Opinion
Nos. 86-1114, 86-1535.
Argued December 3, 1986.
Decided January 16, 1987. As Amended January 16, 1987.
Alexander J. Pires, Jr., with whom John M. Himmelberg and Jeffrey A. Knishkowy were on the brief for petitioners.
Letitia Grishaw, Dept. of Justice, with whom Cara S. Jablon, Michael S. Winer and Stanley H. Abramson, Office of General Counsel, E.P.A. were on the brief, Michael W. Steinberg, Dept. of Justice entered an appearance for respondents.
Petitions for Review of Orders of the Environmental Protection Agency.
Before STARR and WILLIAMS, Circuit Judges, and GREEN, District Judge.
Of the United States District Court for the District of Columbia, sitting by designation pursuant to 28 U.S.C. § 292(a).
Opinion for the Court filed by Circuit Judge STARR.
In the Federal Food, Drug and Cosmetic Act, Congress conferred authority on the Environmental Protection Agency to promulgate tolerances limiting the level of pesticides in raw fruits and vegetables "to the extent necessary to protect the public health." 21 U.S.C. § 346a(b) (1982). Petitioners seek review of EPA's exercise of this authority to fix tolerances for the amount of ethylene dibromide found in imported mangoes. We are constrained to conclude that the agency has engaged in arbitrary and capricious decisionmaking. We therefore grant the petitions for review.
Petitioners are: the National Coalition Against the Misuse of Pesticides (a nonprofit organization the members of which share concerns about pesticide safety); J.R. Brooks Son, Inc. and Ed Mitchell, Inc. (domestic mango growers); and N.P. Brooks and J.K. Mitchell (consumers of foreign and domestic mangoes).
I
Ethylene dibromide (EDB) is a chemical that until recently was widely used as a pesticide on fruits and vegetables in the United States. In the 1970's, evidence accumulated that exposure to EDB increased the risks of cancer, genetic mutations, and adverse reproductive effects in humans. These danger signals prompted EPA to conduct several detailed studies of the risks and benefits of EDB. As a result of its research, EPA in 1983 exercised its authority under the Federal Insecticide, Rodenticide, and Fungicide Act, 7 U.S.C. §§ 136- 136y (1982 Supp. III 1985), to impose bans on the use of EDB on domestic produce. Those bans took effect almost immediately.
EPA embraced a more gradual approach, however, with respect to imported fruit, specifically mangoes, sprayed with EDB in their countries of origin. This particular use received special attention because EDB fumigation was the only effective way then available to insure that imported mangoes were free of fruit flies of various species, including the Mediterranean fruit fly. Acting pursuant to § 346a(b) of the Food, Drug and Cosmetic Act, EPA in January 1985 promulgated a tolerance under which imported mangoes could be treated with EDB as long as residues in the edible pulp of the fruit measured no more than 30 parts per billion (ppb). EPA appended a critical limitation in setting this tolerance, however. The tolerance was to expire September 1, 1985. After that date, a zero tolerance for EDB would take effect, rendering imported mangoes with any trace of EDB "adulterated" under the Food, Drug and Cosmetic Act, 21 U.S.C. §§ 342(a)(2)(B), 346a(a)(1) (1982), and hence prohibited in interstate commerce. EPA summarized its decision as
a prudent determination to remove EDB from the diet long-term while avoiding significant economic disruptions by allowing a year to develop alternative treatment methods.
50 Fed.Reg. 2547, 2550-51 (Jan. 17, 1985), Joint Appendix, No. 86-1114 ("J.A. I") at 15, 17-18.
As the September 1985 expiration date of the 30 ppb tolerance loomed ahead, proponents of continued use of EDB beyond U.S. borders undertook efforts to secure an extension of the interim tolerance. Advocates of the extension advanced two arguments. First, because no alternative to EDB treatment had been developed, expiration of the interim tolerance would amount to a complete ban on imported mangoes; this ban, they feared, would severely damage the fragile economies of mango-producing countries, especially Haiti and Mexico. Second, the interim-extension advocates submitted evidence that EDB dissipated more quickly than EPA had posited when it promulgated the interim tolerance; the proponents thus argued that EDB posed a lower health risk than that which animated EPA in imposing a zero tolerance as of September 1, 1985. This evidence showed, they maintained, that by the time mangoes reached the consumer, the fruit contained no more than a modest 10 to 14 ppb of EDB.
Initially, EPA was unmoved. In fact, the agency expressly rejected both arguments in a memorandum dated August 28, 1985, which was apparently furnished to the State Department to explain the agency's declination of invitations to extend the interim tolerance beyond the September 1, 1985 cutoff. EPA stated that it had considered the impact of a ban on foreign economies but concluded that extending the 30 ppb tolerance would not necessarily bring about a practicable alternative; an extension could, in fact, delay implementation of a substitute treatment method. Second, EPA determined that calculation of the health risk taking into account the effects of EDB's dissipation nonetheless resulted in a risk in the same general range as calculations based on 30 ppb. The agency thus concluded:
Since the risks for the U.S. consumers of treated mangoes are not changed from EPA's 1984 estimates, which found one year of further exposure to be the limit of acceptable continued exposure, the Agency feels that additional exposure to EDB in the diet is not in the public interest.
J.A. I at 19-20. Accordingly, on September 1, 1985, a zero tolerance for EDB on imported mangoes took effect.
Not long after this, however, the agency received entreaties from the State Department and the Department of Agriculture to reconsider its newly imposed ban. Among the importunings were letters from John Whitehead, Deputy Secretary of State, and James Michel, Acting Assistant Secretary for Inter-American Affairs. The Whitehead-Michel communications reiterated the adverse economic impact on friendly countries occasioned by EPA's ban on EDB-fumigated mangoes. In contrast to the agency's reactions to such arguments in late summer, however, the spectre of burdened South and Central American economies now gave EPA considerable pause.
On November 27, 1985, EPA did an about-face. The agency proposed to abandon the zero tolerance and revive the expired 30 ppb tolerance through at least September 30, 1986. EPA further announced that it would consider extending the interim tolerance for an additional year, through September 1987, if by September 1986 it clearly appeared that an alternative to EDB could be in place when the 1987-88 mango harvest began. Without referring to its contrary position of just three months before, outlined in the August 28, 1985 memorandum, EPA cited three reasons for its new course. First, since the zero tolerance had taken effect, EPA had become more keenly aware of the severe impact on the economies of mango-producing nations that the zero tolerance would occasion. Second, the Department of Agriculture expressed optimism that by 1987 or 1988 an alternative to EDB could be available. Third, the health risk of an additional year or two of EDB was deemed "acceptable . . . based on the fact that residues of EDB dissipate fairly rapidly" before mangoes treated with EDB reached United States consumers. See 50 Fed.Reg. 48,799, 48,800 (Nov. 27, 1985), J.A. I at 21, 22.
After receiving comments on its proposal, EPA formally reestablished the 30 ppb tolerance in a final rule published February 14, 1986. 51 Fed.Reg. 5682 (Feb. 14, 1986), J.A. I at 178. The February 1986 decision, like that of the preceding November, never mentioned the tell-tale August 1985 EPA memorandum. EPA justified its action in this latest decision by noting the economic impact of a ban (including potential detriment to the U.S. economy); the Department of Agriculture's assurance of a substitute for EDB treatment by 1987; and the agency's determination that the health risks of EDB were "very low." Id.
A petition for review of the February 14, 1986 Final Rule was filed in this court on the same day. Petitioners challenged EPA's action on four bases: first, that EPA lacked statutory authority to reestablish a tolerance of 30 ppb after having determined a zero tolerance to be necessary to protect public health, see 21 U.S.C. § 346a(b); second, that EPA acted arbitrarily and capriciously by basing the reestablishment of the 30 ppb tolerance primarily on the impact of a total ban on foreign economies; third, that EPA arbitrarily and capriciously shifted its position on the health risk posed by EDB without any new data to justify this reversal; and fourth, that EPA violated due process and the requirements of the Administrative Procedure Act, 5 U.S.C. § 553 (1982), by shortening the time for comment on its proposed rule and by refusing to hold a hearing before promulgating the rule. Petitioners sought and were granted expedited review of their challenge.
Before the date originally scheduled for oral argument of this petition (October 17, 1986), EPA carried out its intention, announced in the February 14, 1986 Final Rule, to extend the interim tolerance through September 30, 1987. See 51 Fed.Reg. 28,603 (Aug. 8, 1986) (proposing extension), J.A. No. 86-1535 ("J.A. II") at 1; 51 Fed.Reg. 34,469 (Sept. 29, 1986) (extension), J.A. II at 3. EPA concluded that the conditions established in February for an extension — substantial progress towards development of an alternative to EDB and proof that such an alternative would likely be in place for the 1987-88 mango harvest — had been satisfied. See 51 Fed.Reg. at 28,603, J.A. II at 1.
This extension of the 30 ppb tolerance through 1987 prompted petitioners to file a second petition for review and to seek to consolidate it with their first. We requested further submissions to determine whether the first petition had been rendered moot by virtue of the fact that the date EPA initially set for the interim tolerance to expire, September 30, 1986, had passed prior to oral argument. We consolidated the petitions for purposes of oral argument and heard the case on an expedited basis. We now conclude that petitioners' earlier petition is not moot and have consolidated the petitions for purposes of disposition.
After reviewing the parties' submissions, we hold that petitioners' first petition, No. 86-1114, challenging EPA's decision on February 14, 1986 to reestablish a 30 ppb tolerance for EDB on imported mangoes, is not moot. Although that decision provided that the tolerance was to expire September 30, 1986, it also provided for an extension of the 30 ppb tolerance through September 1987. Thus, the September 1986 decision extending the 30 ppb tolerance, which is the subject of the second petition, grew out of the decision which is the subject of the first petition. Accordingly, review of the later agency decision requires that we review the earlier one, and the extension of the revived tolerance cannot survive if we conclude that EPA improperly revived it in the first place.
II
Before addressing the merits, we must first examine a threshold issue raised by the agency. Specifically, EPA has interposed the defense that petitioners are not entitled to judicial review because they failed to exhaust administrative remedies. The agency bases its argument upon 21 U.S.C. § 346a(d)(5), which provides in pertinent part:
Within thirty days after publication, any person adversely affected by a regulation published pursuant to paragraph (2) or (3) of this subsection, or pursuant to subsection (e) of this section, may file objections thereto with the Administrator, . . . requesting a public hearing upon such objections. . . . The Administrator shall thereupon, after due notice, hold such public hearing for the purpose of receiving evidence relevant and material to the issues raised by such objections . . . . As soon as practicable after completion of the hearing, the Administrator shall act upon such objections and by order made [sic] public a regulation.
The regulations challenged in this case were promulgated following notice and comment and pursuant to § 346a(e), which permits EPA on its own initiative to establish tolerances for pesticides and provides that "[r]egulations issued under this subsection shall upon publication be subject to paragraph (5) of subsection (d)." EPA interprets the latter provision, § 346a(d)(5), as compelling petitioners to seek a hearing under that subsection before invoking the judicial review provisions of § 346a(i). Alternatively, EPA argues that even if § 346a(d)(5) permits but does not require a hearing, exhaustion principles developed in the case law effectively render § 346a(d)(5) a mandatory provision. We disagree with both contentions. We are persuaded that, under the circumstances of this case, neither the language of the statute nor applicable decisional law requires resort to § 346a(d)(5) prior to seeking review under § 346a(i).
See 49 Fed.Reg. 32,088, 32,089 (1984) (proposing interim tolerance pursuant to 21 U.S.C. § 346a(e)), J.A. I at 166, 167.
First and foremost, the language of the specific provision on which EPA relies (§ 346a(d)(5)), is permissive, not mandatory. Where the language, as here, is clear, that should be the end of the matter. See e.g., Garcia v. United States, 469 U.S. 70, 75 (1984); United Air Lines, Inc. v. CAB, 569 F.2d 640, 647 (D.C. Cir. 1977). But aside from the clear and unambiguous language of the provision, the structure of the statute likewise suggests that Congress did not mandate employment of the § 346a(d)(5) hearing. Section 346a(i)(1) provides for judicial review of "any order under subsection (d)(5), (e), or ( l.)" As we observed previously, EPA's final rule of February 14, 1986 and the extension of September 1986 both qualify as "order[s] under subsection . . . (e)." See supra note 3. If we accept EPA's argument that judicial review under § 346a(i)(1) can only follow formal hearings under subsection (d)(5) of § 364a, then the reference in § 346a(i)(1) to orders under subsection (e) is rendered surplusage. Both the language and structure of § 346a therefore indicate that resort to a formal hearing is unnecessary before repairing to the courts.
The case law does not point to an opposite result. To the contrary, a review of relevant decisions likewise suggests that exhaustion is not appropriately required in this case. First, the orders under challenge took immediate effect and were in that respect "final," see 21 U.S.C. § 346a(e). Since, as we discussed, the applicable statutory provisions by their terms permit review of such orders, conducting review at this stage does not appear to thwart Congressional purposes by encouraging "deliberate flouting of administrative processes," a primary concern underlying the exhaustion doctrine. See Andrade v. Lauer, 729 F.2d 1475, 1484 (D.C. Cir. 1984) (quoting McKart v. United States, 395 U.S. 185, 195, 89 S.Ct. 1657, 1663, 23 L.Ed.2d 194 (1969)). A second concern underlying the doctrine of exhaustion — the need for an adequate factual record — is likewise absent here. See id.; see also Myers v. Bethlehem Shipbuilding Corp., 303 U.S. 41, 50-51, 58 S.Ct. 459, 463, 82 L.Ed. 638 (1938) (exhaustion required where petitioners sought completely to avoid hearing requirement of NLRA); Nader v. Volpe, 466 F.2d 261, 267 n. 36 (D.C. Cir. 1972). EPA's orders followed a notice and comment period in which numerous parties participated. Finally, our review does not deprive EPA of the opportunity to discover and correct its own errors, nor does it otherwise intrude prematurely into matters best left to percolate at the agency, a third factor upon which our decisions have wisely relied as a justification for exhaustion. See Andrade, 729 F.2d at 1484. Indeed, petitioners expressly advanced before EPA both arguments that we find dispositive. See 51 Fed.Reg. at 5638, J.A. I at 179; see also Comments Submitted on Behalf of J.R. Brooks Son, Inc., Ed Mitchell, Inc., N.P. Brooks, and John Mitchell 2-4 (Jan. 10, 1986), J.A. I at 361-63. Thus, while we are quite unprepared to hold that exhaustion may never be required prior to review of a tolerance promulgated under § 346a(e), we are convinced it is unnecessary in the circumstances of this case.
III
Although the petitions raise a variety of substantive and procedural issues, we find it necessary to address in detail only two. Both pertain to petitioners' claim that EPA's decisions to reestablish and extend a 30 ppb tolerance for EDB on imported mangoes amounted to decisionmaking that was arbitrary and capricious and contrary to law, in violation of the APA, 5 U.S.C. § 706(2)(A) (1982). First, petitioners maintain that EPA based its decisions "on factors which Congress has not intended it to consider," Motor Vehicle Manufacturers Association v. State Farm Mutual Automobile Insurance Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 2867, 77 L.Ed.2d 443 (1983), namely, the adverse impact of a zero EDB tolerance on foreign economies. See also Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 411-12, 416, 91 S.Ct. 814, 821, 823, 28 L.Ed.2d 136 (1971). Second, petitioners contend that the agency failed to "supply a reasoned analysis," State Farm, 463 U.S. at 57, 103 S.Ct. at 2874 (quoting Greater Boston Television Corp. v. FCC, 444 F.2d 841, 852 (D.C. Cir. 1970), cert. denied, 403 U.S. 923, 91 S.Ct. 2233, 29 L.Ed.2d 701 (1971)), for reversing its earlier position that a zero tolerance after September 1, 1985 was necessary to protect the public health. We consider each argument in turn.
We may quickly dispose of petitioners' other contentions, which we find lacking in merit. Petitioners argue that EPA lacked statutory authority to promulgate a tolerance of 30 ppb having previously established a zero tolerance. Their reasoning, as we understand it, is that the agency can only promulgate tolerances when " necessary to protect the public health." 21 U.S.C. § 346a(b) (emphasis added). Since by virtue of EPA's January 1985 decision, a zero tolerance went into effect September 1, 1985, no further action was necessary to protect public health against EDB; therefore, no further action was authorized under the statute, certainly not one that resulted in increased exposure to what the agency had already determined to be a hazardous chemical.
While acknowledging the verbal ingenuity of this argument, we reject it, for whatever initial appeal it appears to have when measured against the language of § 346a(b) evaporates when exposed to the light of common sense. To hold EPA statutorily disabled from ever raising a tolerance — the inexorable destination of petitioners' reasoning — would mean that the agency could not act upon changed circumstances or subsequently discovered scientific data even though such considerations might warrant different action. For example, if after EPA established a zero tolerance for EDB on imported mangoes, evidence emerged indicating that this pesticide was less harmful than previously believed, the statute clearly contemplates reassessment of what level of EDB is necessary to protect public health. Because petitioners' argument handcuffs EPA to an extent at odds with the orderly operation of the statutory scheme, we reject it.
Likewise, petitioners' attacks on the procedures EPA followed need not detain us. Petitioners first argue that the 45-day period EPA provided for comment on its November 1985 proposal to revive a 30 ppb tolerance for EDB violated the Administrative Procedure Act, 5 U.S.C. § 553(c), because it was too short to afford a reasonable opportunity for meaningful comment. They do not, however, suggest why this length of time, which is 15 days longer than the EPA usually provided, was insufficient in this case. See In re Surface Mining Regulation Litigation, 627 F.2d 1346, 1354 n. 9 (D.C. Cir. 1980) (attack on agency procedure must be joined with a showing of prejudice). Moreover, § 408(e) of the FDCA permits an agency to publish a final regulation establishing a tolerance only 30 days after publishing a proposed tolerance. See 21 U.S.C. § 346a(e). Petitioners have therefore not established a violation of the APA. Second, petitioners allege that EPA's failure to conduct a hearing prior to publishing the final regulation formally reviving the 30 ppb tolerance violated due process. We believe that petitioners are hardly in a position to argue for a hearing not prescribed in the FDCA in light of their failure to avail themselves of the post-publication hearing that is provided for in the FDCA. See id. § 346a(d)(5). More fundamentally, they have advanced no reason why this situation is not governed by the general rule that due process imposes no constraints on informal rulemaking beyond those imposed by the statute. See, e.g., Sierra Club v. Costle, 657 F.2d 298, 392 n. 462 (D.C. Cir. 1981); Association of Nat'l Advertisers, Inc. v. FTC, 627 F.2d 1151, 1163-64 (D.C. Cir. 1979), cert. denied, 447 U.S. 921, 100 S.Ct. 3011, 65 L.Ed.2d 1113 (1980); Pickus v. Board of Parole, 543 F.2d 240, 242-45 (D.C. Cir. 1976).
A
The source of EPA's authority to promulgate tolerances for pesticides on fruits and vegetables is found in section 408(b) of the Food, Drug and Cosmetic Act, which provides in pertinent part:
The Administrator shall promulgate regulations establishing tolerances with respect to the use in or on raw agricultural commodities of poisonous or deleterious pesticide chemicals . . . to the extent necessary to protect the public health. In establishing any such regulation, the Administrator shall give appropriate consideration, among other relevant factors, (1) to the necessity for the production of an adequate, wholesome, and economical food supply; (2) to the other ways in which the consumer may be affected by the same pesticide chemical or by other related substances that are poisonous or deleterious; and (3) to the opinion submitted with a certification of usefulness under subsection ( l) of this section.
21 U.S.C. § 346a(b). Petitioners maintain that this language makes clear that consideration of foreign economic impact is beyond EPA's lawful authority. In contrast, EPA takes the position that the statute provides the agency with broad discretion to determine what factors are "relevant" to formulation of a tolerance.
Subsection l of section 408 provides for certification by EPA as to the usefulness of a pesticide for which a tolerance is being considered. 21 U.S.C. § 346a( l).
EPA also responds to petitioners' attack by claiming that it considered the effects of an EDB ban on domestic as well as foreign economies. While it is true that the agency referred in its September 1986 decision to "economic hardships to many U.S. and foreign interests," 51 Fed.Reg. at 34,471, J.A. II at 5, as the basis for extending the 30 ppb tolerance through 1987, we are persuaded that its concern for domestic economic interests was entirely subsidiary to its fear that a zero tolerance would cripple foreign mango-growing countries. The agency's prior decisions proposing to reestablish the tolerance (in November 1985) and formally reestablishing the tolerance through 1986 (in February 1986) make no mention of concern for this country's economic well-being, nor does the State Department correspondence that prompted the agency's reversal evidence any such concern. In sum, we believe the agency spoke most candidly when in first announcing its decision to revive the 30 ppb tolerance it stated as the sole reason for its reversal that "the absence of a tolerance rule for EDB on mangoes is likely to result in a severe economic impact on certain Caribbean Basin countries, as well as on other mango producing countries because alternatives have not been developed to replace the use of EDB." 50 Fed.Reg. at 48,799, J.A. I at 21.
The more natural reading of the statutory phrase, "other relevant factors," in § 346a(b) is that EPA can consider only factors related to protection of public health. It is, after all, "the public health" that EPA is charged with protecting through "regulations establishing tolerances." At the same time, we are mindful that reviewing courts are not to substitute their interpretation of a statute for that of an administering agency merely upon concluding that the court's reading is the more natural. See Young v. Community Nutrition Institute, ___ U.S. ___, 106 S.Ct. 2360, 2364, 90 L.Ed.2d 959 (1986). Instead, we are called upon to examine whether Congress' intent can clearly be discerned with respect to the particular matter at hand and, if not, to examine the permissibility of the agency's interpretation. See United States v. City of Fulton, ___ U.S. ___, 106 S.Ct. 1422, 1428, 89 L.Ed.2d 661 (1986); Montana v. Clark, 749 F.2d 740, 745-46 (D.C. Cir. 1984), cert. denied, ___ U.S. ___, 106 S.Ct. 246, 88 L.Ed.2d 255 (1985). It appears to us that in § 346a(b) Congress has not clearly addressed that precise issue whether foreign impact can figure in EPA's setting of tolerances. Nor is Congress' intent on this specific question revealed by other sources of legislative purposes, such as the structure of the statute or its history. Under these circumstances, we are obliged to uphold the agency's construction if we conclude that it is a permissible (or reasonable) one, even if that interpretation is not one which we would have embraced had the decision been ours in the first instance. See Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 844, 104 S.Ct. 2778, 2782, 81 L.Ed.2d 694 (1984).
We note that although the agency based its decisions exclusively on foreign economic impact, see supra note 6, counsel for EPA conceded at oral argument that the statutory duty to "protect the public health" referred to the health of the United States public.
Guided by Chevron principles, we are unprepared to hold that in promulgating tolerances under § 346a(b) EPA may only consider factors related to the health of the U.S. public and may never consider impact on foreign economies. The defect in these proceedings, however, is that EPA has failed entirely to take into account factors that § 346a(b) clearly requires the agency to consider, namely "to the necessity for the production of an adequate, wholesome and economical food supply" and "to the other ways in which the consumer may be affected by the same pecticide chemical or by other related substances." Without purporting definitively to address how prominently consideration of impact on foreign countries can figure in the Administrator's promulgation of a tolerance, we conclude that Congress intended EPA principally to attend to the specifically enumerated factors in the statute, see Environmental Defense Fund, Inc. v. HEW, 428 F.2d 1083, 1086 (D.C. Cir. 1970), and to how nonenumerated factors (such as foreign economic impact) deemed "relevant" by the agency are linked to the statutorily specified factors. Section 346a(b) may permissibly be interpreted to permit the Administrator to assign some relevance to foreign impact, even though unrelated to U.S. public health, but the provision clearly mandates that health factors and issues of foreign impact related to the health of U.S. consumers be at the forefront of EPA's deliberations. Here, EPA fell into error by losing sight of the specific statutory factors in its solicitude for the economic welfare of foreign countries.
See also S.Rep. No. 1635, 83rd Cong., 2d Sess. 7 (1954) (describing factors enumerated in § 346a(b) as "the principal factors which should guide the administrative judgment in setting tolerances").
Counsel for EPA maintains that the agency did explicate the linkage between its concern for the impact on foreign countries of a zero tolerance and the healthbased factors that the agency was statutorily bound to consider. We disagree. The passages in EPA's decision on which counsel relies, see Brief of Respondents, No. 86-1535, at 6-12, do not set forth EPA's justification for the revived 30 ppb tolerance; they merely summarize the comments on EPA's proposals to reestablish a 30 ppb tolerance and extend it through 1987. We have no occasion to say whether the administrative record could support a decision that the economic impact of a zero tolerance could be so severe as to threaten "the public health" — for example, by disrupting the country's mango supply, which could raise concern about the need for an "economical food supply," or encouraging the illicit infiltration of fruit fly-infested mangoes, which could similarly pose a risk to the "adequa[cy]" and "wholesome[ness]" of the food supply. We hold only that EPA has not established any such linkage, nor under settled principles can counsel's post hoc rationalizations cure this deficiency. See State Farm, 463 U.S. at 50, 103 S.Ct. at 2870; Citizens to Preserve Overton Park, 401 U.S. at 419, 91 S.Ct. at 825; cf. Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168-69, 83 S.Ct. 239, 245-46, 9 L.Ed.2d 207 (1962).
We therefore hold that EPA acted arbitrarily and capriciously by relying exclusively on concerns for foreign well-being, albeit with a nod toward domestic interests, see supra note 6, in reviving and extending the 30 ppb tolerance without considering the factors specified in the FDCA and factors of foreign impact related to those factors enjoying express Congressional approbation.
B
Petitioners base their second argument on the well-settled principle that an agency may not shift its position without supplying a reasoned explanation for doing so. See, e.g., State Farm, 463 U.S. at 57, 103 S.Ct. at 2874 (quoting Greater Boston Television Corp. v. FCC, 444 F.2d at 852); Oil, Chemical Atomic Workers International Union v. NLRB, 806 F.2d 269, 273 (D.C. Cir. 1986); Airmark Corp. v. FAA, 758 F.2d 685, 692 (D.C. Cir. 1985).
In this case, as petitioners see it, EPA failed to explain its about-face on the health risk occasioned by EDB. They point out that in establishing an interim tolerance of 30 ppb in January 1985, the agency initially determined that these health dangers required a zero tolerance after September 1, 1985. EPA reaffirmed this determination in August 1985, when the agency expressly found that "additional exposure beyond September 1, 1985] is not in the public interest," J.A. I at 20. And on the appointed date of September 1, 1985, a zero tolerance for EDB on imported mangoes indeed went into effect. Notwithstanding this consistent course of action, the agency in proposing to reestablish the 30 ppb tolerance concluded in November 1985 that the risk posed by an additional year of exposure was "acceptable," 50 Fed.Reg. at 48,800, J.A I at 22; and in adopting the tolerance as a final rule, EPA characterized the risk of exposure through 1986 as "very low," 51 Fed.Reg. at 5684, J.A. I at 180. Finally, in its latest pronouncement, EPA similarly determined that the risk of exposure through yet a third year, 1987, was "very low." 51 Fed.Reg. at 34,471, J.A. II at 5.
The record is virtually barren of reasons for this volte face. The sole justification advanced by the agency for its turnabout on the potential health risk is found in its November 1985 proposal to reestablish a 30 ppb tolerance. There, EPA set forth the following basis for its conclusion that the health risk was acceptable:
[R]esidues of EDB dissipate fairly rapidly from the treated fruit. Therefore, the actual residues at the retail level are likely to be considerably lower than the tolerance limit of 30 ppb which applies to the fruit at the time of entry into the United States.
50 Fed.Reg. at 48,800, J.A. I at 22. On this record, we conclude that this rationale is wholly inadequate. It reflects an unexplained reversal from EPA's position just three months earlier in the memorandum of August 28, 1985. In that document, the agency acknowledged that evidence existed "that mangoes released into commerce have EDB residues measured significantly below the 30 ppb tolerance," J.A. I at 20. Nonetheless, EPA concluded:
[E]ven if the average EDB level in mangoes at the time of consumption is well below the tolerance, for example, at 10 ppb, EPA's assessment is that the dietary risk at that level is still in the same general range as for consuming mangoes with 30 ppb of EDB. Thus, the Agency has not seen any data which alters the 1984 assessment of risks posed by consuming EDB treated mangoes.
Id.
In an effort to shore up the agency's latter-day embrace of the dissipation rationale, EPA's counsel direct us to evidence submitted in the interim that showed, for example, that improved techniques for packing and shipping mangoes further enhanced dissipation. This is, once again, a post hoc rationalization that we are quite unable to accept to buttress agency action. And in any event, it is clear to us that in its November 1985 and subsequent decisions, EPA was recharacterizing the risk of 30 ppb of EDB, not changing its views of the relevant levels of EDB present on mangoes. EPA conceded that it had happened upon no new scientific evidence to justify reassessing the health risk entailed by EDB. In the absence of any such evidence or other explanation, its about-face is inexplicable and, under well-settled principles of law, stands condemned as arbitrary and capricious. Accordingly, if EPA on remand in any respect relies on an assessment of the health risk attributable to consumption of EDB on mangoes that is lower than the assessment underlying its January 1985 determination, it must adduce a legally adequate explanation.
In fact, while its opinion as to the health risk changed, EPA's quantification of the risk remained constant, namely that average levels of consumption of EDB-fumigated imported mangoes for one year increased the risk of cancer to an individual in the range of 1 X 10-[6] to 1 X 10-[7]. See Position Document 4, at 52-54 (Sept. 27, 1983), J.A. I at 56-58; cf. 49 Fed.Reg. at 32,089 (proposal in 1984 to establish 30 ppb tolerance), J.A. I at 167; 51 Fed.Reg. at 5684 (final rule of Feb. 1986 extending tolerance through September 1986), J.A. I at 180.
We have considered petitioners' other objections and, with one exception, find them unnecessary to discuss. That exception is petitioners' complaint that EPA should have included in the public file its memorandum of August 28, 1985, in which it explained its position at that time to allow the 30 ppb interim tolerance to expire September 1, 1985 and a zero tolerance to replace it. Nonetheless, petitioners obtained the memo in time to comment on it in these proceedings, and thus its omission from the public record is harmless error as to them. But we emphasize that the document plainly qualifies as one that EPA should have made public. Congress designed the informal rulemaking procedures of the APA to insure an
exchange of views, information, and criticism between interested persons and the agency. Consequently, the notice required by the APA, or information subsequently supplied to the public, must disclose in detail the thinking that has animated the form of a proposed rule and the data upon which that rule is based.
Home Box Office, Inc. v. FCC, 567 F.2d 9, 35 (D.C. Cir.) (citations omitted) (emphasis in original), cert. denied, 434 U.S. 829, 98 S.Ct. 111, 54 L.Ed.2d 89 (1977). The August 1985 memorandum clearly supplied information critical to informed comment on EPA's proposal to reestablish the 30 ppb tolerance, and, for that matter, to review by this court. Moreover, EPA's description of the memorandum as merely an "unsigned, undated fact sheet" that "did not preclude a reassessment of the risk-benefit analysis at a later date in light of additional information" is disingenuous at best. See Respondent's Brief, No. 86-1114, at 13; Respondent's Brief, No. 86-1535, at 17 n. 10. For one thing, as we have discussed in the text, EPA received no additional scientific information to justify reassessment of the health risk posed by 30 ppb of EDB and expressly rejected in its August memorandum the significance of evidence of dissipation. For another, what EPA now characterizes as the unimportant nature of this document did not prevent it from furnishing the memorandum to the State Department at a pivotal juncture in these proceedings. See Letter from William Mansfield, Deputy Associate Administrator, EPA, to Clifton Metzner, Office of Environmental Affairs, Department of State, Attachment (Aug. 28, 1985), J.A. I at 390.
As opposed to the health risks attributable to elimination of the use of EDB on mangoes.
IV
In summary, we hold that in its challenged orders, EPA acted arbitrarily and capriciously by failing to give appropriate consideration to relevant factors and reversing its position on the health risks of EDB. We therefore grant the petitions for review, reverse EPA's orders of February 14, 1986 and September 29, 1986, and remand this case to the agency for further proceedings.
We are sorely mindful of the severe time constraints present here, which raise the possibility that the September 30, 1987 deadline will arrive before EPA could, in the ordinary course, act on remand and this court would be able, if called upon, to complete its review of the agency's action on remand. Under these unusual circumstances, we remand the case with specific instructions for the agency to address the issues discussed in this opinion and to decide within thirty days from the date hereof whether an interim tolerance of 30 ppb EDB (or any level other than zero) in imported mangoes is justified. To prevent further disruption and uncertainty in this difficult and sensitive matter, we refrain from setting aside immediately the final rule and extension of the 30 ppb tolerance, but permit them to remain in effect until issuance of mandate, which we direct to occur thirty days hence. Cf. Simmons v. ICC, 757 F.2d 296, 300 (D.C. Cir. 1985). This court will then stand ready to take appropriate action on an expedited basis should circumstances so warrant.
Judgment accordingly.