United States v. Abbot

Don R. Willett, Circuit Judge, joined by Jones, Smith, Elrod, Haynes, Duncan, Engelhardt, Oldham, and Wilson, Circuit Judges: In July 2023, the State of Texas installed a 1,000-foot floating barrier in the Rio Grande near Eagle Pass, one of the nation’s busiest hotspots for illegal border crossings. The United States promptly sued, alleging that Texas violated the Rivers and Harbors Appropriation Act of 1899 (RHA). The district court granted a preliminary injunction directing Texas to reposition the barrier to the Texas riverbank. Texas appealed, and a 2–1 panel of our court affirmed. Our full court then granted rehearing en banc, vacated the panel opinion, and stayed the preliminary injunction pending appeal. With its preliminary-injunction dispute on appeal, the parties have continued to march toward trial in the district court.  The district court will soon confront the merits of the United States’ RHA claim on a further-developed factual record. Our focus in this interlocutory appeal, then, is narrow: Did the district court abuse its discretion in granting a preliminary injunction to the United States?  On this record, we conclude it did. The test for whether to grant or deny a preliminary injunction is long-standing and familiar. The district court should deny relief “unless the party seeking it has clearly carried the burden of persuasion” by showing that:  (1) it is likely to succeed on the merits, (2) it is likely to suffer irreparable harm without an injunction, (3) the balance of equities tips in its favor, and (4) an injunction is in the public interest.  The first factor—likelihood of success on the merits—is “the most important.”  For good reason. A district court that grants a preliminary injunction restricts a party’s conduct before its rights are fully litigated—at the risk that the party’s conduct was actually lawful.  It is on this first factor that numerous briefs, two oral arguments, and nearly a year’s worth of discussion and deliberation have largely focused. We ask: Can the United States likely prove during trial that Texas violated the RHA? That question turns on another: Can the United States likely prove that the barrier is located within a navigable stretch of the Rio Grande? Because the RHA extends only to navigable waters,  our answer to this question may—and in our view does—dispose of the first. Cognizant of our role as a court of review  and of the reality that a preliminary injunction is an exceptional remedy,  we hold that the district court clearly erred in finding that the United States will likely prove that the barrier is in a navigable stretch of the Rio Grande. We cannot square the district court’s findings and conclusions with over a century’s worth of precedent, which on a fair and faithful reading renders inapplicable or unpersuasive the evidence on which the district court relies. The United States alone has the “heavy burden” of showing a likelihood of success on the merits,  and the district court (and the dissenting opinions) cannot cure the United States’ evidentiary deficiencies by creatively reinterpreting binding caselaw. Because we conclude that the United States fares no better on the three other preliminary-injunction factors, we hold that the district court abused its discretion by granting the United States a preliminary injunction. Accordingly, we now DISSOLVE the stay pending appeal, REVERSE the district court’s order granting a preliminary injunction, and REMAND with instructions to vacate the preliminary injunction and for further proceedings consistent with this opinion. I About a year ago, Governor Abbott installed a floating barrier along a 1,000-foot stretch of the Rio Grande near Eagle Pass to deter illegal border crossings. The bright orange chain of tethered buoys—running parallel to the riverbank and anchored to the riverbed by concrete blocks—is one of many attempts by Governor Abbott to quell the influx of illegal immigration and drug trafficking into the state. The United States was quick to react. Less than two weeks after Texas installed the barrier, the United States sued the state under the RHA. The United States alleged that Texas violated § 10 of the RHAby (1) obstructing the navigable capacity of the Rio Grande without affirmative congressional authorization and (2) building the barrier without approval from the U.S. Army Corps of Engineers (Corps).  It sought a preliminary injunction to compel Texas to remove the barrier. The district court granted the preliminary injunction, ordering Texas to cease any work on the barrier and to reposition it to the Texas riverbank, but not remove it as the United States requested.  Texas promptly appealed and moved to stay the preliminary injunction. Our court granted an administrative stay, and a panel affirmed the district court’s preliminary injunction.  The full court then ordered rehearing en banc, vacated the panel opinion,  and granted Texas’s motion to stay the preliminary injunction pending our en banc review. II “We review the district court’s grant of [a] preliminary injunction for abuse of discretion, reviewing underlying factual findings for clear error and legal conclusions de novo.”  ”A finding of fact is clearly erroneous ‘when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.’”  A We begin with the first preliminary-injunction factor: whether the United States has shown that it is likely to succeed on the merits of its RHA claim. Success under RHA § 10 requires the United States to show that the barrier sits within navigable water.  Navigability is a question of fact,  and the definition of navigability has been repeated and refined by federal courts since at least 1870.  It was then that theSupreme Court explained that a river is navigable if it is “used, or [is] susceptible of being used, in [its] ordinary condition, as [a] highway[] for [interstate or foreign] commerce, over which trade and travel are or may be conducted in the customary modes of trade and travel on water.”  Later cases have fleshed out this definition. We now know that a river is navigable if non-commercial uses evince its suitability for commercial traffic, even if it is not presently being used for commerce,  or if reasonable improvements, even hypotheticalones, could makeitsuitablefor commercial use.  And it is fine if “the water course is interrupted by occasional natural obstructions or portages” or not suitable for commercial traffic during “all seasons of the year, or at all stages of the water.”  Once a river is found to be navigable, it remains so, even if natural or artificial changes later render it incapable of commercial use.  The definition of navigability is broad—unsurprising, given that Congress’s authority over navigable waters arises from its capacious Commerce Power —but it is not limitless. The Supreme Court’s command that the river “be[] of practical service as a highway of commerce”  is a significant constraint. That the river must be of “practical service” means that the commercial use or susceptibility of use must be more than “sporadic and ineffective”  or “exceptional.”  Navigability therefore does not extend to “every small creek in which a fishing skiff or gunning canoe can be made to float at high water.”  And, as we explain in greater detail later, being useful as a “highway of commerce” means that we must look for evidence of commercial use or susceptibility of use along the river.  Up to this point, we have described what we look for to assess navigability. Now we address where we look. Courts have for decades focused the navigability analysis—when addressing the scope of the government’s regulatory power generally and for the RHA specifically—on disputed segments of the river.  Here, only the 1,000-foot segment in which the barrier sits is disputed. To prove navigability, however, the United States and district court rely on evidence that spans a much further distance. They point to a 1975 Navigability Study by the Corps, which addresses a segment of the river that is over 300 miles long and includes this 1,000-foot stretch, as well as treaties and statutes that discuss the river at Eagle Pass more generally. As we explain below, none of these sources support a navigability finding for that longer 300-mile stretch, let alone for the 1,000 feet at issue. Accordingly, regardless of whether we ask if this 1,000-foot stretch only is navigable or whether this 1,000-foot stretch sits within some longer navigable stretch, as one of the dissenting opinions would prefer us to say,  our answer is the same: The barrier is not within navigable water. Curiously, that dissent resists our approach by arguing that Congress may exercise control over non-navigable stretches of the river to promote commerce on the navigable portions.  Rather than disproving our view that we consider individual segments of the river, this seems to support it. Moreover, the United States does not argue in the alternative that even if this 1,000-foot stretch is not navigable, the RHA still applies because the barrier obstructs commerce on the Rio Grande’s navigable sections. We cannot, as the dissent does, carry the United States’ burden by crafting arguments that the United States does not raise. Having settled on the proper test for navigability, we now turn to whether this stretch of the Rio Grande is navigable. The district court found that it is for two independent reasons: (1) because it was historically susceptible of use in interstate commerce in its natural condition; and (2) because it has been and continues to be susceptible of commercial use with reasonable improvements. The United States argues the same on appeal.  We consider each argument in turn. 1 We begin by considering whether this stretch of the river was historically used or susceptible of use in interstate or foreign commerce in its natural condition. Nothing the United States cites in support—statutes, treaties, a 1975 study by the Corps, or cases describing ferry traffic— convinces us that it is likely to succeed in making this showing. Start with the statutes that authorized various commercial projects across the Rio Grande at Eagle Pass. They provide no basis to find navigability. In Oklahoma v. Texas, the Supreme Court said in no uncertain terms that a similar act that “provided in substance that there should be no interference with navigation . . . was only precautionary and not intended as an affirmation of navigable capacity in that locality.”  The statutes cited by the district court, like the statute in Oklahoma, do not contain the necessary factual findings  that the Rio Grande was used or susceptible of use in commerce; they simply prohibit obstructions to navigation.  In Oklahoma, the Supreme Court explained that this prohibitory language was merely Congress playing it safe by barring obstructions in case that segment of the river turned out to be navigable.  We do not read these statutes to say more than they do. Same for the treaties. The district court pointed to the Treaty of Guadalupe Hidalgo and the Gadsden Treaty, which together require the United States and Mexico to maintain “free and common” “navigation of” the Rio Grande below New Mexico.  The district court gave no reason why these treaties establish navigability and are not merely precatory. The Supreme Court’s opinion in United States v. Rio Grande Dam & Irrigation Co. proves the point.  There, the Court said that these treaties obligated the United States “to preserve . . . the navigability of its navigable waters.”  That is, whether the river is navigable is a preliminary question that is not answered by the treaties themselves. Oklahoma suggests the same. Like the statute in Oklahoma, these treaties do not contain specific factual findings showing navigability and provide only that there should be no interference with “free and common” navigation.  On Oklahoma’s reasoning, then, the treaties are not “affirmation[s] of navigable capacity” but rather “precautionary” statements that navigability should not be obstructed where it exists.  Thus, the district court’s reliance on these treaties only serves to beg the question rather than answer it.  The Corps’s 1975 Navigability Study fares no better. The study assessed navigability between miles 275.5 and 610.0 of the Rio Grande, which includes the disputed 1,000-foot stretch near Eagle Pass. The district court cited the Corps’s study and subsequent determination that the Rio Grande is navigable in support of its finding that this stretch of the river was historically used or susceptible of use in interstate commerce—but the Corps never found navigability based on historical (pre-1975) or then-current (1975) use.  The district court and the United States on appeal conspicuously avoid mentioning that elsewhere in its report, the Corps criticized the evidence on which they rely. And, in doing so, they overlook a key feature of the test for navigability: Use of the river must have been more than “sporadic,” “ineffective,”  or “exceptional.”  Any evidence of past use in the Corps’s study is too sporadic and exceptional to establish historical navigability. The district court cited an anecdote that, in 1850, “[a] keelboat and a skiff, manned by sixteen men, ascended the river by channel to a point a thousand miles above the head of steam travel,” which was Roma, Texas. But the district court omitted that this historical account also described the expedition as “an astonishing penetration for a river with so little water.” And it ignored the Corps’s comment that “[t]here is no showing that substantial items of commerce were shipped from [Roma]” at the time of the expedition. In context, this expedition was at most an “ exceptional” use of the river, which does not suffice to show navigability.  In addition, the Corps stated that “there apparently has never been any ‘practical navigation’ between Roma . . . and El Paso” and that “at normal stages the river apparently was not navigable above Rio Grande City.” Even “during periods of sufficient flow,” only “fishing boats and other shallow draft craft” could navigate the river. And “[a]bove Laredo up to Eagle Pass . . . , navigation was impeded by rocks and ledges at low water stages.” As the Supreme Court explained, navigability does not extend to “every small creek in which a fishing skiff or gunning canoe can be made to float at high water.”  Rather, to be historically navigable, theriver must have been suitable “as [a] highway[] for commerce, over which trade and travel [was] or [could have been] conducted in the customary modes of trade and travel on water.”  There is no evidence of that here. As the Corps stated, any accounts of the river’s use were “sketchy,” and “actual accounts of commercial travel [were] lacking.” The Corps’s study therefore does not show that the river segment was used or susceptible of use in commerce in its natural condition. The Corps instead found navigability based on the treaties and the Supreme Court’s decision in Rio Grande Dam—neither of which can support a navigability finding here. We have already addressed and rejected the treaties as evidence of navigability. We now do the same for Rio Grande Dam. To find navigability, the Corps relied specifically on the Court’s statement that “the Rio Grande, so far as it is a navigable stream, lies as much within the territory of the United States as in that of Mexico, it being, where navigable, the boundary between the two nations.”  The Corps stated that the Court “alluded to the navigable portion” of the Rio Grande. But finding navigability requires much more than a mere allusion. To that end, Rio Grande Dam was careful to underscore a critical point: “[H]ow far up the stream navigability extends . . . should be determined by evidence” if it is not “a matter of general knowledge, or one that ought to be generally known.”  And, the Court acknowledged, “it is not so clear that it can fairly be said . . . that it is . . . a matter of common knowledge at what particular place between its mouth and its source navigability ceases.”  Rio Grande Dam never made those factual findings on the stretch of the river relevant here and in the Corps’s study. Because that case can at most be read to suggest that portions of the river might be navigable, it cannot stand as proof of navigability. We are left, then, with the cases from around the turn of the 19th century that describe ferries transporting cotton across the Rio Grande at Eagle Pass.  The United States argues that these ferries demonstrate the river’s navigability because they were used in commerce. Recall, though, that we are not looking for just any showing of commercial use or susceptibility. Rather, navigability requires evidence that the river was used or susceptible of use as a “highway for commerce,”  a critical qualifier that the United States and the dissenting opinions would have us ignore. That language should narrow our focus from the broader universe of “customary modes of trade and travel on water”  to the subset that uses the river as a highway.  Consistent with the requirement that the river be used as a highway, the Supreme Court has looked for trade or travel “up the river,”  ”down” the river,  and “along” it.  Accordingly, we too look for evidence of trade or travel along the Rio Grande near Eagle Pass. We cannot find navigability without it. Bank-to-bank ferry traffic does not supply the necessary evidence that the stretch of the river can sustain trade or travel along its length. Ferries that travel bank-to-bank, such as those formerly at Eagle Pass, do not leverage the river’s flow and course to transport goods or people along the river, as they would if they used the river as a highway. Rather, the river is an obstacle that the ferry overcomes to connect the land-based transit on either side of the river. Both the Supreme Court and Congress have understood cross-river ferries in this way. The Court has said that a ferry “cross[es]” the highway formed by the navigable river  and is “ a continuation of the [land-based, not river-based] highway from one side of the water over which it passes to the other.”  And about a decade before passing the RHA, Congress likewise recognized that ferries extend land-based transportation when it defined “railroad” in the Interstate Commerce Act to “include all bridges and ferries used or operated in connection with any railroad.”  It follows, then, that evidence of ferry traffic across the river says nothing about the underlying river’s capacity for navigation along its length.  The dissenting opinions disagree for at least three unpersuasive reasons. First, one of them argues that we have removed the Corps’s authority from any river that serves as a state border.  We fail to see how that is so. If the border river satisfies the test for navigability, then it is navigable. Evidence of trade or travel on those other rivers is not before us, and we make no attempt to opine on their navigability. Second, one of the dissenting opinions accuses us of creating a test that is unworkablewhen applied to lakes and other waterways.  How, they ask, can the United States establish navigability of a lake if it cannot show evidence of trade or travel across it? They misunderstand the inquiry. Lakes are obviously not rivers, as one of the dissenting opinions dutifully recognizes, so evidence showing navigability may look different for each. We have said only that a river is not used as a highway when ferries travel across it. We do not opine on whether a lake is used as a continuous highway when, for example, boats travel across the lake from a river entering the lake on one side to a stream exiting the lake on the other.  This case presents no such question. Finally, one of the dissenting opinions argues that Appalachian Electric and our opinion in Puente de Reynosa, S.A. v. City of McAllen  prove that ferry traffic demonstrates navigability because those opinions mentioned its presence.  But in neither opinion did the navigability finding rest solely on the cross-river ferry traffic. Instead, the courts pointed to evidence of use along the rivers and, in Appalachian Electric, of reasonableimprovements.  Thosecases thereforedo not rebut our view that navigability requires some evidence of trade or travel along the river.  Even assuming evidence of cross-river ferry traffic could alone demonstrate navigability, as the United States and the dissenting opinions argue, navigability could exist only in the narrow path where the ferries traveled. The United States and district court have pointed to no evidence that the Eagle Pass ferries crossed within the 1,000 feet in which the barrier now resides. Neither case they cite identifies the points at which the ferries began or ended—other than to say that they passed between Eagle Pass, on one side, and Piedras Negras, on the other.  Although it is theoretically possible that the ferries’ paths overlapped with the current location of the barrier, that something is possible does not mean that it is likely. And only by showing that its success on the merits is likely can the United States satisfy the first of the four preliminary-injunction factors. Such an “extraordinary remedy” cannot be justified by anything less.  Accordingly, without at least some evidence that the historic Eagle Pass ferries crossed where the barrier is now—evidence that is presently lacking—it could not show that it will likely succeed in demonstrating navigability. At bottom, the United States’s argument for historical navigability teeters on only inconsistent and exceptional accounts of past use along this stretch of the Rio Grande. The United States therefore has not carried its burden to show that it is likely to succeed in proving under the RHA that the barrier sitsin a portionof theriver that was historically navigablein its natural condition. 2 Alternatively, the district court concluded, and the United States argues on appeal, that this section of the Rio Grande has been and continues to be susceptible of commercial use with reprioritization and increased flow from nearby dams. But the mere potential for improvements does not prove navigability. As we have explained, any hypothetical improvements must at least be reasonable to support a finding that a river not presently used in or suitable for commerce is nonetheless navigable.  That is, the United States must show “a balance between cost and need at a time when the improvement would be useful.”  The United States and the dissents, however, contend that the United States need only show that improvements are possible, not that they are also reasonable. On this view, the United States could cite any number of improvements—no matter how impractical, improbable, or disfavored—to justify a navigability finding. That defies the Supreme Court’s instruction in Appalachian Electric that “[i]n determining the navigable character of the [river], it is proper to consider the feasibility of interstate use after reasonable improvements which might be made.”  In other words, we must consider reasonableness of hypothetical improvements when we assess navigability.  Thecosts and benefits must, of course,relateto whatever timein which those hypothetical improvements might be made. But, using prospective costs and benefits, we must still analyze reasonableness when faced with whether a river is navigable. One of the dissenting opinions counters that the Supreme Court in AppalachianElectric held thatastretch of theriver was navigable, even though army engineers had at one point described the cost of improvements as “prohibitive.”  This, they argue, shows that reasonableness is irrelevant. Quite the opposite, this part of Appalachian Electric proves rather than disproves our point. The Court concluded that the segment of the river was navigable based on evidence of present use and of “its ready improvability at a low cost for easier keelboat use”—that is, the Court expressly considered costs and benefits.  While the Court concluded that improvements to make the river suitable for “small steamboats” were cost-prohibitive, it reached a different conclusion as to improvements that would “make the section a thoroughfare for the typical, light commercial traffic of the area,” including keelboats.  Because evidence showed that “little was needed in the way of improvements” for that purpose, itfound thatstretch of theriver navigable.  Accordingly, we ask whether the United States has cited sufficient evidence of costs and benefits to assess whether it can likely prove that these hypothetical improvements are reasonable. We conclude that it comes up short. To argue navigability based on reasonable improvements, both the district court and the United States rely on the Corps’s 1975 study and an expedition report from 1850. According to the Corps’s 1975 study, “Improvement of the Rio Grande for navigation [was then] physically possible.” For example, it explained, “the Falcon and Amistad Reservoirs [could be] judiciously used to provide sufficient flow for continuous navigation.” But it made no showing that those improvements would be reasonable. To the contrary, the study noted “serious ecological objections to any channelization” and “doubtful” economic justifications.  The report from 1850, cited by the Corps in its 1975 study, is no better. It concluded that “the Rio Grande was susceptible to be improved for steam navigation up to Babbitt’s Falls” but gavenoindicationof theimprovements’ reasonableness. Without some evidence of costs and benefits, the district court could not find that there were reasonable improvements that would have made this segment of the river susceptible of navigation at the time of the reports in 1850 and 1975. Nor could it look to the Corps’s 1975 study and the 1850 report to find evidence of reasonable improvements that would make this river segment navigable now: Even if these sources offered evidence of costs and benefits then, they would say nothing about the costs and benefits of those improvements today or in the future, whenever those improvements would be useful.  Looking to the modern day, the United States tells us that federal agents operate small watercraft on the river and that a business is conducting kayak tours, seemingly suggesting that the river could facilitate commercial traffic involving bigger watercraft if its flows were increased by water from nearby reservoirs.  But while this may suggest that the river could be improved for commerce, it says nothing tangible about the improvements’ costs or benefits. The United States at most offers a few generalized statements about the use of dams elsewhere in regulating river flow, ultimately failing to explain why repurposing water from the dams would be cost-effective for this river. Because we cannot say that costs are likely justified by benefits on the record before us, the United States has not carried its burden to show that reasonable improvements could facilitate commerce. The district court therefore erred in holding that the United States is likely to prove that this Rio Grande segment is navigable based on reasonable improvements that could facilitate commercial use. * * * By taking a careful eye to the record and over a century of caselaw, we have considered what does and does not support a navigability finding, a task the dissenting opinions seem disinclined to undertake—at least not with the specificity demanded by such a fact-intensive dispute. We arrive at the only conclusion possible on the evidence before us: The United States will likely fail to prove navigability because (1) the at-issue stretch of the river likely was not historically navigable, and (2) the United States has not shown that there are likely reasonable improvements that could render the river navigable. The dissenting opinions, however, reproach us for reaching this point only by overstepping our appellate role, reweighing the evidence in our favor, and thus requiring the United States to show it will undeniably succeed on the merits.  These criticisms are misguided, arising as they do from the dissenting opinions’ misunderstanding about—and our resulting disagreement over—whether certain evidence supports a navigability finding. Where we say no, the district court and dissenting opinions say yes. And where we have held the United States to its burden of citing evidence that demonstrates a likelihood of success, the dissenting opinions are willing to fill evidentiary gaps by stretching and straining precedent to find evidence of navigability where there is none. The dissenting opinions’ impressionistic review of the record and the caselaw leaves unchecked the district court’s clear errors. Because we cannot and do not fill gaps in the United States’ evidence with our own court-supplied conjecture, we are unavoidably left to conclude that the district court clearly erred. Having reviewed the district court’s factual findings for clear error and conclusions of law de novo, we hold that the barrier is not in a navigable stretch of the Rio Grande and thus that the RHA likely does not apply.  Accordingly, the United States has not shown a likelihood of success on the merits—”the most important of the preliminary injunction factors.”  B Wenowturn to thefinal three preliminary-injunctionfactors: whether the United States “is likely to suffer irreparable harm in the absence of preliminary relief, that the balance of equities tips in [its] favor, and that an injunction is in the public interest.”  When the government is a party, “the government’s and the public’s interests merge.”  So we consider the last two factors together.  The United States, however, urges us to forgo all three equitable factors. Citing our decisions in United States v. Marine Shale Processors  and United States v. FDIC,  which both predate Winter, the United States contends that we need not reach them at all because it seeks to enforce a public-interest statuteand has shown that it is likely to succeed on the merits. We disagree. We have already concluded that the United States has not shown a likelihood of success on its RHA claim, so even if Marine Shale Processors and FDIC have not been overtaken by Winter, they do not control here.  That means that we presumptively apply the traditional preliminary-injunction factors from Winter.  As the Supreme Court has long explained and recently confirmed, “we do not lightly assume that Congress has intended to depart from established principles,” including the traditional Winter preliminary-injunction factors .  Accordingly, “absent a clear command from Congress, [we] must adhere to the traditional four-factor test.”  Looking to the text of the RHA, we find nothing that “overcomes the presumption that the four traditional criteria govern a preliminary-injunction request by” the United States.  The RHA provides that “the removal of any structures or parts of structures erected in violation of the provisions [here, § 10] . . . may be enforced by the injunction of [the] district court.”  This language contains no “specific instruction that suggests Congress altered the traditional equitable rules.”  Instead, by stating that the court “may” issue an injunction, the RHA fairly “invokes the discretion that courts have traditionally exercised when faced with requests for equitable relief.”  We therefore turn to the remaining equitable factors. First, the district court erred in concluding that the United States was likely to suffer irreparableharm without thepreliminary injunction. Whilewe acknowledge that “every judge must be sensitive to the Government’s concerns about international affairs,”  it is nonetheless unclear how the injunction alleviates these harms. The record indicates that the alleged diplomatic harms arise from the “construction and presence” of the barrier and that Mexico has demanded the barrier’s “prompt removal.” The preliminary injunction, however, requires Texas to shift the barrier to the Texas bank, not remove it from the river altogether. In ordering the barrier moved, but not removed, the district court’s preliminary injunction is nonresponsive to the United States’ alleged diplomatic harms. So regardless of whether we affirmed or reversed the district court’s judgment granting a preliminary injunction, it appears that the United States would be left in much the same diplomatic position. One of the dissents, however, urges us to conclude that the preliminary injunction is well-tailored to the United States’ alleged harms given the President of Mexico’s positive comment about the injunction during a press conference in September.  Again, the dissent attempts to shoulder the United States’ burden by making arguments the United States has not made itself. The United States cited the President’s statement in its panel brief only once to say—vaguely and without any further detail—that the district court granting a preliminary injunction had “some positive impact” on relations with Mexico. Seemingly abandoning this line of argument, counsel for the United States told the panel at oral argument that there was no record evidence that moving the barrier to the riverbank, as the preliminary injunction requires, would ease the alleged diplomatic tensions.  The United States likewise makes no effort to convince the full court that the preliminary injunction—ordering the barrier moved but not removed—has in any way lessened its ameliorated diplomatic harms. It makes no mention whatsoever of the President of Mexico’s statement in its en banc brief. Instead, the United States’ en banc brief continues to argue that diplomatic harms will persist if the barrier is “not removed.” The United States thus presses a demonstratively inconsistent argument: that we should affirm the preliminary injunction, which does not order the barrier removed, because it will prevent it from suffering diplomatic harm that it has argued will be cured only by the barrier’s removal. The United States cannot justify a preliminary injunction by appealing to harms which the injunction cannot resolve. The United States also fails to show that the balance of the equities and the public interest favor granting the injunction. The district court erroneously identified three reasons that it said favor the United States: the barrier threatens human life and safety, impairs navigation, and violates the RHA. None of them suffices. First, Texas and the district court disagree about the barrier’s safety and usefulness for deterring drug trafficking. Texas argues that the barrier “was designed . . . to save lives and direct migrants to appropriate . . . points of entry while deterring unlawful, dangerous crossings; drug smuggling; human trafficking; and terrorist infiltration” at “one of the most active drug-trafficking and human-trafficking hotspots on the river.” For its part, the United States—which bears the burden on the preliminary injunction factors —has offered no evidence that the barrier threatens safety. The district court tried to spin the river’s naturally treacherous conditions as evidence that the barrier is dangerous. That is not a logical leap we can or should make. The district court again set aside reason by inferring that the barrier caused the deaths of two people found nearby—one at the southern end of the barrier and the other three miles upriver—from news articles that said nothing of the sort.1  Even assuming the district court properly took notice of these articles, they do not support the district court’s conclusion that the barrier is a “threat to human life.” Although the articles reported the deaths, they noted that the causes of death were unknown and even suggested that the individual found at the barrier had drowned elsewhere and floated into the buoys.  On appeal, Texas refutes the district court’s causal inference, contending that the barrier has been “under constant surveillance” and has not harmed anyonewho has tried to cross. Texas gets thebetter of the district court. Given the discovery of one of the individuals three miles up from the barrier and the plausible inference that the other individual also drowned upriver, the district court could do no more than speculate that the barrier was the cause—and “mere speculation” cannot carry the day.  Second, the district court should not have found that the barrier is an “impairment to free and safe navigation” and a “contraindication to the balance of priorities Congress struck in the RHA.” Because the United States is not likely to prove that the barrier is in navigable water and thus is not likely to succeed on its RHA claim, neither of these factors carries weight. In sum, the district court’s analysis of these equitable factors was unpersuasive, unsubstantiated, and incorrect. The United States has not carried its burden. III Because the United States has not “clearly carried the burden of persuasion”  on even one of the requirements to obtain the “extraordinary remedy”  of a preliminary injunction, we DISSOLVE our stay pending appeal, REVERSE the district court’s order granting a preliminary injunction, and REMAND with instructions to vacate the preliminary injunction and for further proceedings consistent with this opinion. Andrew S. Oldham, Circuit Judge, concurring: I join the majority opinion in full. I write separately to underscore that, because the Rivers and Harbors Act (“RHA”) does not apply to Texas’s buoys, “we need not and do not reach Texas’s argument that Article I, § 10, clause 3 of the U.S. Constitution authorizes it to erect the barrier in defense of a border ‘invasion’ even if the barrier violates the RHA.” Ante, at 24 n.86. One of my esteemed colleagues nonetheless contends we must reach theconstitutionalquestion. Post, at63–64(Ho, J., dissenting in relevantpart). I can imagine only three reasons to justify that approach. Each merits a brief response. First, it is well settled that we should not reach constitutional questions if we can instead decide the case on a non-constitutional ground. See Ashwander v. TVA, 297 U.S. 288, 347 (1936) (Brandeis, J., concurring); see also Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 251 (2012). Insofar as my colleagues agree with the Department of Justice that the RHA does apply, see, e.g., post, at 80 (Douglas, J., dissenting), I suppose that would necessitate reaching the constitutional questions in this case. But it is a complete answer to say that the statutory analysis in the majority opinion is correct; the Justice Department is wrong; and that is enough to reverse the district court’s preliminary injunction. Second, myesteemed colleague says “[t]hepolitical questiondoctrine is jurisdictional,” so “we [must] address it first.” Post, at 63 (Ho, J., dissenting in relevant part) (quotation omitted). True, we must establish our jurisdiction before exercising the judicial power. See Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 93–94 (1998). That means to issue or affirm an injunction, we must find the plaintiff is likely to succeed in establishing both jurisdictionand themerits of its claim. See Munaf v. Geren, 553 U.S. 674, 691– 92 (2008). But to reverse an injunction, there is no order of operations. See id. at 691. As the Supreme Court has explained the rule, we must reverse an injunction where there is an “insuperable objection, in point of jurisdiction or merits.” Denver v. New York Trust Co., 229 U.S. 123, 136 (1913) (emphasis added); see also ibid. (assuming jurisdiction and reaching merits of underlying claim in reviewing preliminary injunction); 16 Charles Alan Wright, Arthur R. Miller & Edward H. Cooper, Federal Practice and Procedure § 3921.1 (3d ed.) (“Subject-matter jurisdiction always is open [for appellate review of an injunction]; the issue need not be resolved, however, if it requires determination of factual matters not developed in the record or legal issues not sufficiently defined in the current state of the proceedings.” (footnote omitted)). This rule makes sense because reversing a preliminary injunction prevents the exercise of judicial power over the defendant—regardless of whether we base that reversal on the plaintiff’s inability to establish jurisdiction, to prove the elements of its claim, &c. Third, the only other reason for confronting the constitutional question at this stage is some extrinsic consideration that strays beyond the present appeal. See, e.g., Munaf, 553 U.S. at 691–92. But I cannot imagine such a consideration here. The political question doctrine requires “a series of interpretations of the substantive meaning of particular clauses of the Constitution.” Richard H. Fallon, Jr., John F. Manning, Daniel J. Meltzer & David L. Shapiro, Hart and Wechsler’s The Federal Courts and The Federal System 248 (7th ed. 2015). It is of course true that the invocation of Section 10 constitutes a non-justiciable political question; the parties agree on that, as does every member of our en banc court. See Tex. Letter Br. at 1 (May 22, 2024) (chronicling parties’ agreement); see post, at 38–63 (Ho, J., dissenting in relevant part) (discussing the undisputed issue). But we only get to that invocation if the Justice Department can prove a whole host of other facts and legal elements—none of which it can. And even if we got all the way to the end of this case and had to confront the substantive meaning of Section 10, what would it mean for the buoys in this case? My esteemed colleague offers only a footnote. See post, at 60 n.14 (Ho, J., dissenting in relevant part). And in it he suggests, with nary a word of constitutional analysis, that Section 10 simply incorporates various provisions of customary international law. See ibid. (citing, inter alia, Authority Under International Law to Take Action If the Soviet Union Establishes Missile Bases in Cuba, 1 Op. O.L.C. Supp. 251, 252 (1962)). Even in a case premised on a non-justiciable political question, we have jurisdiction to reverse a preliminary injunction. See, e.g., Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 24 (2008) (citing the political question doctrine as enunciated in Gilligan v. Morgan, 413 U.S. 1, 10 (1973), because the district court’s injunction interfered with the Navy’s military judgments, and then exercising jurisdiction to reverse). The majority correctly does that here and renders an opinion that will end this litigation on the merits. That is a far better course than dodging the merits and allowing the Department of Justice to escape a judgment that will have preclusive effect. Finally, dismissing this case for lack of jurisdiction would have far-reaching consequences. For example, the State of Texas and the Justice Department are currently litigating whether the Biden Administration can remove concertina wire erected by the State of Texas along its border with Mexico. The State brought suit under both state and federal law to enjoin removal of that barrier. See Compl., Texas v. DHS, No. 2:23-CV-55-AM (W.D. Tex. Oct. 24, 2023); see also id. 30(h) (citing Texas’s Section 10 powers). If the political question doctrine bars the buoy suit, would it bar that suit too?*Thankfully, the majority opinion does nothing to interfere with the wire case. Priscilla Richman, Chief Judge, concurring in the judgment: I largely agree with the majority opinion’s thorough analysis of this case. The United States has not presented evidence that commercial traffic has navigated or with reasonable modifications could navigate along, or across, the 1,000-foot stretch of the Rio Grande that contains the floating barrier. Accordingly, there is no evidence in the current record that would supporttheconclusion thatthissegment of the Rio Grandeis navigableunder the Supreme Court’s precedent.1 The United States has not demonstrated a likelihood of success on the merits. However, I disagree with the majority opinion’s conclusion that the United States cannot establish this section of the Rio Grande is navigable through evidence of bank-to-bank ferry traffic.2 The majority opinion’s conclusion that only “evidence that th[is] stretch of the river can sustain trade or travel along its length” can establish navigability is likely incorrect.3 If there were evidence that a commercial ferry—carrying people, vehicles, or goods, for example—had been operating across this portion of the river until the barrier was installed, then it seems clear that, under the RHA, Texas could not obstruct the waterway in which the ferry travelled. Therefore, it also seems clear that if the disputed river segment once served as a ferry crossing or could serve as a ferry crossing in the future with reasonable modifications, then this part of the river is likely navigable. The majority opinion nonetheless argues that because ferries that travel bank-to-bank “do not leverage the river’s flow and course to transport goods or people along the river,” they do not “use[] the river as a highway” as required by Supreme Court precedent.4 True, the Supreme Court has clarified that whether a water course “serve[s] as a highway” for commerce is the “feature that distinguishes between navigability and non-navigability.”5 But the Supreme Court has never interpreted the inquiry to be limited to travel along the length of a river when its navigability was at issue. In Utah v. United States,6 the Supreme Court concluded there was sufficient evidence that the Great Salt Lake—which was “not part of a navigable interstate or international commercial highway”—was navigable.7 The RHA was not at issue in that case, but the Court applied the same test for navigability.8 Indeed, the Court explained that the test for navigability as set forth in The Daniel Ball,9 “applies to all water courses,” including intrastate lakes for purposes of determining public ownership of its bed.10 Applying this test, the Court concluded that “the lake was used as a highway and that is the gist of the federal test.”11 A lake can be navigated by crossing it or hewing to its shores. The inquiry is whether it is used as a “highway” for commerce. A commercial ferry’s pathway across a river fits within the parameters of a “highway” for commerce. As a result, the majority opinion draws the inquiry too narrowly. If the evidence were to establish that this section of the Rio Grande has served, or with reasonable modifications could serve, as a ferry crossing, then this section is likely navigable. James C. Ho, Circuit Judge, concurring in the judgment in part and dissenting in part: A sovereign isn’t a sovereign if it can’t defend itself against invasion. Presidents throughout history have vigorously defended their right to protect the Nation. And the States did not forfeit this sovereign prerogative when they joined the Union. Indeed, the Constitution is even more explicit when it comes to the States. Presidents routinely insist that their power to repel invasion is implied by certain clauses. But Article I, section 10 is explicit that States have the right to “engage in War” if “actually invaded,” “without the Consent of Congress.” U.S. Const. art. I, § 10, cl. 3. See also United States v. Abbott, 92 F.4th 570, 579–80 (5th Cir. 2024) (Ho, J., dissenting).1 Texas Governor Greg Abbott has invoked Article I, section 10 in response to the ongoing illegal immigration crisis. A majority of the Nation’s governors have endorsed that decision. Former senior FBI officials have advised Congress that the illegal immigration crisis constitutes “an invasion of the homeland.” In response, the U.S. House of Representatives has formally recognized illegal immigration as a national security crisis, and a similar resolution is pending in the U.S. Senate. See id. at 578 (collecting authorities). So there is ample support, both among the States and at the National level, that this is a good faith invocation of Article I, section 10. Two established legal principles should compel the conclusion that federal courts lack jurisdiction to review Governor Abbott’s invocation of Article I, section 10, and thus lack jurisdiction to hear this case. To begin with, at various times in our Nation’s history, the United States has recognized that the right to use military force to repel invasion includes attacks by hostile foreign individuals as well as foreign sovereigns. Consider, for example, how Texas responded to foreign incursions during the 19th century. See generally Texas Frontier Troubles, H.R. Rep. No. 44-343, at 164–67 (1876); J. Fred Rippy, The United States and Mexico 1821–1924(1926). For decades, bands of criminals from Mexico crossed the border into Texas to kill Americans and steal cattle. These were private, non-sovereign acts, unauthorized by the Mexican government. Texas sought assistance to stop the attacks, but both Mexico and the United States refused. So a succession of Texas Governors concluded that they had no choice but to take matters into their own hands. See, e.g., Texas Frontier Troubles, supra, at xv (noting efforts of Governor Sam Houston). They ordered state troops to defend the border, and when necessary, to pursue the criminals across the Mexican border. These issues came to a head in 1874. The United States Attorney General informed Texas Governor Richard Coke that his military orders violated a federal statute prohibiting unauthorized hostile actions in other countries. Id. at 164 (citing Act of April 20, 1818, § 6, 3 Stat. 449). Governor Coke responded by invoking his constitutional authority under Article I, section 10. See id. at 164–67 (text of letter available in the Appendix). In doing so, he acknowledged that federal authorities had the “power” to obstruct his actions if they wanted to. Id. at 167. After all, the United States possesses superior military forces, as well as the authority to call statemilitias into federal service (U.S. Const. art. I, § 8, cl. 15–16; id. art. II, § 2, cl. 1). But he explained his “clear conviction[]” that, notwithstanding federal law, States have the same “ right” to self-defense that would “ordinarily reside in the United States.” Id. at 166–67. In response, theAttorney General acquiesced to theGovernor’s claim of authority. Id. at xvi. More recent events further confirm that the war power is not limited to hostile foreign sovereigns. Both before as well as after the attacks of September 11, 2001, Presidents have agreed that military force may be used against terrorists and terrorist groups. See, e.g., The 9/11 Commission Report 132, 485 n.123 (2004) (President Clinton approved applying the law of armed conflict to terrorists); Hamdi v. Rumsfeld, 542 U.S. 507, 510 (2004) (plurality opinion) (“President [Bush] ordered United States Armed Forces . . . to subdue al Qaeda”). See also Curtis A. Bradley & Jack L. Goldsmith, CongressionalAuthorization and the War on Terrorism, 118 Harv. L. Rev.2047, 2066–67 (2005) (noting authorizations of forceagainstpirates and slave traders, President Wilson’s use of U.S. troops to pursue Pancho Villa into Mexico, President Clinton’s missile strikes against al Qaeda targets, and other examples). The use of military force in these contexts continues to be a matter of great controversy. It was controversial before September 11, and it remains controversial after September 11. But that’s the point. These are political controversies, not judicial ones. Which private acts warrant military action are questions for the political branches, not the courts. And that’s the second legal principle that governs this case. Courts have no business deciding which national security threats are sufficiently serious to warrant a military response, and which are not. Supreme Court precedent and longstanding Executive Branch practice confirm that, when a President decides to use military force, that’s a nonjusticiable political question not susceptible to judicial reversal. I see no principled basis for treating such authority differently when it’s invoked by a Governor rather than by a President. If anything, a State’s authority to “engage in War” in response to invasion “without the Consent of Congress” is even more textually explicit than the President’s. Of course, the use of military force—whether by a President or by a Governor—can result in violations of individual rights that are subject to judicial review under governing Supreme Court precedent. But no such claim is presented here. This is a threshold, direct challenge to the State’s invocation of its constitutional authority, full stop. Accordingly, I agree that the preliminary injunction entered against Governor Abbott must be reversed. But I get to that place through a different path. I would instruct the district court to dismiss this case for lack of jurisdiction. I therefore concur in the judgment in part and dissent in part.2 I. The political question doctrine prohibits courts from adjudicating cases that are “outside the courts’ competence and therefore beyond the courts’ jurisdiction.” Rucho v. Common Cause, 588 U.S. 684, 696 (2019). The Supreme Court has identified several contexts in which the political question doctrine applies. For example, a case may present an issue that “lack[s] . . . judicially discoverable and manageable standards for resolving” he dispute. cases. Baker v. Carr, 369 U.S. 186, 217 (1962). This is one of those A. Both the United States and Governor Abbott agree that whether an invasion exists under Article I, section 10 is a nonjusticiable political question—they simply disagree on the implications of that determination. This consensus should surprise no one. To begin with, “there are no manageable standards to ascertain whether or when an influx of illegal immigrants should be said to constitute an invasion.” California v. United States, 104 F.3d 1086, 1091 (9th Cir. 1997). It’s hard to imagine that anyone would conclude that a few border crossings would suffice to justify a military response. On the other hand, numerous officials have concluded that military action was warranted in response to bands of Mexican criminals in the 19th century and terrorist attacks in the 20th and 21st centuries. Determining where the present illegal immigration crisis falls along this spectrum is not a legal question for judges, but a political determination for the other branches of government. Cf. Wu Tien Li-Shou v. United States, 777 F.3d 175, 182 (4th Cir. 2015) (“[I]t is difficult for a court even to define what war is.”); Campbell v. Clinton, 203 F.3d 19, 26 (D.C. Cir. 2000) (Silberman, J., concurring) (noting the lack of “a coherent test for judges to apply to the question what constitutes war”). As the Supreme Court has repeatedly explained, “a controversy involves a political question where there is . . . a lack of judicially discoverable and manageable standards for resolving it.” Zivotofsky ex rel. Zivotofsky v. Clinton, 566 U.S. 189, 195 (2012) (cleaned up). “In such a case, we have held that a court lacks the authority to decide the dispute before it.” Id. This conclusion is also consistent with how courts have approached other, analogous constitutional provisions. Consider, for example, how our sister circuits have treated Article IV, section 4. Under that provision, the United States “shall protect each of [the States] against Invasion.” U.S. Const. art. IV, § 4. Courts across the country have held that determining whether an invasion has occurred for purposes of Article IV, section 4 is a nonjusticiable political question. See, e.g., California, 104 F.3d at 1091 (“[T]he issue of protection of the States from invasion implicates foreign policy concerns which have been constitutionally committed to the political branches.”); Padavan v. United States, 82 F.3d 23, 28 (2nd Cir. 1996) (“[T]he plaintiffs’ Invasion Clause claim is nonjusticiable. The protection of the states from ‘invasion’ involves matters of foreign policy and defense, which are issues that the courts have been reluctant to consider.”); Chiles v. United States, 69 F.3d 1094, 1097 (11th Cir. 1995) (“[W]hether the level of illegal immigration is an ‘invasion’ of Florida and whether this level violates the guarantee of a republican form of government present nonjusticiable political questions.”).3 I see no principled basis for treating Article I, section 10 differently from Article IV, section 4, and other analogous texts. See also, e.g., Hamdi, 542 U.S. at 578 (Scalia, J., dissenting) (noting in the Suspension Clause context that “whether the attacks of September 11, 2001, constitute an ‘invasion,’ and whether those attacks still justify suspension several years later, are questions for Congress rather than this Court”).4 B. Deferring to a State’s determination of an invasion is also consistent with how courts treat the President. See, e.g., Martin v. Mott, 25 U.S. (12 Wheat.) 19, 30 (1827) (holding that, under a statute delegating to the President the authority to call forth the militia to repel invasions, “the authority to decide whether the exigency has arisen, belongs exclusively to thePresident, and . . . his decisionisconclusiveupon all other persons”); The Prize Cases, 67 U.S. (2 Black) 635, 670 (1862) (“Whether the President in fulfilling his duties, as Commander-in-chief, in suppressing an insurrection, has met with such armed hostile resistance, and a civil war of such alarming proportions as will compel him to accord to them the character of belligerents, is a question to be decided by him.”); Hamdi, 542 U.S. at 531 (plurality opinion) (“Without doubt, our Constitution recognizes that core strategic matters of warmaking belong in the hands of those who are best positioned and most politically accountable for making them.”); id. at 585– 86 (Thomas, J., dissenting) (“[W]e lack the information and expertise to question whether Hamdi is actually an enemy combatant, a question the resolution of which is committed to other branches.”).5 Judicial deference to the President is also (unsurprisingly) consistent with the longstanding position of the Executive Branch. See, e.g., Brief for Appellees at 1, Jaber v. United States, 861 F.3d 241 (D.C. Cir. 2017) (No. 16-5093) (“The plaintiffs seek to challenge an alleged drone strike in Yemen. . . . Plaintiffs’ claims would require the courts to second-guess an alleged decision by the Executive to use force against a counter-terrorism target . . . . This Court has repeatedly found that such issues present political questions that are beyond the competence of the courts.”); Brief for the Appellee on Rehearing En Banc at 51, El-Shifa Pharm. Indus. Co. v. United States, 607 F.3d 836 (D.C. Cir. 2010) (en banc) (No. 07-5174) (arguing that claims were “barred by the political question doctrine” because the plaintiffs “have asked this Court to opine on the foreign policy determinations of the United States and pronounce the decision to initiate military hostilities unjustified”); Legality of the Use of Military Commissions to Try Terrorists, 25 Op. O.L.C. 238, 262 (2001) (“[E]ven without any action by Congress to acknowledge a state of war, the President, in his constitutional role as Commander in Chief, and through his broad authority in the realm of foreign affairs, also has full authority to determine when the Nation has been thrust into a conflict that must be recognized as a war and treated under the laws of war.”) (citation omitted); Brief for Appellee at 36, Campbell v. Clinton, 203 F.3d 19 (D.C. Cir. 2000) (No. 99-5214) (“[D]etermining the existence of a ‘war’ for purposes of allocating responsibilities between the political branches requires a political, not a judicial, judgment.”). I see no principled basis for deferring to a President’s determination of an invasion, but not to a State’s. If anything, the Constitution is more explicit in authorizing the States to “engage in War” than it is with the President. The Constitution grants Congress the power to “declare” war. U.S. Const. art I, § 8, cl. 11. And there is no explicit vesting of Presidential power to engage in war—not even in response to an invasion. The Constitution says only that the President “shall be Commander in Chief of the Army and Navy of the United States, and of the Militia of the several States, when called into the actual Service of the United States.” Id. art. II, § 2, cl. 1. So the President’s power to engage in war in response to invasion is implied, not explicit. By contrast, Article I, section 10 expressly authorizes States to “engage in War” in response to an invasion. See, e.g., Arizona v. United States, 567 U.S. 387, 419 (2012) (Scalia, J., concurring in part and dissenting in part) (noting that Article I, section 10 serves as an “acknowledgment of the States’ sovereign interest in protecting their borders” and “leaves intact [States'] inherent power to protect their territory”); John C. Yoo, War and the Constitutional Text, 69 U. Chi. L. Rev. 1639, 1667 (2002) (“[T]he Constitution only allocates to Congress the declare-war power and to the President the commander-in-chief power, without specifically stating—as it does in Article I, Section 10 with regard to the states—how those powers are to interact.”); id. (comparing “[t]he Constitution’s creation of a specific, detailed war powers process at the state level” with “its silence at the federal level”). C. The United States acknowledges—as it must—that there are at least some narrow circumstances in which States may “engage in War” “without the Consent of Congress.” U.S. Const. art. I, § 10, cl. 3. But the United States insists that any right that a State may have to defend itself is a “time-limited emergency authority,” and that the right ceases once the federal government has had theopportunity to respond. The United States rests this position largely on the phrase “as will not admit of delay.” This contention conflicts with the plain text of Article I, section 10. The text makes clear that no consent of Congress is necessary for a State to “engage in War” so long as the State is “actually invaded, or in such imminent Danger as will not admit of delay.” Id. The phrase “as will not admit of delay” modifies “imminent Danger,” not “actually invaded.” Id. Nothing in the text prevents a State from defending itself if it is “actually invaded,” whether or not the United States is also able and willing to protect the State.6 Alternatively, the United States posits that Article IV, section 4 commits the question of whether an “invasion” exists to the federal government, rather than to the States. But Article IV, section 4 focuses on federal action, not State action. This case, by contrast, concerns Article I, section 10. And the text of Article I, section 10 plainly authorizes the States to act “without the Consent of Congress.” The Virginia ratification debates further confirm this understanding. See Yoo, supra, at 1660 (“As Virginia was the critical state in the process of ratification, this evidence directly reflects the original understanding of war powers held by those who ratified the Constitution. These exchanges serve as the most authoritative historical source for interpreting the war power.”). James Madison observed that the federal government’s responsibility to protect the States under Article IV, section 4 is entirely consistent with the State’s own authority to defend itself under Article I, section 10: The safety of the Union and particular states requires that the general government should have power to repel foreign invasions. The 4th section of the 4th article is perfectly consistent with the exercise of the power by the states. 3 Debates in the Several State Conventions on the Adoption of the Federal Constitution424–25 (Jonathan Elliot ed., 1836). After discussing Article IV, section 4, Madison then turned to Article I, section 10: The other clause runs in these words: “No state shall, without the consent of Congress, lay any duty on tonnage, keep troops or ships of war in time of peace, enter into any agreement or compact with another state, or with a foreign power, or engage in war, unless actually invaded, or in such imminent danger as will not admit of delay.” They are restrained from making war, unless invaded, or in imminent danger. When in such danger, they are not restrained. I can perceive no competition in these clauses. They cannot be said to be repugnant to a concurrence of the power. Id. at 425 (second emphasis added). Similarly, John Marshall observed that “Congress may call forth the militia, — as to suppress insurrections and repel invasions,” but that does not deprive States of the power to defend themselves. Id. at 419. To Marshall, it was “unquestionable” that “the state governments can call forth the militia, in case the Constitution should be adopted, in the same manner as they could have done before its adoption.” Id. at 419. He continued: Gentlemen have said that the states cannot defend themselves without an application to Congress, because Congress can interpose! Does not every man feel a refutation of the argument in his own breast? . . . All the restraints intended to be laid on the state governments (besides where an exclusive power is expressly given to Congress) are contained in the 10th section of the 1st article. This power is not included in the restrictions in that section. But what excludes every possibility of doubt, is the last part of it—that “no state shall engage in war, unless actually invaded, or in such imminent danger as will not admit of delay.” When invaded, they can engage in war, as also when in imminent danger. This clearly proves that the states can use the militia when they find it necessary. Id. at 419–20 (emphasis added). In sum, there’s no basis for treating States differently from the President in their ability to respond to an invasion. Of course, as a practical matter, the federal government possesses superior military force that can overwhelm a State if the federal government chooses to use it. The federal government can also exercise extensive control over a State’s military forces. For example, the federal government can federalize a State’s militia. See U.S. Const. art. I, § 8, cl. 15 (“The Congress shall have Power . . . [t]o provide for calling forth the Militia to execute the Laws of the Union, suppress Insurrections and repel Invasions.”). See also, e.g., Perpich v. Dep’t of Def., 496 U.S. 334, 336, 354 (1990) (holding that Congress can authorize the President to require the National Guard to travel abroad for training during peacetime without a Governor’s permission). But as a legal matter, Article I, section 10 does not require a State to obtain the consent of Congress to defend itself against an invasion. D. It’s well established that military force may be used against hostile foreign individuals as well as foreign sovereigns. Consider the federal government’s evolving response to terrorism. Initially, the United States treated terrorism as a matter for our criminal justice system. See 9/11 Commission Report, supra, at 73 (“Legal processes weretheprimary method for responding to . . . early manifestations of a new type of terrorism.”); Robert Chesney & Jack Goldsmith, Terrorism and the Convergence of Criminal and Military Detention Models, 60 Stan. L. Rev. 1079, 1094 (2008) (“Prior to the 1990s, terrorism was addressed primarily through the lens of criminal law.”). That began to change as early as the1980s. See Chesney & Goldsmith, supra, at 1094 (“As early as 1983, . . . awareness that states were using nonstate actors as proxies to inflict significant harm on U.S. personnel overseas led some in the U.S. government to question a pure criminal law model and to endorse military modes of response to terrorism as an alternative.”). President Reagan, for example, described terrorist attacks as “acts of war.”7 So did his Secretary of State, George Shultz.8 Later, the Clinton Administration determined that it could subject members of al Qaeda to the law of armed conflict and thereby capture or kill Osama bin Laden. 9/11 Commission Report, supra, at 132. See also Chesney & Goldsmith, supra, at 1095 (noting that the Justice Department during the Clinton Administration determined “that al Qaeda’s actions constituted aggression that in turn triggered the right of the United States to use armed force in self-defense, up to and including the use of lethal force to kill bin Laden”). But not everyone has agreed with adopting a military approach to combating terrorism. See 9/11 Commission Report, supra, at 94–95 (“Secretary of Defense Caspar Weinberger opposed [Secretary of State] Shultz, who made little headway against Weinberger, or even within his own department.”). See also Ronald J. Sievert, Meeting the Twenty-First Century Terrorist Threat Within the Scope of Twentieth Century Constitutional Law, 37 Hous. L. Rev.1421, 1428 (2000) (“Scholars and political leaders alike are increasingly recognizing that terrorists who indiscriminately kill civilians are criminals and should be treated as such, regardless of their motivation.”); Daniel M. Filler, Values We Can Afford—Protecting Constitutional Rights in an Age of Terrorism: A Response to Crona and Richardson, 21 Okla. City. U. L. Rev.409, 420 (1996) (“[A] war on terrorism is not the only war in town. . . . [W]e must use care to limit the term ‘war’ to those situations that history and experience suggest is, in fact, war. Because we have a domestic problem, does not mean we have a war.”). Opposition to using military force in response to terrorism has continued even after September 11. See, e.g., Bruce Ackerman, This Is Not a War, 113 Yale L.J. 1871, 1873 (2004) (“This is not a war, but a state of emergency.”); Rosa Ehrenreich Brooks, War Everywhere: Rights, National Security Law, and the Law of Armed Conflict in the Age of Terror, 153 U. Pa. L. Rev. 675, 716 (2004) (“In the wake of September 11, many commentators, especially in the human rights law community, insisted that the phrase ‘the war on terrorism’ should be construed only metaphorically, arguing that the September 11 attacks were not part of an ‘armed conflict’ but were simply a crime, albeit a crime of colossal magnitude.”); id. at 716– 17 (“Although al Qaeda is internationalized, it is not a state. . . . Thus, the argument was that al Qaeda was best analogized to global organized crime networks, such as networks of weapons or drug traffickers.”); Mark A. Drumbl, Victimhood in our Neighborhood: Terrorist Crime, Taliban Guilt, and the Asymmetries of the International Legal Order, 81 N.C. L. Rev.1, 5 (2002) (“[A]lthough the attacks may be at one and the same time both armed and criminal, this Article argues that there are consequentialist, communitarian, and deontological reasons why the attacks should be constructed as criminal attacks to which legal responses from the purview of the criminal law are appropriate.”); Mary Ellen O’Connell, When Is a War Not a War? The Myth of the Global War on Terror, 12 ILSA J. Int’l & Comp. L.535, 535 (2006) (“The President’s ‘war on terror’ does not meet the legal definition of war.”). But although it remains controversial, it’s nevertheless established that individual acts of terrorism may be regarded as acts of war. See Hamdi, 542 U.S. at 518, 520 (plurality opinion) (noting that the war on terror is “unconventional,” and that its “national security underpinnings . . . are broad and malleable,” id. at 520, but the detention of certain individuals “is so fundamental and accepted an incident to war as to be an exercise of the ‘necessary and appropriate force’ Congress has authorized the President to use,” id. at 518); Bradley & Goldsmith, supra, at 2066 (“[T]here are indeed differences between this conflict and more traditional interstate conflicts. . . . [W]e do not believe that they affect the conclusion that Congress has authorized the President to fully prosecute a war against the entities covered by the AUMF.”); id. at 2070 (“When . . . both political branches have treated a conflict as a ‘war,’ and that characterization is plausible, there is no basis for the courts to second-guess that determination based on some metaphysical conception of the true meaning of war.”). E. The present border crisis similarly presents significant national security threats. To begin with, “migration can be weaponized by one sovereign to inflict damage on another.” Abbott, 92 F.4th at 579 (Ho, J., dissenting) (collecting sources). In addition, as the FBI Director recently testified, “we are seeing a wide array of very dangerous threats that emanate from the border.” Worldwide Threats: Hearing Before the S. Select Comm. on Intelligence, 118th Cong. (Mar. 11, 2024) (statement of Christopher A. Wray, Director, FBI). Drug cartels, for example, “pose a hybrid threat” beyond ordinary criminal activity that “combin[es] characteristics of organized crime, insurgency, and terrorism.” Christopher J. Curran, Spillover: Evolving Threats and Converging Legal Authorities in the Fight Against Mexican Drug Cartels, 6 Harv. Nat’l Sec. J. 344, 364 (2015). Cartels “[e]mploy[] cadres of military-trained personnel equipped with sophisticated weaponry,” and they “can carry out exceptionally complex operations and apply a degree of force capable of overwhelming the response capacity of civil law enforcement agencies on either side of the border.” Id. at 347–48. Governor Abbott ordered the installation of the buoy barrier in response to various security risks associated with the border. For example, “over 428 million lethal doses of fentanyl” have been seized as part of Governor Abbott’s Operation Lone Star initiative.9 The Governor also cites Congressional testimony that human smuggling networks participate in “other transnational crimes” including “gang activity, identity benefit fraud, money laundering, bulk cash smuggling, narcotics smuggling, arms trafficking, and terrorism and [n]ational security-related crime.”10 Installing the buoy barrier is also consistent with Governor Abbott’s executive order “designating the Mexican drug cartels as foreign terrorist organizations.”11 Moreover, this is not the first time that a Texas Governor has invoked Article I, section 10 in response to a border crisis. In the decades leading up to the 1870s, bands of raiding criminals routinely crossed the Mexican border into Texas, killing residents and taking stolen cattle back across the Rio Grande with them. See Texas Frontier Troubles, supra, at iii–vii.12 At one point, a joint committee of the Texas Legislature “reported 105 murders and a 90% decrease in stock in the region below Eagle Pass.” Rippy, supra, at 292. A special Congressional committee report found the Mexican government “utterly powerless to prevent these evils or to check them.” Texas Frontier Troubles, supra, at xiii. Nor was the United States government willing to take action to assist Texas with its border challenges. Id. See also id. at vii (“This state of things is wholly due to the inactivity of our Government. A few years ago some energetic chastisement, convincing the plunderers that there was some danger connected with their trade, would certainly have checked it, but the impression that our troops dare not cross the river has made them feel at ease.”); Rippy, supra, at 288 (“[N]either government made any serious attempt to apply a remedy.”). The murders and robberies sharply increased in 1874. Rippy, supra, at 290. Governor Cokerepeatedly sought relief from thefederal government, but was refused. Texas Frontier Troubles, supra, at xiii. So the Governor took matters into his own hands and ordered statetroops to protect the frontier. Id. at xv, 165. He authorized the troops, when “in close pursuit of thieves or marauders with their plunder,” to cross the Rio Grande into Mexico if necessary to recover stolen property. Id. at xv. The U.S. Attorney General informed Governor Coke that his actions violated federal law. Id. at xv, 164 (citing Act of April 20, 1818, § 6, 3 Stat. 449). In response, Governor Coke invoked Article I, section 10, and referred to the raids as an invasion: Texas, when forced to assume the unjust burden of defending herself against foreign aggression, and of repelling invasion of her territory, as sheis nowattempting to do, is fully authorized, under article [I], section 10, Constitution of the United States, to use the war powers which ordinarily reside in the United States Government. Id. at 166. Governor Coke defended his right to send troops onto Mexican soil by analogizing Texas’s war powers to the federal government’s war powers: If the forces of the United States have a right to cross the national boundary and continue pursuit of marauders on Mexican soil, of which there can be no doubt, Texas forces, which are doing the duty which ought to be performed by the United States troops, and are doing it because United States troops are not there to do it, and it must be done, have the same right. Id. Governor Coke concluded by observing that, “[w]hile I have clear convictions of my right as governor of Texas, under the Constitution and laws of the United States, and in view of the condition of affairs on the Rio Grande border, to issue the military order of which complaint is made, and have it executed,” he also knew “that if the officers of the United States Government entertain a different view, they have the power to prevent its enforcement, and that no good will result from further effort on my part to execute it.” Id. at 167. After all, the United States possesses stronger military resources, not to mention the constitutional authority to call the State military into federal service. See, e.g., Robert Leider, Federalism and the Military Power of the United States, 73 Vand. L. Rev. 989, 1069 (2020) (“The consolidation of military power in the federal government has largely prevented” states from “[o]bstructing federal authority.”). The Attorney General acquiesced to the Governor’s claim of authority. Id. at xvi. F. To be sure, a state of invasion under Article I, section 10 does not exist just because a State official has uttered a certain magic word. Texas readily acknowledges that a Governor’s declaration of an invasion under Article I, section 10—and any measures he adopts in response—must be done in good faith. See Baker, 369 U.S. at 214 (“[C]learly definable criteria for decision may be available. In such case the political question barrier falls away: A Court is not at liberty to shut its eyes to an obvious mistake.”) (cleaned up); Sterling v. Constantin, 287 U.S. 378, 399–400 (1932) (“The nature of the power [to suppress insurrection and disorder] also necessarily implies that there is a permitted range of honest judgment as to the measures to be taken in meeting force with force.”). No one here has suggested that Governor Abbott is operating in bad faith. The United States has certainly made no such argument. To the contrary, the President has acknowledged that the Nation is facing a “border crisis,”and called for robust government action to “shut down theborder.”13 The U.S. House of Representatives has likewise recognized illegal immigration as a national security crisis, and a similar resolution is pending in the U.S. Senate. Governor Abbott’s declaration of invasion is also endorsed by a majority of the Nation’s governors. See Abbott, 92 F.4th at 578 (Ho, J., dissenting) (collecting authorities). The panel majority described Texas’s invocation of Article I, section 10 as a “plausible defense.” United States v. Abbott, 87 F.4th 616, 631 (5th Cir. 2023) (quotations omitted), vacated on reh’g en banc, 90 F.4th 870 (5th Cir. 2024). And the district court recently allowed Texas’s invasion defense to proceed at trial, noting that it “would be an incredible stretch” to conclude that “there is no possible way for Texas to succeed on its affirmative defense or that its affirmative defense raises no question o[f] fact or law.” Order, United States v. Abbott, No. 1:23-cv-00853, at 11 (W.D. Tex. July 24, 2024). G. The United States contends that, regardless of the political question doctrine, this court should still require Texas to comply with the Rivers and Harbors Appropriation Act of 1899. According to the federal government, ArticleI, section10 “does not purport to excuse States from compliance with all other federal laws, especially on an ongoing basis.” To begin with, however, federal statutes ordinarily must give way to federalconstitutionalrights. If there’s aprincipled reason whyCongressmay enact statutes that violate rights under Article I, but not under the First Amendment, the United States has not offered one. And for good reason. Article I, section 10 makes clear that States may engage in war in response to actual invasion—and that they may do so “without the Consent of Congress.” It would surely violate that provision for Congress to pass a law requiring its approval before States may so act. To be sure, Presidents do not enjoy unlimited war power, and there’s no reason to believe the States should, either. See, e.g., Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 587 (1952) (holding that President Truman’s order to seize most of America’s steel mills “cannot properly be sustained as an exercise of the President’s military power as Commander in Chief”); id. at 635–38 (Jackson, J., concurring) (establishing a framework for when the President acts (1) with Congress’s authorization; (2) with Congress’s silence; or (3) with Congress’s disapproval). But military tactics directed toward hostile foreign enemies fall well within the political question doctrine. See, e.g., Prize Cases, 67 U.S. at 670 (“[The President] must determine what degree of force the crisis demands.”) (quotations omitted); United States v. Curtiss-Wright Export Corp., 299 U.S. 304, 320 (1936) (noting that, in international affairs, Presidents must enjoy “a degree of discretion and freedom from statutory restriction which would not be admissible were domestic affairs alone involved”); Youngstown, 343 U.S. at 644–45 (Jackson, J., concurring) (noting that “the Constitution’s policy [is] that Congress, not the Executive, should control utilization of the war power as an instrument of domestic policy,” but that “I should indulge the widest latitude of interpretation to sustain [the President's] exclusive function to command the instruments of national force, at least when turned against the outside world for the security of our society”); El-Shifa, 607 F.3d at 844 (“If the political question doctrine means anything in the arena of national security and foreign relations, it means the courts cannot assess the merits of the President’s decision to launch an attack on a foreign target.”). This case presents a strategic military decision directed toward a foreign enemy within the State’s authority under Article I, section 10.14 The State has placed a buoy barrier along an international border river, for the express purpose of defending against what Texas views as an invasion by foreign actors. The installation is specifically targeted toward what Texas views as a particular area of concern—a stretch along the Rio Grande near Eagle Pass which Texas asserts is “mostly home to illegal activities, like smuggling drugs and weapons and human trafficking.” So as the district court now recognizes, Texas can certainly prevail on its invasion defense, regardless of how anyone may interpret the RHA. As the district court recently ordered, “Texas could, as a matter of law, succeed in either asserting th[at] invasion is a non-justiciable political question or by presenting evidence that the immigration at the border does amount to the military invasion contemplated by the Founders, making the Court’s prior order to the contrary obsolete.” Order, United States v. Abbott, No. 1:23-cv-00853, at 12 (W.D. Tex. July 24, 2024).15 It’s far from clear that there is in fact any actual conflict between Texas’s actions and federal law. The RHA is a generally applicable statute— not a law that specifically governs how States may respond to an invasion. Courts hesitate to read generally applicable federal laws to intrude on important aspects of state sovereignty. See, e.g., Bond v. United States, 572 U.S. 844, 860 (2014) (“‘Chemical weapon’ is the key term that defines the statute’s reach, and it is defined extremely broadly. But that general definition does not constitute a clear statement that Congress meant the statute to reach local criminal conduct.”); Gregory v. Ashcroft, 501 U.S. 452, 460 (1991) (noting that federal courts must be “certain of Congress’ intent beforefinding that federal lawoverrides” States’ ability to make”decision[s] of the most fundamental sort for a sovereign entity”) (quotations omitted). But in any event, courts routinely apply the political question doctrine to avoid deciding claims that involve generally applicable legal duties. See Spectrum Stores, Inc. v. Citgo Petrol. Corp., 632 F.3d 938, 951 (5th Cir. 2011) (adjudicating the plaintiffs’ antitrust claims alleging a “price-fixing conspiracy involving OPEC member nations” “would require that we review the considered foreign policy of the political branches”); El-Shifa, 607 F.3d at 843 (“[A] statute providing for judicial review does not override Article III’s requirement that federal courts refrain from deciding political questions.”); Def. for Children Int’l-Palestine v. Biden, _ F.4th _, _, 2024 WL 3405631, *4 (9th Cir. 2024) (“Many, if not most, grievances can be styled as the violation of an asserted legal obligation. . . . [T]here is no valid support for the idea that merely alleging the violation of a claimed legal duty means that the political question doctrine does not apply.”). Likewise, determining whether a statute can be constitutionally applied in a given circumstance can also constitute a nonjusticiable political question. See Zivotofsky, 566 U.S. at 208 (Sotomayor, J., concurring in part and concurring in the judgment) (“It is not impossible to imagine a case involving the application or even the constitutionality of an enactment that would present a nonjusticiable issue.”); id. at 211 (Alito, J., concurring in the judgment) (“Under our case law, determining the constitutionality of an Act of Congress may present a political question.”). The United States nevertheless contends that we should ignore Texas’s invasion argument because it constitutes a defense to the federal government’s “otherwise meritorious RHA claim.” But as the United States admitted during oralargument,it would readily taketheexactopposite view, and invoke the political question doctrine, if a plaintiff brought a similar claim against the President. Oral Argument at 38:05–38:55. Notably, Texas does not claim a licenseto violateindividual rights just because it has invoked its authority under Article I, section 10. See, e.g., Sterling, 287 U.S. at 402 (holding that the Texas Governor’s determination of an insurrection did not justify his “attempt to regulate by executive order the lawful use of complainants’ properties in the production of oil”). Cf. Hamdi, 542 U.S. at 536 (plurality opinion) (“[A] state of war is not a blank check for the President when it comes to the rights of the Nation’s citizens.”). And no such claim of individual rights is presented here. * * * Texas’s invocation of Article I, section 10 presents a nonjusticiable political question. Accordingly, the district court lacks jurisdiction to consider the federal government’s claim under the RHA. II. Thepoliticalquestiondoctrineis jurisdictional,so weshould notallow the district court to proceed to trial in any case where the doctrine precludes jurisdiction. See Rucho, 588 U.S. at 695 (“[W]e are asked to decide an important question of constitutional law. But before we do so, we must find that the question is presented in a ‘case’ or ‘controversy’ that is, in James Madison’s words, ‘of a Judiciary Nature.’”) (quotations omitted); Spectrum Stores, 632 F.3d at 943 (“Because the political question doctrine is jurisdictional, we address it first.”). In Lane v. Halliburton, 529 F.3d 548 (5th Cir. 2008), our court concluded that, if a case involves a political question, we can adjudicate that case only if the political question can somehow be separated from the rest of the case. See id. at 557 (evaluating whether “resolving the Plaintiffs’ . . . legal claims invariably require[s]” analyzing a political question). This means determining whether plaintiffs “ can prove any plausible set of facts that would permit recovery . . . without compelling the court to answer a nonjusticiable political question.” Id. So the analysis turns on the nature of the plaintiff’s legal theory—that is, whether the plaintiff can prevail without requiring the court to answer a political question—and not on whether the defendant can prevail on alternative grounds. See, e.g., id. at 567 (“Under [the Plaintiffs'] theory of causation, the district court may be able to resolve the Plaintiffs’ fraud and negligence claims under Texas tort law without second-guessing the acts and decisions of the Army.”). Determining whether a political question is extricable from the rest of a case “requires us to understand just what the Plaintiffs must prove to prevail.” Id. at 561. See also Cooper v. Tokyo Elec. Power Co., 860 F.3d 1193, 1214 (9th Cir. 2017) (“Deciding whether a political question is inextricable from a case necessarily requires us to know what the plaintiff must prove in order to succeed.”). This includes evaluating a defendant’s potential defenses, not just the face of the complaint. See Lane, 529 F.3d at 565 (“We must look beyond the complaint, considering how the Plaintiffs might prove their claims and how KBR would defend.”). See also Cooper, 860 F.3d at 1212 (“Because the political question doctrine is jurisdictional in nature, we must evaluate . . . potential defenses and facts beyond those pleaded in the complaint to determine whether the case is justiciable.”) (quotations omitted); Harris v. Kellogg Brown & Root Servs., Inc., 724 F.3d 458, 465–66 (3rd Cir. 2013) (“[M]ilitary decisions that are textually committed to the executive sometimes lie just beneath the surface of the case. . . . In these situations, the political question appears not from the plaintiff’s claims but from the broader context made relevant by a contractor’s defenses.”). There’s no way for the United States to prevail in this case without defeating the State’s invocation of Article I, section 10. Indeed, the district court has recently said as much. See Order, United States v. Abbott, No. 1:23-cv-00853, at 12 (W.D. Tex. July 24, 2024). At trial, the district court will consider both whether invasion is a nonjusticiable political question and whether Texas has shown that an invasion exists here. Id. That is precisely the type of judicial factfinding and second-guessing that the political question doctrine requires courts to avoid. * * * Iagreethatthepreliminary injunctionshould bereversed. But I would also instruct the district court to dismiss this case for lack of jurisdiction. Accordingly, I concur in the judgment in part and dissent in part. APPENDIX EXECUTIVE OFFICE, STATE OF TEXAS, Austin, August 6, 1874. SIR: Your communication of July 23 reached me in due time, and its contents have been duly considered. You call my attention to an order issued by me, as governor of Texas, to Capt. Refugio Benavides, commanding a minute-company on the Rio Grande border, requiring him, when in close pursuit of Indians, marauders, or cattle-thieves, for the purpose of recovering property taken by them from citizens of Texas, if necessary, to cross the Rio Grande River, and, in connection therewith, refer me to the act of Congress of April 20, 1818, prescribing penalties for breach of the neutrality laws of the United States, and desire to hear from me on the subject, saying that the matter had been referred to you by Hon. Hamilton Fish, Secretary of State. I have to say, in reply, that, in order to determine the propriety of the order to Captain Benavides, it is necessary to understand the circumstances under which it was given, and the condition and necessities of the people for whose protection it was issued. The Rio Grande River, the boundary-line between the United States and the Republic of Mexico, is a narrow, shallow stream, and from Brownsville, thirty miles from its mouth, up, fordable except for a few weeks throughout the year. The country on the Texas side of the river is mostly covered with a dense thicket or chaparral growth. From the mouth of this river up to Fort Duncan, a distance of about four hundred miles, where the depredations hereinafter spoken of are chiefly committed, there are four stations or posts of United States troops, the four having in the aggregate eleven companies of infantry and seven of cavalry. Fort Clark, with eight companies of cavalry and three of infantry, still higher up the river, covers about sixty miles of the river, besides a long line of frontier not on the river. This constitutes the defense provided by the United States Government for the Rio Grande border. Of this force only the cavalry can be used with any effect in repelling and punishing theIndians and Mexican thieves and robbers who depredate on the country. The infantry are useless for that pupose [sic]. For twenty-five years this border has been harassed and depredated on by lawless bands from Mexico of Indians and Mexicans. In 1857 Governor Houston had to send a military force to the Rio Grande to repel an invasion headed by the bandit Cortina. During this time many millions of dollars’ worth of property belonging to citizens of Texas has been destroyed, stolen, and taken with the strong hand by these marauders. Especially since 1865 has it been the case that armed bodies of men from the west side of the Rio Grande are constantly coming into Texas, overawing the people and driving off into Mexico large herds of cattle, the property of citizens of Texas. At other times these parties shoot down and skin large numbers of cattle and carry off the hides, leaving the carcasses to rot on the ground. They not unfrequently, in executing their plans of wholesale robbery, butcher whole families, men, women, and children, and fire their houses. Losses by the people of Texas from this source amount to many hundreds of thousands of dollars annually, and, since 1865, to say nothing of losses previously, to many millions. Life and property between the Nueces and Rio Grande Rivers has from this cause been made so insecure as to threaten its depopulation and an utter destruction of its only producing interest—cattle and horses. The country is sparsely settled, and by the time a few of the citizens can get together for defense and pursuit, these robbers are at or across the Rio Grande with their booty, and, having reached this place of refuge, they taunt and defy the citizens of Texas, whom they have plundered, and leisurely divide the spoils. Within the last six months the invasions of these bandits from Mexico have become so bold and frequent, and their robberies and murders of citizens of Texas so extensive and alarming, that I have been compelled to call out, at a heavy expense to the State, a military force for their protection. On account of the chaparral thickets and brushy character of the country on the Rio Grande, and the great extent of it, the small force the State is able to keep in the field is most effective in patrolling the country immediately on the river, and watching the crossings, where some delay occurs in driving over the stolen stock; and it frequently happens that the robbers are not discovered until they have partially or wholly crossed the river with their plunder. These depredations upon persons and property in Texas by Mexicans, I state, as an undeniable fact, are being committed, and for years past have been committed, with the knowledge and acquiescence of the local municipal authorities on the west side of the Rio Grande; and it is believed to be susceptible of abundant proof that in numerous instances these local authorities have colluded with the marauders, and shared in a division of the booty. Certain it is that in open day-light, without concealment, Mexicans, in full view of the municipal officers on the west side of the river, constantly come across the river into Texas, and return publicly with large herds of cattle and other property belonging to citizens of Texas, and have for years been doing it, without any effort by the authorities for its repression, when it was publicly and notoriously known that they were robbers, and had been plundering the people of Texas. It is equally certain that the central government of Mexico has been, and is now, fully informed of the depredations committed by citizens under its jurisdiction, and having homes and taking refuge from pursuit within its territory, upon the lives and property of citizens of Texas, and that that government, notwithstanding the enormity of the crimes, and the constancy, persistency, and publicity with which, for a number of years, they have been committed, has, up to this time, failed to take any steps to put an end to them, or to punish the criminals, or make reparation to the injured parties in Texas. The force placed by the United States Government on the Rio Grande border is shown by the present condition of that country, the frequency and impunity with which bandit raids are made upon it, to be utterly inadequate for its protection. The facts herein recited are of public and common notoriety, and have been brought to the attention of the Government at Washington time and again by the public press, by the Representatives in Congress from Texas, and by the State authorities of Texas, and can be proved to be true with any amount of testimony. Under these circumstances, and basing my action on these facts, as governor of Texas, in obedience to an imperious necessity, brought about by no dereliction of duty on the part of Texas, or her people, to protect citizens of Texas, as far as possible, from a predatory war being waged on them by foreign desperadoes and robbers, I called into the State service one hundred men, and have posted them in the country between the Rio Grande and Nueces Rovers, and issued to the officers commanding the orders of which complaint is made. The portion of the order to which objection is made is in these words: * * * * “Should the company be in close pursuit of thieves or marauders, with their plunder, it will follow as far as possible, whether on this side of the Rio Grande or the other, having a due regard for its own safety, and the prospect of recovering the stolen property.” * * This order contemplates no “military expedition or enterprise to the carried on” against the territory of Mexico or the people of that country. It simply looks to the employment of thesmall forcetheStateof Texas has been compelled to call out by an inexorable necessity for the protection of her otherwise defenseless people, in the only mode in which it can be used effectively. If these Mexican raids, which this force is called out to repel, were of recent date, of irregular occurrence, and of such character as to elude theeffortsof an ordinarily vigilant and energeticgovernment to suppress, and such efforts were being made in good faith by the authorities of Mexico, I grant in that case, that to allow an armed force to pursue even robbers, for the purpose of recovering their booty, across the Rio Grande, would be a violation of the rights of Mexico, and of well-settled principles of international law. The right of the government of Mexico to immunity for its territory from the incursions of armed forces of Texas, would then be based ona proper dischargeof theduties of that government to Texas, in repressing the lawlessness of its own people, and preventing and punishing their crimes attempted and committed against the people of their neighboring State. It is because each state or nation has undertaken to restrain its people from making war on the people of its neighbors, that the law of nations forbids an armed force from one entering the territory of another. The right of immunity grows out of and depends upon the performance of this duty, which each power owes to the other. No state has surrendered the right of defense of its people in its own way against aggressions from neighboring states or people except upon the promise and performance of the great duty toward itself, which all nations owe each other, of so governing their people as that they shall not depredate or make war upon other nations, or any of their people or territory. Iapprehend thatinternationalcourtesy, comity, and amity has never been required by the law of nations, carried to the romantic extent of surrendering the great natural right of self-defense against the constant inflictionof serious, permanent and wrongfulinjury uponthepeople of one nation by those of another, although the attacks may be unauthorized by the government of the territory from which it comes. The State government of Texas has to deal with the admitted and undeniable fact, that for a series of years a most destructive predatory war has been carried on against the people of Texas and their property, between the Nueces and Rio Grande Rivers, by Indians and Mexicans residing on the west side of the Rio Grande River, and belonging to the jurisdiction of the Mexican government. The fact exists that these people, whether from want of power or of will on the part of that government makes no difference as to results and as to our rights, are not ordered and regulated by the Mexican government, and restrained to a proper line of conduct toward the people of Texas, and that it has become an imperative necessity on Texas, in consequence of this failure of duty on the part of the Mexican government and the inadequacy of the force posted on the Rio Grande by the United States, to provide for the defense of her citizens, and being thus driven to exercise her inherent right of self-defense, it is insisted, if necessary, that she has a clear legal right to send her troops on Mexican soil for the purpose. Only friendly powers havetheright to claim exemption of their territory from armed intrusion; and it is insisted that as to Texas, Mexico is not a friendly power, because Mexican citizens, with the knowledge of the authorities of their government, and unrestrained by them, are making war on the people of Texas and their property, rendering it necessary for their proper defense that Texas troops should pursue the freebooters on Mexican soil. This necessity, if report which has gone the rounds of the newspaper press of the United States uncontradicted, and which is universally believed, be true, has been twice in the last twelve months recognized and acted on by a gallant and able officer of the United States Army. I allude to General McKenzie, who, with troops of his command, pursued bands of marauders from Texas across the national boundary, and on one of these occasions is believed to have inflicted on them merited chastisement on Mexican soil. No word of disapproval has ever been heard here from the Government at Washington of the conduct of this distinguished officer, while the press and the people of the country have loudly applauded it. Texas, when forced to assume the unjust burden of defending herself against foreign aggression, and of repelling invasion of her territory, as she is now attempting to do, is fully authorized, under article [I], section 10, Constitution of the United States, to use the war powers which ordinarily reside in the United States Government; and the constitutional obligation resting on the United States to defend the people of Texas against hostile invasion not having been discharged, would, independent of that provision of the Constitution, have vested in Texas the right to resort to any means for her own defense which might properly have been resorted to by the United States. If the forces of the United States have a right to cross the national boundary and continue pursuit of marauders on Mexican soil, of which there can be no doubt, Texas forces, which are doing the duty which ought to be performed by the United States troops, and are doing it because United States troops are not there to do it, and it must be done, have the same right. It will be observed that the order to Captain Benavides authorizes him to cross theRio Grandeonly when in closepursuit, and for onespecificpurpose, and that is to recapture property stolen or wrested by force from citizens of Texas. He is not authorized to cross the river for purposes of retaliation, nor to make war on the territory or any of the people of Mexico, but only to pursue marauders going out of Texas, and take from them and bring back property found in their possession belonging in Texas. A knowledge by these freebooters that they will be pursued, and that the west bank of the Rio Grande shall no longer be a sanctuary and place of refuge to them, from which they can sally at pleasure, and murder and plunder the people of Texas, will do more to put an end to their operations, and give peace and security to the people of that frontier, than quadruple the present force without authority to cross the Rio Grande. The claims of citizens of Texas for indemnity for property of which they have been plundered by Mexican citizens, already amounting to many millions of dollars, has for some years past been receiving the attention of the Congress of the United States; but no steps have been taken to remove the cause of accrualof these claims, and to-day theStategovernment of Texas is incurring an onerous and most oppressive and unjust expense, which she cannot avoid because necessary for the defense of her people from murder and rapine, which should fall of right upon the Government of the United States. It is to be hoped that the necessities of the people on the Rio Grande frontier will be recognized and appreciated, and cared for by the Government of the United States, and that Texas, as of right she ought to be, will be relieved of the burden now resting on her so unequally, of providing for the defense of a national boundary; but while she is thus taxed, that she will be permitted to use the means of defense she is compelled to employ in the mode she deems most effective. While I have clear convictions of my right as governor of Texas, under the Constitution and laws of the United States, and in view of the condition of affairs on the Rio Grande border, to issue the miliary order of which complaint is made, and have it executed, I at the same time am fully aware that if the officers of the United States Government entertain a different view, they have the power to prevent its enforcement, and that no good will result from further effort on my part to execute it. I therefore have given you the facts upon which the issuance of the order was predicated, by which its propriety may be determined, and if it is decided to be in contravention of the laws of the United States, when notified of the decision, I will revoke the order, but must say that it will be to the last degree unjust to Texas, on the part of the General Government, to refuse to her permission properly to defend herself, unless at the same time adequate provision is made for her defense. Very respectfully, your obedient servant, RICH’D COKE, Governor of Texas. Hon. GEORGE H. WILLIAMS, Attorney-General United States, Washington, D.C. Stephen A. Higginson, Circuit Judge, joined by King, Stewart, Southwick, Graves, Douglas, and Ramirez, Circuit Judges, dissenting: As Judge Douglas highlights in her comprehensive dissent, the majority improperly rejects the factual evidence supporting historical navigability, finding clear error where there is none. I agree with her in full, and write separately only to underscore the legal error the majority commits in announcing its new requirement for river navigability. In so doing, the majority disregards the Supreme Court’s broad understanding of commerce—and the role that ferries have historically played in it—by reading into caselaw a directionality requirement for navigability that not only does not exist, but also lacks a theoretical foothold. The result is a new legaltest thatis ahistorical,unworkable, and contrary to thefederal obligation to guarantee that the waters of the United States remain obstruction-free. First, the majority’s decision runs contrary to the Supreme Court’s expansive understanding of commerce. Tracing back to the founding of this nation, the Supreme Court has declared that the federal government must guarantee that commerce on the water, between states and with foreign nations, remains unimpeded. For example, in Gibbons v. Ogden, the Court explained that Congress’s power to regulate commerce extends to “every species of commercial intercourse between the United States and foreign nations” as well as “among the several States,” which “cannot stop at the external boundary line of each State.” 22 U.S. 1, 193 (1824).1 Crucially, theCourt’s broad understanding of commerceundoubtedly encompasses ferry traffic and transportation. In deeming the “receiving and landing of passengers and freight” as “commerce,” the Supreme Court was clear: It matters not that the transportation is made in ferry-boats which pass between the states every hour of the day. The means of transportation of persons and freight between the states does not change the character of the business as one of commerce . . . . Commerce among the states consists of intercourse and traffic between their citizens, and includes the transportation of persons and property, and the navigation of publicwaters for thatpurpose, as wellas thepurchase, sale, and exchange of commodities . . . . And it needs no argument to show that the commerce with foreign nations and between the states, which consists in the transportation of persons and property between them, is a subject of national character, and requires uniformity of regulation. Congress alone, therefore, can deal with such transportation . . . . Otherwise, there would be no protection against conflicting regulations of different states . . . . Gloucester Ferry Co. v. Pennsylvania, 114 U.S. 196, 203–04 (1885). In Gloucester Ferry, the Supreme Court comprehensively evaluated the scope of national commerce—and addressed, in no uncertain terms, the role that ferries played within commerce: Ferries between one of the states and a foreign country cannot be deemed, therefore, beyond thecontrol of congress under the commercial power. . . . [A]nd if they are not beyond the control of the commercial power of congress, neither are ferries over waters separating states. . . . [T]he fact remains that such a ferry is a means, and a necessary means, of commercial intercoursebetween the states bordering on their dividing waters, and it must, therefore, be conducted without the imposition by the states of taxes or other burdens upon the commerce between them. Id. at 216–17. These explanations demonstrate that ferry traffic constitutes commerce—which, in turn, supplies a basis for navigability: “[R]ivers must be regarded as public navigable rivers in law . . . when they are used, or are susceptible of being used, in their ordinary condition, as highways for commerce, over which trade and travel are or may be conducted in the customary modes of trade and travel on water.” The Daniel Ball, 77 U.S. 557, 563 (1870). As the cases suggest, ferries historically were a “customary mode[]” of both “ trade and travel.” Hence, evidence of historical ferry traffic at Eagle Pass presented in contemporaneous cases, see, e.g., United States v. Weil, 35 Ct. Cl. 42, 77 (1900) (referencing testimony that “[a]t Eagle Pass there were ferryboats in which the cotton was crossed over”) (emphasis added) and Tugwell v. Eagle Pass Ferry Co., 74 Tex. 480, 487 (1888), is evidence of trade and travel on the Rio Grande sufficient to trigger Army Corps statutory authority under Section 10 of the Rivers and Harbors Appropriation Act (“RHA”).2 Overlooking this centuries-old pedigree, the majority isolates the single word “highway” to craft out of whole cloth a new judicially-created rule: That navigability results only from the existence of commerce along the rivers of the United States, not across the rivers of the United States. Maj. Op. at 15–17. Again, this directionality rule runs contrary to both history and precedent. Even assuming arguendo that it does not, there is no actual legal basis for the test the majority seeks to introduce. That “the Supreme Court has looked for trade or travel ‘up the river,’ ‘down’ the river, and ‘along’ it” indicates, merely, that such flow along the river constitutes commerce that would satisfy navigability. Id. at 15. It does not logically follow that flow of commerce “across” a river cannot satisfy navigability—and to be sure, none of the Supreme Court’s cases actually says that. This is why, as Judge Douglas highlights in her dissent, “[t]he lone authority” that the majority cites is one paragraph from an “out-of-circuit district court opinion [that] not only lacks precedential weight, but is also manifestly distinguishable, involving entirely intrastate ferry traffic in Georgia . . . .” Diss. Op., J. Douglas, at 8 (emphasis added). At a minimum, the majority’s new requirement is an insufficient basis upon which to find that the district court abused its discretion in entering the preliminary injunction—a standard of review which the majority consistently overlooks. Id. at 2, 14. Furthermore, the majority’s proffered standard rests on a flawed theory. Themajority offers no explanationfor whythedirectionality inherent to a man-made “highway” on land should constrain the directionality of traffic on water. Due to both traffic laws and custom, going “across” a highway on land seems anathema—those who do so would be in peril from oncoming traffic. But this descriptor, above all an atextual interpretative rule narrowing Congress’s phrase “waters of the United States,” is ill-suited to bodies of water. Whatever superficial appeal exists in re-interpreting the RHA to require a flow of commerce down a “highway” along the length of a river rather than across it rapidly disappears when one considers other bodies of water, such as a bay or a lake. Those too, undoubtedly, can constitute waters of the United States. And the RHA applies not merely to rivers, but to “waters of the United States” more broadly. See 33 U.S.C. § 403 (“The creation of any obstruction not affirmatively authorized by Congress, to the navigable capacity of any of the waters of the United States is prohibited.”); id. (“[I]t shall not be lawful to build or commence the building of any wharf, pier, dolphin, boom, weir, breakwater, bulkhead, jetty, or other structures in any . . . navigable river . . . of the United States . . . except on plans recommended by the Chief of Engineers and authorized by the Secretary of the Army . . . “). Are we to disregard commerce flowing across a lake and insist, instead, upon commerce only around the lake’s perimeter? Or are we to apply different tests for different bodies of water? And even if we are to apply different tests for rivers than other bodies of water, why would cross-border traffic, transportation, boating, and commerce between the United States and Mexico not count as commerce demonstrating navigability? Stated otherwise, the atextual and ahistorical proposition the majority uses to displace the Supreme Court’s centuries-old formulation of navigability threatens to remove Army Corps Section 10 authority from any border river that serves as a state border—which exists in all but four of the United States’ mainland states. Under the majority’s logic, Texas could build a river border wall between itself and Louisiana, and the Army Corps would be powerless to remove that obstruction of cross-river commerce between states. This displacement of authority becomes a diplomatic problem when the border river is also an international border. Necessarily, with an international river separating two countries, any and all commerce between the countries will be across, not along the river. Indeed, the Rio Grande has no fewer than twenty-eight bridges and border crossings.3 Tex. Dep’t of Transp., Texas-Mexico International Bridges and Border Crossings: Freight, International Trade, and Connectivity 9 (2019), https://ftp.dot.state.tx.us/pub/txdot/move-texas-freight/studies/texas-mexico-bridges-crossings-2019.pdf. It is therefore all the more critical that the federal government’s authority not be displaced. The majority’s novel test not only departs from precedent, but also creates a definition of navigability specific to rivers that inexplicably excludes trade and travel between states or foreign countries that crosses rivers. This, in turn, would defeat congressional purpose to prevent obstructions to that commerce—and render the federal government powerless to remove obstructions in rivers that serve as interstate or, like the Rio Grande, international boundaries. Because such a reduction of our definition of the “waters of the United States” is both unsupported and unworkable, I respectfully dissent. Dana M. Douglas, Circuit Judge, joined by King, Stewart, Southwick, Graves, Higginson, and Ramirez, Circuit Judges, dissenting: In July 2023, Texas, at the direction of Governor Greg Abbott, installed a floating barrier in the Rio Grande near Eagle Pass, Texas.1 The United States filed a civil enforcement action against Texas, alleging that installment of the barrier violated the Rivers and Harbors Appropriation Act of 1899 (RHA). The United States moved for a preliminary injunction, which the district court granted, ordering Texas to cease work on the barrier and remove it from the water. The majority concludes that the district court abused its discretion in granting preliminary relief to the United States. It concludes that the United States is unlikely to succeed on the merits because the evidence does not support a finding that the Rio Grande at Eagle Pass is navigable and that the United States had failed to meet its burden to show the remaining preliminary injunction factors favor its position. Despite the majority’s opinion, a review of the record in this case makes clear that the United States is likely to succeed on the merits of its RHA claim. Further, theremaining preliminary injunctionfactors also weigh in favor of the United States. Because I would affirm the district court in full, I respectfully dissent. I Consideration of the standard of review is paramount here. The sole issue is whether the district court abused its discretion in entering a preliminary injunction—requiring Texas to remove its buoy barrier from the waters of the Rio Grande—while it considers the merits of the case. See Janvey v. Alguire, 647 F.3d 585, 591-92 (5th Cir. 2011) (stating that although the standard applied by the district court is “stringent,” the standard of appellate review “is simply whether the issuance of the injunction . . . constituted an abuse of discretion”). The United States has the burden to show, among the usual factors, a likelihood of success on the merits. Winter v. NRDC, 555 U.S. 7, 20 (2008). Importantly, although a preliminary injunction is an “extraordinary remedy,” id. at 24, the United States’ burden at this stage is not nearly as demanding as themajority urges. Thisis because “[a] plaintiffis notrequired to prove its entitlement to summary judgment in order to establish ‘a substantial likelihood of success on the merits’ for preliminary injunction purposes.” Byrum v. Landreth, 566 F.3d 442, 446 (5th Cir. 2009); see also Univ. of Tex. v. Camenisch, 451 U.S. 390, 395 (1981) (“A party thus is not required to prove his case in full at a preliminary-injunction hearing.”); Netflix, Inc. v. Babin, 889 F.4th 1080, 1096 (5th Cir. 2023) (“The standard to show likely success on the merits. . . is obviously lower than that for establishing actual success on the merits during the hearing for a permanent injunction.”) (quoting Smith v. Hightower, 693 F.2d 359, 367 n.19 (5th Cir. 1982)). Findings of fact are reviewed for clear error, while conclusions of law are reviewed de novo. Restaurant Law Center v. U.S. Dep’t of Labor, 66 F.4th 593, 597 (5th Cir. 2023). A key dispute in this case—whether the Rio Grande is a navigable waterway as defined in the RHA—is the district court’s finding of fact, which we review for clear error. United States v. Appalachian Elec. Power Co., 311 U.S. 377, 405 (1940). “Clear error exists when although there may be evidence to support it, the reviewing court on the entire record is left with the definite and firm conviction that a mistake has been committed.” Black v. SettlePou, P.C., 732 F.3d 492, 496 (5th Cir. 2013) (quoting Hollinger v. Home State Mut. Ins. Co., 654 F.3d 564, 569 (5th Cir. 2011)). Clear error is applied to factual findings at thepreliminary injunctionstage for good reason. These determinations are made “on the basis of procedures that are less formal and evidence that is less complete than in a trial on the merits.” Camenisch, 451 U.S. at 395. Faithfully abiding by the standard of review leads to only one conclusion—the district court must be affirmed. II Section 10 of the RHA provides in pertinent part: The creation of any obstruction not affirmatively authorized by Congress, to the navigable capacity of any of the waters of the United States is prohibited; and it shall not be lawful to build or commence the building of any wharf, pier, dolphin, boom, weir, breakwater, bulkhead, jetty, or other structures in any . . . navigable river . . . of the United States . . . except on plans recommended by the Chief of Engineers and authorized by the Secretary of the Army . . . . 33 U.S.C. § 403. The first clause prohibits the construction of any obstruction in navigable waters without the consent of Congress. Id., cl. 1. The second clause prohibits the construction of specified and other structures in those navigable waters absent permission from the Corps. Id., cl. 2. The district court made two alternative, independent findings that the at-issue segment of the Rio Grande is navigable: first, that it had been used or was susceptible of use in commerce in the past and second, that it was susceptible of future use in commerce with reasonable improvements. Both clauses require the waterway to be navigable, so we begin there. We then turn to the evidence supporting both the district court’s alternative holdings. I emphasize the majority’s statement that the “definition of navigability is broad”—a well-established proposition backed by decades of legal authority. Indeed, over the centuries, the Supreme Court has stressed this. For example, the use need not be continuous, “[s]mall traffic compared to the available commerce of the region is sufficient,” and “absence of use over long periods of years . . . does not affect the navigability of rivers in the constitutional sense.” Appalachian Elec., 311 U.S. at 409-10. “Nor is lack of commercial traffic a bar to a conclusion of navigability where personal or private use by boats demonstrates the availability of the stream for the simpler types of commercial navigation.” Id. at 416 (emphasis added). Courts have distilled those precepts into a practical test. A navigable waterbody is one that must be or have been used or susceptible of use in the customary modes of trade and travel on water as a highway for interstate commerce. Miami Valley Conservancy Dist. v. Alexander, 692 F.2d 447, 449-50 (6th Cir. 1982). A The vacated panel opinion properly held that the evidence put forth by the United States at this stage supports the district court’s conclusion that the Rio Grande was historically navigable or historically susceptible of use in commerce. United States v. Abbott, 87 F.4th 616, 623 (5th Cir. 2023); see also Econ. Light & Power Co. v. United States, 256 U.S. 113, 117 (1921) (finding the Desplaines River navigable despite “no evidence of actual navigation within thememory of living men” yet historic accounts of fur trading); The Montello, 87 U.S. 430, 441 (1874) (“The capability of use by the public for purposes of transportation and commerce affords the true criterion of the navigability of a river, rather than the extent and manner of that use.”). The United States presented ample evidence supporting navigability that the district court relied on in reaching its conclusions.2 This included a number of studies conducted by the Army Corps and the Coast Guard. For example, a 2011 Army Corps Study titled “Navigable Waters of the United States in the Fort Worth, Albuquerque, and Tulsa Districts Within the State of Texas,” which specifically included the Rio Grande from “the Zapata-Webb county line upstream to the point of intersection of the Texas-New Mexico state line and Mexico” as “navigable waters of the United States” falling within the Corps’ jurisdiction. Likewise, the United States Coast Guard published a 1984 “Navigability Determination, Rio Grande River, TX” indicating that the Rio Grande “was listed among the navigable waters of the United States pursuant to treaties with Mexico and for Coast Guard regulatory purposes” which was determined to be “still in effect.” The studies continue: the United States also pointed to a 1975 Navigability Study of the Army Corps that concluded “[t]he Rio Grande River between River Mile 275.5 and 610.0, on the United States side from the centerline of the normal channel, is a navigable water of the United States,” expressly encompassing the portion of the channel at issue. Then, there are the more than half a dozen historical references to navigationin theregionthat specifically discuss navigationontheRio Grande throughout history, which the majority does not attempt to address despite the Supreme Court’s statement that this type of evidence is sufficient in and of itself to show navigability. In Appalachian Electric, the Supreme Court found “[f]ourteen authenticated instances of use in a century and a half by explores and trappers, coupled with general historical references to the river as an early water route for fur traders” sufficed to show navigability. 311 U.S. at 416.3 But that’s not the only evidence the district court considered at this early stage of the litigation. Instead, the United States pointed to treaties between the United States and Mexico, including the Treaty of Guadalupe Hidalgo, which agreed that navigation on the Rio Grande “divided in the middle between the two Republics . . . shall be free and common to the vessels and citizens of both countries,” an agreement the United States surely cannot uphold should Texas persist in leaving its obstruction in the middle of the river. As discussed further below, the district court also properly considered contemporaneous court cases showing ferry traffic of expressly commercial goods, such as cotton, between Eagle Pass and Piedras Negras and four different acts of Congress preserving the navigability of the river. All this ample evidence is outlined in great detail in the vacated panel opinion. See Abbott, 87 F.4th at 623-28. It is hard to imagine a stronger preliminary showing of historical use in commerce supporting the district court’s navigability determination, which, to reiterate, is reviewed for clear error. It is harder still to imagine what the United States could put forth that would satisfy themajority. This is especially so considering that theSupreme Court itself has weighed in on this precise issue. In Rio Grande Dam, the Court reasoned “[t]hat the Rio Grande, speaking generally, is a navigable river.” United States v. Rio Grande Dam & Irrigation Co., 174 U.S. 690, 698 (1899). This understanding, according to the Supreme Court, was “a matter of such common knowledge” that “courts may properly take judicial notice of that fact.” Id.4 The majority raises a few arguments seeking to undercut this evidence, specifically regarding the cross-river ferry traffic, the four acts of Congress, and the requirement to consider the Rio Grande’s navigability in a narrow, segment-by-segment manner. I address each in turn. i. Cross-River Navigation To begin, the majority dismisses historical evidence of commercial ferry traffic between Eagle Pass and Piedras Negras. Specifically, the majority asserts that cross-river traffic does not support a finding of navigability because the traffic must run along the river, not across it, to be a “highway[] for commerce.” But such a distinction has no basis in fact, law, or common sense: commerce must overcome many physical obstacles— mountains, lakes,rivers—buteven passing through thesebarriers,commerce is occurring. And it is occurring up and down stream, as well as across it, despite the majority’s interesting linguistics argument. Two cases relied upon by the district court confirm the historical presence of competitive commercial navigation at Eagle Pass, as they address disputes over ferries between Eagle Pass and Piedras Negras. United States v. Weil, 35 Ct. Cl. 42, 77 (1900) (“At Eagle Pass there were ferryboats in which the cotton was crossed over.”); Tugwell v. Eagle Pass Ferry Co., 74 Tex. 480, 9 S.W. 120 (1888) (resolving dispute between rival ferry companies operating between Eagle Pass and Piedras Negras). As the United States argues, these cases involved expressly commercial enterprises: transporting commercial goods across an international boundary. And as the Supreme Court has declared, “[s]mall traffic compared to the available commerce of theregionis sufficient” to establish navigabilityunder theRHA. Appalachian Elec., 311 U.S. at 409. In any event, nothing supports the novel proposition that commercial navigation cannot be conducted cross-river, particularly where that crossing is an international boundary. The lone authority cited as support—an out-of-circuit district court opinion—not only lacks precedential weight, but is also manifestly distinguishable, involving entirely intrastate ferry traffic in Georgia on the Chattahoochee River. United States v. Crow, Pope & Land Enterprises, Inc., 340 F. Supp. 25, 35 (N.D. Ga. 1972). The majority likewise finds support in Crow, but its use is unpersuasive.5 If anything, the Supreme Court has made clear that using ferry boats in commerceis a “customary mode[] of trade and travel onwater,” satisfying navigability under the RHA. Econ. Light, 256 U.S. at 122; see also New York Cent. & H.R.R. Co. v. Bd. of Chosen Freeholders of Hudson Cnty., 227 U.S. 248, 264 (1913) (“[A]ll business of the ferries between the two states was interstate commerce within the power of Congress to control.”). Indeed, in Appalachian Electric, the Supreme Court expressly acknowledged ferry traffic in performing a navigability analysis. 311 U.S. at 413 n.46 (“At different times before 1935 ferries crossed the river at no less than ten points along the Radford-Wiley’s Falls stretch.”). We havedonethe same by concluding that the Rio Grande at the Hidalgo-Reynosa Bridge is a navigable water because, inter alia, “[d]efendant concedes that ferrying and minor fishing activities still are carried out. The Supreme Court repeatedly has sustained a conclusion of navigability on this kind of evidence.” Puente de Reynosa, S.A. v. City of McAllen, 357 F.2d 43, 51 (5th Cir. 1966). Moreover, in The Daniel Ball, the Supreme Court made clear that navigable waters are brought within the sphere of the sovereignty of the United States when those waters implicate interstate or foreign commerce. The Daniel Ball, 77 U.S. 557, 563 (defining waters that are navigable in fact as those “over which commerce is or may be carried on with other State or foreign countries in the customary modes in which such commerce is conducted on water”) (emphasis added). Here, it cannot be overstated that the Rio Grande is an international boundary and that cross-river commerce between theUnited States and Mexico hashistorically occurred in theregion. And importantly, Supreme Court authority requires that we look at the “character of the region,” when considering the navigability of a waterway. Appalachian Elec., 311 U.S. at 409. The majority in no way attempts to grapple with the Rio Grande’s position as the international boundary between theUnited States and Mexico—adefining characteristicof theriver. Accordingly, the historic ferry usage here—involving expressly commercial enterprises transporting goods across an international boundary—is satisfactory evidence of historical navigability. ii. Acts of Congress Next, the majority dismisses four separate acts of Congress as irrelevant to a determination of navigability. Its dismissal of these statutes relies on Oklahoma v. Texas. There, the Supreme Court stated that unrelated acts of Congress were “only precautionary and not intended as an affirmation of navigable capacity in that locality.” Oklahoma v. Texas, 258 U.S. 574, 586 (1922). But, as the United States persuasively argues in its en banc brief, the acts in Oklahoma are readily distinguishable from those cited in the instant matter. In Oklahoma, Congress delegated the navigability questions to the Secretary of War, giving him “authority” to “secure free and complete navigation” by requiring alterations to the project. See, e.g., Act of May 15, 1886, c. 332, 24 Stat. 28. The Eagle Pass statutes here, by contrast, go further: They expressly preclude interference with navigation and provide redress in federal district court for obstructions. 23 Stat. 29 (1884); 26 Stat. 495 (1890); 26 Stat. 502 (1890) (all enabling redress in federal court “in case the free navigation of the river shall at any time be substantially or materially obstructed”). These acts also protect the Rio Grande’s navigability in connection with authorizing bridges and other projects involving foreign commerce between Eagle Pass, Texas, and Piedras Negras.6 By their terms, the laws specifically mandate “free navigation” on this precise segment of the Rio Grande. At the very least, the statutes reflect a congressional judgment that the Rio Grande near Eagle Pass is susceptible to use in commerce satisfying navigability at this stage of the litigation. iii. Segment-by-Segment Approach Finally, the majority errs in its navigability analysis in one final way: accepting Texas’s position (without explanation) that navigability must be assessed based upon a segment-by-segment approach. Texas relies on PPL Montana, LLC v. Montana for this proposition. The question there was “whether discrete, identifiable segments of [three] rivers in Montana were nonnavigable, as federal law defines that concept for purposes of determining whether the State acquired title to the riverbeds underlying those segments, when the State entered the Union in 1889.” 565 U.S. 576, 580 (2012). There is no question that this is not the same standard applied in RHA cases, as PPL Montana specifically involved title to disputed riverbeds under the constitutional equal-footing doctrine. Id. at 581. In fact, the Supreme Court specifically stated in PPL Montana, “that the test for navigability is not applied in the same way in these distinct types of cases.” Id. at 592. Significantly, “[t]o determine title to a riverbed under the equal-footing doctrine, this Court considers the river on a segment-by-segment basis to assess whether the segment of the river, under which the riverbed dispute lies, is navigable or not.” Id. at 593 (emphasis added). I find no support for such a narrow segment-by-segment approach under theRHAnor does themajority offer one. Instead, “[t]hequestion here is not with respect to a short interruption of navigability in a stream otherwise navigable, or of a negligible part, which boats may use, of a stream otherwise nonnavigable. We are concerned with long reaches with particular characteristics of navigability or nonnavigability.” Utah, 283 U.S. at 77 (emphasis added). Further, the Supreme Court has indicated that “Congress may exercise its control over non-navigable stretches of a river in order to preserve or promote commerce on the navigable portions.” United States v. 531.13 Acres of Land, More or Less, in Oconee Cnty., State of S.C., 366 F.2d 915, 921 (4th Cir. 1966) (quoting Oklahoma ex rel. Phillips v. Guy F. Atkinson Co., 313 U.S. 508, 525-26 (1941)). Finally, the Supreme Court has made clear that a barrier in an unnavigable stretch of a water may nevertheless be an obstruction of a navigable water if it impacts the navigable capacity of a downstream water area that is navigable. See, e.g. Rio Grande Dam, 174 U.S. at 708 (remanding for a determination of whether proposed construction of a dam in an unnavigable portion of the Rio Grande [near New Mexico] would impact the navigability of a navigable portion of the Rio Grande); Econ. Light, 256 U.S. at 122 (“Navigability, in the sense of the law, is not destroyed because the water course is interrupted by occasional natural obstructions or portages; nor need the navigation be open at all seasons of the year, or at all stages of the water.”). This, too, supports a broader reading of RHA navigability than that urged in a segment-by-segment approach. In sum, the historical evidence provided here sufficiently encompasses the stretch of the river at issue. Moreover, even limited to a segment-by-segment approach, the evidence of ferry usage and the specific acts of Congress regulating this precise location are sufficient to withstand clear error review at the preliminary injunction stage. B Next, I address the district court’s alternative holding regarding future navigability with reasonable improvements. I am not persuaded by the majority’s reasoning on this point and would again affirm the district court’s conclusion. Much discussion hinges on whether improvements would be “reasonable.” For example, Texas argues that a cost-benefit analysis must be performed to show that any improvements would be practicable, and that the United States has made no effort to “quantify” or “describe” what improvements could be performed to make this portion navigable, and the majority agrees. The district court concluded that this portion of the river remains capable of increased commercial navigation with reprioritization and increased flow from the Amistad Dam. The caselaw supports the district court on this point. In Appalachian Electric, the Supreme Court stated that “a balance between cost and need at a time when the improvement would be useful” is required, which suggests that this analysis is not required at the present time. 311 U.S. at 407-08 (emphasis added). In Appalachian Electric, the cost of possible improvements to make the New River navigable were described by the Army Corps as “prohibitive” but the Court concluded that the river was capable of future use in commerce with reasonable improvements, and therefore navigable, nonetheless. Id. at 418. Further, it is not “necessary that the improvements should be actually completed or even authorized.” Id. at 408. “The plenary federal power over commerce must be able to develop with the needs of that commerce which is the reason for its existence.” Id. at 409. As the United States argues, “[w]hether the costs justify the necessary improvements is ultimately Congress’s call but does not impact the legal analysis of navigability.” Here, the district court concluded that reprioritization of water from nearby reservoirs or even modest changes in flow due to natural causes would make the river susceptible to use for commercial navigation. This conclusion is not clear error. Even accepting that a cost-benefit analysis is required on the merits, the demand for this analysis on the evidence offered at the preliminary injunction stage is improper. See Byrum, 566 F.3d at 446. And we must take the nature of the district court’s task in a preliminary injunction proceeding into account when performing a clear error review. Camenisch, 451 U.S. at 395. At this stage, and with the broad reading demanded by the Supreme Court in assessing navigability under the RHA, the United States has adequately shown that the Rio Grande is susceptible to future use in commerce with reasonable improvements. The majority, ever forgetting the standard of review, again demands far more (an entire cost-benefit analysis) from the United States than this stage of litigation requires. Considering both thedistrictcourt’s findings onhistoricalnavigability and future navigability with reasonable improvements, I am not left with a “definitive and firm conviction” that the district court erred. F.2d at 501-02 (quoting U.S. Gypsum Co., 333 U.S. at 395). Clark, 693 Quite the opposite—the evidence relied upon by the district court is more than sufficient to affirm its findings under a clear error standard. The majority attempts to poke holes in the evidence to render it insufficient. But as this opinion makes clear, the evidence was proper to rely on and demonstrates navigability on the Rio Grande. Instead of properly accepting these findings on a preliminary injunction, the majority instead reweighs the evidence to support its desired result. This is not the purview of an appellate court. And a preliminary injunction is certainly not the place for this. Camenisch, 451 U.S. at 395 (“[I]t is generally improper for a federal court at the preliminary-injunction stage to give a final judgment on the merits.”). Had it held off on taking a preliminary injunction en banc, the majority could have weighed in with greater force, as it would not be constrained by this standard of review.7 But instead, we find ourselves reviewing and re-reviewing the grant of a preliminary injunction over the past year. Having not clearly erred in either its navigability analysis or its conclusion, the district court should be affirmed. III Having concluded that the district court correctly determined that the United States is likely to show that the Rio Grande at Eagle Pass is navigable, we turn to whether the floating barrier constitutes an obstruction. The first clause of RHA’s Section 10 bars the “creation of any obstruction not affirmatively authorized by Congress, to the navigable capacity of any of the waters of the United States.” 33 U.S.C. § 403. After finding navigability, the district court found obstruction, concluding that “the floating barrier interferes with or diminishes the navigable capacity of the Rio Grande and creates a hazard.” Whether there is an obstruction is likewise a question of fact reviewed for clear error. Rio Grande Dam, 174 U.S. at 709. The Supreme Court has defined an obstruction as tending to “interfere with or diminish[] the navigable capacity of a stream.” Id. It has also emphasized that its own cases define “obstruction” as used in Section 10 as “broad enough to include diminution of the navigable capacity” of the waterway at issue. United States v. Republic Steel Corp., 362 U.S. 482, 489 (1960).8 Under this reading, the Supreme Court has previously found matter described as “fine particles” from an iron mill to be an obstruction under Section 10. Id. at 483. For our part, we have acknowledged the Supreme Court’s broad interpretation of obstruction. Vieux Carre Prop. Owners, Residents & Assocs., Inc. v. Brown, 875 F.2d 453, 462-63 (5th Cir. 1989) (“The Supreme Court has encouraged a broad interpretation of a section 10 ‘obstruction’. . ..”). And it has construed the term flexibly, as the district court noted, without a size or positional limit. See United States v. Raven, 500 F.2d 728, 731 (5th Cir. 1974) (finding a sunken schooner to be an obstruction “[i]f floating particles can be an obstruction”); United Tex. Transmission Co. v. U.S. Army Corps of Eng’rs, 7 F.3d 436, 438 (5th Cir. 1993) (finding a pair of gas pipelines running under the bed of a bayou an obstruction). Here, the district court began its analysis by noting that Texas designed and deployed the floating barrier to literally obstruct lateral movement across the river.9 It then looked to the credible testimony and evidence before it, including the declaration of Mario Gomez, Acting Area Operations Manager for the Amistad Dam Field Office of the United States Section of the International Boundary and Water Commission (“IBWC”), who indicated that “[n]ormally, the Mexico or U.S. Section of the Commission can go into any location of the Rio Grande independently and do surveying and other engineering work that the Commission Sections carry out.” But the floating barrier, he stressed, is “an impediment to the Sections crossing independently in this part of the river,” including a planned survey by the Mexican Section of the Commission that was unable to proceed because of the obstruction. Additionally, the Chief of the U.S. Border Patrol (“USBP”), Jason D. Owens, declared that border patrol agents rescue individuals in distress in the Rio Grande, utilizing “small watercraft to quickly respond as the incidents unfold.” He noted that “[a]ny obstructions in the water could naturally impair the freedom of movement and potentially delay response times.” “From the beginning of fiscal year 2018 through July 23, 2023 there were 249 water-related rescues and 89 water-related deaths of individuals whose rescue or death occurred in or around the Rio Grande throughout the Eagle Pass Station [Area of Responsibility].”10 Further, in the declaration of Joseph L. Shelnutt, the Regulatory Project Manager in the Compliance and Enforcement Branch for the Corps, he indicates that the “placement and tandem configuration of the buoys, which allows them to move somewhat independently even though they are connected, present a structural barrier to cross-river navigation and would force a vessel to maneuver around the structure to avoid collision or entanglement at this location.” Moreover, the floating barrier is not just a couple floating buoys: “Photographs show these grey concrete anchors standing from the bed of the river, with nomarkings to identify them as hazards. These concrete obstacles present a serious risk to watercraft of any kind.” This is because the anchors are not easily seen by oncoming watercraft but are at a level that would cause damage to a vessel of any size that came upon them. This evidence, coupled with the Supreme Court’s command to interpret “obstruction” within Section 10 broadly, supports the district court’s finding that the floating barrier is an obstruction to the navigable capacity of the Rio Grande, and I again see no clear error. Having made the requisite showing that Texas is likely in violation of the first clause of Section 10 because its obstruction was not “affirmatively authorized by Congress,” 33 U.S.C. § 403, the United States has shown that it is likely to succeed on the merits of its first claim.11 IV Having reached this point, I must nowengage with Texas’s invocation of the political question doctrine. The majority, having concluded that the district court abused its discretion in granting the United States relief, notes that it need not and does not reach Texas’s argument that Article 1, § 10, clause 3 of the U.S. Constitution authorizes it to erect the floating barriers in defense of a border “invasion,” even if the barrier violates the RHA. Texas claims that it erected the floating barrier pursuant to its constitutional authority to “engage in War” without consent of Congress, if “actually invaded, or in such imminent Danger as well not admit of delay.” U.S. Const. art. I, § 10, cl. 3. The issue of whether an influx of migrants constitutes an “invasion” is robustly debated. See, e.g., California v. United States, 104 F.3d 1086, 1091 (9th Cir. 1997) (ruling that the Invasion Clause “afford[s] protection in situations wherein a state is exposed to armed hostility from another political entity” and “not intended to be used” to combat illegal migration)); New Jersey v. United States, 91 F.3d 463, 468-69 (3d Cir. 1996) (reaching the same conclusion); Padavan v. United States, 82 F.3d 23, 28 (2d Cir. 1996) (“In order for a state to be afforded the protections of the Invasion Clause, it must be exposed to armed hostility from another political entity, such as another state or foreign country that is intending to overthrow the state’s government.”); but see Chiles v. United States, 69 F.3d 1094, 1097 (11th Cir. 1995) (“[W]e conclude that whether the level of illegal immigration is an ‘invasion’ of Florida and whether this level violates the guarantee of a republican form of government present nonjusticiable political questions.”). Even if one were to agree that the issue presents a nonjusticiable political question, however, that is not dispositive of the court’s jurisdiction. Texas’s argument posits that even if Texas violated federal law in erecting the floating barrier, this court could not decide the issue of whether Texas acted properly pursuant to its limited war powers, as this would present a nonjusticiable political question. But this argument is misleading. Contrary to Texas’s understanding, Clause 3 gives Texas a limited right to self-defense without congressional approval that is temporary. And even if we accept the premise that Texas had a limited right to engage in self-defense, the window for Texas to act without Congress’s consent has certainly closed. Clause 3 provides that a state may engage in war without consent of Congress only when it is “actually invaded, or in such imminent Danger as will not admit of delay.” The language of Clause 3, in addition to history and precedent, indicates that a state’s emergency powers under Clause 3 activate when it would be infeasible for a state, when faced with imminent danger, to receive consent from Congress to engage in self-defense. See, e.g., Articles of Confederation of 1781, art. VI, para. 5 (limiting a state’s power to engage in war “till the united states in congress assembled can be consulted”); Robert G. Natelson & Andrew T. Hyman, The Constitution, Invasion, Immigration, and the War Powers of States, 13 Brit. J. Am. Legal Stud.1, 17 (2024) (noting that, in regard to state war powers, the Constitution resulted in “a balance between federal and state prerogatives roughly similar to that under the Articles of Confederation”); Virginia v. Tennessee, 148 U.S. 503, 518 (1893) (“[I]t would be the height of absurdity to hold that the threatened states could not unite in providing means to prevent and repel the invasion of [a] pestilence without obtaining the consent of congress, which might not be at the time in session.” (emphasis added)).12 Because a state’s Clause 3 powers emerge when receiving consent from Congress is infeasible, it rationally follows that these powers lapse once Congress can be notified and respond. See United States v. Abbott, 690 F. Supp. 3d 708, 729 n.29 (W.D. Tex. 2023) (“[The] authority to act independently only lasts until resources of the federal government can reach the invasion.”). In other words, the relevant issue in a Clause 3 inquiry is not whether the federal government has responded meaningfully to the state’s alleged crisis, but whether Congress has had the opportunity to be consulted and respond. After all, the Guarantee Clause ensures that the United States “shall protect each of [the states] against Invasion.” U.S. Const. art. IV, § 4. And if the federal government responds in a manner that Texas finds inadequate, then Texas is taking issue with the United States’ protection of the states against “invasion” pursuant to the Guarantee Clause. But Texas cannot have it both ways: If this court cannot adjudicate Texas’s response to an “invasion” under Clause 3, then it also cannot adjudicate the federal government’s response to the same “invasion” under the Guarantee Clause. Additionally, if the window of Texas’s limited war powers depended on the adequacy of the federal government’s response, then Congress’s failure to respond to an invasion, even when it had time torespond, would enable Governor Abbottto engage in acts of war in perpetuity. Such agrant of power to a state governor would seriously undermine federal supremacy in the areas of war, foreign policy, immigration, and protection against invasion. See The Federalist No. 42 (James Madison) (suggesting that the Framers did not want to “leave it in the power of any indiscreet member to embroil the Confederacy with foreign nations”). The district court got this exactly right: The State War Clause, read consistently with other provisions of the Constitution, authorizes states to respond to invasion only as an emergency interim measure before the federal government may respond. Put simply, if Texas is engaging in war . . . it must cede authority to the federal government to conduct that war once the federal government has had time to respond to the purported invasion. United States v. Texas, 2024 WL 861526, at *35 (W.D. Tex. Feb. 29, 2024). Here, nearly a year after erecting the floating barrier, it cannot reasonably be argued that the United States has not had adequate time to respond to the purported “invasion” of Texas. This case does not require the court to precisely nail down the temporal parameters of Texas’s limited self-defense powers; no matter how wide that window was open, that window has clearly closed. It would be difficult to accept the argument that Texas’s continued exercise of its war powers is a political question in this instance, where no reasonable person could contend that the federal government has not had an adequate opportunity to respond to any purported “invasion.” Under any reasonable standard, the United States has had ample opportunity to respond to the purported “invasion.” It has been a year since Governor Abbott first announced Texas’s intentionto erect thefloating barrier. If indeed there was a limited window for Texas to unilaterally engage in acts of war without Congress’s consent, that window is closed. Accordingly, we may properly (and rather simply) dispose of Texas’s attempt to invoke a defense to its illegal actions by pointing to Clause 3. V Concluding that the United States has shown a likelihood of success on the merits, I turn to the remaining preliminary injunction factors: whether the United States has shown a likelihood of irreparable harm absent preliminary relief, that the balance of equities tips in its favor, and that an injunction is in the public interest. Winter, 555 U.S. at 20. The district court concluded these factors weigh in favor of the United States and that determination was no abuse of discretion. The district court emphasized the strain on United States–Mexico relations in assessing irreparable harm. For example, the floating barrier “has been the subject of a series of correspondence from the Mexican section of the IBWC” and is “interfering with the ability of the IBWC to fulfill its mission” to implement the core provisions of the 1944 Treaty between the United States and Mexico, crucial to the allocation of waters in the Rio Grande. On July 14, 2023, Mexico’s Section of the IBWC “objected to the placement of the buoys and requested intervention of the United States Section to remove the buoys.” It further reported that, “as a result of the floating barriers in Eagle Pass, Mexico was cancelling a July 24, 2023, meeting concerning water releases to the United States from the Rio Conchos River in Mexico.” In a July 14, 2023, meeting between the Texas Commission on Environmental Quality and the IBWC to discuss delivery by Mexico to the United States of water from the Rio Grande, Mexico’s Section of the IBWC indicated that Texas’s unilateral actions “could affect cooperation between the two countries going forward.” “[I]f the proposed [structure] and appropriation of waters of the Rio Grande constitute a breach of treaty obligations or of international duty to Mexico, they also constitute an equal injury and wrong to the people of the United States.” Rio Grande Dam, 174 U.S. at 701. On July 24, 2023, Hillary Quam, the U.S.-Mexico Border Coordinator at the U.S. Department of State and acting Director of the Office of Mexican Affairs, declared that if the barrier “is not removed expeditiously, its presence will have an adverse impact on U.S. foreign policy,” and that Mexico “[o]n a number of different occasions beginning in late June, 2023″ has “protested to the United States the deployment of a floating barrier within the Rio Grande.” Mexico has protested the installation of the floating barrier, “asserting thatitcauses obstructionand deflectionof theriver as well as possible runoff into Mexican territory” “in contravention of the provisions of the 1970 Treaty.” The President of Mexico, Andrés Manuel López Obrador, in his daily press conference, has discussed the matter of the floating barrier at least six times since July 25, 2023. Crediting this evidence to find a likelihood of irreparable harm was not erroneous.13 Turning to the balance of equities and the public interest, when the United States is a party, the third and fourth elements of the traditional preliminary injunction analysis merge. Cf. Nken v. Holder, 556 U.S. 418, 435 (2009). The district court found that “the barrier’s threat to human life, its impairment to free and safe navigation, and its contraindication to the balance of priorities Congress struck in the RHA outweigh Texas’s interest in implementing its buoy barrier in the Rio Grande River.” The Supreme Court would agree. In Sanitary District of Chicago v. United States, 266 U.S. 405 (1925), the Supreme Court stated: “There is no question that this power [to remove obstructions to interstate and foreign commerce] is superior to that of the States to provide for the welfare or necessities of their inhabitants.” Id. at 426. More pointedly, in Arizona v. United States, 567 U.S. 387 (2012), the Supreme Court noted that while Arizona “may have understandable frustrations with the problems caused by illegal immigration,” as Texas has alleged, “the State may not pursue policies that undermine federal law.” Id. at 416. Further, the district court emphasized that the “balance of priorities Congress struck in the RHA” outweighed Texas’s asserted interests. Courts may look at the statute at issue for guidance in determining whether the issuance of a preliminary injunction would be in the public interest. See, e.g., Amoco Production Co. v. Village of Gambell, AK, 480 U.S. 531, 544 (1987) (finding the public interest promoted by the Alaska National Interest Lands Conservation Act was “to protect Alaskan subsistence resources from unnecessary destruction,” rather than preventing the actions the plaintiff sought to enjoin). Congress has spoken to thepublic interest through passage of the RHA, and the Supreme Court has emphasized the same: “We are dealing here with the sovereign powers of the Union, the Nation’s right that its waterways be utilized for the interests of the commerce of the whole country.” Appalachian Elec., 311 U.S. at 405.14 The district court relied on all the evidence discussed herein to find that the balance of hardships favors the United States. It considered the threat to navigation and federal government operations on the Rio Grande, as well as the potential threat to human life the floating barrier created. All the district court’s findings of fact were well supported by the record, and its conclusion that the equities favor issuance of a preliminary injunction was not an abuse of discretion. VI The majority errs in reversing the district court’s decision. In doing so, it disregards the relevant standard of review, rejects the ample evidence put forth by the United States, and relies on unsupported reasoning. Accordingly, I respectfully dissent.