Wages and White Lion Invst v. FDA ( 2021 )


Menu:
  • Case: 21-60766      Document: 00516068813         Page: 1   Date Filed: 10/26/2021
    United States Court of Appeals
    for the Fifth Circuit
    United States Court of Appeals
    Fifth Circuit
    FILED
    October 26, 2021
    No. 21-60766                       Lyle W. Cayce
    Clerk
    Wages and White Lion Investments, L.L.C., doing business as
    Triton Distribution,
    Petitioner,
    versus
    United States Food and Drug Administration,
    Respondent.
    Petition for Review of an Order of the
    Food and Drug Administration
    Before Elrod, Oldham, and Wilson, Circuit Judges.
    Andrew S. Oldham, Circuit Judge:
    The Food and Drug Administration denied Triton’s application to
    market flavored e-cigarettes. Triton moved for a stay pending disposition of
    its petition for review. We grant the stay.
    I.
    A.
    In 2009, Congress enacted the Family Smoking Prevention and
    Tobacco Control Act (“TCA”) to regulate tobacco products. Pub. L. No.
    111-31, 
    123 Stat. 1776
     (2009). The TCA authorizes the Secretary of Health
    Case: 21-60766     Document: 00516068813          Page: 2   Date Filed: 10/26/2021
    No. 21-60766
    and Human Services to implement the Act through the Food and Drug
    Administration (“FDA”). See 21 U.S.C. §§ 387a(b), 393(d)(2). The TCA
    prohibits manufacturers from selling any “new tobacco product” without
    authorization. See id. § 387j(a). In 2016, the FDA deemed electronic nicotine
    delivery systems (“ENDS”)—colloquially called “electronic cigarettes” or
    “e-cigarettes”—a “new tobacco product.” 
    81 Fed. Reg. 28,973
     (May 10,
    2016) (“Deeming Rule”); see also Big Time Vapes, Inc. v. FDA, 
    963 F.3d 436
    ,
    443 (5th Cir. 2020) (“In the TCA, Congress delegated to the Secretary the
    power to ‘deem’ which tobacco products should be subject to the Act’s
    mandates.”). Thus, the TCA and the Deeming Rule generally prohibited the
    marketing of e-cigarettes.
    This created a serious and obvious problem because, by the time the
    FDA got around to issuing the Deeming Rule, manufacturers were widely
    marketing e-cigarettes throughout the United States. To avoid an overnight
    shutdown of the entire e-cigarette industry, the FDA delayed enforcement of
    the Deeming Rule. Then the FDA forced e-cigarette makers to meet a series
    of requirements and staggered deadlines to keep their products on the
    market.
    As relevant here, the FDA required e-cigarette manufacturers to
    submit premarket tobacco applications (“PMTAs”). The PMTA process is
    “onerous,” to put it mildly. See Big Time Vapes, 963 F.3d at 439 (“The
    PMTA process is onerous, requiring manufacturers to gather significant
    amounts of information.”). A manufacturer must submit to the FDA
    information on the product’s health risks, ingredients, and manufacturing
    process. The manufacturer also must include samples of the product and its
    proposed labeling. 21 U.S.C. § 387j(b)–(c).
    In the months and years following the Deeming Rule, the FDA moved
    its regulatory goalposts in at least two important ways. First, it moved the
    2
    Case: 21-60766      Document: 00516068813          Page: 3   Date Filed: 10/26/2021
    No. 21-60766
    PMTA deadline. Originally, the FDA demanded that all PMTAs must be
    filed within 24 months of the Deeming Rule—i.e., by 2018. The FDA later
    purported to extend the PMTA deadline to 2022. But then, in response to
    litigation from anti-smoking groups, the FDA moved the deadline up to
    September 9, 2020. Second, and crucial to this case, the FDA changed the
    regulatory requirements for PMTAs. Initially, the FDA’s guidance stated
    that “in general, FDA does not expect that applicants will need to conduct
    long-term studies to support an application.” A.74; see also A.92 (same). As
    Triton’s case illustrates, however, the FDA later changed its mind and
    required the very thing it said it would not—namely, long-term studies of e-
    cigarettes.
    B.
    Wages and White Lion Investments, LLC, doing business as Triton
    Distribution (“Triton”), is a Texas-based manufacturer of e-cigarettes.
    Some of its e-cigarette products have been on the market since August 4,
    2016—before the Deeming Rule’s effective date. Triton submitted a timely
    PMTA for certain flavored e-cigarettes. So did many other e-cigarette
    manufacturers.
    On August 26, 2021, the FDA announced that it would deny the
    PMTAs for 55,000 flavored e-cigarettes. In its press release, the FDA
    explained that it would do so because it “likely” needed evidence from long-
    term studies to grant a PMTA for flavored e-cigarettes. Less than a week after
    the FDA changed its regulatory requirements, Triton submitted a letter
    stating that it intended to conduct long-term studies of its products.
    About two weeks later, on September 14, the FDA issued a marketing
    denial order (“Order”) to Triton. See 21 U.S.C. § 387j(c)(2). The FDA
    acknowledged that it did not consider Triton’s letter in its determination
    because the FDA “received [the letter] near the completion of scientific
    3
    Case: 21-60766        Document: 00516068813             Page: 4      Date Filed: 10/26/2021
    No. 21-60766
    review.” A.14–15. The “key basis” for the denial, wrote the FDA, was that
    Triton’s PMTA lacked “robust and reliable evidence” from long-term
    studies, such as a “randomized controlled trial,” a “longitudinal cohort
    study,” or “other evidence . . . evaluat[ing] the impact of the new flavored
    vs. Tobacco-flavored products on adult smokers’ switching or cigarette
    reduction over time.” A.49.
    Triton then petitioned for review and moved to stay the Order
    pending that review.1 We granted a temporary administrative stay to prevent
    the FDA from shutting down Triton’s business. Now we enter a full stay
    pending disposition of Triton’s petition.
    II.
    For a stay pending review, we must consider four factors: (1) whether
    the requester makes a strong showing that it’s likely to succeed on the merits;
    (2) whether the requester will be irreparably injured without a stay;
    (3) whether other interested parties will be irreparably injured by a stay; and
    (4) where the public interest lies. Nken v. Holder, 
    556 U.S. 418
    , 426 (2009).
    “The first two factors are the most critical.” Valentine v. Collier, 
    956 F.3d 797
    , 801 (5th Cir. 2020) (per curiam). “‘The party seeking the stay bears the
    burden of showing its need.’” Tex. League of United Latin Am. Citizens v.
    Hughs, 
    978 F.3d 136
    , 143 (5th Cir. 2020) (quoting Clinton v. Jones, 
    520 U.S. 681
    , 708 (1997)); see also Nken, 
    556 U.S. at
    433–34 (“The party requesting a
    stay bears the burden of showing that the circumstances justify an exercise of
    that discretion.”). Triton has met its burden: The first three factors support
    a stay, while the fourth is at worst neutral.
    1
    Triton did not first ask the FDA for a stay. But it’s common ground that it would
    have been “impracticable” for Triton to do so. See Fed. R. App. P. 18(a)(2)(i).
    4
    Case: 21-60766       Document: 00516068813           Page: 5     Date Filed: 10/26/2021
    No. 21-60766
    A.
    First, likelihood of success. The Administrative Procedure Act
    (“APA”) directs courts to “hold unlawful and set aside agency action[s]”
    that are “arbitrary, capricious, an abuse of discretion, or otherwise not in
    accordance with law.” 
    5 U.S.C. § 706
    (2). “The APA’s arbitrary-and-
    capricious standard requires that agency action be reasonable and reasonably
    explained.” FCC v. Prometheus Radio Project, 
    141 S. Ct. 1150
    , 1158 (2021). We
    must not “substitute” our “own policy judgment for that of the agency.”
    
    Ibid.
     Still, we must ensure that “the agency has acted within a zone of
    reasonableness and, in particular, has reasonably considered the relevant
    issues and reasonably explained the decision.” Ibid.; see also Motor Vehicle
    Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 
    463 U.S. 29
    , 43
    (1983) (“[T]he agency must examine the relevant data and articulate a
    satisfactory explanation for its action including a ‘rational connection
    between the facts found and the choice made.’” (quoting Burlington Truck
    Lines v. United States, 
    371 U.S. 156
    , 168 (1962))). “Put simply, we must set
    aside any action premised on reasoning that fails to account for ‘relevant
    factors’ or evinces ‘a clear error of judgment.’” Univ. of Tex. M.D. Anderson
    Cancer Ctr. v. HHS, 
    985 F.3d 472
    , 475 (5th Cir. 2021) (quoting Marsh v. Or.
    Nat. Res. Council, 
    490 U.S. 360
    , 378 (1989)).
    In reviewing an agency’s action, we may consider only the reasoning
    “articulated by the agency itself”; we cannot consider post hoc
    rationalizations. State Farm, 
    463 U.S. at 50
    ; see also DHS v. Regents of the
    Univ. of Cal., 
    140 S. Ct. 1891
    , 1909 (2020) (“An agency must defend its
    actions based on the reasons it gave when it acted.”). Our review is “not
    toothless.” Sw. Elec. Power Co. v. EPA, 
    920 F.3d 999
    , 1013 (5th Cir. 2019). In
    fact, after Regents, it has serious bite. See 140 S. Ct. at 1907–15; see also, e.g.,
    Texas v. Biden, 
    10 F.4th 538
    , 552–57 (5th Cir. 2021) (per curiam); Biden v.
    Texas, No. 21A21, 
    2021 WL 3732667
    , at *1 (U.S. Aug. 24, 2021).
    5
    Case: 21-60766     Document: 00516068813           Page: 6   Date Filed: 10/26/2021
    No. 21-60766
    Triton has shown a strong likelihood of success on the merits. That’s
    because the FDA failed to “reasonably consider[] the relevant issues and
    reasonably explain[]” the Order. Prometheus, 141 S. Ct. at 1158; see also
    Michigan v. EPA, 
    576 U.S. 743
    , 750, 752 (2015) (“[A]gency action is lawful
    only if it rests on a consideration of the relevant factors” and “important
    aspect[s] of the problem.” (quotation omitted)). The relevant factors the
    FDA inadequately addressed or explained include: (1) Triton’s marketing
    plan; (2) Triton’s reliance interests; (3) less disruptive alternatives;
    (4) device-type preferences; and (5) evidence on the potential benefits of
    flavored e-cigarettes. The FDA’s counterarguments (6) are unavailing.
    1.
    The FDA failed to reasonably consider Triton’s proposed marketing
    plan. The FDA repeatedly stated that a marketing plan is “a critical factor
    in[] FDA’s statutorily required determination.” Premarket Tobacco Product
    Applications and Recordkeeping Requirements, 
    86 Fed. Reg. 55,300
    , 55,324
    (Oct. 5, 2021) (“Final Rule”); see also 
    84 Fed. Reg. 50,566
    , 50,581 (Sept. 25,
    2019) (“Proposed Rule”) (“The applicant’s marketing plans . . . will provide
    input that is critical to FDA’s determination of the likelihood of changes in
    tobacco product use behavior, especially when considered in conjunction
    with other information contained in the application.” (emphasis added));
    A.45 n.xix (“Limiting youth access and exposure to marketing is a critical
    aspect of product regulation.” (emphasis added)); A.45 (Premarket
    “assessment includes evaluating the appropriateness of the proposed
    marketing plan.”). Here, however, the FDA simply ignored Triton’s plan. It
    stated: “[F]or the sake of efficiency, the evaluation of the marketing plan in
    applications will not occur at this stage of review, and we have not evaluated
    any marketing plans submitted with these applications.” A.45 n.xix.
    6
    Case: 21-60766      Document: 00516068813          Page: 7    Date Filed: 10/26/2021
    No. 21-60766
    The FDA’s excuses for ignoring the “critical factor” of Triton’s
    marketing plan are unpersuasive. First, the FDA says it didn’t evaluate
    Triton’s plan for “the sake of efficiency.” 
    Ibid.
     But “efficiency” is no
    substitute for “reasoned decisionmaking.” Michigan, 576 U.S. at 750; see also
    Judulang v. Holder, 
    565 U.S. 42
    , 64 (2011) (emphasizing that “cheapness
    alone cannot save an arbitrary agency policy”).
    Second, the FDA claimed that its purported expertise and experience
    showed that no marketing plan would be sufficient, so it stopped looking:
    It is theoretically possible that significant mitigation efforts
    could adequately reduce youth access and appeal such that the
    risk for youth initiation would be reduced. However, to date,
    none of the ENDS PMTAs that FDA has evaluated have
    proposed advertising and promotion restrictions that would
    decrease appeal to youth to a degree significant enough to
    address and counter-balance the substantial concerns, and
    supporting evidence, discussed above regarding youth use.
    Similarly, we are not aware of access restrictions that, to date,
    have been successful in sufficiently decreasing the ability of
    youth to obtain and use ENDS.
    A.45 n.xix. This statement is insufficient. For one thing, it’s unreasonable for
    the FDA to stop looking at proposed plans because past ones have been
    unpersuasive. That’s like an Article III judge saying that she stopped reading
    briefs because she previously found them unhelpful.
    For another, reliance on expertise and experience, like efficiency, is
    no substitute for “reasoned decisionmaking.” Michigan, 576 U.S. at 750. Of
    course, “[a]gencies . . . have expertise and experience in administering their
    statutes that no court can properly ignore.” Judulang, 
    565 U.S. at 53
    . But
    here that hurts, not helps, the FDA. That’s because experience and expertise
    bring responsibility:
    7
    Case: 21-60766        Document: 00516068813              Page: 8       Date Filed: 10/26/2021
    No. 21-60766
    [A]n agency’s “experience and expertise” presumably enable
    the agency to provide the required explanation, but they do not
    substitute for the explanation, any more than an expert
    witness’s credentials substitute for the substantive
    requirements applicable to the expert’s testimony under
    [Federal Rule of Evidence] 702. The requirement of
    explanation presumes the expertise and experience of the
    agency and still demands an adequate explanation in the
    particular matter.
    CS Wind Viet. Co., Ltd. v. United States, 
    832 F.3d 1367
    , 1377 (Fed. Cir. 2016)
    (citations omitted).
    The FDA did not meet its obligation. Its statement on marketing plans
    is conclusory, unsupported, and thus wholly insufficient. See, e.g., United
    Techs. Corp. v. U.S. Dep’t of Def., 
    601 F.3d 557
    , 562 (D.C. Cir. 2010) (“We
    do not defer to the agency’s conclusory or unsupported suppositions.”
    (quotation omitted)); Texas v. Biden, 10 F.4th at 556 (collecting cases).2 This
    “omission alone [likely] renders [the FDA’s] decision arbitrary and
    capricious.” Regents, 140 S. Ct. at 1913.
    2.
    The FDA also failed to reasonably consider Triton’s legitimate
    reliance interests. Between the Deeming Rule’s effective date and the
    deadline for PMTAs, the FDA held public meetings and issued guidance on
    2
    The FDA’s failure to meaningfully consider Triton’s marketing plan is even more
    unreasonable because part of Triton’s plan was endorsed by a former FDA commissioner.
    See Statement from FDA Commissioner Scott Gottlieb, M.D., On Proposed New Steps to
    Protect Youth by Preventing Access to Flavored Tobacco Products and Banning Menthol
    in Cigarettes (Nov. 15, 2018) (“The changes I seek would protect kids by having all flavored
    ENDS products (other than tobacco, mint and menthol flavors or non-flavored products)
    sold in age-restricted, in-person locations and, if sold online, under heightened practices
    for age verification.”); ibid. (calling some of Triton’s proposed marketing restrictions
    “best practices”).
    8
    Case: 21-60766      Document: 00516068813          Page: 9   Date Filed: 10/26/2021
    No. 21-60766
    how e-cigarette manufacturers could get premarket authorization. In its
    “final guidance,” the FDA stated that it did not “expect” that tobacco
    manufacturers would need to conduct long-term studies to support their
    PMTA. See, e.g., A.73–74; A.92; see also Nicopure Labs, LLC v. FDA, 
    944 F.3d 267
    , 282 (D.C. Cir. 2019) (“The FDA has expressed willingness to
    accept scientific literature reviews instead of commissioned studies in
    support of e-cigarette applications in appropriate circumstances.”). The
    FDA’s expectation did not deviate in its Proposed Rule issued before the
    Order or the Final Rule issued a couple weeks after the Order. See Final Rule,
    86 Fed. Reg. at 55,387 (“FDA does not expect that long-term clinical studies
    will need to be conducted for each PMTA; instead, it expects that it should
    be able to rely on other valid scientific evidence to evaluate some PMTAs.”);
    Proposed Rule, 84 Fed. Reg. at 50,619 (similar). Many e-cigarette companies
    relied on the FDA’s repeated insistence that it did “not expect that
    applicants will have to conduct long-term studies to support an application”
    and did not perform or submit such evidence. A.74.
    Then the FDA “pull[ed] a surprise switcheroo on regulated entities.”
    Env’t Integrity Project v. EPA, 
    425 F.3d 992
    , 996 (D.C. Cir. 2005) (Sentelle,
    J.); accord Azar v. Allina Health Servs., 
    139 S. Ct. 1804
    , 1810 (2019) (citing
    the “surprise switcheroo” doctrine). Almost a year after the PMTA
    deadline, the FDA issued its first marketing denial orders for various flavored
    e-cigarettes and announced that it required the very studies it originally
    expected it didn’t need. See Press Release, FDA Denies Marketing
    Applications for About 55,000 Flavored E-Cigarette Products for Failing to
    Provide Evidence They Appropriately Protect Public Health (Aug. 26, 2021).
    It explained: “[T]he evidence of benefits to adult smokers for such products
    would likely be in the form of a randomized controlled trial or longitudinal
    cohort study, although the agency does not foreclose the possibility that other
    types of evidence could be adequate if sufficiently robust and reliable” and
    9
    Case: 21-60766     Document: 00516068813            Page: 10   Date Filed: 10/26/2021
    No. 21-60766
    performed over time. 
    Ibid.
     About two weeks later, the FDA maintained its
    long-term-study requirement in the Order denying Triton premarket
    authorization. See A.49; A.37 (materially identical language to Press Release).
    Despite the radical difference, the FDA never mentioned, let alone
    reasonably considered, whether e-cigarette manufacturers, like Triton,
    could’ve reasonably relied on the FDA’s prior meetings and guidance.
    The law requires more. “When an agency changes course, . . . it must
    be cognizant that longstanding policies may have engendered serious reliance
    interests that must be taken into account.” Regents, 140 S. Ct. at 1913
    (quotation omitted). This does not mean that the FDA could not have
    “determine[d], in the particular context before it, that other interests and
    policy concerns outweigh any reliance interests. Making that difficult
    decision was the agency’s job, but the agency failed to do it.” Id. at 1914. This
    reinforces that the Order was likely arbitrary, capricious, or otherwise
    unlawful.
    3.
    The FDA insufficiently addressed alternatives to issuing the Order as
    well. “[W]hen an agency rescinds [or alters] a prior policy[,] its reasoned
    analysis must consider the alternatives that are within the ambit of the existing
    policy.” Regents, 140 S. Ct. at 1913 (emphasis added) (quotation omitted).
    While considering less disruptive alternatives, the FDA “was required to
    assess whether there were reliance interests, determine whether they were
    significant, and weigh any such interests against competing policy concerns.”
    Id. at 1915. The FDA did not consider alternatives when changing from its
    no-long-term-studies-necessary policy to its apparent long-term-studies-
    required policy.
    And even if the FDA did, it failed to adequately assess reliance
    interests. “So it would be impossible for the [Order] to properly weigh the
    10
    Case: 21-60766     Document: 00516068813             Page: 11   Date Filed: 10/26/2021
    No. 21-60766
    relevant interests against competing policy concerns while considering
    alternatives.” Texas v. Biden, 10 F.4th at 555.
    4.
    The FDA also failed to adequately address Triton’s contention that
    its reusable e-cigarette will reduce youth popularity compared to disposable
    e-cigarettes. In January 2020 guidance, the FDA found that “youth
    overwhelmingly prefer [disposable] ENDS products” because they “are easy
    to conceal” and “can be used discreetly.” Enforcement Priorities for
    Electronic Nicotine Delivery Systems and Other Deemed Products on the
    Market Without Premarket Authorization; Guidance for Industry;
    Availability, 
    85 Fed. Reg. 720
    , 722 (Jan. 7, 2020). By contrast, the FDA found
    in the Order that the type of system didn’t matter. Specifically, the FDA
    found that “preference for device types and popularity of certain styles is
    likely fluid and affected by the marketplace” and “that the removal of one
    flavored product option prompted youth to migrate to another ENDS type
    that offered the desired flavor option, underscoring the fundamental role of
    flavor in driving appeal.” A.42.
    Because its “new policy rest[ed] upon factual findings that contradict
    those which underlay its prior policy,” the FDA had to provide “a more
    detailed justification.” FCC v. Fox Television Stations, Inc., 
    556 U.S. 502
    , 515
    (2009). The FDA initially said that disposable e-cigarettes pose risks to
    youths. When Triton said that concern doesn’t apply to its reusable e-
    cigarettes, the FDA turned around and ignored its prior disposable-reusable
    distinction. The FDA failed to adequately explain this change. This further
    reinforces that the Order is likely arbitrary, capricious, or otherwise unlawful.
    5.
    In announcing its rule that the manufacturer must provide long-term
    studies to get approval for flavored e-cigarettes, the FDA resorted entirely to
    11
    Case: 21-60766     Document: 00516068813           Page: 12   Date Filed: 10/26/2021
    No. 21-60766
    experience and expertise from reviewing applications other than Triton’s
    PMTA. See A.45. In so doing, the FDA used “generalized language to
    reject” Triton’s PMTA. See Siddiqui v. Holder, 
    670 F.3d 736
    , 744 (7th Cir.
    2012) (“Where, as here, the agency uses only generalized language to reject
    the evidence, we cannot conclude that the decisions rest on proper
    grounds.”). The consequence is that the FDA failed to reasonably consider
    relevant issues that Triton brought up in its PMTA but that others might not
    have.
    The FDA responded to much of Triton’s evidence for the first time
    before our court. But “[i]t is a fundamental precept of administrative law that
    an administrative agency cannot make its decision first and explain it later.”
    Texas v. Biden, 10 F.4th at 558–59; see also Sherley v. Sebelius, 
    689 F.3d 776
    ,
    784 (D.C. Cir. 2012) (Sentelle, C.J.) (“The failure to respond to comments
    is significant only insofar as it demonstrates that the agency’s decision was
    not based on a consideration of the relevant factors.” (quotation omitted));
    Circus Circus Casinos, Inc. v. NLRB, 
    961 F.3d 469
    , 476 (D.C. Cir. 2020)
    (“New rules set through adjudication must meet the same standard of
    reasonableness as notice and comment rulemaking.” (citing Allentown Mack
    Sales & Serv., Inc. v. NLRB, 
    522 U.S. 359
    , 374 (1998))).
    For example, Triton urged the FDA to consider a 2015 survey of
    20,000 e-cigarette users showing that nearly a third of the respondents
    “started out using tobacco or menthol flavors” and then began using other
    flavored e-cigarettes. A.296. Similarly, Triton asserted that flavored e-
    cigarettes “could serve an important role in transitioning existing adult users
    away from more harmful, combustible cigarette products.” 
    Ibid.
     But in the
    Order, the FDA ignored the first point altogether and gave the second short
    shrift. The FDA cannot cure those deficiencies by offering post hoc
    rationalizations before our court. The very fact that the FDA perceived the
    need to rehabilitate its Order with new and different arguments before our
    12
    Case: 21-60766     Document: 00516068813            Page: 13   Date Filed: 10/26/2021
    No. 21-60766
    court underscores that the Order itself omitted a reasoned justification for
    the agency’s action. This further confirms that the Order is likely arbitrary,
    capricious, or otherwise unlawful.
    6.
    The FDA makes four other counterarguments. They fail.
    First, the FDA argues that its consistency “in reviewing other
    manufacturers’ similar applications to market flavored e-cigarette products
    is a hallmark of good government, not a reason to fault the agency.” Opp. at
    23 (citation omitted). Consistency is great—but only when the agency is
    consistently following the law. As the Supreme Court has made clear:
    “Arbitrary agency action becomes no less so by simple dint of repetition.”
    Judulang, 
    565 U.S. at 61
    ; see also 
    ibid.
     (“[L]ongstanding capriciousness
    receives no special exemption from the APA.”).
    Second, the FDA insists that the reasoning in the Order is consistent
    with its prior guidance. According to the FDA, it didn’t make a rule requiring
    long-term studies because it left open that “other types of evidence could be
    adequate[] and will be evaluated on a case-by-case basis.” A.37.
    But the administrative record makes clear that the FDA now requires
    direct evidence through studies performed “over time” for flavored e-
    cigarettes. A.46; see also, e.g., A.37 n.vi; A.47 n.xxiii. And it’s clear the FDA
    expressly rejected reliance on evidence it approved of in its pre-Order
    guidance, such as observational and consumer-perception studies. Compare
    A.46–47, with A.99. The FDA did not have to completely flip flop for there
    to be a change in position. Cf. Sw. Airlines Co. v. Fed. Energy Regul. Comm’n,
    
    926 F.3d 851
    , 856 (D.C. Cir. 2019) (“A full and rational explanation becomes
    especially important when, as here, an agency elects to shift its policy or
    depart from its typical manner of administering a program.” (quotation
    omitted)). It is enough that the FDA’s guidance indicated long-term studies
    13
    Case: 21-60766     Document: 00516068813             Page: 14   Date Filed: 10/26/2021
    No. 21-60766
    were likely unnecessary, while the FDA’s Order at the very least created a
    strong presumption that such evidence is required.
    Plus, if we accepted the FDA’s current position that it did not
    acknowledge a change in policy in the Order, then the Order would obviously
    be arbitrary and capricious. That’s because “[w]hen an agency changes its
    existing position, it . . . must at least display awareness that it is changing
    position and show that there are good reasons for the new policy.” Encino
    Motorcars, LLC v. Navarro, 
    136 S. Ct. 2117
    , 2125–26 (2016) (quotation
    omitted); see also id. at 2126 (explaining that an “unexplained inconsistency
    in agency policy is a reason for holding an [action] to be an arbitrary and
    capricious change from agency practice” (quotation omitted)); Fox, 
    556 U.S. at 515
     (“[T]he requirement that an agency provide reasoned explanation for
    its action would ordinarily demand that it display awareness that it is
    changing position. An agency may not . . . depart from a prior policy sub
    silentio.”). It would be impossible for the FDA to display awareness that it
    was changing position if it believed it wasn’t.
    Third, the FDA argues that Triton should not have relied on the
    agency’s pre-Order guidance. This is because, the FDA claims, 21 U.S.C.
    § 387j(c)(5) “directs FDA to make that finding based on ‘clinical
    investigations by experts qualified by training and experience to evaluate the
    tobacco product’ or other ‘valid scientific evidence’ that FDA determines is
    sufficient.” Opp. at 19; see also id. at 20 (The “2019 guidance does not and
    could not relax the statute’s requirements.”). Of course, an agency cannot
    issue guidance on the meaning of a statute, encourage its regulated entities to
    rely on the guidance, and then blame the statute for pulling the rug out from
    under the entities. And in any event, the FDA mischaracterizes § 387j(c)(5).
    Paragraph (5) does not require the FDA to base all of its appropriate-for-the-
    protection-of-the-public-health findings on long-term studies; instead, it
    requires the FDA to base its decision on “well-controlled investigations”
    14
    Case: 21-60766     Document: 00516068813           Page: 15    Date Filed: 10/26/2021
    No. 21-60766
    “when appropriate” and provides that those investigations “may include 1 or
    more clinical investigations.” 21 U.S.C. § 387j(c)(5)(A) (emphases added).
    And the consideration of other “valid scientific evidence” is likewise
    discretionary. See id. § 387j(c)(5)(B) (“may authorize”). The FDA’s “final
    guidance” reflected its “expect[ation]” that, at the time, it would not deem
    it “appropriate” to base its decision on long-term studies. A.74; A.92. The
    guidance also stated that the FDA would consider the type of evidence Triton
    presented “valid scientific evidence.” So of course, the statute might have
    permitted the FDA to demand the evidence it ultimately did. But it does not
    follow that the statute required the FDA to jettison the guidance it previously
    offered regulated entities.
    Fourth and last, the FDA argues that Triton’s reliance interests
    shouldn’t matter because Triton has been breaking the law and the FDA’s
    non-enforcement was entirely discretionary. Regents squarely forecloses this
    argument. There, the Department of Homeland Security (“DHS”) tried to
    rescind the Deferred Action for Childhood Arrivals (“DACA”) program
    because of “the Attorney General’s conclusion that DACA was unlawful.”
    Regents, 140 S. Ct. at 1910. The United States argued that justified ignoring
    potential reliance interests. Id. at 1913–14. The Supreme Court rejected that
    argument. Ibid. The Court instead required reasonable consideration of the
    relevant issues and the “important aspects of the problem.” Id. at 1910
    (quotation omitted). That was because, the Court explained, “deciding how
    best to address a finding of illegality moving forward can involve important
    policy choices.” Ibid. The same is true here. The FDA was free to make that
    policy choice, but it had to address Triton’s reliance interests in a reasonable
    and reasonably explained decision.
    For these reasons, Triton has shown a likelihood of success based on
    its APA challenge. So this critical factor favors granting a stay. We therefore
    need not address Triton’s argument that the FDA violated the Due Process
    15
    Case: 21-60766     Document: 00516068813           Page: 16   Date Filed: 10/26/2021
    No. 21-60766
    Clause for not giving “fair warning” of its change in position on what
    evidence would be required in its PMTA.
    B.
    Next, irreparable injury. Triton alleges that because of the Order, it
    “has stopped production of all of its flavored ENDS products, representing
    90 percent of its annual revenue, thereby requiring the company to make
    plans to lay off its employees within approximately two weeks and
    threatening the company’s very existence.” Stay Mot. at 21; see also A.15–16
    (Declaration of Triton’s General Manager). The FDA does not contest that
    allegation.
    Triton’s alleged injury is irreparable for two independent reasons.
    First, we’ve explained that “substantial financial injury” may be “sufficient
    to show irreparable injury.” Texas v. EPA, 
    829 F.3d 405
    , 433 (5th Cir. 2016).
    Triton’s alleged financial injury “threatens the very existence of [its]
    business.” Id. at 434. Even assuming the financial costs are recoverable, this
    suffices to show irreparable injury. See id. at 434 n.41 (“Even recoverable
    costs may constitute irreparable harm where the loss threatens the very
    existence of the movant’s business.” (quotation omitted)).
    Second, the costs are likely unrecoverable. “Indeed, complying with
    [an agency order] later held invalid almost always produces the irreparable
    harm of nonrecoverable compliance costs.” Id. at 433 (quotation omitted).
    The FDA does not contend that Triton has an avenue to recover costs from
    complying with the Order. That’s probably because federal agencies
    generally enjoy sovereign immunity for any monetary damages. See, e.g.,
    Alabama-Coushatta Tribe of Texas v. United States, 
    757 F.3d 484
    , 488 (5th Cir.
    2014); Louisiana v. United States, 
    948 F.3d 317
    , 320 (5th Cir. 2020); Muniz-
    Muniz v. U.S. Border Patrol, 
    741 F.3d 668
    , 671 (6th Cir. 2013) (“Sovereign
    immunity extends to agencies of the United States.” (quotation omitted)).
    16
    Case: 21-60766     Document: 00516068813           Page: 17   Date Filed: 10/26/2021
    No. 21-60766
    At bottom, Triton’s lack of a “guarantee of eventual recovery” is another
    reason that its alleged harm is irreparable. Alabama Ass’n of Realtors v. Dep’t
    of Health & Hum. Servs., 
    141 S. Ct. 2485
    , 2489 (2021).
    The FDA makes no developed argument contesting irreparable harm.
    See Opp. at 11, 13 (mentioning “irreparable injury” in passing). So such
    arguments are forfeited. See, e.g., DeVoss v. Sw. Airlines Co., 
    903 F.3d 487
    ,
    490 n.1 (5th Cir. 2018) (concluding that an argument was “forfeited”
    because it wasn’t “structured”); Texas v. EPA, 829 F.3d at 435 (“Because
    EPA offers nothing beyond this cursory comment, it has waived any
    argument about the scope of the stay.”).
    In these circumstances, given Triton’s uncontested allegations of
    injury and the FDA’s failure to make a developed argument challenging this
    factor, we conclude that Triton has met its burden of showing irreparable
    harm. Thus, the two most critical factors favor granting a stay.
    C.
    Now, the balance of harms and public interest.
    The balance of the harms favors a stay. We’ve explained that “the
    maintenance of the status quo is an important consideration in granting a
    stay.” Barber v. Bryant, 
    833 F.3d 510
    , 511 (5th Cir. 2016) (quotation omitted).
    And staying the Order will preserve the status quo ante. Cf. Turning Point
    Brands, Inc. v. FDA, No. 21-3855, ECF No. 19 at 9–10 (6th Cir. Oct. 8, 2021)
    (FDA letter rescinding a marketing denial order and stating the “FDA has
    no intention of initiating an enforcement action against any of your tobacco
    products identified in” the relevant PMTA). “Given the great likelihood that
    [Triton] will ultimately succeed on the merits, combined with the
    undeniable, irreparable harm that [the Order] would inflict on” Triton and
    the FDA’s failure to make a developed argument on this factor, we conclude,
    17
    Case: 21-60766        Document: 00516068813         Page: 18   Date Filed: 10/26/2021
    No. 21-60766
    in these circumstances, “that the balance of harms weighs in favor of”
    Triton. Tex. Democratic Party v. Abbott, 
    961 F.3d 389
    , 412 (5th Cir. 2020).
    The public-interest factor is at worst neutral. The “public interest is
    in having governmental agencies abide by the federal laws that govern their
    existence and operations.” Texas v. Biden, 10 F.4th at 559 (quotation
    omitted). “And ‘there is generally no public interest in the perpetuation of
    unlawful agency action.’” Id. at 560 (alteration omitted) (quoting League of
    Women Voters of U.S. v. Newby, 
    838 F.3d 1
    , 12 (D.C. Cir. 2016)). Although
    the FDA fails to argue this factor, amici curiae do. They argue that the public
    interest cuts against a stay because continued sale of flavored e-cigarettes will
    endanger the youth much more than it might help adults. “But our system
    does not permit agencies to act unlawfully even in pursuit of desirable ends.”
    Alabama Ass’n of Realtors, 141 S. Ct. at 2490. So we conclude that this factor
    is at best neutral, or, in all events, outweighed by the three other factors
    favoring a stay.
    III.
    Finally, the FDA argues that Triton requests relief we cannot give.
    We have no authority, says the FDA, to permit Triton to continue marketing
    and selling the products denied in the Order. But again, the APA says
    otherwise. Under 
    5 U.S.C. § 705
    , we may, under certain “conditions[,] . . .
    and to the extent necessary to prevent irreparable injury, . . . issue all
    necessary and appropriate process to postpone the effective date of an agency
    action or to preserve status or rights pending conclusion of the review
    proceedings.”
    The immigration context is instructive. Consider an alien that is
    unlawfully present in the United States. Suppose the Government attempts
    to remove the alien. Then the alien argues that he should not be removed
    because he deserves asylum, and he asks us to stay the removal pending our
    18
    Case: 21-60766     Document: 00516068813              Page: 19    Date Filed: 10/26/2021
    No. 21-60766
    review of his petition. Under the FDA’s logic, we couldn’t do anything. After
    all, we couldn’t order the Board of Immigration Appeals to grant the alien
    asylum or otherwise adjust his immigration status to make his presence
    lawful. But of course, we could grant a stay of the removal, giving the alien
    interim relief. See generally Tesfamichael v. Gonzales, 
    411 F.3d 169
     (5th Cir.
    2005) (granting a stay of removal pending the court of appeals’ consideration
    of the party’s petition for review); see also Nken, 
    556 U.S. at 429
     (“An alien
    seeking a stay of removal pending adjudication of a petition for review does
    not ask for a coercive order against the Government, but rather for the
    temporary setting aside of the source of the Government’s authority to
    remove. Although such a stay acts to bar Executive Branch officials from
    removing the applicant from the country, it does so by returning to the status
    quo—the state of affairs before the removal order was entered.” (quotation
    omitted)).
    Triton’s request is not materially different. It merely seeks to preserve
    the status quo ante, before the FDA issued the Order. In other words, “the
    relief sought here would simply suspend administrative alteration of the status
    quo.” Nken, 556 U.S. at 430 n.1. So we reject the FDA’s argument that we
    lack authority to grant a stay that provides interim relief.
    *        *         *
    Three factors—including the two most critical—favor granting a stay,
    while one factor is at worst neutral. Triton has thus met its burden. Contrary
    to the FDA’s suggestion, we have the authority to give Triton relief pending
    review. For the foregoing reasons, Triton’s motion for a stay pending review
    of its petition is GRANTED.
    19