Petition for Review of a Decision of the Department of Transportation, NTSB Agency No. DOT-OST-2015-0070

Before ROSENBAUM, BRANCH, and TJOFLAT, Circuit Judges.

BRANCH, Circuit Judge:

The U.S. Department of Transportation (“DOT”) is responsible for approving or denying the formation of international airline joint ventures that operate in the United StatesPage 2 based on its finding of the competitive and public interest effects the joint venture will have.

Delta Air Lines, Inc., and Aerovias de México, S.A. de C.V. (collectively, “the petitioners”) applied for approval of, and a grant of antitrust immunity for, a joint venture (“the joint venture”) to provide integrated airline services between the United States and Mexico. In 2016, DOT approved the joint venture and granted it antitrust immunity.

But DOT's role does not end once it approves a joint venture-it can terminate approval of a joint venture after a subsequent review of the joint venture’s impact on competition, including review of whether the joint venture is providing important public benefits and whether alternatives that are materially better for competition can meet transportation needs. So, in 2024, DOT issued show-cause orders that proposed ending the joint venture's approval and antitrust immunity because it believed the competitive landscape had worsened when the Mexican government restricted take-off and landing slots and banned all-cargo carriers from operating at one Mexico City airport. And in 2025, nine years after DOT approved the joint venture, it issued a final order that ended approval of the joint venture and withdrew its antitrust immunity.

The petitioners petitioned for review of the final order in this Court, arguing that the final order was arbitrary and capricious, and we stayed the final order pending our review. After careful review and with the benefit of oral argument, we agree with the Page 3petitioners. DOT did not reasonably explain why it conducted a far more limited market analysis in this case than it has always done in the past or why it imposed a requirement for approval of the joint venture that it did not require of similar joint ventures it approved in Japan. Accordingly, we vacate the final order.

Joint ventures between airlines that operate in international markets are common and often necessary to provide travelers with a broad network of connections because many jurisdictions prohibit foreign ownership of their airlines. Brian Pearce & Gary Doernhoefer, The Economic Benefits Generated by Alliances and Joint Ventures , Int’l Air Transp. Ass’n, Nov. 28, 2011, at 1. The parties to a joint venture can coordinate flight schedules, expand their capacity and offerings, and reduce travel time for consumers. Id. at 7. Some joint ventures, including the joint venture in this case, are considered “metal neutral,” which means that the partners to the joint venture share revenue regardless of which airline flies the passenger. Id. at 2.

In addition to the private joint ventures that facilitate international air travel, the United States negotiates “open skies” agreements with the governments of various countries to liberalize aviation markets. Defining Open Skies , DOT Order 92-8-13, 1992 WL 204010, at *1 (Aug. 5, 1992). The “basic elements” of an open skies agreement include, for example, “[o]pen entry on all routes,” “[u]nrestricted capacity and frequency on all routes,” andPage 4 '[u]nrestricted route and traffic rights.' Id. at *5. The United States and Mexico entered into an open skies agreement in 2015 (“the open skies agreement'). Air Transport Agreement Between the Government of the United States of America and the Government of the United Mexican States , U.S.-Mex., Dec. 18, 2015. Despite open skies agreements that promote the freedom of airlines to fly all routes free from government restrictions, extremely congested airports nonetheless need to use a 'slot' system to manage traffic. Slot Administration - Slot Definition , Federal Aviation Administration, https://www.faa.gov/about/office_org/headquarters_offices /ato/service_units/systemops/perf_analysis/slot_administration /slot_definition [https://perma.cc/K7AF-DN5V]. A slot is an authorization to take-off or land at a particular airport on a particular day during a specified time period. Id.

American air carriers can apply to DOT for approval of their joint ventures with foreign air carriers, which DOT “shall approve” if the joint venture “is not adverse to the public interest.” 49 U.S.C. § 41309(a)-(b). In addition to approving a joint venture, DOT “may exempt a [joint venture] affected by [an approval] order from the antitrust laws to the extent necessary to allow the person to proceed with the transaction specifically approved by the order.” Id. § 41308(b). DOT shall disapprove, or “after periodic review” can end approval of, a joint venture “that substantially reduces or eliminates competition” unless the joint venture “is necessary to meet a serious transportation need or to achieve important public benefits (including international comity and foreign policy considerations); and the transportation need cannot be met or Page 5those benefits cannot be achieved by reasonably available alternatives that are materially less anticompetitive.'¹ Id. § 41309(b).

In March 2015, the petitioners filed with DOT an application for approval of, and antitrust immunity for, the joint venture. The joint venture, the application explained, would coordinate on routes between the United States and Mexico, be metal neutral, and combine the petitioners’ respective networks to create a new competitor in the U.S.-Mexico aviation market. The joint venture would include “all nonstop transborder routes” operated by the petitioners between the United States and Mexico, 'as well as behind and beyond connecting flights within [the United States or Mexico].” The petitioners’ application identified 22 routes that it predicted would see increased traffic if the joint venture was approved, with eight routes having a predicted increase of 20% or more. Ultimately, the application predicted that the joint venture would lead to increased competition and would deliver public benefits like a broader network of flights between the United States and Mexico, increased flight frequency, and more seamless scheduling.

In November 2016, DOT 'grant[ed] approval of and antitrust immunity . . . for' the joint venture. In conducting itsPage 6 analysis pursuant to 49 U.S.C. §§ 41308, 41309, DOT first examined the 'country-pair' market, 2 which is the broader U.S.-Mexico market. Next, DOT analyzed all “city-pair” markets (for example, the Chicago-Los Cabos market).³ There are 1,687 city-pairs in the U.S.-Mexico market. Two of the largest city-pair markets within the U.S.-Mexico market are Los Angeles-Guadalajara (LAX-GDL) and New York JFK-Mexico City (JFK-MEX).

DOT’s country-pair analysis included statistics regarding the number of passengers and percentage share of the market that all airlines controlled. DOT then predicted how those figures would change if it approved the joint venture, agreeing that the joint venture would benefit the public by allowing the petitioners to join American Airlines and United Airlines as the primary U.S.-Mexico market competitors. And on the city-pair level, DOT conducted a “share shift analysis,” which predicted which city-pairs would see a reduction in the number of competitors following approval of the joint venture. DOT’s city-pair share shift analysis also factored in the size of the markets that would see a reduction in competition. DOT concluded that the joint venture would result in increasing the number of viable competitors in 15 markets, while only reducing the number of competitors in small markets that make up less than one percent of all U.S.-Mexico passengers.

Page 7But DOT expressed concern about one Mexican airport, specifically 'the [Mexican government's] non-transparent slot allocation regime . . . at Mexico City’s Benito Juarez International Airport [(“MEX”)],” which it thought could impede competition, particularly because the petitioners, through the joint venture, would control almost half of the slots at MEX.

To attempt to address its concerns over slot allocation at MEX, DOT conditioned its approval as follows: petitioners must divest 24 slot-pairs at MEX, and DOT’s grant of antitrust immunity to the joint venture would be limited to five years. The petitioners complied with DOT's condition, divested the slots, and began operating the joint venture with DOT's approval.

In March 2022, the petitioners applied to renew the joint venture's approval and antitrust immunity. The petitioners' renewal application included their assessment of how the joint venture affected the country-pair market and the city-pair markets since its approval and grant of antitrust immunity. They reported that the joint venture had increased U.S.-Mexico flight capacity by 761 seats per day, “significantly expanded capacity” on 26 city-pair routes, and 'launch[ed] entirely new service on six [city-pair] routes.” DOT extended the joint venture’s approval and antitrust immunity while its renewal application was pending.

Then, in January 2024, DOT issued a show-cause order that proposed terminating the joint venture's approval and antitrust immunity, focusing again on one airport-MEX. DOT explained that 'recent actions taken by the Government of Mexico' as toPage 8 MEX were inconsistent with the open skies agreement between the United States and Mexico, and under longstanding DOT policy, “ de facto or de jure implementation” of the open skies agreement was a “necessary precondition” for antitrust immunity of a joint venture.⁴ In addition to its previously stated concerns regarding slot allocation at MEX, DOT noted that '[t]he Government of Mexico issued a decree in which all-cargo operations were prohibited at MEX,'⁵ which is inconsistent with the '[l]iberal cargo regime' that is a part of DOT's definition of open skies. See Defining Open Skies , 1992 WL 204010, at *6. And DOT, '[d]espite further formal engagement with the Mexican Government,” claimed that it had not made progress toward resolving its concerns at MEX. So DOT “tentatively dismisse[d] without prejudice” the petitioners’ renewal application.

In response, the petitioners filed an objection to the showcause order. The petitioners argued that in the show-cause order DOT (1) failed to apply the relevant statutory standard; (2) did not sufficiently engage with the research and analysis in their renewal application; (3) made an unsupported finding that regulatory conditions at MEX caused material harm to the aviation market;Page 9 (4) applied a more stringent standard for approval than it had applied to other joint ventures; (5) ignored several reasonable alternatives to terminating approval of the joint venture; and (6) did not provide interested parties the opportunity to comment.

DOT then issued a supplemental show-cause order. The supplemental show-cause order primarily focused on the “changed circumstances” since DOT initially approved the joint venture in 2016 and how those changed circumstances affected the public benefits the joint venture could provide. What were those changed circumstances? DOT reiterated its concern that the Mexican government imposed “anticompetitive measures” at one Mexican airport—MEX—which were “distorting the marketplace.” DOT explained its stance that the Mexican government began breaching the open skies agreement in 2022 when it 'arbitrarily reduced capacity” at MEX and “confiscated slots from U.S. carriers at MEX,” which continued in 2023 when it 'ordered all-cargo carriers to vacate MEX.' According to DOT, slot allocation at MEX was no longer transparent, it was discriminatory, and it lacked coherence. As DOT described it, the problematic slot administration at MEX included “confiscat[ing] slots on spurious and unsupported grounds [and] refus[ing] to implement reforms suggested by Mexico's competition authority.' Following those regulatory changes at MEX, DOT alleged that by 2024, the petitioners began to deliver more growth at MEX than their competitors. As DOT saw it, the petitioners' 'dominant slot holdings and favored positioning at MEX' allowed them capacity to announce new routes between the United States and MEX, while other carriers,Page 10 including those whose slots were confiscated, were forced to cut capacity between the United States and MEX. And the petitioners gained an advantage from the all-cargo ban given their ability to carry 'belly cargo' on their passenger flights that land at MEX.

The petitioners objected to the supplemental show-cause order too. The petitioners’ objection included its own country-pair analysis and predictions on how termination of the joint venture’s approval would affect the U.S-Mexico competitive landscape. For example, the petitioners predicted that termination of the joint venture’s approval would risk eliminating nonstop service on 21 city-pair routes. Beyond providing an analysis of the relevant markets, the petitioners largely reiterated their objections to the initial show-cause order.

In September 2025, over the petitioners’ objections and after fielding comments from interested parties, DOT issued a final order terminating approval of the joint venture and its grant of antitrust immunity. DOT largely reiterated the reasoning it provided in its show-cause orders, again focusing its concerns on the Mexican government’s slot-allocation and all-cargo policies at one Mexican airport-MEX.⁶ In the final order, DOT concluded Page 11that based on these policies at MEX, the “joint venture substantially reduces or eliminates competition in [the] markets [in which it participates] and is adverse to the public interest.”

Unlike its comprehensive market analysis in 2016 that evaluated competition in the U.S.-Mexico market and 1,687 citypair markets, DOT simply highlighted four general competitive concerns in the final order, each of which related to conditions at MEX.⁷ DOT determined that those four concerns warranted terminating approval of the joint venture given DOT’s obligation to 'avoid[] conditions that would tend to allow at least one air carrier . . . to unreasonably increase prices, reduce services, or exclude competition in air transportation.” See 49 U.S.C. § 40101(a)(10).Notably, the final order did not include (1) updated country-pair figures reflecting each competitor’s number ofPage 12 passengers transported and share of the market; (2) its prediction for how those numbers would change if it continued to approve of the joint venture; or (3) a share shift analysis at the city-pair level. Instead, DOT applied its findings, which were based on its four competitive concerns at MEX only, to the statutory language of 49 U.S.C. § 41309 and found that the joint venture was not necessary to 'meet a serious transportation need or to achieve important public benefits.”

The petitioners timely petitioned for review of the final order in this Court and sought a stay pending review, which we granted.

The petitioners argue that the final order was arbitrary and capricious because DOT failed to reasonably justify (1) how it analyzed whether to terminate approval of the joint venture and its grant of antitrust immunity and (2) its ultimate decision to terminate approval of the joint venture and its grant of antitrust immunity. We agree with the petitioners that the final order was arbitrary and capricious for two reasons, each related to DOT's method of analysis. First, DOT did not reasonably explain its reliance on market conditions at a single airport, which was a departure from its uniform practice of conducting far broader market analyses. Second, DOT did not reasonably explain why it Page 13imposed an open skies requirement to approve the joint venture when it has not imposed such a requirement in other similar cases.⁸

The petitioners’ challenge to the final order proceeds under the Administrative Procedure Act (“APA”). See 5 U.S.C. §§ 701– 706. Under the APA, reviewing courts must “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,” which is a deferential standard. FCC v. Prometheus Radio Project , 592 U.S. 414, 423 (2021). Under the APA, we are deferential not just to the agency’s ultimate decision, but also to its “drafting decisions like how much discussion to include on each topic, and how much data is necessary to fully address each issue.” Black Warrior Riverkeeper, Inc. v. U.S. Army Corps of Eng’rs , 833 F.3d 1274, 1285 (11th Cir. 2016).

Because our review pursuant to the APA is deferential, we are 'not to substitute [our] judgment for that of the agency.' Motor Vehicle Mfrs. Ass’n U.S., Inc. v. State Farm Mut. Auto. Ins. Co. , 463 U.S. 29, 43 (1983). Under the APA, “we ask only whether ‘the agency came to a rational conclusion.’” Gray Television, Inc. v. FCC , 130 F.4th 1201, 1212 (11th Cir. 2025) (quoting Sierra Club v. Van Page 14Antwerp , 526 F.3d 1353, 1360 (11th Cir. 2008)). But “we are not a rubber stamp-'courts retain a role, and an important one, in ensuring that agencies have engaged in reasoned decisionmaking.’” In re Gateway Radiology Consultants, P.A. , 983 F.3d 1239, 1263 (11th Cir. 2020) (quoting Judulang v. Holder , 565 U.S. 42, 53 (2011)).

The petitioners argue that DOT’s final order was arbitrary and capricious because it relied on its concerns about competition at MEX, a single airport, in a manner that was inconsistent with DOT’s precedent, which has historically involved a more complete market analysis. DOT argues that the final order can be squared with its precedent because it conducted a more complete evaluation of the relevant markets when it approved the joint venture in 2016 and this case, unlike the precedent the petitioners point to, is distinguishable because it involved revoking a joint venture's approval. We agree with the petitioners that DOT abruptly departed from its uniform precedent of conducting a comprehensive market analysis without reasonably explaining such a departure.

An agency action is arbitrary and capricious if “it is an abrupt and unexplained departure from agency precedent.' LopezMartinez v. U.S. Att’y Gen. , 149 F.4th 1202, 1210 (11th Cir. 2025). While an agency can depart from its precedent, it must “explain its departure . . . and adequately explain the rationale of its decision.” McHenry v. Bond , 668 F.2d 1185, 1192-93 (11th Cir. 1982). For example, the Supreme Court has held that the Federal Page 15Communications Commission reasonably departed from its precedent when it “acknowledged that its recent actions [increasing enforcement against broadcasting expletives] broke new ground' and reasonably determined that technological advances made it easier to “bleep out” certain words and justified such increased enforcement. F.C.C. v. Fox Television Stations, Inc. , 556 U.S. 502, 517-18 (2009). But in another case, the Supreme Court held that the National Highway Traffic Safety Administration did not reasonably explain its departure from requiring new motor vehicles to be equipped with passive restraints (seatbelts or airbags). Motor Vehicle Mfrs. Ass’n , 463 U.S. at 33–34. There, the agency relied on its determination that many people may not use seatbelts even if they are installed, but it did not adequately consider relevant information like the safety benefits of airbags and quickly dismissed the fact that certain types of seatbelts were more commonly used than others. Id. at 47, 55– 56.

We find that DOT departed from its precedent without providing a reasonable explanation when it issued the final order. Until this case, when evaluating joint ventures between airlines pursuant to 49 U.S.C. §§ 41308, 41309, DOT’s uniform practice has been to 'weigh[] both pro- and anti-competitive effects' of a venture 'as a whole,' 'across a number of different markets,' which included a 'broad assessment' 'at the network, countrypair, and city-pair levels.” Am. Airlines, Inc. , DOT Order 2010-7-8 at 9 (July 20, 2010). For example, when considering approval of a joint venture between KLM Dutch Airlines and Northwest Page 16Airlines, DOT analyzed the 'U.S.-Europe, the U.S.-Netherlands, and the Detroit/Minneapolis-St. Paul-Amsterdam markets.” Northwest Airlines, Inc. , DOT Order 93-1-11 at 9–10 (Jan. 11, 1993). DOT does not contest that it undertakes such comprehensive reviews when evaluating joint ventures seeking approval and antitrust immunity; in fact, DOT has never before departed from this broad method of market analysis. And DOT has explicitly rejected requests for it to rely on narrower analyses, like assessing potential competitive harm a joint venture may cause in certain city-pair markets without considering the severity of that harm or the competitive benefits of the joint venture as a whole. Am. Airlines, Inc. , DOT Order 2010-7-8 at 9. Instead, DOT said that it conducts a “broad assessment” of the relevant country-pairs and city-pairs even if it finds that a joint venture will result in reduced competition in subsets of the broader market. Id.

Here, when reassessing the joint venture in 2025, DOT abruptly departed from its uniform precedent of comprehensively analyzing all of the relevant markets and instead relied on its finding that there were anticompetitive conditions at MEX, a single airport. The final order analyzed none of the 1,687 city-pairs in the U.S.-Mexico market, let alone engaged in a comprehensive share shift analysis like DOT did in 2016. Nor did it conduct a countrypair analysis of the U.S.-Mexico market. Unlike its initial analysis, DOT did not survey the share of the market each competitor controlled and predict how that landscape would evolve if the joint venture continued to operate. Far from conducting a broad Page 17assessment across a number of different markets, DOT based its decision on competitive concerns at MEX, a single airport.

DOT's attempts to square its focus on MEX with its precedent fail. First, it argues that the precedent the petitioners point to involved approval of joint ventures, while the final order addressed whether to terminate previous approval of the joint venture. But the fact that the final order decided whether to continue approval of the joint venture does not reasonably explain why DOT departed from its uniform method of analysis. The statutory standard DOT must apply for initial approval and continuing approval of joint ventures is the same. See 49 U.S.C. § 41309(b). Yet DOT conducted a drastically more limited market analysis when evaluating the petitioners’ application for continued approval than it does when evaluating initial applications for approval.

Second, DOT argues that it did not need to replicate the more robust analysis of the relevant markets it conducted in 2016 because when it initially approved the joint venture it raised concerns about the competitive conditions at MEX. This explanation of its rationale is inadequate. While DOT did conduct a more comprehensive analysis when it first approved the joint venture in 2016, that analysis was nine years old when it issued the final order. DOT does not explain why its 2016 analyses of the U.S.Mexico market and various city-pairs were still accurate nine years later. Nor does it explain why anticompetitive conditions at MEX (which makes up only a 21% share of U.S.-Mexico flights) were Page 18severe enough to warrant ignoring the broader market. DOT’s failure to provide an adequate explanation dooms the final order.

Accordingly, the final order was arbitrary and capricious because DOT drastically departed from its uniform practice of analyzing country-pairs and city-pairs without a reasonable explanation of why such analyses were not necessary in this case.⁹

The petitioners also argue that DOT held the joint venture to a higher standard than other similarly situated joint ventures when it made the Mexican government’s compliance with an open skies agreement a prerequisite to renewing approval of the joint venture. DOT tries to distinguish the other joint ventures the petitioners point to by arguing that the diplomatic situation in Mexico is unique. Again, we agree with the petitioners.

To survive arbitrary and capricious review, agencies must “treat like cases alike.” Westar Energy, Inc. v. FERC , 473 F.3d 1239, 1241 (D.C. Cir. 2007); see also Black Warrior Riverkeeper , 833 F.3d at 1289 (analyzing whether an agency articulated sufficient rationale for treating similar cases differently). But an agency can treat similarly situated parties differently if it “provide[s] a ‘satisfactory Page 19explanation for its action.’” Black Warrior Riverkeeper , 833 F.3d at 1290 (quoting Motor Vehicle Mfrs. Ass'n , 463 U.S. at 42-43). In reviewing that explanation, we consider whether there is a “rational connection between the facts found and the choice made” and “whether there has been a clear error of judgment.” Motor Vehicle Mfrs. Ass’n , 463 U.S. at 43.

In the 2025 final order, DOT first stated that compliance with “[a]n Open Skies regulatory framework is necessary under the competition and public interest analysis required by Sections 41309 and 41308 but not sufficient to obtain approval of and maintain a grant of [antitrust immunity]” (emphasis added). DOT acknowledged that an open skies “preliminary test” was not in the relevant statutes, but it justified such a requirement by stating that a foreign government’s compliance with an open skies agreement informs the required public interest and competition analysis of a joint venture.

We agree with the petitioners that DOT’s requirement that the Mexican government implement its open skies agreement at MEX was arbitrary and capricious because DOT failed to treat like cases alike. DOT has approved other joint ventures where there was no implementation of an open skies agreement at an airport where those joint ventures operate, and it did not reasonably explain why it imposed such a requirement here. See Black Warrior Riverkeeper , 833 F.3d at 1289. In 2009, the United States and Japan entered into an open skies agreement, but that agreement carved out Tokyo’s Haneda International Airport by prohibiting all-cargoPage 20 20 25-13546

service at Haneda, and Japan later limited American carriers to 18 daily slot-pairs at Haneda. Air Transport Agreement, Memorandum of Understanding, Annex § 3, U.S.-Japan (Dec. 14, 2009); Delta Air Lines, Inc. , DOT Order 2023-6-24 at 1 (June 30, 2023). Despite the lack of an implemented open skies agreement between the U.S. and Japan that covered Haneda, DOT granted approval of, and antitrust immunity for, two joint ventures between American and Japanese airlines to operate at Haneda. U.S.-Japan All. , DOT Order 2010-11-10 at 1 (Nov. 10, 2010).

DOT did not treat like cases alike because it held the petitioners to a higher standard for approval than the Japanese joint venture applicants when it made compliance with an open skies agreement at MEX necessary for approval in this case but approved two Japanese joint ventures despite the lack of an open skies agreement that included Haneda. And the final order does not adequately explain why the lack of open skies at MEX (an all-cargo ban and restrictive slot policies) mandated terminating approval of the joint venture when DOT has approved two joint ventures in Japan despite the lack of open skies at Haneda.

DOT resists our conclusion. DOT does not contest that it held the petitioners to a higher standard than other joint venture applicants. Instead, DOT attempts to distinguish this case, reasoning that only Mexico “is unacceptably distorting competition.' According to DOT, the United States and Japan “productively collaborated” before and after approval of the two joint ventures operating at Haneda and “competitive conditions at Page 21Haneda have improved” such that those joint ventures met the statutory factors in 49 U.S.C. § 41309. But DOT's argument is nonresponsive to the petitioners’ concern that DOT imposed an open skies requirement at MEX but not when evaluating approval of other joint ventures that operate at a foreign airport that is carved out of the relevant open skies agreement. Even if DOT is correct that Mexico, unlike Japan, has not productively collaborated with the United States on aviation issues, DOT made open skies a necessary step for approval of this joint venture while it was not necessary for the two similar joint ventures in Japan. Because DOT did not treat like cases alike when it held the petitioners and the U.S.-Japan joint venture applicants to different standards for approval, the final order was arbitrary and capricious. See Black Warrior Riverkeeper , 833 F.3d at 1289.

For the above reasons, the final order was arbitrary and capricious. Accordingly, we vacate the final order.

VACATED.

Page 22Rosenbaum, Circuit Judge, Concurring in the Judgment:

I agree that we must grant Delta Air Lines, Inc., and Aerovias de México, S.A. de C.V.'s petition for review of the Department of Transportation's final order ending approval of Petitioners' joint venture and withdrawing the joint venture's antitrust immunity. But I agree solely because DOT reasoned, in significant part, that Mexico City's Benito Juarez International Airport did not comply fully with Mexico’s open-skies agreement with the United States, and compliance with open-skies agreements was “necessary . . . to obtain approval of and maintain a grant of [antitrust immunity].' Delta Airlines, Inc., Aerovias de Mexico, S.A. de C.V. , Docket DOT-OST-2015-0070 (DOT Sept. 15, 2025) (“DOT Final Order”), at 17. As it turns out, that’s not entirely accurate.

During oral argument, DOT conceded that Haneda Airport in Tokyo, Japan, enjoys a carveout from the United States’s openskies agreement with Japan.1 Yet DOT approved, and granted antitrust immunity to, two joint ventures between American and Japanese airlines to operate at Haneda. U.S.-Japan All. , DOT Order 2010-11-10 at 1 (Nov. 10, 2010). To be sure, DOT had its reasons for exempting the joint venture at Haneda from the open-skies requirement. But that doesn't matter. The Haneda exemption shows that DOT’s supporting statement that an airport’s compliance with an open-skies agreement is 'necessary' wasPage 23 wrong. In fact, complete compliance with an open-skies agreement with the United States is not “necessary.” Certainly, it’s strongly preferred, But the Haneda arrangement shows that strict compliance with an open-skies agreement is not “necessary.”

Although the Haneda experience shows compliance with an open-skies agreement is not “necessary,” DOT used MEX’s failure to strictly comply with the open-skies agreement as a sufficient basis, in and of itself, to withdraw its approval of Petitioners’ joint venture. So for the reasons the Majority Opinion discusses at Part II.B., I agree with my colleagues that the DOT’s revocation here was “arbitrary and capricious.”

That said, most respectfully, I disagree with the Majority Opinion’s decision to grant the Petition because “DOT [allegedly] abruptly departed from its uniform precedent of conducting a comprehensive market analysis without reasonably explaining such a departure.” Maj. Op. at 14. DOT did no such thing.

Approvals of new joint ventures and withdrawals of approvals of existing joint ventures that DOT initially approved are not both apples. So the Majority Opinion errs in comparing DOT’s treatment of its withdrawal of approval of Petitioners' joint venture to DOT’s treatment of applications for new joint ventures.

When DOT receives an application for a new joint venture, it starts from ground zero. It has no preexisting knowledge of how the proposed joint venture functions in the air-traffic system and no preexisting analysis of potential antitrust problems that the proposed joint venture may present. So of course, DOT must Page 24conduct a comprehensive analysis of all the factors that go into its decision about whether to grant a new joint venture and antitrust immunity.

That’s not the case, though, when DOT considers whether to withdraw approval of an existing joint venture that it has previously fully analyzed and continues to monitor. In that case, as here, DOT knows what competition problems the joint venture presents because DOT has previously identified those concerns and because DOT can measure a joint venture’s ongoing progress on those parameters. If problems DOT identified when it initially approved the joint venture for a limited period have worsened, and the joint venture can’t sufficiently explain why that doesn’t present an unacceptable competition issue, there's nothing arbitrary or capricious about DOT’s decision to withdraw approval for the joint venture. And that’s so whether DOT conducts a “comprehensive market analysis” like it did with the initial joint-venture request or a more limited analysis that continues to home in on the same problems the DOT identified when it approved the initial application.

So DOT’s treatment of applications for new joint ventures does not tell us what “DOT’s uniform practice” is for withdrawing approval of an existing joint venture that is failing to satisfy concerns DOT identified when it approved the joint venture. See Maj. Op. at 15–16. And we can’t accurately characterize DOT’s different review practice for withdrawing approval of a joint venture as a “depart[ure] from its precedent” when the precedent Page 25we are considering is limited to that involving DOT’s review of applications for new joint ventures.

Nor does the fact that “[t]he statutory standard DOT must apply for initially approving and continuing approval of joint ventures is the same,” Maj. Op. at 17, somehow mean that DOT must necessarily engage in precisely the same scope of analysis on an initial application and a withdrawal of approval of a joint venture. After all, under the Administrative Procedure Act, DOT has discretion to decide 'how much data is necessary to fully address each issue.' Black Warrior Riverkeeper, Inc. v. U.S. Army Corps of Eng’rs , 833 F.3d 1274, 1285 (11th Cir. 2016). So as long as DOT’s reasons for not engaging in as comprehensive an evaluation when withdrawing approval for a joint venture as it did when approving the joint venture in the first place are not arbitrary and capricious (and for the reasons I've discussed, they aren't here), DOT need not perform precisely the same extent of analysis on a withdrawal of approval as it does when granting a joint-venture approval.

In short, DOT precedent involving withdrawals of approval of joint ventures provides the appropriate universe for comparison of DOT’s actions here. But the Majority Opinion identifies no such precedent from which DOT’s analysis here departs.

And when we evaluate the analysis DOT undertook when it withdrew approval of Petitioners’ joint venture, we can’t say it’s arbitrary and capricious in its own right (except to the extent that DOT disqualified Petitioners continuing approval because of thePage 26 open-skies problem I discussed earlier). So I do not join the Majority Opinion’s decision to grant the petition because of DOT’s failure to conduct as comprehensive an analysis of DOT's withdrawal of approval of Petitioners’ joint venture as it did when it granted the original application for joint venture.