dissenting:
In holding that we can review withdrawal of proposed regulations an agency had no duty to adopt, my colleagues overlook the sea-change in administrative law wrought by Heckler v. Chaney, 470 U.S. 821, 831-32, 105 S.Ct. 1649, 84 L.Ed.2d 714 (1985), which held that we have no authority to review an agency’s discretionary decision not to act. See The Supreme Court — Leading Cases: Judicial Review of Agency Inaction, 99 Harv. L.Rev. 264, 269 (1985) (hereafter “Judicial Review of Agency Inaction ”) (“Chaney represents a significant departure from the trend of recent administrative law.”). Failure to appreciate the fundamental distinction between agency action and inaction permeates every aspect of the majority’s analysis, starting with the threshold question of standing.
Standing
1. The majority wastes much toner trying to show that USDA’s Draft Policy at issue here would, if adopted, ameliorate the conditions plaintiffs complain about. Maj. op. at 834-836. Even if this were true, which it is not, see pp. 845-847 infra, it doesn’t matter. In determining whether plaintiffs have standing, the question is not whether the agency’s proposed action would have benefitted them, but whether *845this lawsuit will do so. The Draft Policy only matters if the court can order, or at least move the agency towards, its adoption.
Certainly we cannot order USDA to adopt the Draft Policy.1 No court I am aware of has ever ordered an agency to adopt a particular regulation, and for good reason.2 Adoption of regulations is an executive function that no court has authority to perform. Even where the agency is required by law to act, and a court finds that it failed to do so, the court cannot tell the agency what regulation to adopt; rather, it must remand for further action. See, e.g., Envtl. Def. Ctr., Inc. v. EPA, 344 F.3d 832, 863 (9th Cir.2003); Defenders of Wildlife v. Norton, 258 F.3d 1136, 1146-47 (9th Cir.2001). In such cases it can at least be said that the court nudged the agency towards a result favorable to the plaintiffs by correcting the agency’s misinterpretation of the statute and thus narrowing its discretion. But the agency here has no duty to act; it fully discharged its statutory responsibility by adopting the regulations approved by the DC Circuit. See Animal Legal Def. Fund, Inc. v. Glickman, 204 F.3d 229, 235 (D.C.Cir.2000) (“Glickman II ”). The most we can do is tell the agency that its reason for rejecting the Draft Policy is insufficient. See maj. op. at 844. This won’t ensure that the agency will adopt the Draft Policy or narrow its discretion in any meaningful way. The agency is free to open up the proceedings for further comment, adopt a different policy, or again reject the policy, this time giving better reasons.3
I suppose that plaintiffs could then bring another lawsuit and we could, once again, disapprove the agency’s reasons. But no matter how often the kabuki is repeated, it can never give plaintiffs what they want— an order that materially limits the agency’s discretion to reject the Draft Policy. This is because the agency has no legal duty to act, and there is thus no law we can bring to bear in narrowing its discretion. The agency can refuse to adopt the Draft Policy again and again, citing reasons of policy and convenience that we may not second-guess. See pp. 847-850 infra. Plaintiffs are pushing a long and very limp string, and no matter how hard they try, they can never get the object at the far end to move.
2. Let’s assume, nevertheless, that plaintiffs could obtain an order forcing the agency to adopt the Draft Policy. Would the policy, if adopted, redress their injuries, as that term is understood in our standing jurisprudence? See Allen v. *846Wright, 468 U.S. 737, 751, 104 S.Ct. 3315, 82 L.Ed.2d 556 (1984). Plaintiffs have the burden to show they have standing and cannot rely on “[sjpeculative inferences ... to connect their injury to the challenged actions.” Simon v. E. Ky. Welfare Rights Org., 426 U.S. 26, 45, 96 S.Ct. 1917, 48 L.Ed.2d 450 (1976).
A favorable result that hinges on a third party’s response is generally too speculative to support standing. In Linda R.S. v. Richard D., 410 U.S. 614, 618, 93 S.Ct. 1146, 35 L.Ed.2d 536 (1973), for instance, the Supreme Court held that a mother didn’t have standing to enforce a statute that imprisoned dead-beat dads; while such prosecution might have resulted in future child-support payments, the result was “speculative.” Likewise, Allen v. Wright held that plaintiffs had no standing to challenge the tax exempt status of segregated private schools because there was no assurance the schools would desegregate out of fear of losing their exemptions. 468 U.S. at 758-59, 104 S.Ct. 3315. In applying these precedents we have held that an injury is not redressed if any benefit to plaintiffs is “conditioned on the discretionary behavior of any third party.” Fernandez v. Brock, 840 F.2d 622, 627 (9th Cir.1988).
Of course, if the third party has no choice, its response is not discretionary and may provide a basis for standing. But this is only so if the regulation in question is sufficiently coercive. Maj. op. at 835-836. The Draft Policy here is not coercive; it is, as the majority recognizes, a “safe harbor.” See id. at 830, 831, 835, 836, 840, 841. Entities that house primates need not follow the Draft Policy; they may lawfully choose alternate means of statutory compliance. If fear of incarceration in Linda R.S. and loss of tax benefits in Allen v. Wright were insufficiently coercive to give plaintiffs there standing, I don’t see how a wholly permissive regime — one that expressly allows for alternate means of compliance — can give these plaintiffs standing.
The majority mistakenly relies on the representation of the intervenor, the National Association for Biomedical Research (NABR). NABR takes this view, however, based on its interpretation of the Draft Policy as mandatory — a position the majority correctly rejects. Maj. op. at 839. Once we have determined that the Draft Policy is merely a safe harbor, NABR’s members will be free, just like everyone else, to abide by the policy or not, as they see fit. Certainly, nothing NABR has said in this litigation can bind its members to follow the policy, should it be implemented. Because there is no way of knowing whether any particular facility will comply, plaintiffs cannot show that they will benefit from the policy because the facilities they are concerned with may not.
What intervenor says doesn’t matter anyhow, because it cannot speak — nor does it purport to speak — for all facilities that house primates. “NABR is an association of more than 300 universities, colleges, medical schools, veterinary schools, research institutes, research companies, and pharmaceutical manufacturers, as well as individual biomedical researchers, engaged in the care and use of animals in biomedical research.” Yet the single example on which the majority bases standing involves Terry, a chimpanzee housed in a zoo. Zoos differ from the research-industry entities that make up NABR’s membership. They’re typically cash-strapped non-profit organizations with significant public support in the local community. They need not fear that sensitive and expensive research projects will be disrupted if they fall out of compliance. Their risk calculus might be quite different from that of organizations represented by NABR. The Southern Ne*847vada Zoological-Botanical Park where Terry is housed, a facility already certified by USDA as compliant, may well choose not to comply with the Draft Policy.
But let’s assume, for the sake of argument, that Terry’s zoo did feel it had to follow the Draft Policy. I don’t see how we can be sure that such compliance would ameliorate Buchanan’s injury. The gravamen of her complaint is that Terry is lonely because he is housed by himself. The majority concludes that compliance with the Draft Policy would alleviate Terry’s suffering, pointing to the following language: “[i]n order to address the social needs of nonhuman primates under § 3.81(a), the [zoo’s environmental enhancement] plan must provide for each primate of a species known to be social in nature to be housed with other primates whenever possible.” Maj. op. at 834 (quoting Draft Policy on Environmental Enhancement for Nonhuman Primates, 64 Fed.Reg. 38,145, 38,147 (proposed July 15, 1999)) (alterations in original). The majority zooms in on “must” but gives no effect to “whenever possible.” It also overlooks the language that exempts facilities from this requirement if there is “[d]oeumented unavailability of compatible individuals.” 64 Fed.Reg. at 38,147. In other words, the zoo must only house Terry with another chimpanzee if it has another chimpanzee to house him with. Buchanan could easily have alleged that Terry’s zoo had a compatible primate, but she didn’t. For all we know — for all the record reflects — Terry is the only chimp at the Southern Nevada Zoological-Botanical Park, or the zoo may have other chimps that can’t be moved without disruption to themselves or their current companions. Even if the zoo believed it needed to comply with the Draft Policy, it could do so by documenting the unavailability of another compatible animal. To conclude that Buchanan has established standing is to make speculation the hallmark of redressability.4
Reviewability
1. The majority concludes that the agency’s failure to adopt the Draft Policy is reviewable because the Draft Policy itself, if adopted, would have been reviewable. Maj. op. at 840-841. This doesn’t follow; adoption and nonadoption of regulations are asymmetrical events. Regulations change the law, and thus can sharply affect the legal interests of private parties. Failure to adopt regulations leaves rights and responsibilities unchanged; it leaves the status quo in place. See Kennecott Utah Copper v. U.S Dep’t of the Interior, 88 F.3d 1191, 1207 (D.C.Cir.1996) (“Be*848cause [the agency’s] decision to withdraw the [proposed regulation] did not alter substantive legal obligations under previously published regulations, the agency’s decision to withdraw the document did not constitute a ‘regulation’ within the meaning of § 113 of CERCLA.”). Because parties are not entitled to an agency’s discretionary regulation, they can’t claim to be aggrieved when the agency fails to adopt it.
When an agency adopts regulations, a court can judge whether those regulations are consistent with the agency’s grant of authority (under the broad limits of Chevron USA Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 844-45, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984)), and whether they constitute an abuse of discretion. There is, in the words of the Supreme Court, “law to apply.” Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 410, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971). But an agency may choose not to adopt discretionary regulations for a variety of reasons, many of which a court can’t review: a change in policy; a lack of enforcement resources; a lack of scientific expertise to address the problem at this time; a change in direction based on a determination that the problem is better addressed some other way. All of these matters are, in the words of the Administrative Procedure Act, “committed to agency discretion by law.” 5 U.S.C. § 701(a)(2). There is nothing for a court to review.
There was a time when the Supreme Court (like the majority here) disdained the distinction between agency action and inaction. See, e.g., Rochester Tel. Corp. v. United States, 307 U.S. 125, 143, 59 S.Ct. 754, 83 L.Ed. 1147 (1939). But the Court turned its back on this caselaw in Heckler v. Chaney, 470 U.S. 821, 105 S.Ct. 1649, 84 L.Ed.2d 714 (1985), which held that an agency’s decision not to act is, under ordinary circumstances, unreviewable.
Chaney involved death row inmates who claimed that the FDA had shirked its duty by failing to prevent states from using drugs for lethal injections. According to the inmates, these drugs had not been proven “safe and effective” for use in human executions, as required by the applicable statute. Id. at 824, 105 S.Ct. 1649. In rejecting this contention, the Court closely examined the text of the APA, and concluded that an agency’s failure to act is insulated from judicial review in two circumstances. First, under 5 U.S.C. § 701(a)(1), an agency decision may not be judicially reviewed “when Congress has expressed an intent to preclude judicial review.” 470 U.S. at 830, 105 S.Ct. 1649. Neither here nor in Chaney is this provision applicable. But the Court went on to hold that judicial review is also precluded under 5 U.S.C. § 701(a)(2) where “no judicially manageable standards are available for judging how and when an agency should exercise its discretion.” 470 U.S. at 830, 105 S.Ct. 1649. The Court held that an agency’s discretionary decision not to enforce a statute — even though it has the authority to do so — is unreviewable.
In reaching this conclusion, the Court noted that the agency’s decision not to exercise its enforcement powers is “generally committed to [its] absolute discretion.” Id. at 831, 105 S.Ct. 1649. This, the Court said, “is attributable in no small part to the general unsuitability for judicial review of agency decisions to refuse enforcement.” Id. “The reasons for this general unsuitability,” the Court continued, “are many.” Id. Among them, the Court listed the following:
[T]he agency must not only assess whether a violation has occurred, but whether agency resources are best spent on this violation or another, whether the *849agency is likely to succeed if it acts, whether the particular enforcement action requested best fits the agency’s overall policies, and, indeed, whether the agency has enough resources to undertake the action at all....
In addition to these administrative concerns, we note that when an agency refuses to act it generally does not exercise its coercive power over an individual’s liberty or property rights, and thus does not infringe upon areas that courts often are called upon to protect. Similarly, when an agency does act to enforce, that action itself provides a focus for judicial review, inasmuch as the agency must have exercised its power in some manner. The action at least can be reviewed to determine whether the agency exceeded its statutory power.
Id. at 831-32, 105 S.Ct. 1649.
While Chaney spoke directly to an agency’s enforcement actions, its teachings sweep more broadly. Regulations implicate precisely the same concerns addressed in Chaney; one can properly view the adoption — or non-adoption — of regulations as an enforcement decision.5 In adopting regulations, the agency implements the mandate of its governing statute by coercing the conduct of private parties. Regulations channel the agency’s enforcement efforts and commit agency resources to a particular interpretation of the statute. A decision to promulgate regulations, just like a decision to enforce, amounts to an exercise of agency authority, which “at least can be reviewed to determine whether the agency exceeded its statutory powers.” Id. at 832, 105 S.Ct. 1649. Failure to adopt discretionary regulations, by contrast, may reflect considerations besides the applicable statute, such as a change in policy, a lack of resources, a decision to pursue a different course of action or a decision to gather additional information before acting. Such failure to adopt cannot normally be tested against the applicable statute, because it may be based on non-statutory considerations of policy and convenience that are within the agency’s inherent authority.
The failure to adopt regulations, like failure to enforce, may be reviewed in certain narrow circumstances. The first, and most common, is where the agency has a legal duty to act. Id. at 832-35, 105 S.Ct. 1649. The typical such case is where Congress has directed the agency to take a particular regulatory action — such as listing species as endangered. See Defenders of Wildlife v. Norton, 258 F.3d 1136, 1137-38 (9th Cir.2001); see also Greater L.A. Council on Deafness, Inc. v. Baldrige, 827 F.2d 1353, 1361 (9th Cir.1987). The second is where the agency refuses to act based solely on a belief that it lacks jurisdiction. Chaney, 470 U.S. at 833 n. 4, 105 S.Ct. 1649; see, e.g., Transpacific Westbound Rate Agreement v. Fed. Mar. Comm’n, 938 F.2d 1025, 1026-27 (9th Cir.1991) (“The Commission’s order declined jurisdiction over parts of shipping agreements that regulate wholly foreign transportation.... Instead, the Commission held that it had jurisdiction only over the United States-foreign portion of the agreements.”). The third is where the agency’s refusal to act violates constitutional rights. Chaney, 470 U.S. at 838, 105 S.Ct. 1649. Finally, a court might have jurisdiction where “the agency has ‘consciously and expressly adopted a general policy’ that is so extreme as to amount to an abdication of its statutory responsibilities.” Id. at 833 n. 4, 105 S.Ct. 1649 (quoting Adams v. Richardson, 480 F.2d 1159 (D.C.Cir.1973) (en banc)).
*850None of these exceptions apply. There is certainly no claim that the agency believed it lacked jurisdiction to adopt the Draft Policy. Nor is there any suggestion that the agency abdicated its statutory responsibilities or violated constitutional rights. And, most significantly, we know for a fact that the agency had no statutory duty to act because the DC Circuit found the regulations the agency had adopted sufficient. Glickman II, 204 F.3d at 235. The agency need do no more. Adoption of the Draft Policy was an entirely discretionary decision, one the agency could choose to eschew based on a variety of non-legal considerations.
The majority remands for a determination “whether the USDA’s withdrawal of the Draft Policy was arbitrary and capricious in violation of § 706(2) of the APA,” maj. op. at 844, but does not say how that determination can be made. The implication seems to be that review should be on the record of the rulemaking proceedings. Id. at 40. But it’s not clear why this should be so. Absent a statutory duty to act, an agency need not adopt regulations, even if all public comments submitted favor them (which is certainly not the case here). The agency may decide not to adopt regulations because of a change in administrations, or some other change in policy.6 Or, the agency may decide to withdraw a proposed regulation in response to a change in law.7 The agency might also choose a different path.8 An agency’s decision not to proceed with a regulation — just like a decision not to enforce — can properly be based on a variety of judicially unreviewable considerations.
The majority’s implication that withdrawal of the policy must be justified by the public comments is alarming. It locks an agency into a course of action — a course it cannot abandon, unless it be for reasons a court finds sufficient on the administrative record. This not only improperly constrains agency discretion under Chaney, it discourages agencies from proposing discretionary regulations, lest they be stuck with them if they cannot convince a federal court that the record supports abandonment. If this was ever the law, it’s certainly not the law today.
The majority relies on several cases from the DC Circuit, see maj. op. at 841-844, but those cases have been overtaken by Chaney.9 Indeed, the case on which *851my colleagues principally rely, Natural Resources Defense Council, Inc. v. SEC, 606 F.2d 1031 (D.C.Cir.1979) (“NRDC”), is listed in Justice Marshall’s concurrence as repudiated by the majority.10 Chaney, 470 U.S. at 850 n. 7, 105 S.Ct. 1649 (Marshall, J., concurring); see also Judicial Review of Agency Inaction, supra, at 264 n. 1 (listing NRDC among the cases inconsistent with Chaney). Justice Marshall strongly disagreed that agency refusals to act are presumptively non-reviewable, but wrote only for himself; eight Justices joined in the opinion.
2. Even on its own terms, the majority is wrong. Assuming standing, assuming reviewability, assuming the cases the majority relies on are still good law, judicial review is still not justified. The majority extracts from the DC Circuit cases two requirements for reviewing non-adoption of a regulation and finds both satisfied. Maj. op. 842-844. In fact, there are two additional requirements. First, the agency’s regulatory process must have come to an end. See Ctr. for Auto Safety v. Nat’l Highway Traffic Safety Admin., 710 F.2d 842, 847-48 (D.C.Cir.1983) (action not ripe if the agency “has left open the possibility that the agency will initiate new rulemak-ing proceedings at some point in the future”). Second, the agency must have “explained in detail its reason for not adopting those rules.” NRDC, 606 F.2d at 1047. Neither requirement is satisfied here.
a. It is perfectly clear that the agency’s decision-making process has not come to rest. Dr. Gipson’s announcement that the rulemaking had been terminated adds as follows: “USDA and NIH had agreed to develop best management practice guidelines concerning nonhuman primates.” The agency has not abandoned its effort to give further guidance as to the management of primates. Rather, it has shifted focus toward other means of implementing that objective.
In Center for Auto Safety, on which the majority relies, the agency terminated rulemaking, but “left open the possibility that the agency [would] initiate new rule-making proceedings at some point in the future.” 710 F.2d at 847 (emphasis added). Because the agency still “ha[d] a chance ... to amend” its existing regulations, the DC Circuit concluded that judicial review of the refusal to do so at that time would have “entangl[ed] [the courts] in abstract disagreements over administrative policies.” Id. at 848 (quoting Abbott Labs. v. Gardner, 387 U.S. 136, 148, 87 S.Ct. 1507, 18 L.Ed.2d 681 (1967)). Here we have not merely a possibility that the agency will resume the regulatory process at some unspecified future time; we know it’s doing so now. If review was prema*852ture in Center for Auto Safety, it is previa-ble here.11
The majority’s ruling also conflicts with Ecology Center, Inc. v. United States Forest Service, 192 F.3d 922, 926 (9th Cir.1999), which held that the monitoring of forest activity was not reviewable. The Forest Service there published reports that were not interlocutory or tentative, but the monitoring itself was one of the “steps leading to an agency decision, rather than the final action itself.” Id. at 925. Thus, it did not “mark the consummation of the agency’s decision making process,” id., and review was premature. Here, Dr. Gipson’s declaration is not the end point of the agency’s decision making, but only a step along the way. It is no more reviewable than the reports in Ecology Center.
Review of interlocutory agency statements hopelessly entangles the judiciary in the administrative process. The agency here cannot have two policies on the same issue: It can either adopt something like the Draft Policy, or it can adopt best management practice guidelines like the ones it is now considering. By allowing this lawsuit to proceed, the majority leaves open the possibility that the agency will be forced or coaxed into adopting the Draft Policy.12 Even if the agency prevails before the district court, further agency proceedings will be delayed at least until the next appeal — which may be years away. Until this lawsuit is over, the agency cannot implement best practice guidelines, which may take the agency in a different direction from the Draft Policy. The majority thus interferes with the agency’s ongoing consideration of a live issue.
b. The majority ignores an additional requirement in the DC Circuit case law— that the agency “explain[] in detail its reasons for not adopting those rules.” NRDC, 606 F.2d at 1047.
In the cases the majority cites, see maj. op. at 841-844, the agencies all provided reasons a court could evaluate. NRDC, the cornerstone of the majority’s argument, involved “lengthy explanatory statements,” id. at 1039, that the DC Circuit repeatedly referenced in affirming the agency’s decision. Id. at 1058-59, 1061-62. In Professional Drivers Council v. Bureau of Motor Carrier Safety, 706 F.2d 1216 (D.C.Cir.1983), the agency published its reasons for terminating its rulemaking, id. at 1220, and the DC Circuit upheld the agency by reviewing the “relevant, permissible factors” upon which the agency relied. Id. at 1222. In Center for Auto Safety, the agency also published its reasons for withdrawing the policy, 710 F.2d at 844, but the court never reached the issue. See p. 851 supra. And in International Union, the agency provided three reasons, which the court then reviewed. 358 F.3d at 44.
By contrast, the agency here hasn’t provided a detailed explanation as to why it abandoned the Draft Policy, or any reason at all. We have only Dr. Gipson’s an*853nouncement that “the USDA was not going forward with the Draft Policy on Environmental Enhancement for Nonhuman Primates.” Without an explanation for the agency’s refusal to adopt the Draft Policy, it is impossible to determine whether its decision is arbitrary or capricious. Even if we assume that the agency did so for reasons contained in the record, we don’t know what those reasons are. Judicial review would thus “have an undesirably abstract and hypothetical quality.” NRDC, 606 F.2d at 1047.
What are the district judge and the parties to do on remand? The judge can certainly review the record for comments supporting and opposing the Draft Policy, but how is he to determine which of them are legitimate and which aren’t? What weight is he to give the various arguments? Considering and weighing public comments, then deciding whether to adopt, reject or modify a proposed rule is a quintessentially administrative function. It calls for the kind of expertise a district judge does not have and the kind of policy judgment a district judge is not competent to make. It is simply not a judicial function. Reviewing an agency decision where there is no agency decision to review is much like making ham and eggs without ham or eggs. It is precisely for this reason that the DC Circuit cases on which the majority relies require an explicit and detailed statement of reasons from the agency as a condition for judicial review. By ignoring this crucial requirement, the majority sends the parties and the district court on a fool’s errand.
* * *
The majority expands the law of standing beyond recognition. It unmoors administrative law from sound principles of judicial review, and insinuates the federal courts into sensitive policy judgments that are the exclusive province of the Executive Branch. It ignores the teachings of the Supreme Court and misapplies the precedents it relies on. It will cause no end of mischief. Count me out.
. An agency’s decision to adopt regulations is thus fundamentally different from a failure to adopt. When an agency adopts regulations that turn out to be contrary to law, a court can order the agency to set them aside. See, e.g., Ecology Ctr., Inc. v. Austin, 430 F.3d 1057, 1071 (9th Cir.2005). This is concrete relief that will benefit the plaintiff directly (presuming he can show that he would be injured by adoption of the regulations). A court can grant no commensurate relief when a party complains of an agency’s failure to adopt a discretionary regulation.
. Whether one views the proposed policy as an interpretive regulation or a safe harbor, there is no doubt that, once adopted, the policy would tie the agency's hands. Those who comply with the policy will be entitled to count on it if the agency attempts to bring an enforcement action in the future. See, e.g., Cmty. Nutrition Inst. v. Young, 818 F.2d 943, 948 (D.C.Cir.1987) (per curiam). Adoption of the regulation thus has consequences far beyond the immediate parties to the dispute.
.The case is therefore also quite different from an agency adjudication where a court can, for example, order an agency to treat an immigrant as eligible for asylum. See, e.g., He v. Ashcroft, 328 F.3d 593, 604 (9th Cir.2003). In an adjudication, the agency has a legal duty to act where the record compels a particular action vis-a-vis a particular party.
. The majority also claims plaintiffs’ injuries would be redressed because forcing the agency to provide “a rationale for not going forward with the Draft Policy could influence [USDA’s] decision.” Maj. op. at 836. The majority lifts this "procedural redressability” theory from our NEPA case law, but NEPA is a statute that spells out procedural requirements designed to influence the agency's decision-making process. The injury thus arises when the agency makes a decision without complying with those procedures. For example, in Citizens for Better Forestry v. U.S. Department of Agriculture, 341 F.3d 961 (9th Cir.2003), on which the majority relies, plaintiffs claimed that the agency had published a national forest management policy that “was not accompanied by any environmental or endangered-species analysis,” and "did not entertain comments regarding the rule's environmental impact.” Id. at 967. Because the agency bypassed statutory procedures designed to influence its deliberations, we held that plaintiffs need not show that the agency would make a different decision, had the procedures been followed; it was sufficient that the procedures could influence the decision-making process. Plaintiffs here can point to no statutory requirement that the agency bypassed by withdrawing the Draft Policy without comment. The procedural redressability theory has no application where there is no procedural requirement the agency failed to comply with.
. Indeed, part of the relief sought by the inmates in Chaney was that the agency "adopt procedures for seizing the drugs from state prisons.” 470 U.S. at 824, 105 S.Ct. 1649.
.Withholding proposed regulations that are final but for publication in the Federal Register seems to be common when there is a change in administrations. See Dabney v. Reagan, 542 F.Supp. 756, 760 (S.D.N.Y.1982) ("Shortly after taking office, President Reagan directed the heads of all Executive Departments to postpone all pending regulations.”); Kennecott Utah Copper Corp., 88 F.3d at 1200-01 (Clinton administration withdrawal of G.H.W. Bush administration regulation); Chen v. INS, 95 F.3d 801, 803-04 (9th Cir.1996) (same); Kootenai Tribe of Idaho v. Veneman, 142 F.Supp.2d 1231, 1236 n. 6 (D.Idaho 2001) (G.W. Bush administration postponed publication of all pending rules and regulations upon first entering office). The majority's opinion casts serious doubt on this venerable practice.
. This appears to be what happened here. The agency proposed the Draft Policy in response to the district court’s ruling in Animal Legal Defense Fund, Inc. v. Glickman, 943 F.Supp. 44 (D.D.C.1996), that the regulations it had adopted were insufficiently specific. After that ruling was reversed, see Glickman II, 204 F.3d at 235, the agency no doubt deemed the Draft Policy unnecessary.
. This too seems to have played a role here. Dr. Gipson, in his statement, noted as follows: "USDA and NIH had agreed to develop best management practice guidelines concerning nonhuman primates.”
. The majority also relies on a post-Chaney case from the DC Circuit, International Union, United Mine Workers of America v. U.S. Department of Labor, 358 F.3d 40 (D.C.Cir.*8512004), but the case is inapposite. Contrary to the majority’s assertions, International Union is not a case where the agency was embarking on a "discretionary course of action.” Maj. op. at 843. Rather, the agency there was acting pursuant to a statutory obligation imposed by the Mine Safety and Health Act, which states that "[t]he Secretary shall ... develop, promulgate, and revise ... improved mandatory health or safety standards for the protection of life and prevention of injuries in coal or other mines.” 30 U.S.C. § 811(a) (emphasis added); see also 358 F.3d at 41. As already noted, when an agency has an obligation to act, its failure to do so is reviewable. See pp. 845, 849-850 supra. International Union says nothing whatsoever about the situation presented here, where the agency has no obligation to act.
. While Justice Marshall's opinion is styled a concurrence, he vigorously dissented from the majority's central holding that an agency’s decision not to enforce is presumptively non-reviewable. 470 U.S. at 853-55, 105 S.Ct. 1649 (Marshall, J., concurring).
. Action may be final even if some degree of uncertainty lingers. But the majority cites no case reviewing a policy currently undergoing transformation. In Her Majesty the Queen in Right of Ontario v. EPA, 912 F.2d 1525, 1532 (D.C.Cir.1990), the court found that the agency’s statement was “devoid of any suggestion that it might be subject to subsequent revision.” Here, of course, USDA's primate policy is still up in the air. Likewise, in American Petroleum Institute v. EPA, 906 F.2d 729, 739-40 (D.C.Cir.1990) (per curiam), the possibility of "unforeseen amendments” did not render the decision unripe. Here, the change in policy is not only foreseeable; it’s in progress.
. As I note above, we have no way of doing this. See pp. 845-846 supra. But the agency can’t be sure until the litigation comes to an end. It must therefore delay implementation of best practice guidelines so long as litigation over the Draft Policy continues.