Texas Department of Family and Protective Services; Stephanie Muth, in Her Official Capacity as Dfps Commissioner; Texas Health and Human Services; Cecile Erwin Young, in Her Official Capacity as Hhsc Executive Commissioner; Corrections Corporation of America; And the Geo Group, Inc. v. Grassroots Leadership, Inc.; Gloria Valenzuela; E.G.S., for Herself and as Next Friend for A.E.S.G.; F.D.G., for Herself and as Next Friend for N.R.C.D.; Y.E.M.A., for Herself and as Next Friend for A.S.A.

• Court: Texas Supreme Court
• Decided: May 30, 2025
• Docket: 23-0192
• Status: Published

Opinion

Supreme Court of Texas ══════════ No. 23-0192 ══════════

Texas Department of Family and Protective Services; Stephanie Muth, in her Official Capacity as DFPS Commissioner; Texas Health and Human Services; Cecile Erwin Young, in her Official Capacity as HHSC Executive Commissioner; Corrections Corporation of America; and The GEO Group, Inc., Petitioners, v. Grassroots Leadership, Inc.; Gloria Valenzuela; E.G.S., for herself and as next friend for A.E.S.G.; F.D.G., for herself and as next friend for N.R.C.D.; Y.E.M.A., for herself and as next friend for A.S.A., Respondents

═══════════════════════════════════════ On Petition for Review from the Court of Appeals for the Third District of Texas ═══════════════════════════════════════

Argued October 31, 2024

JUSTICE YOUNG delivered the opinion of the Court.

Chief Justice Blacklock and Justice Sullivan did not participate in the decision.

We must decide whether Texas courts are constitutionally authorized to adjudicate moot cases that raise questions of considerable public importance. The court of appeals below invoked this so-called “public-interest exception” and proceeded to resolve the case on its merits. Under the text, structure, and history of our Constitution, however, the court of appeals exceeded its authority. There is no such thing as a public-interest exception to mootness in Texas. This case stems from a Department of Family and Protective Services rule that authorized state licenses for two residential facilities at which the federal government has detained mothers and children after their illegal entry into the United States. Under a prior federal consent decree, such state licenses were necessary for more than very brief detentions of immigrant children. A cadre of detained mothers, on behalf of themselves and their minor children, together with Grassroots Leadership, Inc., and a children’s day-care operator, challenged the department’s authority under state law to adopt the rule. Their goal: to prohibit the detention of children at the newly licensed facilities. Long before their challenge to the rule reached the court of appeals, however, the mothers and their children were all released from the facilities. The court of appeals thus concluded that their claims were moot. Ordinarily, that would lead to dismissal of the case. But due to its application of “the public-interest exception” to mootness, the court proceeded to address the merits and held the rule to be invalid under the Administrative Procedure Act. The court of appeals was correct to view the case as moot, and it should have proceeded no further. Mootness, which refers to when a case no longer presents a live controversy, is simply one branch of the larger doctrine of justiciability. The justiciability requirements collectively ensure that courts exercise only “[t]he judicial power of this State.” Tex.

2 Const. art. V, § 1. The judicial power of the Texas courts does not include the rendition of advisory opinions. So unyielding is this principle that only by separate constitutional authorization may this Court answer certified questions from federal appellate courts. See id. art. V, § 3-c. All other advisory opinions remain prohibited. The inherent consequence of deciding a moot case, however, is the rendition of an advisory opinion. It naturally follows that the only proper judgment in a moot case is one of dismissal for lack of jurisdiction. Our cases have described several “exceptions” to the mootness doctrine, but there are no exceptions to the fundamental constitutional requirement that courts may reach the merits of only live disputes. Each recognized mootness “exception” complies with that mandate by identifying disputes that seem to have ended but in fact remain live and thus are not truly moot at all in a constitutional sense. Each exception carefully ensures that the parties retain a genuine stake in the case and that a judgment resolving the dispute would still afford genuine relief. They are not exceptions to the Constitution’s limitations and, particularly, its prohibition of advisory opinions. The “public-interest exception,” by contrast, would be a true exception—one that would allow courts to openly render advisory opinions despite the constitutional ban on doing so. Courts must steadfastly accept the constitutional limitations on our authority. That these limitations are the law’s mandate should be enough. But beyond that, we can hardly expect the other branches and the public to respect constitutional boundaries if the courts are anything less than punctilious in doing so, particularly if we are perceived as aggrandizing our own power.

3 We accordingly reverse the court of appeals’ judgment to the extent the court held that it had jurisdiction to adjudicate the merits of a moot dispute. And because that determination was erroneous, we vacate the court of appeals’ judgment on the merits along with the orders and judgment of the trial court, and we render judgment dismissing the case for lack of subject-matter jurisdiction. I In a 1997 consent decree, a California federal court approved a class-action settlement in which, among other things, the federal government agreed not to house illegal-immigrant minors in residential facilities that lacked a valid state license. In 2014, the federal government began using two Texas facilities, called Dilley and Karnes, to detain families with children who had illegally entered the United States. Shortly thereafter, class members returned to federal court to file a motion to enforce the 1997 consent decree on the ground that the Dilley and Karnes facilities lacked a valid state license. The court held that the federal government had breached the settlement agreement by housing mothers and their children in secure, unlicensed facilities. See Flores v. Johnson, 212 F. Supp. 3d 864, 880 (C.D. Cal. 2015). The Texas Department of Family and Protective Services responded by promulgating a rule in March 2016 (after initially adopting it the prior year on an emergency basis) that established licensing requirements for “family residential centers”—facilities like Dilley and Karnes and, as far as we know, only those facilities. 26 Tex. Admin. Code § 748.7. To simplify things, family residential centers are essentially a subset of a preexisting category, “general residential operations,” which

4 includes specifically defined child-care facilities that provide full-time care for a specified number of children. Tex. Hum. Res. Code § 42.002(4). To qualify for a license, a new “family residential center” must satisfy the rules governing “general residential operations,” with a few exceptions. One exception allows qualifying facilities to house adults and children in the same bedroom, aiming to prevent the nighttime separation of children from their mothers. 26 Tex. Admin. Code § 748.7(c). To avoid splitting sibling groups, another exception allows more than four occupants per bedroom in certain circumstances. Id. Grassroots Leadership, Inc., a nonprofit civil-rights organization, sued the department in September 2015 to challenge the rule. Grassroots later amended its petition to add several detainee mothers and a day-care operator as plaintiffs. The petition alleges that, in reliance on the rule, Dilley and Karnes allowed unrelated adults and children to share bedrooms and that, because of that action, one of the mother’s children was sexually assaulted. The rule, they allege, increased the safety risk to detainees and children and resulted in longer detention periods. The plaintiffs asked the district court for a declaration that the department lacked the authority under the Administrative Procedure Act to adopt the rule; they also sought a permanent injunction. If the rule is invalid, then so are the licenses; if the licenses are invalid, then the consent decree bars the federal government from detaining children at Dilley and Karnes. Corrections Corporation of America (known as CoreCivic) and The GEO Group, Inc., the respective operators of Dilley and Karnes, intervened to defend the rule. The department, CoreCivic, and GEO Group filed pleas to the jurisdiction in the trial court asserting that the

5 plaintiffs lacked standing to challenge the rule. In relevant part, the trial court denied the pleas and eventually rendered summary judgment on the claim for declaratory relief under the Administrative Procedure Act. It declared the rule invalid and enjoined the department from granting licenses under it. The Third Court of Appeals reversed. It held that the detainee- mothers’ injuries were not traceable to the rule because the rule does not allow children to share a bedroom with unrelated adults. The court also held that any increase in the length of detention was not traceable to the rule. With no individual plaintiff who had standing, the court concluded, Grassroots likewise lacked standing. We reversed, holding that the detainee mothers had adequately established standing. Grassroots Leadership, Inc. v. Tex. Dep’t of Fam. & Protective Servs., 646 S.W.3d 815, 820–21 (Tex. 2022). Specifically, we concluded that the rule does allow children to share bedrooms with unrelated adults, which the “general residential operations” requirements prohibit, and so the injury was sufficiently traceable to the new rule. Id. at 819–20. We remanded for the court of appeals to consider the remaining jurisdictional issues and, if appropriate, the merits. Id. at 821. (Only the plaintiff-mothers’ claims remain, see id. at 819 & n.4, but for convenience, we collectively refer to “Grassroots” as making the plaintiffs’ shared arguments.) On remand, the court of appeals held the claims “moot by definition” because the detainees were no longer residents at the facilities. 665 S.W.3d 135, 141 (Tex. App.—Austin 2023). The period of detention at the facilities was quite short; the court of appeals explained,

6 and Grassroots agrees, that “the evidence establishes that the average length of detention [at the facilities] is eleven days, a period too short to complete litigation.” Id. For example, the only concrete evidence presented as to any particular plaintiff’s detention shows that one mother was released eight days after being added as a plaintiff to this lawsuit. It is undisputed that none of the plaintiff mothers remains detained. The parties do dispute the conditions of any release, while the court of appeals held that “there is no evidence explaining the circumstances or conditions, if any, of the detainees’ release.” Id. The court also held that the “capable of repetition yet evading review” exception did not save the claims from mootness. Grassroots cited legal authority allowing re-detention of the plaintiffs generally but did not cite evidence “that these same former detainees are reasonably likely to be detained at Dilley or Karnes again.” Id. Therefore, the court could not “conclude that there [was] more than a mere physical or theoretical possibility that they will be detained in one of these two centers—or any center—again.” Id. at 142. The court then, however, explained that its precedent allowed it to invoke the so-called “public-interest exception” to mootness, under which it could reach the merits despite having no live dispute involving the parties to the litigation. Stating that this Court has not yet considered the “viability of [the] public-interest exception,” the court of appeals explained that the exception “allows appellate review of a question of considerable public importance if that question is capable of repetition between either the same parties or other members of the public but for some reason evades appellate review.” Id. (quoting Univ. Interscholastic

7 League v. Buchanan, 848 S.W.2d 298, 304 (Tex. App.—Austin 1993, no writ)). The court deemed the rule’s validity to be a matter of great public importance. Id. After reaching the merits, the court held that the rule was invalid. See id. at 147–48. The department filed a petition for review. It contends that the court of appeals rightly found the claims moot, wrongly employed the public-interest exception as a basis to reach the merits, and then reached the wrong merits result. Grassroots defends the judgment below but argues that the court of appeals was mistaken to think that it needed to invoke the public-interest exception to mootness. According to Grassroots, the claims are not moot at all because the plaintiff-mothers’ release from detention does not foreclose the possibility of re-detention. In any event, it continues, even if this argument were wrong, the capable- of-repetition-yet-evading-review exception would save the claims from mootness. And if even that exception were not enough, Grassroots concludes, the court of appeals’ invocation of the public-interest exception would be justified, and the court correctly invalidated the rule. II The Texas Constitution imposes important justiciability limits on the judiciary’s power. “Justiciability doctrines” anchored in the Constitution’s text, structure, and history help courts assess their authority to adjudicate a given dispute. Central to this case is the doctrine of mootness. Unlike in the many cases where a doctrine’s underlying source does not much matter, whether and to what extent the Constitution defines and requires dismissal of moot cases is of the utmost importance here. The court of appeals’ decision to reach the merits of a

8 moot case relies entirely on the notion that a court may choose to do so despite acknowledging the lack of a live dispute. Implicit in that notion is a view of the Constitution’s understanding of justiciability. Specifically, if a court may decide a moot case that the court thinks raises issues of public significance, then the Constitution must have authorized that choice all along, for if the Constitution requires a live dispute at every stage of litigation, then no court could make an “exception” to it. See, e.g., Travelers’ Ins. Co. v. Marshall, 76 S.W.2d 1007, 1011 (Tex. 1934) (“[Courts] are without power to write . . . an exception into the organic law.”). The judgment of the court of appeals, therefore, rests on the premise that the doctrine of mootness is at least in large measure a matter of judicial administration or prudence rather than of constitutional command. The court of appeals never expressly confronted that question; the Constitution makes no appearance in its opinion. In fairness, the court did not devise the “exception” in this case. It followed its own 1993 precedent that expressly authorized the “public-interest exception.” See 665 S.W.3d at 142 (tracing the authority to Buchanan, 848 S.W.2d at 304). As we describe below, this Court’s cases have repeatedly reaffirmed, including after the Third Court embraced the public-interest exception in 1993, that mootness is a constitutional limitation on judicial authority. Thus, “[h]owever much we may desire to provide answers in these now-moot . . . proceedings, the constitution prohibits us from doing so, and we must respect that prohibition.” ERCOT, Inc. v. Panda Power Generation Infrastructure Fund, LLC, 619 S.W.3d 628, 631 (Tex. 2021) (emphasis added). Notably, we have at least heavily suggested that the

9 public-interest exception is an unauthorized basis for jurisdiction: “We do not have power to decide moot cases, whether they ‘involve a matter of public concern’ or not.” Morath v. Lewis, 601 S.W.3d 785, 789 (Tex. 2020) (citation omitted). These and other holdings should have cast doubt on the court of appeals’ reliance on its precedent that embraced the public-interest exception. See Mitschke v. Borromeo, 645 S.W.3d 251, 256–58 (Tex. 2022) (recognizing that courts of appeals must adhere to the principles of horizontal stare decisis, which includes not following precedents that have been disturbed by higher authority, including decisions of this Court). Today’s case, at the very least, illustrates the need for this Court to remove all doubt about the matter. We must, in other words, assess whether the Constitution truly imposes the core requirement of a live dispute or whether the courts may depart from it. We confirm our repeated assertions that mootness is one aspect of the constitutional concept of justiciability—a limitation on judicial power that neither the judiciary nor any other authority subject to the Constitution may set aside or disregard. To resolve the case, therefore, we proceed in two steps. First, we explain why mootness—as one of the core justiciability doctrines—is in fact a manifestation of the text, structure, and history of the Texas Constitution. Second, we apply the mootness doctrine to the circumstances of this case. A “Justiciability” is a formal word for a fundamental concept: the line separating lawsuits that courts may adjudicate from those they may not. A case may be justiciable yet not fall within a particular court’s subject-

10 matter jurisdiction, but courts never have subject-matter jurisdiction over cases that are nonjusticiable. Over time, therefore, justiciability doctrines have been developed to readily identify when adjudicating a case would exceed judicial authority, and especially when doing so would violate the limitations on the judiciary imposed by the Constitution. One fundamental example is “adversity between parties” because “without such adversity there is no justiciable controversy.” Paxton v. Longoria, 646 S.W.3d 532, 538 (Tex. 2022). The most familiar justiciability doctrines are standing, mootness, and ripeness, all of which “help ensure that courts do not issue advisory opinions,” Bienati v. Cloister Holdings, LLC, 691 S.W.3d 493, 498 (Tex. 2024), and which aid the resolution of “a real and substantial controversy involving genuine conflict of tangible interests and not merely a theoretical dispute,” Sw. Elec. Power Co. v. Lynch, 595 S.W.3d 678, 685 (Tex. 2020) (quoting Bonham State Bank v. Beadle, 907 S.W.2d 465, 467 (Tex. 1995)). The Texas Constitution does not mention mootness, standing, ripeness, or any other justiciability doctrine by name, but those doctrines work together to ensure that at every stage of litigation, a live dispute exists that is proper for judicial resolution, which in turn ensures that the eventual judgment resolves the dispute. Unsurprisingly, given that there are only “subtle differences between mootness and related justiciability concepts, such as ripeness and standing,” Travelers Ins. Co. v. Joachim, 315 S.W.3d 860, 865 (Tex. 2010), a case failing one of those tests often will fail another, or even all three, cf., e.g., MedImmune, Inc. v. Genentech, Inc., 549 U.S. 118, 128 n.8 (2007) (noting that the “standing and ripeness” issues in the case “boil[ed] down to the same question”).

11 Cases that focus on one doctrine often, therefore, deploy constitutionally based justiciability analysis that is applicable to the others. Beyond delineating their constitutional dimensions, Texas courts have sometimes noted that the familiar justiciability doctrines also have a “prudential” aspect, meaning that the courts themselves have erected justiciability barriers beyond what the Texas Constitution may demand. See, e.g., King St. Patriots v. Tex. Democratic Party, 521 S.W.3d 729, 735 & n.21 (Tex. 2017) (noting that “even in a case raising only prudential concerns, the question of ripeness may be considered on a court’s own motion” (citation omitted)). The question before us, however, is the extent to which the Constitution provides the underlying basis for these interrelated justiciability requirements. In other words, we need not address the extent to which the judiciary or the legislature may require more than the Constitution does; the point is that they may not authorize courts to proceed with any lesser showing of justiciability than the Constitution demands. As this Court’s cases reflect, at least five specific provisions of the Constitution’s text implicate justiciability: Article V, § 1’s “judicial power” provision; Article V, § 3(a)’s “cases” requirement; Article II, § 1’s “separation of powers” clause; Article IV, § 22’s “legal advice” provision; and Article I, § 13’s “open courts” provision. These provisions also illustrate how our constitutional structure likewise limits the judiciary to the exercise of only judicial power, and they are bound up in the history of how our People have allocated governmental authority. We address them in turn. The judicial-power provision. First, the Constitution confines

12 Texas courts to exercising only “judicial power” and no other kind. Article V, which governs the “Judicial Department,” begins this way: “The judicial power of this State shall be vested in one Supreme Court . . . .” Tex. Const. art. V, § 1; see also id. art. V, § 3(a) (providing for this Court to “exercise the judicial power of the state except as otherwise provided in this Constitution”). As we recently observed, “the judicial power” referenced in the Constitution includes two broad categories: “jurisdictional power” and “administrative powers.” See Webster v. Comm’n for Law. Discipline, 704 S.W.3d 478, 489 (Tex. 2024). In Webster, we focused on the administrative category and explained that “the original public meaning of the ‘judicial power’ created by the Texas Constitution” includes, among other things, “the judiciary’s authority to regulate the practice of law.” Id. at 490–91; accord id. at 507 n.5 (Boyd, J., dissenting). Today’s case, by contrast, involves the far more familiar “jurisdictional power” of the courts—the power “to adjudicate cases or liquidate law.” Id. at 489. In this context, the original public meaning of the term “judicial power” is well recognized. It implicates the authority to resolve actual, non-collusive legal disputes brought by adverse parties who have a genuine legal interest and a live stake in the outcome, which can be reduced to an enforceable judgment. This power includes, for example, “the distinctly judicial duties of rendering judgment, imposing sentence, and adjudicating any appellate or collateral challenges that may be raised.” In re Tex. House of Representatives, 702 S.W.3d 330, 342 (Tex. 2024). In this context, “ ‘[j]udicial power’ is the power of a court to decide and pronounce a judgment and carry it into effect between persons

13 and parties who bring a case before it for a decision.” Morrow v. Corbin, 62 S.W.2d 641, 644 (Tex. 1933); see also, e.g., Panda Power, 619 S.W.3d at 637 n.17. The traditional understanding of “judicial power” goes far beyond Texas, of course. The text of Article III of the U.S. Constitution begins exactly as Article V of our Constitution does: “The judicial Power of the United States, shall be vested in one supreme Court . . . .” U.S. Const. art. III, § 1. The U.S. Supreme Court has emphasized that the judicial power granted in the U.S. Constitution “gives the Federal Judiciary the power, not merely to rule on cases, but to decide them, subject to review only by superior courts in the Article III hierarchy—with an understanding, in short, that ‘a judgment conclusively resolves the case’ because ‘a “judicial Power” is one to render dispositive judgments.’ ” Plaut v. Spendthrift Farm, Inc., 514 U.S. 211, 218–19 (1995) (citation omitted); see also, e.g., Stern v. Marshall, 564 U.S. 462, 494 (2011). Confining courts to the exercise of “judicial power,” therefore, circumscribes the kinds of disputes that courts can resolve. We have long held, for example, that “the rendition of advisory opinions has generally been held not to be the exercise of judicial power.” Morrow, 62 S.W.2d at 643–44 (emphasis added); see also, e.g., Firemen’s Ins. Co. of Newark, N.J. v. Burch, 442 S.W.2d 331, 334 (Tex. 1968) (“[T]he giving of advice as to proposed or possible settlements is not a judicial function.” (emphasis added)), overruled on other grounds by Farmers Tex. County Mut. Ins. Co. v. Griffin, 955 S.W.2d 81, 83 (Tex. 1997). To be clear, advisory opinions are not inherently pejorative or improper—they can be quite valuable. Clients depend on lawyers to provide high-quality legal advice. But

14 providing advisory opinions is not a species of judicial power, as it exists outside “the conception of those who framed” the judicial-power provision. Morrow, 62 S.W.2d at 644. Thus, Morrow rejected authority conferred by statute precisely because it called for “the rendition of an advisory opinion” that was “not within the appellate power of our revisory courts.” Id. at 647. A dispute must remain live until final judgment to be one that fits within the judicial power. It can be entirely appropriate for the other branches of government to resolve disputes that the judiciary cannot, even in some legal contexts. Every time a law is passed, an executive order is issued, a nomination is approved or rejected, or a policy choice is otherwise enacted, a dispute of some sort is thereby resolved. But such decisions are not judicial—they would not fit within “the judicial power”—unless they actually and finally resolve a live, concrete, particularized dispute among genuinely adverse parties whose legal rights will be liquidated by a judgment. “Decisions” or “opinions” lacking those features might be useful or even urgent, but they would not be “judicial” because they would be impermissibly advisory. The Constitution’s grant of the “judicial power,” in other words, itself reflects a justiciability limit that the courts may not transgress. This Court has been so committed to this limitation that it is only because of a constitutional amendment that we may answer certified questions of law from federal courts. See United Servs. Life Ins. Co. v. Delaney, 396 S.W.2d 855, 859–64 (Tex. 1965) (explaining that Texas courts could not even adjudicate a declaratory-judgment action to supply a legal answer necessary for resolving pending-but-abated federal cases because,

15 in so doing, the Texas court could not resolve the case but could only supply an advisory opinion for another court’s use). A certified question still constitutes a request for an advisory opinion—but it is a kind, indeed the only kind, that the Constitution authorizes us to issue. In 1985, the People amended the Constitution to allow us to answer questions certified by “a federal appellate court.” Tex. Const. art. V, § 3-c(a); see Panda Power, 619 S.W.3d at 637 n.17 (describing this provision as a constitutionally authorized “exception” to the otherwise-binding prohibition of issuing “advisory opinions”). Article V, § 3-c is thus yet another constitutional provision illustrating the requirements of justiciability. If advisory opinions were already available, we would not have needed that provision, and its exacting specificity makes clear that, outside the context of federally certified questions, advisory opinions fall outside the judicial power, no matter how important or desirable. See, e.g., Panda Power, 619 S.W.3d at 641. And even in that context, we have never said that the other justiciability principles are irrelevant or have been eliminated. In short, “the judicial power” authorizes Texas courts only to render judgments that bind live parties rather than give useful but abstract legal answers. Mootness is simply one tool of complying with the constitutional directive that courts exercise only the judicial power. The “cases” requirement. Second, and like the federal Constitution, ours not only limits courts to “the judicial power” but then defines the courts’ adjudicative authority by reference to “cases.” Article III describes “cases” or “controversies,” whereas our Constitution refers to “cases” when describing the adjudicative functions authorized

16 under “the judicial power.” Both formulations play the same limiting role. Our Constitution, for example, states that this Court’s authority “shall extend to all cases except in criminal law matters and as otherwise provided in this Constitution or by law.” Tex. Const. art. V, § 3(a) (emphasis added). It repeatedly describes the other courts’ authority in terms of deciding “cases” as well. See id. art. V, §§ 4(b), 5(a), (b) (court of criminal appeals); id. art. V, §§ 5(b), 6(a) (courts of appeals); id. art. V, §§ 7(d), 8 (district courts); id. art. V, § 11 (describing recusal of judges based on varying relationships between a judge and a “case”). No less than the federal Constitution, therefore, the Texas Constitution repeatedly invokes the textual concept of “cases” as the judiciary’s proper domain. “Cases” provide the form, in other words, in which the adjudicative power of the Texas courts may operate. The concept of a “case” goes hand-in-hand with the concept of “the judicial power.” For in a “case,” the court “hear[s] the facts,” “decide[s] the issues of fact made by the pleadings,” “decide[s] the questions of law involved, and possess[es] the power to enter a judgment on the facts found in accordance with the law as determined by the court.” Delaney, 396 S.W.2d at 861; accord, e.g., Holmes v. Morales, 924 S.W.2d 920, 923 (Tex. 1996). And that “determination once made, in [that] particular case,” cannot be reversed by, say, the legislature, which only “may prescribe a new rule for future cases.” Plaut, 514 U.S. at 222 (emphasis added) (quoting The Federalist No. 81, at 545 (Alexander Hamilton) (J. Cooke ed. 1961)). This Court has treated our “cases” provision as identical to the federal “case or controversy” provision to make the point: “A judicial decision reached without a case or controversy is an advisory opinion,

17 which is barred by the separation of powers provision of the Texas Constitution.” Brooks v. Northglen Ass’n, 141 S.W.3d 158, 164 (Tex. 2004) (citing Tex. Const. art. II, § 1); see also Sw. Elec. Power Co. v. Grant, 73 S.W.3d 211, 222–23 (Tex. 2002); Tex. Ass’n of Bus. v. Tex. Air Control Bd., 852 S.W.2d 440, 444 (Tex. 1993). The textual reference to “cases” bolsters the core point that advisory opinions are beyond the judicial sphere. A “case” must be understood as one within the judicial power, and as Brooks put it, an advisory opinion in contravention of the separation of powers would result if the courts resolved a dispute that was not properly deemed a “case.” 141 S.W.3d at 164. The separation-of-powers clause. Third, “[l]ike the United States and our sister states, ours is a tripartite system of government” with three branches: legislative, executive, and judicial. Webster, 704 S.W.3d at 487 (first quoting Tex. Const. art. II, § 1; and then citing id. arts. III–V). But unlike the federal Constitution, ours goes beyond structure and implication by expressly commanding the separation of powers. Beginning in 1845, each Constitution of the State of Texas has provided that “[t]he powers of the Government of the State of Texas shall be divided into three distinct departments,” granting powers that “are Legislative to one” department, “those which are Executive to another, and those which are Judicial to another,” and then forbidding those departments from “exercis[ing] any power properly attached to either of the others, except in the instances herein expressly permitted.” Tex. Const. art. II, § 1; see also Webster, 704 S.W.3d at 487. Article V requires courts to exercise only “the judicial power” and

18 to do so by resolving only “cases,” while Article II requires that only courts may exercise that power. “[T]he lines which separate the powers of the three great departments of our government are not always clearly drawn,” of course, but “we find no difficulty in concluding that no power is more properly or certainly attached to the judicial department than that which determines controverted rights to property by means of binding judgments.” Bd. of Water Eng’rs v. McKnight, 229 S.W. 301, 304 (Tex. 1921). The separation of powers thus protects the judicial role from incursion, but at the same time, it imposes a “limit on courts’ jurisdiction.” Tex. Ass’n of Bus., 852 S.W.2d at 444 (emphasis added). To prevent improper judicial intrusion into non-judicial terrain, “we have construed our separation of powers article to prohibit courts from issuing advisory opinions because such is the function of the executive rather than the judicial department.” Id. We have frequently noted that the separation-of-powers clause limits courts to resolving live disputes, which advisory opinions by definition cannot do. “The constitutional roots of justiciability doctrines such as ripeness, as well as standing and mootness, lie in the prohibition on advisory opinions, which in turn stems from the separation of powers doctrine.” Patterson v. Planned Parenthood of Hous. & Se. Tex., Inc., 971 S.W.2d 439, 442 (Tex. 1998) (citing Tex. Const. art. II, § 1); see NCAA v. Jones, 1 S.W.3d 83, 86 (Tex. 1999) (“Appellate courts are prohibited from deciding moot controversies. This prohibition is rooted in the separation of powers doctrine in the Texas and United States Constitutions that prohibits courts from rendering advisory opinions.” (internal citation omitted)); see also, e.g., In re Guardianship of Fairley, 650 S.W.3d 372,

19 379 (Tex. 2022); Valley Baptist Med. Ctr. v. Gonzalez, 33 S.W.3d 821, 822 (Tex. 2000). The U.S. Supreme Court has also repeatedly linked justiciability to the separation of powers. See, e.g., Powell v. McCormack, 395 U.S. 486, 512 (1969) (“[T]he doctrine of separation of powers is more properly considered in determining whether the case is ‘justiciable.’ ”); Allen v. Wright, 468 U.S. 737, 752 (1984) (“[T]he law of Art. III standing is built on a single basic idea—the idea of separation of powers.”). Notably, we have recognized that the separation of powers, and its concomitant justiciability limitation, is an innovation that distinguishes American government from our English roots: “In England chancery courts exercise nonjudicial, as well as judicial, powers; but our equity courts possess only judicial powers.” Ex parte Hughes, 129 S.W.2d 270, 273 (Tex. 1939); accord Allred v. Beggs, 84 S.W.2d 223, 228 (Tex. 1935). The English courts were the king’s courts, after all. If our courts were mere adjuncts of the other branches, rather than a purposefully independent branch, it would not matter much whether our courts ventured into non-judicial territory. But the People of Texas have instead delineated judicial authority with precision, both to protect the independence and accountability of the judiciary and to ensure that the other branches remain independent and accountable for their own actions. Resolving live rather than theoretical disputes is a justiciability limit that ensures that all branches properly exercise their own, and only their own, authority. The legal-advice provision. Fourth, and again like the federal Constitution, the Texas Constitution punctuates the requirement for live and concrete decisions to qualify as justiciable by expressly assigning

20 advisory opinions elsewhere: “The Attorney General shall . . . give legal advice in writing to the Governor and other executive officers, when requested by them.” Tex. Const. art. IV, § 22. The federal Constitution more broadly provides that “[t]he President . . . may require the opinion, in writing, of the principal officer in each of the executive departments, upon any subject relating to the duties of their respective offices.” U.S. Const. art. II, § 2, cl. 1. If anything, the Texas provision is more focused on the proper authority to give advisory legal opinions, but both this Court and the U.S. Supreme Court have recognized the respective provisions as reflecting an important justiciability limitation. Our decision in Morrow relied heavily on this textually expressed delegation of legitimate advisory- opinion authority to explain why the judiciary lacked such authority under the original public meaning of our Constitution: “the Attorney General, a member of the Executive Department, is the only state officer expressly authorized to render such opinions.” 62 S.W.2d at 644 (emphasis added). Only after making that textual point did Morrow emphasize that, wholly aside from Article IV, § 22, “the rendition of advisory opinions has generally been held not to be the exercise of judicial power.” Id. (citing a wide variety of scholarship and judicial authorities). We have reaffirmed these holdings. See, e.g., Cal. Prods., Inc. v. Puretex Lemon Juice, Inc., 334 S.W.2d 780, 782–83 (Tex. 1960) (rejecting authority to issue “merely advisory opinions” via declaratory judgments because “[i]n government this is a duty of the executive branch” and because “[i]n private business it is the function of the legal profession”). And nearly from the beginning of its history, the U.S. Supreme

21 Court has invoked the analogous provision from Article II of the federal Constitution to demarcate judicial authority to only live disputes. The courts, it said, could not render advisory opinions because “the Power given by the Constitution to the President of calling on the Heads of Departments for opinions, seems to have been purposely as well as expressly limited to executive Departments.” 6 Documentary History of the Supreme Court of the United States, 1789–1800, at 755 (M. Marcus ed., 1998) (reprinting letter dated Aug. 8, 1783, from the justices to President Washington). In that instance, Secretary of State Thomas Jefferson, on behalf of President George Washington, turned to the Supreme Court with a list of serious questions about the United States’ legal obligations. It was a dangerous time for the fledgling republic—in the midst of a war ranging between Britain and France—and the president “would . . . be much relieved” to receive answers to those questions, which “would secure us against errors dangerous to the peace of the [United States].” Id. at 747 (reprinting letter dated July 18, 1793, from Thomas Jefferson to the justices). If the Supreme Court declined to provide urgently requested assistance to George Washington because only live disputes between parties presented justiciable cases and because advisory opinions were purposefully assigned to other branches, it is hard to imagine the kind of urgency that would warrant this Court yielding to the temptation to do so today. The open-courts provision. Fifth, and unlike the federal Constitution, our Bill of Rights contains an open-courts provision: “All courts shall be open, and every person for an injury done him . . . , shall have remedy by due course of law.” Tex. Const. art. I, § 13. Justiciability

22 limits flow in part from “the open courts provision, which contemplates access to the courts only for those litigants suffering an injury.” Tex. Ass’n of Bus., 852 S.W.2d at 444 (emphases added). We have had many occasions to confirm that the “access” that “the open-courts provision guarantees” is for “those who have suffered actual injury, not to provide a forum for general injuries or hypothetical complaints.” City of Houston v. Williams, 353 S.W.3d 128, 145 (Tex. 2011); see also, e.g., Data Foundry, Inc. v. City of Austin, 620 S.W.3d 692, 700 (Tex. 2021); In re Abbott, 601 S.W.3d 802, 807 (Tex. 2020); Garcia v. City of Willis, 593 S.W.3d 201, 206–07 (Tex. 2019); Heckman v. Williamson County, 369 S.W.3d 137, 147 (Tex. 2012); S. Tex. Water Auth. v. Lomas, 223 S.W.3d 304, 307 (Tex. 2007). As Justice Gonzalez put it in a concurring opinion quoting Texas Association of Business, “[w]e held that [the open-courts] provision ‘contemplates access to the courts only for those litigants suffering an injury.’ This provision, which authorizes the courts to remedy injuries . . . , implicitly defines the bounds of potentially justiciable issues.” Patterson, 971 S.W.2d at 445 (Gonzalez, J., concurring) (quoting 852 S.W.2d at 444). An injury that has not yet ripened, or that is not concrete or genuine, or that has ceased to be remediable, is not an injury that a litigant “is suffering.” We therefore reaffirm that the open-courts provision expects courts to act as courts—to remedy actual injuries without fear or favor, but not to address speculative or theoretical disputes or disputes that, while once live, no longer are. * * * These textual provisions of the Texas Constitution provide the roots of our justiciability doctrines. The fact that our Constitution

23 contains every textual and structural justiciability provision found in the U.S. Constitution—and then adds even more through our separation-of- powers and open-courts clauses—explains why, in case after case, this Court has found federal doctrine instructive on justiciability’s minimum requirements. See, e.g., Heckman, 369 S.W.3d at 154; Brown v. Todd, 53 S.W.3d 297, 305 (Tex. 2001); Tex. Ass’n of Bus., 852 S.W.2d at 444. True, we are not in any sense bound by the U.S. Supreme Court’s conception of the federal judicial power. See, e.g., Perry v. Del Rio, 66 S.W.3d 239, 250 (Tex. 2001) (finding federal principles to be “consistent with our own ripeness jurisprudence”). Article III directly governs only the federal courts, not those of Texas. But given its additional textual and structural limitations, any justiciability differences under our Constitution are likely to be more restrictive, not less. Reserving the demarcation of any such differences for future cases, we find it enough to note here that the law-articulating work of both the federal and Texas courts is available only if a judgment would redress a genuine injury. Stated differently, but stated in our cases very frequently, courts must refrain from issuing advisory opinions—not because such opinions would be useless, but because issuing them would not resolve a live dispute between actually adverse parties and therefore would not constitute an exercise of the judicial power. The justiciability doctrines help ensure that a case will not lead the judiciary into forbidden terrain. Accordingly, we reaffirm that mootness, one of the core justiciability doctrines, is rooted in the Texas Constitution. When a case “becomes moot, and the issues no longer justiciable,” the case “should be dismissed.” Sterling v. Ferguson, 53 S.W.2d 753, 760 (Tex. 1932). As we

24 have emphasized, “our lack of jurisdiction over moot cases is a mandate of the constitution, not a matter of convenience.” Panda Power, 619 S.W.3d at 641. We have repeatedly described mootness in mandatory constitutional terms, as a constitutional rather than merely prudential or administrative limitation on the judicial power. See, e.g., State ex rel. Best v. Harper, 562 S.W.3d 1, 6 (Tex. 2018); Matthews v. Kountze ISD, 484 S.W.3d 416, 418 (Tex. 2016). Thus, “[a]ny ruling on the merits of a moot issue constitutes an advisory opinion, which we lack jurisdiction to issue.” In re J.J.R.S., 627 S.W.3d 211, 225 (Tex. 2021) (emphasis added). With this confirmation of justiciability’s constitutional minimum— a live dispute whose resolution will not generate an advisory opinion— we proceed to examine whether the case before us is moot. B Assessing mootness generally proceeds in two steps. First, we ask if the case is moot on its face—that is, has the live controversy come to an end. See Joachim, 315 S.W.3d at 865. If the answer is yes, we then ask if any “exception” to mootness applies. See Williams v. Lara, 52 S.W.3d 171, 184 (Tex. 2001). Given that the Constitution requires a live dispute between genuinely adverse parties, see Joachim, 315 S.W.3d at 865, this second step does not mean that the court may adjudicate a case lacking those features. It instead means that some cases that are moot on their face actually remain live—such as when collateral consequences flow from the resolution of a seemingly moot dispute, see, e.g., Carrillo v. State, 480 S.W.2d 612, 616–18 (Tex. 1972), or when circumstances causing the injury, despite having ended, are likely to recur as between the same parties yet evade review because the injury is of short duration, see, e.g.,

25 Blum v. Lanier, 997 S.W.2d 259, 264 (Tex. 1999) (applying the capable- of-repetition exception). Determining whether the court of appeals properly exercised jurisdiction in this case requires us to examine both steps because Grassroots contends that the case is not moot on its face but that if it is, the capable-of-repetition exception saves it. But we also examine a third step that the court of appeals added to the analysis: whether a “public- interest exception” saves an otherwise moot dispute. Grassroots argues that if we must reach the public-interest exception at all, we should adopt it and then affirm the court of appeals’ judgment on the merits. We address all three steps in turn. 1 The court of appeals held that this case became moot because, as is undisputed, every plaintiff has long since been released from detention. See 665 S.W.3d at 141. Grassroots disputes this conclusion on the ground that the federal government may re-detain the plaintiffs and could do so at Dilley or Karnes. The court of appeals rejected that basis for avoiding mootness as unsupported by any record evidence about “the circumstances or conditions, if any, of the detainees’ release.” Id. We agree with the court of appeals. The legal claim here is that the department promulgated its rule in violation of the Administrative Procedure Act. To assess a claim’s justiciability, a court must begin with a clear-eyed view of how the claim specifically affects the rights and interests of the parties themselves. To anyone not held or imminently subject to being held within a relevant facility, the procedural regularity of a rule leading to licensure raises only

26 a theoretical question. When we last addressed this case, we held that these plaintiffs’ standing was secure because the plaintiff mothers could allege specific and concrete injuries traceable to the rule. Grassroots, 646 S.W.3d at 820–21. The same kind of specificity is now essential to show why the legal issue has not become abstract and theoretical because of the mothers’ release. “Justiciability is a matter of concern in every civil case, and remains a live concern from the first filing through the final judgment.” Heckman, 369 S.W.3d at 147. One way mootness arises is if a ruling cannot “affect the parties’ rights or interests” such that it “would be without practical effect.” Panda Power, 619 S.W.3d at 639 (emphasis added). If the prospective relief that plaintiffs demand could not affect their interests in a non-speculative way, resolving the case would generate an advisory opinion. “[I]t is axiomatic that appellate courts do not decide cases in which no controversy exists between the parties.” Camarena v. Tex. Emp. Comm’n, 754 S.W.2d 149, 151 (Tex. 1988). To decide whether the case is moot, we ask whether Grassroots (or, more precisely, one of the plaintiffs it represents) retains, as the U.S. Supreme Court has put it, “a ‘personal stake in the outcome of the lawsuit,’ ” or, in other words, a “concrete interest, however small.” Campbell-Ewald Co. v. Gomez, 577 U.S. 153, 161 (2016) (first quoting Genesis Healthcare Corp. v. Symczyk, 569 U.S. 66, 72 (2013); and then quoting Chafin v. Chafin, 568 U.S. 165, 172 (2013)). Our statement in Panda Power—that relief must have a “practical” effect, 619 S.W.3d at 639—reflects the need for a judgment to actually affect the plaintiff, not merely vindicate a favored legal position. Another way our cases have

27 articulated the same point is that throughout litigation, there must remain a “real and substantial controversy involving genuine conflict of tangible interests and not merely a theoretical dispute.” Bonham State Bank, 907 S.W.2d at 467. To be clear, “substantial” does not mean “a big part of the case” or “important to the law.” A claim does not become moot if only a small part of the original amount in controversy remains disputed or if a judge deems the legal rights that remain disputed to be relatively insignificant compared to those asserted at a lawsuit’s inception. Rather, a “real and substantial controversy” is one where the dispute is genuine, concrete, and tangible rather than speculative, contingent, or hypothetical. Accordingly, if a concrete dispute has become only theoretical, such as when a party seeking prospective relief no longer would be affected if that relief is granted, the claim is presumptively moot. That is why the court of appeals was correct to conclude that the undisputed evidence of these plaintiffs’ release from detention, which ended the injury essential to justiciability, rendered their claims moot on their face absent some non-speculative basis to show that the dispute nonetheless remains live. 665 S.W.3d at 141. The injury that would support jurisdiction to assess the rule’s validity cannot merely be the risk of re-detention or even the possibility of re-detention at Dilley or Karnes. Rather, the relevant injury to the plaintiff-mothers is re-detention in an unlawfully licensed facility with minor children and for an amount of time that would violate the federal consent decree. Mootness cannot be defeated without showing that all this is reasonably likely rather than speculative. Grassroots resists this conclusion and asserts that its claims are

28 not moot even on their face. Its argument comes in two parts. First, because mootness is not easily established, it contends that courts should not find mootness until the party asserting it (here, the department) proves it. Second, Grassroots argues that, even if it must show why the case is not moot on its face, the federal government’s statutory authority to re-detain these plaintiffs and Grassroots’s expert’s deposition testimony achieve that goal. We address both points in turn. As to the first one, Grassroots unduly expands the circumstances in which mootness will not be found. To the extent the department has a “burden” to establish mootness, pointing out that the plaintiff mothers have all been released met that burden. As to its second point, nothing Grassroots identifies constitutes a non-speculative basis to think that any plaintiff in this case is reasonably likely to be re-detained at Dilley or Karnes (much less with children and for a time that would exceed the consent decree’s limits). a According to Grassroots, the plaintiff mothers are subject to re-detention at Dilley or Karnes and so their claims never became moot. Grassroots asserts that the department must affirmatively establish mootness and that it cannot do so in light of these circumstances. Grassroots supports these assertions by emphasizing our oft- repeated admonition that mootness should be found reluctantly rather than readily. As we put it last summer: “[M]ootness is difficult to establish. The party asserting it must prove that intervening events make it ‘impossible for a court to grant any effectual relief whatever to the prevailing party.’ ” In re Dallas County, 697 S.W.3d 142, 151 (Tex.

29 2024) (quoting Abbott v. Mexican Am. Legis. Caucus, 647 S.W.3d 681, 689 (Tex. 2022)). Grassroots says that it is not “impossible” that any of the plaintiff mothers might be re-detained at Dilley or Karnes with minor children and for a lengthy period, so the department cannot meet its burden to establish mootness. Grassroots is at least correct that there is no presumption in favor of mootness. Jurisdiction once established should not lightly be negated. But the assessment of whether it is “impossible” is premised on the antecedent requirement that such relief must be “effectual.” That requirement cannot be speculative or contingent because it is key to the constitutional mandate that a court may only resolve a dispute that remains genuine and live, not one that has become theoretical or abstract. Once it is established that the relief would be effectual if granted—that it would resolve a live dispute and affect the plaintiff’s rights—then we will refuse to find a case moot unless it is “impossible” for the court to award that kind of relief. Grassroots’s error, in other words, is to expand this principle to circumstances where it is doubtful that there is any actual dispute to resolve in the first place. The cases that Grassroots cites, and others, make this point. In Mexican American Legislative Caucus, we held that an electoral- redistricting claim that was unquestionably live when we issued our opinion was not moot even though it would likely be overtaken by events before the litigation’s end. See 647 S.W.3d at 689–90. Even more recently, when pandemic-era governmental authority was the subject of a case in this Court, the legislature enacted a statute that eliminated such authority. See Abbott v. Harris County, 672 S.W.3d 1, 7 n.19 (Tex. 2023).

30 We did not dismiss the case as moot because the statute, despite having been enacted, would not take effect until several months after we decided the case. Id. The dispute was live until that supervening event occurred. Id. Thus, it was not “impossible” to resolve a live dispute in those cases— the stakes in each, although diminishing, remained present. As we observed in Dallas County, “[c]ourts do not act in anticipation of potential mootness,” and “only after a case becomes moot does a court lose jurisdiction.” 697 S.W.3d at 151. The Supreme Court’s decision in Campbell-Ewald is consistent with this premise because it required a “concrete” interest for the plaintiff—perhaps “small,” but not attenuated, doubtful, speculative, hypothetical, contingent, or theoretical. See 577 U.S. at 161. We added further definition to the contours of when a court’s ability to grant effectual relief is too speculative or contingent in Harper, where we reiterated that a case was not moot because it was not “impossible for the court to grant the relief requested or otherwise ‘affect the parties’ rights or interests.’ ” 562 S.W.3d at 6 (emphasis added) (quoting Heckman, 369 S.W.3d at 162). But we highlighted two practical ways that a moot issue does not necessarily moot the case. For one, “a case ‘is not rendered moot simply because some of the issues become moot during the appellate process’ ”; rather, “[i]f only some claims or issues become moot, the case remains ‘live,’ at least as to other claims or issues that are not moot.” Id. (emphasis added) (quoting In re Kellogg Brown & Root, Inc., 166 S.W.3d 732, 737 (Tex. 2005)); see also, e.g., Heckman, 369 S.W.3d at 167 (“[I]f even one issue remains live between defendants and the putative class, the suit as a whole is not moot.”); Commonwealth Bank

31 & Tr. Co. v. Morris, 77 S.W.2d 871, 871 (Tex. 1934) (reviewing a challenge to a statute and explaining that “since this question [was] the sole one certified in this case, the cause as it exist[ed] in this court [was] moot”). Courts may continue to adjudicate any parts of a case that remain within their subject-matter jurisdiction and are otherwise justiciable. We did so just this term, where we concluded that a challenge to a divorce decree was not moot as it usually is following the death of one of the spouses “because whether the marriage ended by divorce or by death substantially affect[ed] the wife’s asserted property interests,” thus providing a basis to resolve the legality of the divorce. In re Marriage of Benavides, ___ S.W.3d ___, 2025 WL 1197404, at *1 (Tex. Apr. 25, 2025). A second way arises when “in some cases—but not all—a claim for attorney’s fees ‘breathes life’ into a suit that has become moot in all other respects.” Harper, 562 S.W.3d at 7 (citation omitted). In other words, the issue that animated a lawsuit may become moot, but if a statute entitles a litigant to fees based on the merit of the claim, what seems like a moot case may remain live. See, e.g., Marshall v. Hous. Auth. of City of San Antonio, 198 S.W.3d 782, 790 (Tex. 2006) (“[I]n some instances a case is not moot even though the only issue presented relates to court costs.”). Beyond Harper’s two examples, our cases identify “voluntary cessation” as another way that a case is not moot on its face despite the seeming end of the underlying injury. See, e.g., In re Cont. Freighters, Inc., 646 S.W.3d 810, 812–14 (Tex. 2022) (holding that the mandamus petition remained live when plaintiffs unilaterally withdrew discovery demands after the Court indicated interest in reviewing the petition because the withdrawal lacked any enforceable guarantee that similar

32 demands could not be made again). Voluntary cessation typically is not a basis for mootness because it often represents not a defendant’s surrender but its attempt to avoid a binding loss. If voluntary cessation required dismissal, a defendant unilaterally “could control the jurisdiction of courts with protestations of repentance and reform, while remaining free to return to their old ways.” Matthews, 484 S.W.3d at 418. Importantly, however, as both Contract Freighters and Matthews made clear, “voluntary cessation” can lead to mootness “when subsequent events make absolutely clear that the [challenged conduct] could not reasonably be expected to recur.” Id. (internal quotation marks and citation omitted); accord Cont. Freighters, 646 S.W.3d at 814. In such a situation, the dispute necessarily has ended, and no live claim remains. A good example of this principle, and of the need for non- speculative showings of continuing injury, is the U.S. Supreme Court’s decision in DeFunis v. Odegaard, 416 U.S. 312 (1974). DeFunis alleged that a state law school denied him admission because of his race. Id. at 314. He was nonetheless admitted provisionally after obtaining an injunction from a state court, but the state supreme court later reversed the injunction. Id. at 314–15. He was not expelled because the Supreme Court stayed the state supreme court’s judgment and set the case for argument. Id. at 315. When the case was argued, DeFunis was in his last term. Id. at 315–16. The law school represented that whether it won or lost the appeal, it would allow DeFunis to complete that term and graduate; graduation, of course, would eliminate the injury of being wrongly denied admission. Id. at 316. The Court dismissed the case as moot, reasoning that even if the challenged conduct (denial of admission)

33 had in a sense voluntarily ceased, the school’s representation satisfied the principle that it was not reasonably likely to recur as to DeFunis. Id. at 316–20. The kind of “voluntary cessation” that would not lead to mootness, the Court observed, would have existed if the law school had simply (and not irrevocably) changed its admission procedures, leaving it free upon dismissal of the case to restore those procedures and eject DeFunis. Id. at 318. DeFunis reflects that at bottom, mootness poses a practical test, not one that turns on speculative, theoretical, contingent, or unlikely events that might happen. After all, a ruling for DeFunis would have given him some additional protections. Justice Brennan’s dissent pointed out that “[a]ny number of unexpected events—illness, economic necessity, even academic failure—might prevent his graduation at the end of the term.” Id. at 348 (Brennan, J., dissenting). The case was not moot, Justice Brennan argued, because the law school did not guarantee DeFunis any terms after this one—so reversal “could insure that, if for some reason [DeFunis] did not graduate this spring, he would be entitled to re- enrollment at a later time on the same basis as others who have not faced the hurdle of the University’s allegedly unlawful admissions policy.” Id. Justice Brennan’s approach accords with the mootness theory that Grassroots advances, but this Court’s cases share the DeFunis majority’s functional understanding of mootness. Hence our conclusion in Panda Power that the case was moot because a judgment would have no “practical effect,” despite the still-raging legal dispute and the strong desire for an advisory opinion. See 619 S.W.3d at 639. Time and again, we have taken this approach. In Texas A & M

34 University–Kingsville v. Yarbrough, a professor sued over an allegedly improper negative evaluation that she thought would harm her tenure prospects—but then was granted tenure. 347 S.W.3d 289, 291 (Tex. 2011). She claimed that the dispute remained live because the evaluation would remain in her file and, under the university’s rules, could affect “future employment decisions” about her. Id. We found her claim to be moot on its face because it “d[id] not present a substantial controversy, between parties having adverse legal interests, of sufficient immediacy and reality to warrant the issuance of a declaratory judgment.” Id. (first emphasis added) (internal quotation marks and citations omitted). It is not “impossible,” of course, that the negative evaluation could ever cause later harm, but it was entirely speculative. See id. Similarly, in Glassdoor, Inc. v. Andra Group, LP, we found facial mootness where Andra hoped to identify and then sue anonymous online commenters through a Rule 202 proceeding against Glassdoor, the online forum where the allegedly defamatory statements had been posted. 575 S.W.3d 523, 525 (Tex. 2019). But the limitations period for any defamation claim expired during the Rule 202 litigation, leading us to conclude that the Rule 202 litigation had been rendered moot. Id. at 527, 530. True, we agreed, it was not literally impossible for the Rule 202 proceeding to benefit Andra, for “the statute of limitations is an affirmative defense,” id. at 527 n.3, that any defendants identified through the Rule 202 proceeding could have waived or forfeited. That theoretical possibility, however, was too remote to satisfy our mootness standards. In short, as with the law of standing, our approach to mootness insists on a close tether to reality and rejects indulging “an ingenious

35 academic exercise in the conceivable.” United States v. SCRAP, 412 U.S. 669, 688 (1973). Yes, it is conceivable that one of the previous detainees may be re-detained, with children, at Dilley or Karnes, and for the requisite amount of time. But under our cases, that mere possibility is not enough, without more, to overcome facial mootness. The “impossible to grant relief” test does not refer to metaphysical impossibility, as if we were inviting lawyers to devise the equivalent of a Rube Goldberg machine for justiciability—some theoretical but highly unlikely path to converting what otherwise would be an advisory opinion into a genuinely effectual judgment. “[O]ne can never be certain that findings made in a decision concluding one lawsuit will not some day . . . control the outcome of another suit. But if that were enough to avoid mootness, no case would ever be moot.” United States v. Juv. Male, 564 U.S. 932, 937 (2011) (per curiam) (quoting CFTC v. Bd. of Trade of Chi., 701 F.2d 653, 656 (7th Cir. 1983) (Posner, J.)). Accordingly, the department carried any burden it bore to show that all the claims in this case are moot on their face. Without dispute, it conveyed that none of the individual plaintiffs remained detained in an unlicensed facility—not because of any actions of the department but because of third-party decisions of the federal government. Likewise, any return to detention would require the federal government’s independent decisions. To avoid dismissal, Grassroots must identify a non-speculative basis for concluding that the detainees’ release did not facially moot the claims. We now turn to the arguments Grassroots has raised on that point.

36 b Grassroots disputes any burden to negate mootness, but it identifies two grounds for why the claims are not moot, even on their face. First, the plaintiff mothers are subject to being re-detained at the discretion of the Attorney General of the United States. See, e.g., 8 U.S.C. § 1226(b) (“The Attorney General at any time may revoke a bond or parole[,] . . . rearrest the alien under the original warrant, and detain the alien.”). Second, Grassroots claims that its expert’s deposition testimony shows that the re-detention authority is real. Thus, judicial relief is necessary to prevent re-detention at an unlicensed facility. We find these contentions insufficient to overcome mootness and address them in turn. First, we assume that the statute’s scope is as broad as Grassroots asserts. If we were adjudicating the plaintiffs’ right to be in the United States free from any threat of removal or detention, the existence of that statutory authority might well prevent conditional release from mooting the claims. Grassroots cites several federal cases involving direct federal claims under federal immigration law that suggest that result. See Clark v. Martinez, 543 U.S. 371, 376 n.3 (2005) (reasoning that because the parolee’s release was “subject to the [Secretary of Homeland Security’s] discretionary authority to terminate” parole, he “ ‘continue[d] to have a personal stake in the outcome’ of his petition” (citation omitted)); Rosales- Garcia v. Holland, 322 F.3d 386, 395–96 (6th Cir. 2003) (en banc) (reasoning that because Rosales’s “immigration parole can be revoked by INS at any time for almost any reason,” his “appeal [was] not moot”); Rodriguez v. Hayes, 591 F.3d 1105, 1117–18 (9th Cir. 2010) (relying on Clark to deem the claim not moot); In re Hutto Fam. Det. Ctr., No. A-07-

37 CA-164-SS, 2007 WL 9757682, at *1 (W.D. Tex. May 29, 2007) (same). But even assuming that a conditional release would not moot a challenge to one’s immigration status under federal law, whether that release has mooted the state-law rule challenge here raises a quite different question. For one thing, the relief in the two circumstances is materially different. A successful plaintiff in the federal cases would obtain a judicial determination against the federal government concerning her immigration status, including whether, for example, she could be removed or detained at all. Even if a plaintiff is never again to be detained, determining the lawfulness of her presence in the country itself may itself have consequences as to her ability to obtain employment or housing and could thus provide the basis for her claim to remain live (assuming that the federal government remains adverse). Grassroots, of course, does not ask for any immigration-related relief here. The premise of this case is that the plaintiffs can be subjected to continued detention, yet that, for distinct reasons, particular kinds of detention require a valid state license. All agree, as Grassroots puts it, that “[t]he propriety of the rule [the department] issued is unquestionably and solely a matter of state law” under our Administrative Procedure Act. For the state-law question to remain live, therefore, the continuing threat of federal enforcement is a necessary but not sufficient requirement. It is not sufficient because, unlike obtaining some form of federal relief, invalidating the department’s rule requires the nonspeculative continuing threat not just of some adverse federal action, but all four of these specific events: a governmental choice to re-detain any of these plaintiffs; re- detention at Dilley or Karnes; re-detention with a minor child; and re-

38 detention for a duration that exceeds what the consent decree allows. The statutory re-detention authority that Grassroots cites, therefore, cannot prevent mootness of this far-more-attenuated claim unless it is accompanied by evidence suggesting more than a speculative likelihood of those four developments for the particular plaintiffs in this case. Otherwise, the federal government’s statutory authority standing alone is comparable to a bare theoretical possibility that a decision might eventually prove useful in cases like DeFunis, Yarbrough, or Glassdoor— hypothetically possible, but insufficient to overcome facial mootness. Thus, while we assume that each of the four steps is possible, mere possibility is not enough. “[U]nder our Constitution,” the judiciary may “not give advice or decide cases upon speculative, hypothetical, or contingent situations.” Coalson v. City Council of Victoria, 610 S.W.2d 744, 747 (Tex. 1980); accord Camarena, 754 S.W.2d at 151 (citing Coalson, 610 S.W.2d at 747). It is at least as true in the context of justiciability as elsewhere in the law that “we cannot ‘pile speculation on speculation and inference on inference.’ ” Raoger Corp. v. Myers, ___ S.W.3d ___, 2025 WL 1085173, at *4 (Tex. Apr. 11, 2025) (quoting Marathon Corp. v. Pitzner, 106 S.W.3d 724, 729 (Tex. 2003)). The claims here are therefore moot on their face unless Grassroots can show that reaching the merits will not be an abstract ruling but will directly affect these plaintiffs’ interests under the same “reasonable likelihood” or “reasonable expectation” standard typically used for justiciability inquiries. Cf., e.g., Robinson v. Parker, 353 S.W.3d 753, 755–56 (Tex. 2011) (discussing ripeness); Williams, 52 S.W.3d at 184–85 (discussing mootness). The deposition testimony of Erica Schommer, Grassroots’s expert

39 witness, comes into play here because it supplies the only other evidence on which Grassroots relies. Schommer stated that U.S. Immigration and Customs Enforcement (ICE) has broad discretion to re-detain released immigrants for any reason that it deems appropriate. But Schommer’s testimony fails to show any non-speculative, heightened, or imminent likelihood as to even one named plaintiff. Instead, the testimony achieves the opposite. As we read it, it reveals that the mothers’ re-detention is only a remote possibility rather than the reasonable likelihood that this Court’s precedent requires. See, e.g., Williams, 52 S.W.3d at 184. Indeed, in 2015 alone, some 16,000 individuals resided in Dilley, but Schommer only knew of “at least over [twenty]” families in her entire career that had been re-detained at either Dilley or Karnes, and she personally had only “a few clients” ever in that situation. Even more significantly, Schommer confirmed that her re-detained clients fit within three basic groups: those who violated their conditions of parole, those who committed a crime, or those who received an adverse decision in their federal case. The first and second grounds—the possibilities of violating conditions of parole or committing crimes— cannot overcome mootness, and it is of course improper to assume that anyone would commit those actions. See, e.g., Williams, 52 S.W.3d at 185 (observing that ex-detainees are “required by law to prevent their own recidivism”). Said another way, the notion that a former detainee would violate a condition of parole or commit a criminal offense is a speculative leap that blocks rather than establishes justiciability. See, e.g., O’Shea v. Littleton, 414 U.S. 488, 496–97 (1974) (“[A]ttempting to anticipate whether and when these respondents will be charged with crime . . .

40 takes us into the area of speculation and conjecture.”). The third-cited reason for re-detention—receiving an adverse immigration decision—is equally unlinked to any of these plaintiffs. Indeed, nothing in the testimony (or elsewhere in the record as presented to us) suggests that any of them is especially likely to receive an adverse determination or that, even if such a determination comes, it would make re-detention a likely consequence. The testimony only works the other way: that if re- detention occurs, the re-detained individual would likely have fit in one of the three categories Schommer listed. None of this remotely supports Grassroots’s contention that the case never became moot in the first place. At best, the evidence Grassroots has presented “is hypothetical, ‘iffy’ and contingent,” which amounts to no evidence at all. Burch, 442 S.W.2d at 333. The court of appeals rightly held that the record before it did nothing to show how these plaintiffs could benefit in more than a speculative way by continuing the litigation. 665 S.W.3d at 141. * * * We reiterate that “under our Constitution, [courts] do not give advice nor decide cases upon speculative, hypothetical, or contingent situations.” Coalson, 610 S.W.2d at 747. No plaintiff is currently experiencing the alleged injury here—detention in an allegedly unlawfully licensed facility—and their return to any affected facility is entirely “speculative, hypothetical, or contingent.” Nothing here prevents this case from being deemed moot on its face. Unless an “exception” applies—the question to which we next turn—the case must be dismissed without prejudice for lack of subject-matter jurisdiction. At the same time, however, we cannot and need not hold that none of the plaintiffs will ever be re-detained at either of these facilities. Our

41 case law addresses this scenario, too. In another case, which we dismissed for lack of ripeness, we wrote: “We note that the [plaintiffs] are not irrevocably harmed by this dismissal. Because the case is dismissed without prejudice, if they choose, they can re-file and develop a record demonstrating that the claims have ripened, allowing a new suit to proceed.” Waco ISD v. Gibson, 22 S.W.3d 849, 853 (Tex. 2000). The same is true for mooted disputes, of course. That a claim cannot be adjudicated because of mootness does not foreclose litigation if the injury does recur. It would no longer be speculative or attenuated, and the very fact of recurrence as to a particular plaintiff would likely make it harder to again dismiss that plaintiff’s claim as moot. 2 Grassroots argues that the claims are saved by the capable-of- repetition-yet-evading-review “exception” even if we find them moot on their face. The court of appeals rejected that argument, 665 S.W.3d at 141–42, and we again agree. To explain why, we first discuss the concept of “exceptions” to mootness and then assess whether any of them saves Grassroots’s claims from mootness. This Court has recognized two primary and generally available “exceptions” to mootness: (1) the “collateral consequences” exception and (2) the “capable of repetition” exception. See FDIC v. Nueces County, 886 S.W.2d 766, 767 (Tex. 1994). 1 The deployment of the word “exception”

1 Class actions present a separate context that is not at issue here. Without addressing all the justiciability complexities that class litigation can engender, it is sufficient to note that courts cannot adjudicate a class’s claims unless “there remains a live interest between the class of affected individuals— thereby satisfying constitutional justiciability concerns.” Heckman, 369 S.W.3d

42 can be misleading because what we have referred to as mootness “exceptions” are not really exceptions at all. They do not allow courts to disregard the boundaries of the judicial power and adjudicate cases that are actually moot. Rather, the exceptions elucidate when a case that seems moot actually remains live for reasons that might not be immediately apparent. The collateral-consequences exception allows courts to decide cases “when vacating the underlying judgment will not cure the adverse consequences suffered by the party seeking to appeal that judgment.” Marshall, 198 S.W.3d at 789. Criminal cases are frequent examples. Convicts may be subject to burdens or obligations that linger even after any sentence has ended, affecting their right to vote, possess firearms, or enjoy other freedoms generally available to those without a conviction. This Court has long recognized, for example, that “[a] juvenile’s appeal of his adjudication is not moot simply because his disposition has ended when . . . potential collateral consequences remain,” such as required sex- offender registration or the inability to seal juvenile records. In re T.V.T., 675 S.W.3d 303, 307 (Tex. 2023) (citing Carrillo, 480 S.W.2d at 617). In such cases, the parties remain adverse and have the requisite stakes to pursue the case. The government’s interest is in defending its conviction and allowing any continuing consequences to be imposed. The defendant’s interest is not merely in clearing his name but in avoiding those consequences. “In no practical sense, therefore, can [a] case be said

at 165. Notably, this case was not brought as a class action, ostensibly for legitimate reasons relating to litigation funding. We express no view of whether this case, if brought as a class, could avoid mootness; it is enough that it cannot avoid mootness in the form in which it has been brought.

43 to be moot” under those circumstances. Fiswick v. United States, 329 U.S. 211, 222 (1946). We have made the same point regarding various kinds of mental-health adjudications that entail continuing consequences. See, e.g., In re A.R.C., 685 S.W.3d 80, 83 (Tex. 2024) (“An involuntary- commitment order imposes collateral consequences under federal and state law.”); State v. Lodge, 608 S.W.2d 910, 912 (Tex. 1980) (similar). But a dispute remains live under the collateral-consequences exception only if the otherwise-moot claim itself is the source of a sufficiently concrete collateral consequence. If the consequence would exist regardless, then the case remains moot. When a juvenile defendant was required to register as a sex-offender under state law regardless of the underlying federal adjudication, no concrete interest justified maintaining the litigation: “True, a favorable decision in this case might serve as a useful precedent for respondent in a hypothetical lawsuit challenging Montana’s registration requirement on ex post facto grounds. But this possible, indirect benefit in a future lawsuit cannot save this case from mootness.” Juv. Male, 564 U.S. at 937. The capable-of-repetition exception likewise applies only in rare circumstances. Williams, 52 S.W.3d at 184. To invoke it, a plaintiff must prove that “(1) the challenged action was too short in duration to be litigated fully before the action ceased or expired; and (2) a reasonable expectation exists that the same complaining party will be subjected to the same action again.” Id. (emphasis added). In other words, the same dispute still divides the same parties despite the seeming termination of the dispute’s initial cause, thus generating a “reasonable expectation” of that cause’s recurrence. In this sense, the capable-of-repetition exception

44 dovetails with two circumstances that we have described as not leading to mootness in the first place: mere voluntary cessation and when there is a non-speculative basis for finding that adjudicating the case and granting the desired relief would give the plaintiff a tangible and concrete benefit. The capable-of-repetition exception recognizes genuine, ongoing disputes as not really moot at all, but it screens out cases where any resulting judgment would be unlikely to directly affect the actual parties’ interests. See, e.g., In re Uresti, 377 S.W.3d 696, 696 (Tex. 2012) (“Uresti has not shown a reasonable expectation that he will be subjected to the same action again.”). The central point is that both exceptions only help determine whether a case that seems moot at first glance really is—essentially addressing the same question we considered in response to Grassroots’s assertion that the case was not even facially moot. The fundamental requirement, to which there is no exception, remains unchanged: the presence of a live dispute between parties with continuing adverse interests that can be settled by a judgment that will actually affect the parties’ rights. We are not alone in recognizing that the term “exception” might be a bit of a misnomer. See, e.g., Eden, LLC v. Justice, 36 F.4th 166, 170 n.4 (4th Cir. 2022) (noting that “ ‘exception’ in th[e] [mootness] context is a colloquialism” (citing Lighthouse Fellowship Church v. Northam, 20 F.4th 157, 162 n.4 (4th Cir. 2021))). The point is important because if the “exceptions” are perceived as allowing courts to adjudicate cases that are not actually live, it would be tempting to embrace the erroneous view that mootness is a purely “prudential,” rather than a distinctly constitutional,

45 doctrine. Contra, e.g., State v. Roat, 466 P.3d 439, 446 (Kan. 2020) (concluding that “[i]f mootness were jurisdictional, we could not have such court-created exceptions”). Whatever we call them, therefore, it is important to correctly view the “exceptions” as in no way purporting to be exemptions from the constitutional mandate that we have described. With that foundation, we now turn to Grassroots’s contention that the capable-of-repetition exception saves its case from mootness. The court of appeals correctly concluded that the first prong of the exception was met because the period for detention at Dilley and Karnes is eleven days on average, a period far too short to complete litigation. 665 S.W.3d at 141. But it also correctly identified the problem with the second prong. As the U.S. Supreme Court has put it, a “mere physical or theoretical possibility” is insufficient to invoke the capable-of-repetition exception. Murphy v. Hunt, 455 U.S. 478, 482 (1982). Because we do not presume mootness, when there are circumstances like those discussed above in Part II.B.1, we will not regard a claim as moot on its face. But as Grassroots has acknowledged, we do not presume that one of the “exceptions” to mootness applies—the exception must be established. The evidence required is essentially the same as what a plaintiff must identify to rebut a defendant’s establishment of facial mootness: a basis that is not speculative or hypothetical for why continuing the litigation would tangibly and directly affect the plaintiffs’ rights. As both we and the U.S. Supreme Court have put it in the context of the capable-of-repetition exception, “there must be a ‘reasonable expectation’ or a ‘demonstrated probability’ that the same controversy will recur involving the same complaining party.” Id. (emphasis added); see also Uresti, 377 S.W.3d at

46 696 (observing that “a reasonable expectation must exist that the ‘same complaining party will be subjected to the same action again’ ” (quoting Williams, 52 S.W.3d at 184)). But to support its capable-of-repetition argument, Grassroots points to the same deposition testimony of its expert, who stated that ICE has broad discretion to re-detain released immigrants for any reason. We have already explained why that testimony cannot satisfy the required standards. See supra Part II.B.1.b. Grassroots cannot show a “reasonable likelihood” that any plaintiff in the case will be subjected to any of the steps leading to a restoration of the injury that a judgment could redress, much less all of them, and thus cannot establish the capable-of-repetition exception. 3 After concluding that Grassroots’s claims are moot and that the capable-of-repetition exception did not apply, the court of appeals then addressed “whether appellees’ second asserted exception to mootness applies: the public-interest exception.” 665 S.W.3d at 142. According to the court, this exception “expands the capable-of-repetition exception to include parties other than those involved in the current case.” Id. In one sense, it is hard to blame courts for pushing the envelope. If several “exceptions” to mootness already exist, after all, why not one more? But the public-interest exception differs from the existing exceptions not just in degree but in kind. Every recognized exception carefully ensures that the minimum requirements for constitutional justiciability are satisfied and never leads to an advisory opinion. But the public-interest exception is a true exception. It allows adjudication of

47 admittedly non-live disputes when a judge finds an issue to be of broad public importance. This “exception” does not steer clear of constitutional impediments. It hurtles toward them. If invoking it is necessary for a court to reach the merits, an advisory opinion is the guaranteed outcome. Each of the constitutional provisions that we have identified as relevant to justiciability confirms our analysis. “The judicial power” has never been understood to encompass rendering judgments that could only benefit non-parties. A “case” has always been understood to require genuine adversity by those who are party to it. The separation-of-powers clause leaves to the other branches any dispute that lacks the features of a “case” that can be adjudicated by a court. The open-courts provision allows individuals to bring suit if a judgment can remedy an actual injury or prevent one that is threatened and imminent. And the Constitution expressly leaves advisory opinions about legal issues that are not reduced to adversary litigation to the attorney general, at least until a justiciable dispute arises. Grassroots argues that a number of other states have embraced the public-interest exception and that we should adopt it too. The U.S. Supreme Court rejected a similar argument in DeFunis. 416 U.S. at 316. The Court acknowledged that the public-interest exception would save the case from mootness in Washington’s state courts. Id. But it refused to follow suit because the exception was inconsistent with the federal Constitution’s limitations on federal jurisdiction. Id. at 319–20. Neither has this Court been unaware that some other states’ judiciaries can issue various forms of advisory opinions. See, e.g., Delaney, 396 S.W.2d at 859. Each state is free to chart its own course

48 based on its own constitutional text and tradition, which may include allowing various kinds of advisory opinions, such as those generated by the public-interest exception. See, e.g., Duhon v. Gravett, 790 S.W.2d 155, 156 (Ark. 1990). After all, some states’ constitutions have, for centuries, not only expressly allowed but have compelled certain advisory opinions. Well before the U.S. Constitution’s ratification, several state constitutions directed state high courts to answer requests for advice. In Massachusetts, for example, “[e]ach branch of the Legislature, as well as the Governor and Council, shall have the authority to require the opinions of the Justices of the Supreme Judicial Court, upon important questions of law, and upon solemn occasions.” Mass. Const. of 1780, part II, ch. III, art. II. That obligation remains intact to this day. Nearly verbatim text was considered but rejected at the federal convention, see, e.g., 2 The Records of the Federal Convention of 1787, at 341 (M. Farrand ed., 1911) (“Each branch of the Legislature, as well as the Supreme Executive shall have authority to require the opinions of the supreme Judicial Court upon important questions of law, and upon solemn occasions.”). But as ratified, the Constitution leaves federal courts with only unadorned “judicial power,” which excludes advisory opinions. We do not purport to instruct our colleagues on our sister states’ high courts how to best interpret their own law. But the roots of our justiciability doctrines lie in the text of the Texas Constitution, not in the traditions, experiences, or choices of other states. We therefore reiterate that “[w]hile other jurisdictions possibly having different constitutional provisions may hold differently from our present holding,” “[i]n the

49 absence of a constitutional provision authorizing the Texas courts to render advisory opinions, such power does not exist.” Burch, 442 S.W.2d at 335. The framers and ratifiers of the Texas Constitution were students of the federal Constitution and those of other states, borrowing language from them for use in our own Constitution. Texans were no less aware than those who framed and ratified the federal Constitution that people in other states had authorized their judiciaries to sometimes issue advisory opinions. But the constitutional limitations adopted by the People of Texas are far closer to the federal model. Our Constitution contains every justiciability limit in the federal Constitution and then adds more of its own. See supra Part II.A. The People have amended the Constitution once and only once to grant us authority that did not lie within “the judicial power” by allowing us to answer questions certified by federal appellate courts—that is, to provide a kind of advisory opinion when we could not adjudicate the case itself. See Tex. Const. art. V, § 3-c. The fact that only such advisory opinions have been constitutionally authorized underscores that all other advisory opinions remain as firmly inconsistent with “the judicial power” of our State as they have ever been. Of course, should they deem it expedient, the People may again amend the Constitution to authorize advisory opinions of whatever sort and to whichever courts they choose. We note again that this Court has already at least forecast our decision today. In Morath, we explained: We do not have power to decide moot cases, whether they “involve a matter of public concern” or not. Indeed, the need for courts to mind their jurisdictional bounds is perhaps at

50 its greatest in cases involving questions of public importance, where the potential for undue interference with the other two branches of government is most acute. If courts were empowered to ignore the usual limits on their jurisdiction, such as mootness, when matters of public concern are at stake, then we would no longer have a judiciary with limited power to decide genuine cases and controversies. We would have a judiciary with unbridled power to decide any question it deems important to the public. That is not the role assigned to the courts by our constitution. 601 S.W.3d at 789 (internal citations omitted). Whether that statement was necessary for the decision in Morath or not, we think it accurately summarizes the law, and to the extent it was not already so understood, we adopt that statement from Morath as a holding. 2 A genuine controversy must exist between the parties at each stage of proceedings, including on appeal. Williams, 52 S.W.3d at 184. There can be no exception to that foundational justiciability principle. We hold that the “public-interest exception” violates the Texas Constitution’s justiciability limitations. No court in Texas may invoke that doctrine as a basis to reach the merits of a case that otherwise is not

2 Arguably, an even older authority rejects the public-interest exception

for moot cases. In General Land Office v. OXY U.S.A., Inc., the State argued that “the ‘collateral consequences’ exception is applicable because of both the public interest in resolving this important question of administrative law, and the ruling’s effect upon the numerous administrative hearings which are pending.” 789 S.W.2d 569, 572 (Tex. 1990). With some understatement, we observed that “[t]his is not the type of case which was envisioned when this exception was created.” Id. And we held that “the fact that an important question of administrative law is involved, the resolution of which would aid the agency, is not sufficient impetus for this court to render an advisory opinion.” Id. (emphasis added).

51 justiciable. We expressly disapprove any case to the extent that it has relied upon the public-interest exception or has acknowledged the public- interest exception’s validity as a method to assess subject-matter jurisdiction. 3 We express no other views about anything stated in those opinions. III A court always has jurisdiction to decide its own and the lower courts’ jurisdiction. The court of appeals concluded that this case was justiciable. This holding was erroneous because Grassroots’s claims are moot, the “capable of repetition” doctrine does not save them, and there is no “public-interest exception” that authorizes Texas courts to resolve

3 We accordingly disapprove of the jurisdictional holdings in at least

these cases: In re Guerra, 235 S.W.3d 392, 432 n.198 (Tex. App.—Corpus Christi–Edinburg 2007, pet. denied); Securtec, Inc. v. County of Gregg, 106 S.W.3d 803, 810–11 (Tex. App.—Texarkana 2003, pet. denied); Nueces County v. Whitley Trucks, Inc., 865 S.W.2d 124, 126 (Tex. App.—Corpus Christi–Edinburg 1993), dismissed sub nom. FDIC, 886 S.W.2d at 766; Buchanan, 848 S.W.2d at 304; Tex. Dep’t of Pub. Safety v. LaFleur, 32 S.W.3d 911, 914 (Tex. App.— Texarkana 2000, no pet.). We likewise disapprove of the recognition of the availability of the public- interest exception in at least the following cases: City of Georgetown v. Putnam, 646 S.W.3d 61, 73 (Tex. App.—El Paso 2022, pet. denied); Port of Corpus Christi, LP v. Port of Corpus Christi Auth. of Nueces County, No. 13-19-00378-CV, 2021 WL 2694772, at *8 (Tex. App.—Corpus Christi–Edinburg July 1, 2021, no pet.); NextEra Energy, Inc. v. PUC, No. 03-19-00425-CV, 2020 WL 4929778, at *4 (Tex. App.—Austin Aug. 21, 2020, pet. denied); Gates v. Tex. Dep’t of Fam. & Protective Servs., No. 03-15-00631-CV, 2016 WL 3521888, at *6 (Tex. App.—Austin June 23, 2016, pet. denied); Fiske v. City of Dallas, 220 S.W.3d 547, 550 (Tex. App.— Texarkana 2007, no pet.); Hatten v. Univ. Interscholastic League, No. 13-06- 00313-CV, 2007 WL 2811833, at *5 (Tex. App.—Corpus Christi–Edinburg Sept. 27, 2007, pet. denied); In re Guardianship of Keller, 171 S.W.3d 498, 501–02 (Tex. App.—Waco 2005), rev’d sub nom., Zipp v. Wuemling, 218 S.W.3d 71 (Tex. 2007); Ngo v. Ngo, 133 S.W.3d 688, 691–92 (Tex. App.—Corpus Christi– Edinburg 2003, no pet.).

52 moot cases. We therefore reverse the court of appeals’ judgment with respect to its jurisdictional conclusion. Likewise, because the court of appeals lacked authority to proceed to the merits, we vacate its judgment with respect to its decision on the merits of the challenge to the rule, and we also vacate the judgment and orders of the district court. We accordingly render a judgment of dismissal without prejudice for lack of subject-matter jurisdiction.

Evan A. Young Justice OPINION DELIVERED: May 30, 2025