Larry Leon Chaney v. Margaret M. Heckler, as Secretary of Health and Human Services

SCALIA, Circuit Judge,

dissenting:

The majority converts a law designed to protect consumers against drugs that are unsafe or ineffective for their represented use into a law not only permitting but mandating federal supervision of the manner of state executions. This implausible result is achieved by rewriting the law with regard to enforcement discretion and ignoring it with regard to FDA jurisdiction. I dissent from what seems to me a clear intrusion upon powers that belong to Congress, the Executive Branch and the states. I would affirm for the reasons set forth in the district court’s opinion.

I. Enforcement Discretion

Even if it were correct that the Food, Drug, and Cosmetic Act makes the challenged drug use illegal, the majority’s disposition of this case would not be supportable under well established law governing judicial review of agencies’ enforcement discretion. Generally speaking, enforcement priorities are not the business of this Branch, but of the Executive — to whom, and not to the courts, the Constitution confides the responsibility to “take Care that the Laws be faithfully executed,” U.S. Const, art. II, § 3. Preserving that sound allocation of responsibility was one of the purposes, and has been perhaps the primary application, of that provision of the Administrative Procedure Act which excludes from judicial review “agency action ... committed to agency discretion by law.” 5 U.S.C. § 701 (a)(2) (1976). As this court has accurately stated, “[a]n agency’s decision to refrain from an investigation or an enforcement action is generally unreviewable.” Kixmiller v. SEC, 492 F.2d 641, 645 (D.C.Cir.1974). See also Action on Safety and Health v. FTC, 498 F.2d 757, 762 (D.C.Cir.1974). The reason is well explained in a Supreme Court case that could more easily have been regarded as appropriate for judicial control of agency enforcement policy than the present one, since it involved non-enforcement against some companies combined with enforcement against others:

[I]n the shaping of its remedies within the framework of regulatory legislation, an agency is called upon to exercise its specialized, experienced judgment. Thus, the decision as to whether or not an order against one firm to cease and desist from engaging in illegal price discrimination should go into effect before others are similarly prohibited depends on a variety of factors peculiarly within the expert understanding of the Commission.... [T]he Commission alone is empowered to develop that enforcement policy best calculated to achieve the ends contemplated by Congress and to allocate its available *1193 funds and personnel in such a way as to execute its policy efficiently and economically.

Moog Industries v. FTC, 355 U.S. 411, 413 , 78 S.Ct. 377, 379 , 2 L.Ed.2d 370 (1958). See also Vaca v. Sipes, 386 U.S. 171, 179 , 87 S.Ct. 903, 911 , 17 L.Ed.2d 842 (1967).

The majority opinion gives its view of enforcement discretion the appearance of reality by quoting from Supreme Court cases that assert a “strong presumption” of reviewability rebuttable only by “clear and convincing” legislative intent to negate review. Maj.Op. at 1183-1184 n. 23. These excerpts are misleading, since they are taken from cases or portions of cases that do not involve review of enforcement discretion. Indeed, all but one of them come from cases or portions of cases that do not even deal with the “committed to agency discretion” provision of the APA (§ 701(a)(2)), but rather with the provision that denies review “to the extent that .. . statutes preclude judicial review” (§ 701(a)(1)). The citation to Dunlop v. Bachowski, 421 U.S. 560 , 567 n. 7, 95 S.Ct. 1851 , 1858 n. 7, 44 L.Ed.2d 377 (1975), is illustrative. The court of appeals in that case, Bachowski v. Brennan, 502 F.2d 79 (3d Cir.1974), considered and rejected separate arguments by the Secretary of Labor that (1) the Labor-Management Reporting and Disclosure Act of 1959, 29 U.S.C. § 401 et seq. (1976) (“L-MRDA”), precluded judicial review of his failure to institute a civil action to set aside a tainted union election, 502 F.2d at 84-86 , and (2) his decision not to prosecute was committed to agency discretion, id. at 86-88. The Supreme Court addressed the first of those arguments in detail, Dunlop, supra, 421 U.S. at 566-68 , 95 S.Ct. at 1857-58 . In the context of its preclusion discussion, the Court established that “[i]n the absence of an express [statutory] prohibition,” there exists “the strong presumption that Congress did not mean to prohibit all judicial review,” id. at 567 , 95 S.Ct. at 1857 . It is this statement upon which the majority here relies. In response, however, to the Secretary’s second contention, which is the contention at issue in the present case, the Court

agree[d] with the Court of Appeals, for the reasons stated in its opinion, 502 F.2d 79, 86-88 (CA 3 1974), that there is no merit in the Secretary’s contention that his decision is an unreviewable exercise of prosecutorial discretion.

Id. at 567 n. 7, 95 S.Ct. at 1858 n. 7 (emphasis added). An examination of the cited pages in the appellate court’s opinion produces no language even hinting at a “strong presumption” of reviewability, but rather the contrary implication in its observation that “[n]ot every refusal by a Government official to take action to enforce a statute, however, is unreviewable,” Bachowski, supra, 502 F.2d at 87 (emphasis added). The court of appeals finds it necessary to carefully distinguish the type of enforcement responsibility given the Secretary under the L-MRDA from the more typical case in which a decision not to enforce is unreviewable. Bachowski, supra, 502 F.2d at 87 -88 & n. 11. Thus, Dunlop reaffirms the principle of general unreviewability of enforcement decisions.

The other Supreme Court cases cited by the majority also do not support its position. Abbott Laboratories v. Gardner, 387 U.S. 136 , 87 S.Ct. 1507 , 18 L.Ed.2d 681 (1967), has even less to do with the present case than the first portion of Dunlop , involving neither the “agency discretion” exemption ( 5 U.S.C. § 701 (a)(2)) nor agency enforcement action. The Court held that Congress did not intend to “preclude judicial review” ( 5 U.S.C. § 701 (a)(1)) of agency rule-making prior to the enforcement stage. 387 U.S. at 142 , 87 S.Ct. at 1512. Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 410 , 91 S.Ct. 814, 820 , 28 L.Ed.2d 136 (1971), at least involved the relevant provision of the APA; but it did not involve enforcement discretion, and therefore the Court’s dictum to the effect that the exception for action “committed to agency discretion” is to be construed narrowly does not establish what the majority suggests. Unless one assumes that the Supreme Court was casually setting forth a rule wildly out of accord with prior case law, the obvious meaning of “narrow construction” was an *1194 application that would limit the exception to enforcement discretion and to a few other matters where either a strong tradition (and therefore congressional expectation) of judicial abstention exists, see, e.g., Curran v. Laird, 420 F.2d 122 (D.C.Cir.1969) (en banc), or a specific legislative commitment to discretion is evident, see Dunlop v. Bachowski, supra, 421 U.S. at 595 , 95 S.Ct. at 1870 (Rehnquist, J., concurring in the result in part and dissenting in part). 1

Cases in this circuit no more support narrowing the “agency discretion” exception by an across-the-board application of a “presumption of reviewability” than do the decisions of the Supreme Court. 2 We have sat en banc to consider the subject on two occasions. In the first case, Curran v. Laird, supra, we held that the Executive’s decision not to withdraw ships from the national defense reserve fleet fell within the “span of executive actions — pertaining to the fleet’s establishment, expansion, curtailment, maintenance and use — that are ‘committed to agency discretion’ within the meaning of [ 5 U.S.C. § 701 (a)(2)].” 420 F.2d at 128 . Our opinion made no mention of a “presumption of reviewability,” though the dissent (per Wright, J.) did. Id. at 140 . Our second en banc consideration of this general subject was Adams v. Richardson, 480 F.2d 1159 (D.C.Cir.1973), which not only involved the “agency discretion” exception, but considered its precise application to enforcement discretion. We again did not invoke any “presumption of reviewability” of enforcement decisions, but to the contrary found it necessary to distinguish the prosecutorial discretion cases on the grounds that (1) those cases “do not support a claim to absolute discretion,” (2) the assertion in the case before the court was not merely deficient enforcement “with regard to a few school districts in the course of a generally effective enforcement program,” but rather “conscious[] and express[] [adoption of] a general policy which is in effect an abdication of ... statutory duty,” and (3) the nonenforcement consisted not merely of inaction but of “affirmatively continuing] to channel federal funds to defaulting schools.” Id. at 1162 (emphasis in original). (The centrality of the second point was emphasized in our recent en banc decision of Adams v. Bell, 711 F.2d 161, 166 , 167 & n. 35 (D.C.Cir.1983).)

Nor are the panel opinions of this and other circuit courts cited by the majority any more probative of the novel assertion of a “presumption of reviewability” of enforcement decisions. None of them sets forth such a startling proposition, and all of them display special circumstances overcoming the usual presumption of nonreviewability. Thus, the case requiring an agency to enforce the restrictions of the Davis-Bacon Act (Carpet, Linoleum & Resilient Tile Layers, Local Union No. 419 v. Brown, 656 F.2d 564 (10th Cir.1981)) involved the third factor that we found operative in Adams v. Richardson, supra: the agencies were channeling federal funds to those in violation of the law. The split decision of a panel of this court in Environmental Defense Fund, Inc. v. Ruckelshaus, 439 F.2d 584 (D.C.Cir.1971), relied upon a statement by the agency making the very finding that triggered mandatory enforcement action under the governing statute; and while reviewing the generalized standard of proof which the agency had estab *1195 lished for the exercise of its discretion, the court pointedly declined to review its finding as to whether that standard had been met. 439 F.2d at 595-96 . (In the instant case, by contrast, the agency has made a finding that there is no “serious danger to the public health dr ... blatant scheme to defraud,” Letter from the Commissioner at 3, Jt.App. at 88, and it is this finding which the majority claims power to review.) The panel decisions of this court in Natural Resources Defense Council, Inc. v. SEC, 606 F.2d 1031 (D.C.Cir.1979), and WWHT, Inc. v. FCC, 656 F.2d 807 (D.C.Cir.1981), are not apposite, since they involved refusal to conduct rulemaking, rather than enforcement discretion in the sense of a refusal to proceed against alleged law violators.. And even in that context, the dicta of these cases do not support a presumption of reviewability in the sense here at issue. Natural Resources Defense Council states that the totality of § 701(a) (which reads “This chapter applies, according to the provisions thereof, except to the extent that — (1) statutes preclude judicial review; or (2) agency action is committed to agency discretion by law”) creates a “strong presumption of reviewability,” 606 F.2d at 1043 . Undoubtedly so. But that does not suggest that the presumption applicable to the totality of agency action is applicable to the narrow category of action consisting of refusal to adopt rules. To the contrary, the opinion makes it clear that with respect to that category “the relevant factors incline against reviewability,” id. at 1047 . As for WWHT, its dicta (which again are applicable only to the weaker case of refusal to adopt rules) are based upon a terminology which resolves the problems at issue here under the rubric of “scope of review” rather than “reviewability.” 3

The short of the matter is, that far from there being a “presumption of reviewability” with regard to enforcement determinations, the well known presumption is precisely the contrary. As the author of the majority opinion noted in dissenting from our en banc decision in Curran v. Laird, supra:

This “presumption of reviewability” is often regarded as reversed for certain types of Executive actions. Actions relating to the conduct of foreign or military affairs have not normally been reviewed by the courts, absent exceptional circumstances.... Thus we would not normally review the failure of the Executive to exercise emergency powers granted it for the purpose of meeting the military needs of the nation.

420 F.2d at 141 (Wright, J., dissenting). Enforcement decisions are unquestionably one of those “certain types of Executive action” to which this analysis properly applies. Even those who favor general review of decisions not to undertake enforcement action acknowledge that it will constitute a change in the law, since “reluctance to review remains entrenched.” Note, Judicial Review of Administrative Inaction, 83 Col.L.Rev. 627, 658 (1983). The majority seeks to dismiss established law by citing Professor Davis for the proposition that “the real ‘key lies in recognizing that the law has been in transition and that the case law * * [is now] strongly on the side of review-ability.’ ” Maj.Op. at 1187-1188, quoting from 2 K. Davis, Administrative Law Treatise § 9.6 at 239-40 (2d ed. 1979). This also is misleading. The quoted statement was made in the context of a discussion dealing largely with the enforcement discretion of criminal prosecutors, which had previously been asserted to be not merely (as the cases establish with regard to the enforcement *1196 discretion of administrative agencies) generally unreviewable, but entirely so. As a refutation of that bald proposition, whether applied to criminal or administrative enforcement, Davis’s statement is entirely accurate. But neither the statement, nor the cases, upon which it relies, support any general presumption of reviewability in the sense at issue here.

Proceeding, then, from the premise that an agency’s exercise or (a fortiori) nonexercise 4 of its enforcement discretion is generally unreviewable, one must consider whether there are any special circumstances justifying departure from that general rule in the present case. I am aware of none, and the majority opinion has assuredly not suggested any. Its reliance (footnote 29) upon the mandatory terms of the statute (“any person who violates § 331 ‘shall’ be imprisoned or fined,” citing 21 U.S.C. § 333 (a) (1976)) is half-hearted, and properly so. Most of the criminal code is cast in such mandatory terms, and yet prosecutors’ discretion not to indict is the archetype of unreviewable enforcement discretion. Besides which it is the Justice Department, rather than the FDA, that executes § 333(a). See 21 U.S.C. § 335 .

The opinion places its major reliance upon what it calls the FDA’s “Policy Statement,” the relevant portion of which is set forth at page 1186. Even if an agency’s failure to follow its rules regarding enforcement policy is reviewable (which may not be the case, see e.g., United States v. Snell, 592 F.2d 1083 (9th Cir.1979)); and even if the statement in question constituted an agency rule (which it did not, as will be discussed below); it is impossible to see how that “rule” has been violated here. The statement is full of flexible terms, the precise application of which was obviously intended to be, and could properly be, left to the discretion of the agency — for example, whether an unapproved use has become “widespread”; whether it “endangers the public health”; what particular action “is warranted to protect the public”; the meaning of the phrase “when necessary.” Legal Status of Approved Labeling for Prescription Drugs; Prescribing for Uses Unapproved by the Food and Drug Administration, 37 Fed.Reg. 16503, 16504 (1972). Moreover, the last paragraph of the quoted excerpt, which is apparently intended to give more precise content to the first paragraph, suggests that the intended actions, as discretionary as they are, are only to be taken where “any person in the chain of distribution ... does anything that directly or indirectly suggests to the physician or to the patient that an approved drug may properly be used for unapproved uses.” Id. It applies, in other words, where misbranding has occurred — which (as will be discussed in the next part) is not the ease here.

More fundamentally, however, the quoted statement is not an agency rule, and is indeed not even an authoritative policy statement. The majority’s opinion gives it the title “Policy Statement” (which it never bore), see Maj.Op. at 1176 & n. 3, and uses that conclusory designation throughout. In fact, however, the statement was part of the policy justification set forth in a Notice of Proposed Rulemaking, with respect to a proposal that was never adopted. The paragraphs quoted by the majority, id. at 1186 , and particularly the first of them, recite principles that are more specifically embodied in the subsequently recited text of the proposed amendment to 21 C.F.R., see 37 Fed.Reg. at 16504. It is remarkable to suggest that, although the text of the rule was rejected, the substance of that text was authoritatively adopted by the mere recital of it in the Notice of Proposed Rulemaking. Even if one is to accept (as I am not inclined to) the expansive dictum in two of our cases referred to by the majority that the term “rule” is “broad enough ‘to include nearly every statement an agency may make,’ ” Center for Auto Safety v. NHTSA, 710 F.2d 842, 846 (D.C.Cir.1983), *1197 quoting Batterton v. Marshall, 648 F.2d 694, 700 (D.C.Cir.1980), surely the exceptions must at least include statements of principles that the agency is thinking of adopting but ultimately decides not to adopt. It is the height of irrationality to equate the prologue of a rejected rule with (what was involved in Center for Auto Safety) an agency’s authoritative statement of its reason for withdrawing a notice of proposed rule. As Center for Auto Safety observed, quoting from Columbia Broadcasting System, Inc. v. United States, 316 U.S. 407, 416 , 62 S.Ct. 1194, 1199 , 86 L.Ed. 1563 (1942):

[I]t is the substance of what the [agency] has purported to do and has done which is decisive.

710 F.2d at 846 . Here, what the agency purported to do and did by its proposal and nonadoption was nothing — unless it was to disapprove the principles which the proposal and its prologue contain.

The opinion seeks to remedy this patent deficiency by observing that the “FDA still considers [the so-called Policy Statement] binding and to have substantive effect,” Maj.Op. at 1186. That is not so. The majority’s citation to support that conclusion is to a portion of the agency’s Memorandum in Support of the Defendant’s Motion to Dismiss, which (in turn referring to the Commissioner’s letter response to the Appellants’ request) only concedes that the Notice sets forth the agency’s position on “the legal status of approved labeling for prescription drugs.” Jt.App. at 131 n. 10. That legal status has nothing to do with the rigid and detailed enforcement obligations which the opinion suggests the agency has imposed upon itself.

Thus, there is ultimately no special factor to support the extraordinary assertion of authority to control the agency’s enforcement discretion in the present case — nothing, except the majority’s disagreement with the agency’s determination that no serious danger to the public health exists. But we are not the only public officials endowed with intelligence and worthy of trust, and our system of laws has committed the relative evaluation of public health concerns to others. Moreover, even if the clear erroneousness of an agency’s reason for not conducting enforcement activities were a proper basis for our intervention (which it is not), there is no clear error here. Without belittling the humane concerns associated with the present complaint, it must be acknowledged that the public health interest at issue is not widespread death or permanent disability, but (at most) a risk of temporary pain to a relatively small number of individuals (200, which the majority swells to 1,100 by including prisoners under sentence of death in states that have not adopted lethal injection statutes). Moreover, it is not a matter of pain versus no pain, but rather pain of one sort substituted for pain of another — and in all likelihood substitution of a lesser pain, since that is the principal purpose of the lethal injection statutes, 5 In these circumstances, it is *1198 hardly clear error to determine that lethal injection statutes do not pose a “serious danger to the public health,” justifying the diversion of FDA enforcement resources from other projects. Indeed, it seems to me the conclusion that they do pose a serious danger is fanciful. 6

As for the “anachronistic ring” which the majority hears in all this (Maj.Op. at 1188 n. 36): If the clear statements of this court, see Kixmiller and Action on Safety and Health, supra, become downright anachronisms in nine years, we should perhaps stop publishing our opinions. I fancy, however, that the sound which the majority heard was not an anachronistic ring at all, but the stifled cry of smothered stare decisis, or perhaps the far-off shattering of well established barriers separating the proper business of the executive and judicial branches.

II. Jurisdiction

If, however, the discretion issue is found against the FDA, it becomes necessary to establish the existence of FDA authority to take the requested action. The majority opinion fails to carry this point as well. It properly demolishes the agency’s reliance upon state action, see Maj.Op. at 1179-1181, but does not address the more serious obstacle: FDA jurisdiction depends upon the existence of misbranding under § 331(k), which cannot be established under the facts of this case.

The majority opinion concludes and summarizes its discussion on the jurisdictional point as follows:

The states’ lethal injection statutes purport to mandate the use of certain prescription drugs for a purpose not listed on their label.... FDA therefore must have jurisdiction to regulate such activity-

Maj.Op. at 1182 (emphasis added). This progression of thought does not follow unless the opinion establishes the intermediate proposition that “using prescription drugs for a purpose not listed on their label *1199 is unlawful under the FDCA.” It makes no attempt to establish that proposition — principally I think, because it is impossible to do so. The FDCA is directed at the sale and distribution of drugs rather than their use. The only provision that could remotely apply here is the prohibition against “the doing of any ... act with respect to ... a ... drug ... [which] results in such article being ... misbranded,” 21 U.S.C. § 331 (k). Even if one adopts the extraordinary notion that a person causes an article to be misbranded by simply using it for a purpose not stated on the label, § 331(k) would still not apply, since — in accordance with the Act’s focus upon sale and distribution rather than drug use — it requires that the misbranding occur “while such article is held for sale (whether or not the first sale) after shipment in interstate commerce.” Id. Here the drugs are in the possession of the states’ penal authorities. Under no conceivable interpretation of the English language could they be deemed “held for sale.”

The majority opinion points out, quite correctly, that § 331(k) was added to the Act with the intent of expanding its application. Given the fact that the Act’s primary application was to commercial transactions, it was a vast expansion to extend it to sales within a state after the first sale following shipment in interstate commerce. Nonetheless, that expansion was still limited to misbranding which occurs “while such article is held for sale.” This is a statute that bears criminal penalties, and it is simply not possible to disregard this qualification entirely, as the majority would do. It is sometimes improper to “look only to the statutory language,” Maj.Op. at 1181; but it is always improper — and opens the gates much more widely to judicial abuse — to ignore intentional statutory language entirely. Thus, I would acknowledge that the expansive intent of the amendment may be invoked to extend the statutory text to all “holding” that occurs while the drug is in the stream of commerce from manufacturer to ultimate consumer, see United States v. Wiesenfeld Warehouse Co., 376 U.S. 86, 92 , 84 S.Ct. 559, 563 , 11 L.Ed.2d 536 (1964). But not beyond that point. Given the statutory language “held for sale,” it is clear that when the House Report, quoted by the majority (p. 1181), said that the intent of the 1938 amendment was to “extend the protection of consumers contemplated by the law to the full extent constitutionally permissible,” it was using the word “consumer” in the sense of “purchaser” rather than “ingester” — just as “consumer protection” refers to shoppers rather than gourmets, and the consumer price index to purchasers rather than eaters. Thus, in the context of discussion of this statute, the majority’s notion of an “unwilling consumer,” Maj.Op. at 1182, does not compute.

None of the cases relied upon by the majority to support the principle of expansive interpretation involves an expansiveness beyond the field of actual sale. United States v. Sene X Eleemosynary Corp., 479 F.Supp. 970 (S.D.Fla.1979), involved the issue of whether the drug was “held for sale” when it was compounded with another substance for sale; United States v. 10 Cartons Labeled, in Part, “Hoxsey”, 152 F.Supp. 360 (W.D.Pa.1957), held (erroneously, as later case law has established) that a physician can misbrand a drug by using it, as part of a paid-for therapy, for an unapproved use. Even those sharing the majority’s apparent disdain for the plain meaning of statutory language would be hard pressed to explain why, to express the concept of any transmission to the ultimate ingester (as opposed to the ultimate purchaser) of the drug, Congress should choose the peculiarly inapt phrase “held for sale.” And the consequences that ensue are as implausible as the interpretation that produces them: It cannot seriously be thought that the householder who administers a drug to his child for an unapproved use, or who places it in an unmarked bottle on his medicine shelf to administer later for an approved use, violates the criminal provisions of the Food, Drug, and Cosmetic Act. In short, even giving the phrase “held for sale” the minimum content it can possibly contain without reading it out of the statute entirely, it does not cover the holding by a state for purposes of coerced administration of the *1200 drug to execute a sentence of capital punishment. In that context, the state is as much the ultimate consumer of the drug as it was of the electricity previously used for the same purpose; and the condemned prisoner executed by injection is no more the “consumer” of the drug than is the prisoner executed by firing squad a consumer of the bullets. Thus, misbranding has not occurred in the present case.

The majority opinion attempts, once again unsuccessfully, to condemn the agency out of its own mouth. The majority asserts that “[i]n both [prison clinical investigations and veterinary practices] FDA has rejected arguments that it does not have authority to regulate unapproved uses of approved drugs.” Maj.Op. at 1180. But examination of the citations shows that in the first of these situations the Commission was proceeding not under the more general portions of the statute at issue here, but under its specific authority to regulate clinical testing of drugs, see 21 U.S.C. §§ 355 (i), 357(d); 21 C.F.R. § 50.44 (1982); and that the second situation involved assertion not of authority to regulate use, but of authority to require proper labeling of drugs sold for a particular use. See United States v. Beuthanasia-D. Regular, [1979 Transfer Binder] Food Drug Cosm.L.Rep. (CCH) 1f 38265 (D.Neb.1979). Of course even if the FDA had- acknowledged what the opinion suggests, we are not in the habit of permitting an agency, to expand its authority by simply acknowledging the legitimacy of such expansion. Except for special provisions such as that alluded to above pertaining to clinical or investigative drug use, there is nothing in the statute giving the Commission “authority to regulate unapproved use of approved drugs.”

sk * *

The majority discerns that “the FDA is clearly refusing to exercise enforcement discretion because it does not wish to become embroiled in an issue so morally and constitutionally troubling as the death penalty.” Maj.Op. at 1192. I doubt that. As discussed above, there were plenty of well supported reasons for the Commission to refuse to act. If speculations as to motivation are to be indulged in, I would suggest that the agency was properly refusing to permit its powers and the laws it is charged with enforcing from being wrongfully enlisted in a cause that has less to do with assuring safe and effective drugs than with preventing the states’ constitutionally permissible imposition of capital punishment. This court should have done the same. It is our embroilment, rather than the FDA’s abstention, that is remarkable.

I would affirm on the basis of the opinion below.

. The majority asserts (footnote 25) that the Supreme Court itself has not interpreted Over-ton Park in this fashion, since it cited that case in Dunlop to support the proposition that the Secretary of Labor’s exercise of enforcement discretion was renewable. The majority overlooks, however, that both the portion of Dunlop in which the citation appears and the portion of Overton Park to which the citation refers, deal with the possible bar to review posed — not by § 701(a)(2), the “committed to agency discretion” provision — but by § 701(a)(1), which relates to statutory preclusion of judicial review. As explained above, I do not dispute that the latter provision requires a clear expression of legislative intent.

. The majority’s response (footnote 23) fashions a mixture out of this dissent’s quite separate assertions regarding Supreme Court case law, on the one hand, and the case law of this and other courts of appeals, on the other hand. I stand by the assertions, unmingled, and must leave it to the reader to dialyze the majority opinion.

. When particular agency action can be set aside by the courts only for a narrow category of alleged infractions (e.g., complete abdication of enforcement responsibility, see Adams v. Richardson, supra) it is possible to describe that situation by saying either that the agency action is not generally reviewable or that it is reviewable but “the scope of review [is] very narrow,” WWHT, Inc. v. FCC, supra, 656 F.2d at 809 . In this opinion I have used the former formulation — not only because I prefer it but because that is what the majority has used, as is evident from the fact that all of the issues here discussed are addressed in that portion of its opinion entitled “Judicial Review of Enforcement Discretion,” Maj.Op. at 1183, rather than the portion entitled “Scope of Judicial Review,” Maj.Op. at 1188.

. It is interesting to note that Professor Davis’s statement discussed in text applied only to decisions to prosecute. Even in the limited sense in which that statement is to be understood (availability of review in some instances) it is not extended to “prosecutors’ decisions not to prosecute,” as to which “the law is still somewhat unclear.” 2 K. Davis supra, § 9.6 at 244.

. See Gimlin, Administering the Death Penalty, Editorial Research Rep., Weekly Reminder Service, August 19, 1977:

The public hanging, with its circus atmosphere, disappeared in America early in this century. But then the electric chair, installed as a more humane device for snuffing out life, has produced a host of grim stories from witnesses at the death scene. The gas chamber, usually regarded by penal officials as an improvement over the electric chair, requires a sealed room and a strapped-in victim ....

Oklahoma and Texas have since moved in a new direction, providing for execution by lethal injection. “I think and I hope this will provide some dignity with death,” Texas Gov. Dolph Briscoe said last May when he signed the injection bill into law, ... Briscoe could have added that for years veterinarians have considered the lethal injection the most “humane” way of putting hopelessly crippled or diseased animals out of their misery.

The Report cited repeatedly by the majority, Royal Commission on Capital Punishment, 1949-1953 Report (1953), said that “intravenous injection, if practicable, would fulfill our three requisites [of humanity, certainty and decency] better than any other method.” Id. at 257. The Report doubted, however, the method’s practicability, on the basis of medical knowledge and technique thirty years ago, and therefore rejected it for the time being, but recommended, “unanimously and emphatically, that the question should be periodically examined, especially in the light of progress made in the science of anaesthetics, with a view to a change of system being proposed to Parliament *1198 as soon as it can be shown that there are no longer any grounds for the doubts that now deter us from recommending it.” Id. at 261. The Report’s verdict concerning the alternative means of execution that these 200 prisoners will undergo if lethal drugs are not used is a good deal less qualified: “[W]e cannot recommend that either electrocution or the gas-chamber should replace hanging as the method of judicial execution in this country. In the attributes we have called ‘humanity’ and ‘certainty’ the advantage lies, on balance, with hanging.” Id. at 256.

. The majority opinion seeks to preclude consideration of the realities of the matter by asserting that the Commissioner “irrebuttably presume[d] that ‘duly authorized statutory enactments * * * [which further] proper State functions’ cannot, as a matter of law, pose such a danger to the public.” Maj.Op. at 1190. That is not so. What the Commissioner wrote was that “[w]e cannot conclude that those dangers [of a serious risk to the public health or a blatant scheme to defraud] are present under State lethal injection laws, which are duly authorized statutory enactments in furtherance of proper State functions.” Letter to the Commissioner at 3, Jt.App. at 88 (emphasis added). If, in that excerpt, the phrase “lethal injection” were omitted, or the word “which” were replaced with the words “since they,” it would be possible to assert that the Commissioner was maintaining that no serious health hazard was presented only by virtue of the fact that these were state laws. As written, however, the excerpt refers to the specific nature of these laws (“lethal injection laws”) — and therefore does not exclude from the Commissioner’s calculus the factors discussed in text. The descriptive phrase, “which are duly authorized, etc.,” reflects an important element in the Commissioner’s analysis, but not necessarily the only dis-positive one. To interpret the statement as an assertion that whatever is provided by state law will not be challenged by the FDA would render it quite incompatible with the Commissioner’s statement in the preceding paragraph that “[w]e do not say, of course, that provisions of State law that are contrary to the Federal Food, Drug, and Cosmetic Act are lawful.” Id.

There is no authority for the majority’s assertion that the Commissioner’s statement of “no serious danger to the public health” is not enough, but that he must “give rational reasons” why that is so. Maj.Op. at 1190 n. 43. Indeed, I know of no authority for the proposition that an agency must provide any ground for its refusal to comply with a request for exercise of its enforcement discretion, much less that it must provide the analysis underlying that ground. This seems to me one of those novel procedural requirements we have been told not to invent. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519 , 98 S.Ct. 1197 , 55 L.Ed.2d 460 (1978).