Two fn 11. The term encompasses what are considered to be the goals of regulatory government: "good institutions, decent processes, and just outcomes" and also means the prevention of mistakes and the increased likelihood "that officials will pursue the commonweal efficiently and effectively." (lbid, at pgs. x-xi).  134 these conflicts by making more transparent the risk analysis aspect of environmental regulation. If Canadian environmental protection law is to obtain its objectives and to be respected, the inevitably tough choices that the government must make must be made through open and legitimate processes. This is, after all, what we expect from our democracy.  135 BIBLIOGRAPHY BOOKS AND REPORTS  Berman, Morris. The Reenchantment of the World. (Ithaca: Cornell University Press, Bantam Edition, 1984; original publication, 1981). Boardman, Robert, ed. Canadian Environmental Policy: Ecosystems, Politics and Process. (Toronto: Oxford University Press, 1992). Bowers, James R. Regulating the Regulators: An Introduction to the Legislative Oversight of Administrative Rulemaking. (New York: Praeger Publishers, 1990). British Columbia Ministry of Environment, Waste Management Branch. "Criteria for Managing Contaminated Sites in B.C.". Draft #6 dated November 21, 1989. British Columbia Ministry of Environment, Land and Parks. New Approaches to Environmental Protection in British Columbia: A Legislation Discussion Paper (April 27, 1992).  Brown-John, Lloyd C. Canadian Regulatoty Agencies. (Toronto: Butterworths & Co. Ltd., 1981). Brunk, Conrad G., Haworth, Lawrence, and Lee, Brenda. Value Assumptions in Risk Assessment: A Case Study of the Alachlor Controversy. (Waterloo: Wilfrid Laurier University Press, 1991). Burton, I., Fowle, C.D., and McCullough, R.S., eds. Living With Risk: Environmental Risk Management in Canada. (Toronto: The Institute for Environmental Studies, 1982). Canadian Environmental Protection Act, Report for the Period April 1991 to March 1992. (Ottawa: Minister of Supply and Services, Canada, 1992).  Chilton, Kenneth and Warren, Melinda, eds. Environmental Protection: Regulating for Results. (Boulder: Westview Press, Inc., 1991). Doern, G. Bruce. The Peripheral Nature of Scientific Controversy in Federal Policy Formulation. (Ottawa: Science Council of Canada, 1981).  136  Doern, G. Bruce. Getting It Green: Case Studies in Canadian Environmental Regulation. (Ottawa: Renouf Publishing Company Limited for the C. D. Howe Institute, 1990). Doern, G. Bruce. Canadian Environmental Policy: Why Process is Almost Everything, Commentary #19 for the C.D. Howe Institute. (Ottawa: Renouf Publishing Company Limited, July 1990). Doem, G. Bruce. Green Diplomacy: How Environmental Policy Decisions Are Made. (Ottawa: Renouf Publishing Company Limited for the C. D. Howe Institute, 1993). Dussault, Rene and Borgeat, Louis. Administrative Law: A Treatise, 2nd Edition. (Toronto: Carswell, 1985; Translated by Murray Rankin). Edley, Christopher, Jr. Administrative Law: Rethinking Judicial Control of the Bureaucracy. (New Haven: Yale University Press, 1990). Galligan, Denis J. Discretionary Powers: A Legal Study of Official Discretion. (afford: Clarendon Press, 1986). Gellhorn, Ernest and Levin, Ronald M. Administrative Law and Process In a Nutshell, Third Edition. (St. Paul: West Publishing Co., 1990). Government of Canada. Canada's Green Plan for a Healthy Environment. (Ottawa: Minister of Supply and Services Canada, 1990). Government of Canada, Treasury Board Secretariat, Administrative Policy Branch, Regulatory Affairs. How Regulators Regulate: A Guide to Regulatory Processes in Canada. (Ottawa: Minister of Supply and Services, 1992). Grad, Frank P. Treatise on Environmental Law. (New York: Matthew Bender & Co., Inc., 1973; 1992 Supplement). Graham, John D., ed. Harnessing Science for Environmental Regulation. (New York: Praeger Publishers, 1991). Greenwood, Ted. Knowledge and Discretion in Government Regulation. (New York: Praeger Publishers, 1984). Hoberg, George. Pluralism By Design: Environmental Policy and the American Regulatory State. (New York: Praeger Publishers, 1992). Hogg, Peter W. Constitutional Law of Canada, Second Edition. (Toronto: The Carswell Company Limited, 1985).  137 Holland, Denys C. and McGowan, John P. Delegated Legislation in Canada. (Toronto: The Carswell Co. Ltd., 1989). Howell, Dorothy J. Scientific Literacy and Environmental Policy: The Missing Prerequisite for Sound Decision Making. (New York: Quorum Books, 1992). Hughes, E., Lucas, A. and Tilleman, W., eds. Environmental Law and Policy, Volume I, Preliminaiy Edition. (Toronto: Emond Montgomery Publications Limited, 1992). Law Reform Commission of Canada. Report on Independent Administrative Agencies: A Framework for Decision Making (Report 26). (Ottawa: Law Reform Commission of Canada, 1985). MacDonald, Doug. The Politics of Pollution: Why Canadians Are Failing Their Environment. (Toronto: McClelland & Stewart Inc., 1991). Melnick, Shep R. Regulation and the Courts: The Case of the Clean Air Act. (Washington: The Brookings Institution, 1983). Ontario Ministry of the Environment. Report of the Task Force on the Ontario Environmental Bill of Rights (July, 1992). Paehlke, Robert and Torgerson, Douglas, eds. Managing Leviathan: Environmental Politics and the Administrative State. (Peterborough: Broadview Press, 1990). Saxe, Dianne. Environmental Offences: Corporate Responsibility and Executive Liability. (Aurora: Canada Law Book, 1990). Schrecker, Ted F. Political Economy of Environmental Hazards. Protection of Life Series, A Study Paper prepared for the Law Reform Commission of Canada. (Ottawa: Minister of Supply and Services, 1984).  Science Council of Canada. Regulating the Regulators: Science, Values and Decisions. (Ottawa: Minister of Supply and Services, 1982). Science Council of Canada. Annual Report for 1988-89. Shrader-Frechette, K.S. Risk Analysis and Scientific Method: Methodological and Ethical Problems with Evaluating Societal Hazards. (Dordrecht: D. Reidel Publishing Company, 1985). Shrader-Frechette, K.S. Risk and Rationality: Philosophical Foundations for Populist Reforms. (Berkeley: University of California Press, 1991).  138 Stanbury, William T. "Reforming the Federal Regulatoiy Process in Canada, 19711992." Report to the Sub-Committee on Regulations and Competitiveness of the  Standing Committee on Finance, published as Appendix SREC-2 in the Minutes of Proceedings and Evidence of the Sub-Committee, House of Commons, Third Session of the Thirty-Fourth Parliament 1991-1992.  Stanbury, William T. Business-Government Relations in Canada: Influencing Public Policy. (Scarborough: Nelson Canada, A Division of Thomson Canada Limited, 1993). Tardi, Gregory. The Legal Framework of Government: A Canadian Guide. (Aurora: Canada Law Book Inc., 1992). Tingley, Donna, ed. Into the Future: Environmental Law and Policy for the 1990's. (Edmonton: Environmental Law Centre, proceedings of a conference, October 1989). Webb, K. Pollution Control in Canada: The Regulatoly Approach in the 1980s. Administrative Law Series, A Study Paper prepared for the Law Reform Commission of Canada. (Ottawa: Minister of Supply and Services, 1988). Whyte, Anne V. and Burton, Ian, eds. Environmental Risk Assessment, SCOPE 15. (New York: John Wiley & Sons, 1980). Published on behalf of SCOPE = The Science Committee on Problems of the Environment. ARTICLES  Anderson, Elizabeth L. "Scientific Developments in Risk Assessment: Legal Implications". 14 Columbia Journal of Environmental Law 411 (1989). Bazelon, David. "Risk and Responsibility". 205 Science 277 (1979). Bazelon, David. "Science and Uncertainty: A Jurist's View". 5 Harvard Environmental Law Review 209 (1981). Bryden, Philip. "Canadian Administrative Law: Where We've Been". 16 Queen's Law Journal 7 (1992). Carter, Stephen. "Separatism and Skepticism". 92 The Yale Law Journal 1334 (1983). Chard, III, Leslie F. "The 1990 Clean Air Act Amendments: Section 112 Comes of Age". 59 University of Cincinnati Law Review 1253 (1991).  139 Clairborne, Margaret. "The New Air Toxics Program". 7 Natural Resources and Environment 21 (Fall 1992). Clarke, David. "Looking at Risk". 8 The Environmental Forum 12 (March/April 1991). Costle, Douglas M. "Brave New Chemical: The Future Regulator), Histoty of Phlogiston". 33 Administrative Law Review 195 (1981). Daggett, Christopher J., Hazen, Robert E., and Shaw, Judith Auer. 'Advancing Environmental Protection Through Risk Assessment". 14 Columbia Journal of Environmental Law 315 (1989). "Dioxins: time for tough limits". The Vancouver Sun, Letters in the OP/ED page, pg. A13.  Elliott, Donald E. "Environmental Law at a Crossroad". 20 Northern Kentucky Law Review 1 (1992). Elgie, Stewart A.G. "Environmental Groups and the Courts: 1970-1992" in Thompson, G., McConnell, M.L. and Huestis, L., eds. Environmental Law and Business in Canada. (Toronto: Canada Law Book, 1993). Farber, Daniel A. "Risk Regulation in Perspective: Reserve Mining Revisited." 21 Environmental Law 1321 (1991). Garrett, Theodore L. and Winner, Sonya. 'A Clean Air Act Primer: Part l". 22 Environmental Law Reporter 10159 (March 1992); 'A Clean Air Act Primer: Part II". 22 Environmental Law Reporter 10235 (April 1992); and 'A Clean Air Act Primer: Part III". 22 Environmental Law Reporter 10301 (May 1992). Goldstein, Bernard D. "Risk Assessment and the Interface Between Science and Law". 14 Columbia Journal of Environmental Law 343 (1989). Graham, John 0. "The Failure of Agency-Forcing: The Regulation of Airborne Carcinogens under Section 112 of the Clean Air Act". 1985:1 Duke Law Journal 100 (1985). Grey, J. H. "Discretion in Administrative Law". 17 Osgoode Hall Law Journal 107 (1979). Hornstein, Donald T. "Reclaiming Environmental Law: A Normative Critique of Comparative Risk Analysis". 92 Columbia Law Review 562 (1992).  140 Jasanoff, Sheila. "Science and the Limits of Administrative Rule-Making: Lessons from the OSHA Cancer Policy". 20 Osgoode Hall Law Journal 536 (1982). Latin, Howard. "The Significance of Toxic Health Risks: An Essay on Legal Decisionmaking". 10 Ecology Law Quarterly 339 (1982). Latin, Howard. "Good Science, Bad Regulation, and Toxic Risk Assessment". 5 Yale Journal on Regulation 89 (1988). Latin, Howard. "Regulator), Failure, Administrative Incentives, and the New Clean Air Act". 21 Environmental Law 1647 (1991). Lave, Lester B. "Risk Assessment and Regulator), Priorities". 14 Columbia Journal of Environmental Law 307 (1989). Leape, James P. "Quantitative Risk Assessment in Regulation of Environmental Carcinogens." 4 Harvard Environmental Law Review 87 (1980). Lyndon, Mary L. "Risk Assessment, Risk Communication and Legitimacy: An Introduction to the Symposium". 14 Columbia Journal of Environmental Law 289 (1989). McGarity, Thomas 0. "The Internal Structure of EPA Rulemaking." 54 Law and Contemporary Problems 57 (1991). McQuaid, Janet L. "Risk Assessment of Hazardous Air Pollutants under the EPA's Final Benzene Rules and the Clean Air Act Amendments of 1990". 70 Texas Law Review 427 (1991). Melnick, R. Shep. 'Administrative Law and Bureaucratic Reality". 44 Administrative Law Review 245 (1992). Paustenbach, Denis J. "Health Risk Assessments: Opportunities and Pitfalls". 14 Columbia Journal of Environmental Law 379 (1989). Ruckelshaus, William D. "Science, Risk and Public Policy". 221 Science 1026 (1983). Russell, Milton and Gruber, Michael. "Risk Assessment in Environmental PolicyMaking". 236 Science 286 (1987). Sagoff, Mark. "Ethics, Ecology, and the Environment: Integrating Science and Law". 56 Tennessee Law Review 77 (1988). Slovic, Paul. "Perception of Risk". 236 Science 280 (1987).  141 Tuohy, Carolyn J. "Regulation and Scientific Complexity: Decision Rules and Processes in the Occupational Health Arena". 20 Osgoode Hall Law Journal 567 (1982). Wilson, Richard and Crouch, E.A.C. "Risk Assessment and Comparisons: An Introduction". 236 Science 267 (1987). Yellin, Joel. "Science, Technology and Administrative Government: Institutional Designs for Environmental Decisionmaking". 92 The Yale Law Journal 1300 (1983). CASES  Industrial Union Department, AFL-CIO v. American Petroleum Institute et. al. 448 US 607 1980 (USSC).  Kennecott Copper Corporation v. EPA 462 F.2d 846 (D.C. Cir. 1972). Natural Resources Defense Council v. EPA 824 F.2d 1146 (D.C. Cir. 1987). R. v. Consolidated Mayburn Mines Limited (1992), 73 C.C.C. (3d) 268 (Ontario Prov. Div.).  Re Reese et. al. v. The Queen in right of Alberta (1992), 87 D.L.R. (4th) 1 (Alta Q.B.). Sierra Club v. Ruckelshaus 602 F.Supp. 892 (N.D. Cal. 1984). STATUTES  Administrative Procedure Act, 5 U.S.C. s.500 et. seq. Canadian Environmental Protection Act, R.S.C. 1985, c. 16 (4th Supp.). Clean Air Act, 42 U.S.C. s.7410 et. seq. Commissioner on Resources and Environment Act, S.B.C. 1992, c. 34. Environment Management Act, S.B.C. 1981, c. 41. Environmental Protection and Enhancement Act, S.A. 23, 1992, C. E-13.3. Waste Management Act, S.B.C. 1981, c. 41.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items