* (c) 2003, Vern R. Walker, Professor of Law, Hofstra University; Ph.D., University of Notre Dame, 1975; J.D., Yale University, 1980. The author is grateful for the financial support provided by a research grant from Hofstra University.
1 For examples of risk findings concerning food safety, see Section III.A infra.
2 For an example involving the precautionary principle of European law, see Section III.B infra. For a general discussion of triggering precautionary measures under various treaties, see generally James E. Hickey, Jr. & Vern R. Walker, Refining the Precautionary Principle in International Environmental Law, 14 Va. Envtl. L. J. 423 (1995).
3 See, e.g., discussion infra Section III.C.
4 For examples of vague formulations in use today, see infra Section III.
5 See, e.g., The Random House Dictionary of the English Language 1660 (2d ed. 1987) (defining “risk” as “exposure to the chance of injury or loss”); Webster’s Ninth New Collegiate Dictionary 1018 (1991) (defining “risk” as “possibility of loss or injury”); Oxford English Dictionary Online (defining “risk” as “exposure to mischance or peril”), at http://www.oed.com (last visited Apr. 8, 2003).
6 For discussions of modal logic, see, e.g., Clarence I. Lewis & Cooper H. Langford, Symbolic Logic 153–66, 199–234 (2d ed. 1959); Gerald J. Massey, Understanding Symbolic Logic 163–222 (1970); Mark Sainsbury, Logical Forms: An Introduction to Philosophical Logic 13–19, 220–90 (1991).
7 For a more detailed treatment of these five types, see Vern R. Walker, The Siren Songs of Science: Toward a Taxonomy of Scientific Uncertainty for Decisionmakers, 23 Conn. L. Rev. 567, 574–618 (1991) [hereinafter Walker, Siren Songs].
8 For discussions of types of measurement variables, see, e.g., Edwin E. Ghiselli, et al., Measurement Theory for the Behavioral Sciences (1981); Herman J. Loether & Donald G. McTavish, Descriptive and Inferential Statistics: An Introduction 16–24 (4th ed. 1993).
9 See discussion of causal uncertainty, infra Section II.A.5.
10 See, e.g., Edward G. Carmines & Richard A. Zeller, Reliability and Validity Assessment 11–13 (John L. Sullivan ed., 1979); Ghiselli et al., supra note 8, at 184, 191.
11 See, e.g., Theodore Peters, Jr. & James O. Westgard, Evaluation of Methods, in Textbook of Clinical Chemistry 410, 412 (Norbert W. Tietz ed., 1986).
12 See, e.g., Carmines & Zeller, supra note 10, at 12–13; Ghiselli et al., supra note 8, at 266.
13 Peters & Westgard, supra note 11, at 412; David H. Kaye & David A. Freedman, Reference Guide on Statistics, in Federal Judicial Center, Reference Manual on Scientific Evidence 83, 103–04 (2d ed. 2000).
14 See, e.g., Peters & Westgard, supra note 11, at 412; Kaye & Freedman, supra note 13, at 104.
15 See, e.g., William L. Hays, Statistics 224–25 (5th ed. 1994).
16 See, e.g., Kaye & Freedman, supra note 13, at 373–93.
17 See, e.g., Carl F. Cranor, Regulating Toxic Substances: A Philosophy of Science and the Law 29–48 (1993); Neil B. Cohen, Confidence in Probability: Burdens of Persuasion in a World of Imperfect Knowledge, 60 N.Y.U. L. Rev. 385, 409–17 (1985); Michael Cowles & Caroline Davis, On the Origins of the .05 Level of Statistical Significance, 37 Am. Psychologist 553 (1982); Hays, supra note 15, at 267–82, 302–03.
18 Walker, Siren Songs, supra note 7, at 598–608 (linear regression model); Vern R. Walker, The Concept of Baseline Risk in Tort Litigation, 80 KY. L.J. 631, 651–62 (1992) (relative risk model); see Michael D. Green et al., Reference Guide on Epidemiology, in Federal Judicial Center, supra note 13, at 348–54 (relative risk and odds ratio in epidemiologic studies); David L. Rubinfeld, Reference Guide on Multiple Regression, in Federal Judicial Center, supra note 13, at 179–227.
19 See, e.g., Green et al., supra note 18, at 348–54.
20 See discussion supra Section II.A.1.
21 See, e.g., James A. Davis, The Logic of Causal Order 25–27 (1985).
22 See generally National Research Council, Risk Assessment in the Federal Government: Managing the Process (1983) [hereinafter NRC (1983)]; National Research Council, Science and Judgment in Risk Assessment (1994) [hereinafter NRC (1994)]; 1 The Presidential/Congressional Commission on Risk Assessment and Risk Management, Framework for Environmental Health Risk Management (1997).
23 This notion of risk assessment is now incorporated into an international treaty. See Agreement on the Application of Sanitary and Phytosanitary Measures, Apr. 15, 1994, WTO Agreement, Annex 1A, Legal Instruments--Results of the Uruguay Round vol. 1 (1994), available at http://docsonline.wto.org [hereinafter SPS Agreement].
24 See, e.g., NRC (1983), supra note 22, at 19–27; NRC (1994), supra note 22, at 26, 57–66; Commission Directive 93/67, 1993 O.J. (L 227) 9 (“laying down the principles for assessment of risks to man and the environment of substances notified in accordance with Council Directive 67/548/EEC”).
25 NRC (1983), supra note 22, at 20, 27–28; NRC (1994), supra note 22, at 26–27, 43–55; Commission Directive 93/67, supra note 24.
26 See, e.g., Vern R. Walker, Keeping the WTO from Becoming the “World Trans-science Organization”: Scientific Uncertainty, Science Policy, and Factfinding in the Growth Hormones Dispute, 31 Cornell Int’l L.J. 251, 263–66 (1998).
27 One option for risk assessment scientists, of course, is merely to communicate all of the residual uncertainties to others, and to leave the ultimate task of factfinding about risk to those others. I express no opinion here about the advantages and disadvantages of such an approach. The analysis here applies to conclusions or findings about the existence or non-existence of a triggering risk, regardless of whether scientists or non-scientists make those ultimate findings. In addition, the analysis in Section II of this article applies even to findings about causation, which provide the evidentiary grounds for ultimate findings about triggering risks.
28 See, e.g., EPA, Guidelines for Carcinogen Risk Assessment, 51 Fed. Reg. 33,992 (1986); EPA, Proposed Guidelines for Carcinogen Risk Assessment, 61 Fed. Reg. 17,960 (1996); NRC (1983), supra note 22; NRC (1994), supra note 22; Walker, supra note 26, at 258–63.
29 Science policies are also said to be “inference guidelines” or to result in “default assumptions.” See NRC (1983), supra note 22, at 28–37; NRC (1994), supra note 22, at 27, 85–90; EPA, Proposed Guidelines for Carcinogen Risk Assessment, supra note 28, at 17,964.
30 See, e.g., EPA, Memorandum: EPA Risk Characterization Program (1995); EPA, Policy for Risk Characterization (1995); EPA, Science Policy Council, Guidance for Risk Characterization (1995); Walker, supra note 26, at 260–62.
31 For a more complete review of food safety law from the standpoint of taking precautions, see Vern R. Walker, Some Dangers of Taking Precautions Without Adopting the Precautionary Principle: A Critique of Food Safety Regulation in the United States, 31 ENVTL. L. REP. 10040 (2001).
32 21 U.S.C. � 342 (2000).
33 The word “marketed” here refers to any acts that are prohibited under 21 U.S.C. � 331.
34 Id. � 342(a)(1).
35 Id. �� 342(a)(2)(A), 346.
36 See, e.g., United States v. 449 Cases Containing “Tomato Paste,” 212 F.2d 567, 569–73 (2d Cir. 1954) (surveying various factual predicates that serve as regulatory triggers under the FFDCA); United States v. 24 Cases, More or Less, 87 F. Supp. 826, 828 (D. Me. 1949) (holding, in a libel in rem proceeding relating to cases of canned herring roe, that “in order for a product to be subject to condemnation as unfit for food, on account of it’s [sic] tough and rubbery consistency, the product must be proved to be so tough and rubbery that the average, normal person, under ordinary conditions, would not chew and swallow it”).
37 21 U.S.C. � 342(a)(1); 21 C.F.R. �� 109.3(c), (d) (2000). Cf. Peter Barton Hutt & Richard A. Merrill, Food and Drug Law, Cases and Materials 948–50 (2d ed. 1991) (carcinogenic natural foods and ingredients).
38 21 U.S.C. �� 321(q), 342(a)(2)(B), 346a(a)(1). The statute declares any pesticide chemical residue in or on a food to be “unsafe” unless its use conforms to a regulatory tolerance or an exemption. Id. �� 342(a)(2)(B), 346a(a).
39 21 U.S.C. �� 321(s), 342(a)(2)(C)(i), 348(a). The statute declares any food additive (other than a food contact substance) to be “unsafe” unless its use conforms to an exemption or a use regulation. Id. �� 342(a)(2)(C)(i), 348(a).
A precondition for being a “food additive” is that the substance not be “generally recognized, among experts qualified by scientific training and experience to evaluate its safety, as having been adequately shown . . . to be safe under the conditions of its intended use”—that is, that the substance is not “generally recognized as safe” (GRAS). Id. � 321(s). Being GRAS is therefore based on “scientific procedures”—unless the substance was used in food prior to 1958, in which case an alternative basis is “experience based on common use in food.” Id. If a substance is GRAS, then it is not a food additive and does not trigger the pre-marketing approval required for food additives. 21 C.F.R. � 170.30(g) (2000).
40 21 U.S.C. �� 321(t), 342(c), 379e(a). The statute declares any color additive to be “unsafe” unless its use conforms to a use regulation or an exemption. Id. �� 342(c), 379e(a).
41 Id. �� 321(v), 342(a)(2)(C)(ii), 360b(a). The statute declares any new animal drug to be “unsafe” unless its use conforms to an approved application. Id. �� 342(a)(2)(C)(ii), 360b(a)(1). Excluded from this category are drugs “generally recognized . . . as safe and effective” by qualified experts and other “grandfathered” drugs. See id. � 321(v).
42 Other pre-marketing triggers also exist, such as the inspection-and-condemnation provisions for meat and meat food products, poultry products, and egg products. Id. �� 601 ff, 610(c)(2), 451 ff, 458(a)(2)(B), 1031 ff, 1037(b)(2). For such products, inspection measures are triggered categorically for the relevant animals or products, or for establishments that process them. See, e.g., id. �� 455, 603, 604, 608, 1034. If the required inspection results in an administrative determination that a particular product is “adulterated,” then that product can be condemned and destroyed. See, e.g., id. �� 604, 672, 673, 455(c), 467a, 467b, 1034(c), 1048, 1049. The relevant agency also has the authority to prosecute judicial proceedings, and the court may condemn and seize any covered product that is found to be adulterated and capable of being used as human food. See, e.g., id. �� 610(c)(1), 673(a)(1)(B), 458(a)(2)(A), 467b(a)(1)(B). “Adulteration” is generally defined as bearing or containing any poisonous or deleterious substance that may render the product either injurious to health or unfit for human food. See, e.g., id. �� 601(m), 453(g), 1033(a).
43 21 C.F.R. � 70.3(i) (2000).
44 Id. � 170.3(i).
45 See Final OECD-U.S. Summary Precaution Annex, Precaution in U.S. Food Safety Decisionmaking: Annex II to the U.S. National Food Safety System Paper (Mar. 2000), �� B.1, B.3 & � 82 [hereinafter Precaution Annex].
46 The normally applied safety factor for a color additive is “1/100th of the maximum no-effect level for the most susceptible experimental animals tested.” 21 C.F.R. � 70.40 (2000). The tolerance for a food additive usually will not exceed “1/100th of the maximum amount demonstrated to be without harm to experimental animals.” Id. � 170.22. The requirement to consider “safety factors” for food and color additives is statutory. 21 U.S.C. �� 348(c)(5)(C), 379e(b)(5)(A)(iii). For similar use of a 100-fold safety factor by EPA in setting tolerance levels for pesticide residues, see H.R. Rep. No. 104–669, pt. 2, at 41 (1996).
47 21 U.S.C. � 346a(b)(2)(C). EPA is authorized to “use a different margin of safety for the pesticide chemical residue only if, on the basis of reliable data, such margin will be safe for infants and children.” Id.
48 Compare note 46 supra with 21 U.S.C. � 346a(b)(2)(C); see H.R. Rep. No. 104–669, pt. 2, at 41, 43 (1996); Frank B. Cross, The Consequences of Consensus: Dangerous Compromises of the Food Quality Protection Act, 75 Wash. U.L.Q. 1155, 1166–67 (1997).
49 Precaution Annex, supra note 45, at �� 87–92.
50 21 U.S.C. � 348(c)(3)(A).
51 Id. � 379e(b)(5)(B).
52 Id. � 360b(d)(1)(I).
53 See Precaution Annex, supra note 45, �� 87, 89. Because a food ingredient that is GRAS does not fall within the definition of “food additive,” it is not technically subject to the Delaney Clause. Hutt & Merrill, supra note 37, at 338. Moreover, the Delaney Clauses are not triggered unless the agency makes several determinations that normally involve science policies. See, e.g., 21 C.F.R. � 70.50 (2000) (color additives).
54 Congress can, of course, exempt food products from the Delaney Clause prohibitions, as it did in the case of saccharin. See Hutt & Merrill, supra note 37, at 922–28.
55 See FDA, Policy for Regulating Carcinogenic Chemicals in Food and Color Additives, 47 Fed. Reg. 14,464 (Food & Drug Admin., Advance Notice of Proposed Rulemaking, Apr. 2, 1982); Scott v. FDA, 728 F.2d 322, 325 (6th Cir. 1984) (upholding FDA’s determination that the color additive as a whole did not cause cancer in test animals, although the p-toluidine present in minute quantities is carcinogenic when tested separately). Between the Scott decision and 1990, FDA approved more than thirty color additives and indirect food additives containing a variety of trace carcinogenic constituents. Hutt & Merrill, supra note 37, at 922.
56 Pub. Citizen v. Young, 831 F.2d 1108, 1113 (D.C. Cir. 1987), cert. denied sub nom. Cosmetic, Toiletry and Fragrance Ass’n v. Pub. Citizen, 485 U.S. 1006 (1988) (rejecting FDA’s de minimis re-interpretation of the color additives Delaney Clause); Les v. Reilly, 968 F.2d 985, 990 (9th Cir. 1992), cert. denied sub nom. Nat’l Agric. Chems. Ass’n v. Les, 507 U.S. 950 (1993) (rejecting EPA’s de minimis re-interpretation of the food additives Delaney Clause, in a case involving carcinogenic pesticides that concentrated during processing and therefore became food additives in processed food under the FFDCA as it was then written).
57 Les v. Reilly, supra note 56.
58 See Food Quality Protection Act of 1996, Pub. L. no. 104–170, 110 Stat. 1489, 1514–35, � 405 [hereinafter FQPA]; 21 U.S.C. � 346a(b)(2)(A)(ii); H.R. Rep. No. 104–669, pt. 1 at 36, pt. 2 at 40 (1996); Cross, supra note 48, at 1160–62. Under the prior law, pesticide residues that concentrated in food processing triggered the regulatory provisions for food additives, including the food additive Delaney Clause; see Les v. Reilly, supra note 56.
59 See supra text accompanying notes 43–44 ; Cross, supra note 48, at 1162–63.
60 See H.R. Rep. No. 104–669, pt. 2, at 41 (1996), which states:
In the case of a nonthreshold effect which can be assessed through quantitative risk assessment, such as a cancer effect, the Committee expects, based on its understanding of current EPA practice, that a tolerance will be considered to provide a “reasonable certainty of no harm” if any increase in lifetime risk, based on quantitative risk assessment using conservative assumptions, will be no greater than “negligible.” It is the Committee’s understanding that, under current EPA practice, . . . EPA interprets a negligible risk to be a one-in-a-million lifetime risk. The Committee expects the Administrator to continue to follow this interpretation.
Id. See Cross, supra note 48, at 1164–65. This quantitative level of protection of one in one million is precisely the risk that FDA had announced to be negligible or de minimis. See e.g., Hutt & Merrill, supra note 37, at 899–904; Pub. Citizen v. Bowen, 833 F.2d 364, 365 (D.C. Cir. 1987).
61 Case T-13/99, Pfizer Animal Health SA/NV v. Council, 2002 E.C.R. II __ (2002).
62 Case T-70/99, Alpharma, Inc. v. Council, 2002 E.C.R. II __ (2002).
63 For discussions of that evolving law, see, e.g., Theofanis Christoforou, The Origins, Content and Role of the Precautionary Principle in European Community Law, in Le principe de pr�caution: Aspects de droit international et communautaire 205–30 (Charles Leben & Joe Verhoeven eds., 2002); Theofanis Christoforou, The Precautionary Principle, Risk Assessment, and the Comparative Role of Science in the European Community and the United States Legal Systems, in Green Giants? Environmental Policy of the United States and the European Union (N. Vig and M. Faure eds.) (forthcoming 2003).
64 Council Regulation 2821/98, 1998 O.J. (L 351).
65 Alpharma, 2002 E.C.R. II __, �� 3–6, 17–20, 83, 93–95; Pfizer, 2002 E.C.R. II __, �� 3–6, 17–19, 91, 101–03.
66 Treaty Establishing the European Community, art. 37, Nov 10, 1997 O.J. (C 340) 3 (1997) [hereinafter EC Treaty]; Alpharma, 2002 E.C.R. II __, � 132; Pfizer, 2002 E.C.R. II __, � 111.
67 Council Directive 70/524/EEC, art. 3, 1970 O.J. (L 270).
68 Pfizer, 2002 E.C.R. II __, �� 91, 101–03, 112.
69 Alpharma, 2002 E.C.R. II __, �� 83, 93–95.
70 Pfizer, 2002 E.C.R. II __, � 126.
71 Id. � 127.
72 Id. � 112; Council Directive 70/524/EEC, supra note 67, art. 11(3).
73 Pfizer, 2002 E.C.R. II __, � 112.
74 Id. � 401; see id. �� 315–22, 359–61, 393.
75 See Council Directive 70/524/EEC, supra note 67, art. 11(1), (3).
76 Pfizer, 2002 E.C.R. II __, � 138.
77 Id. �� 312–24, 342–401.
78 Id. �� 147, 165.
79 Id. � 401.
80 Id. � 143.
81 Pfizer, 2002 E.C.R. II __, � 140.
82 Id. �� 139, 141, 146, 360, 379, 384–87, 388.
83 Id. �� 144, 161, 168, 323, 389, 393. The following finding by the Court suggests a large number of focal points for non-scientific decisions:
[T]he Court finds that the Community institutions did not exceed the bounds of the discretion conferred on them by the Treaty when they took the view that the various experiments and observations referred to in . . . the contested regulation were not mere conjecture but amounted to sufficiently reliable and cogent scientific evidence for them to conclude that there was a proper scientific basis for a possible link between the use of virginiamycin as an additive in feedingstuffs and the development of streptogramin resistance in humans.
Id. � 389.
84 Id. �� 371–76, 379–80, 382–83, 388.
85 Pfizer, 2002 E.C.R. II __, �� 354, 370.
86 Id. �� 160–62, 382–83.
87 Id. �� 382–83.
88 Communication from the Commission on the Precautionary Principle, COM(2000) 1, 2–4 [hereinafter Communication].
89 Id. at 9. Another statement of trigger in the Communication is: “The precautionary principle is relevant only in the event of a potential risk, even if this risk cannot be fully demonstrated or quantified or its effects determined because of the insufficiency or inconclusive nature of the scientific data.” Id. at 13.
90 In addition to the discussion above, see Pfizer, 2002 E.C.R. II __, � 153.
91 There was the possibility of irreversible harm at stake in Pfizer and Alpharma. See Pfizer, 2002 E.C.R. II __, � 334.
92 The Commission acknowledges that in the face of scientific uncertainty, risk assessment itself can take a “prudential approach” and incorporate prudential aspects, but seems ambivalent about calling this risk management. See Communication, supra note 88, at 12. The better approach is to state clearly that science policies for risk assessors are risk management decisions and that findings of risk are not purely scientific.
93 SPS Agreement, supra note 23.
94 Id. art. 2.2.
95 Id. art. 5.1.
96Id. annex A.1. For example, a “sanitary or phytosanitary measure” might be a governmental measure applied “to protect human or animal life or health within the territory of the Member from risks arising from additives, contaminants, toxins or disease-causing organisms in foods, beverages or feedstuffs.” Id. annex A.1(b). Thus, the SPS Agreement applies to health risks from food and animal feed.
97 WTO, Appellate Body, Report on EC Measures Concerning Meat and Meat Products (Hormones), WT/DS26/AB/R & WT/DS48/AB/R, � 181 (Jan. 16, 1998) [hereinafter Hormones Report].
98 SPS Agreement, supra note 23, pmble., arts. 2.1, 3.3, 5, annex A.5.
99 See id. annex A.5; Communication, supra note 88, � 5.2.1.
100 See, e.g., SPS Agreement, supra note 23, art. 5.5, annex A.5; Walker, supra note 26, at 268–71.
101 Hormones Report, supra note 97, � 2.
102 SPS Agreement, supra note 23, annex A.1(b).
103 Id.
104 Hormones Report, supra note 97, �� 182–84; SPS Agreement, supra note 23, annex A.4.
105 Hormones Report, supra note 97, � 186; WTO, Appellate Body, Report on Australia—Measures Affecting Importation of Salmon, WT/DS18/AB/R, � 125 (Oct. 20, 1998) [hereinafter Salmon Report].
106 Hormones Report, supra note 97, � 186.
107 Id. �� 186, 193.
108 Id. �� 187, 198–201.
109 Id. � 187.
110 See id. � 194.
111 Salmon Report, supra note 105, � 1.
112 Id. �� 112, 121; SPS Agreement, supra note 23, annex A.1(a).
113 Salmon Report, supra note 105, � 125.
114 Id. � 123.
115 Id. � 125.
116 Id. � 130.