PAXTON v. AMERICAN OVERSIGHT (2025)

This case raises the age-old question of the judiciary's authority to issue a writ of mandamus against an executive officer. See, e.g., Marbury v. Madison, 5 U.S. 137, 1 Cranch 137, 2 L.Ed. 60 (1803). A private, nonprofit group called American Oversight filed a petition for writ of mandamus in district court against the Governor and the Attorney General, alleging violations of the Public Information Act. As we observed thirty years ago, however, the relevant statutes and constitutional provisions do not empower district courts to issue writs of mandamus against constitutional executive officers like the Governor and the Attorney General. See A & T Consultants, Inc. v. Sharp, 904 S.W.2d 668, 672–73 (Tex. 1995). The Legislature could have provided otherwise in the three decades since Sharp, but we conclude that it has not done so. As a result, without addressing any other issues, we reverse the judgment of the court of appeals and direct the district court to dismiss the mandamus petition for lack of jurisdiction.

This result, which arises from the peculiarities of mandamus jurisdiction, does not mean that constitutional executive officers cannot be sued if they violate the Public Information Act. In addition to the PIA's criminal penalties, section 552.3215 authorizes suits for declaratory or injunctive relief “against a governmental body that violates this chapter.” Tex. Gov't Code § 552.3215(b). American Oversight could have proceeded under that provision, which requires the participation of the Travis County district attorney, but it did not. In addition to that available remedy, the Legislature could have attempted to authorize district courts to issue mandamus relief against constitutional executive officers under the PIA, but we conclude that it has not. It has, instead, limited mandamus jurisdiction over constitutional executive officers to this Court. See Tex. Gov't Code § 22.002(c). We are bound to follow that statutory directive, just as all governmental bodies in Texas, including those headed by constitutional executive officers, are bound to follow the Public Information Act.

I.

In 2022, American Oversight sent three Public Information Act requests to the Governor's office. The first sought official communications involving any non-governmental email address attributed to the Governor. The second sought all text messages sent or received by the Governor regarding official business. The third sought emails between the Governor's office and the National Rifle Association, the Texas State Rifle Association, and similar groups.

The Governor's office identified documents responsive to the first request, but it considered them exempt from disclosure for various reasons, including attorney-client privilege. It requested a ruling from the Attorney General's office, as contemplated by the PIA. See generally Tex. Gov't Code § 552.301. The AG's office determined that all the responsive documents were privileged or related to pending litigation. For the second request, the Governor's office again sought to withhold the responsive information and submitted a request for a ruling to the AG's office. The AG's office agreed that some of the responsive information must be or could be withheld, but it also ruled that some of the information must be turned over. The Governor's office provided American Oversight with roughly 100 pages of text messages in response to the second request. For the third request, the Governor's office determined that it had no responsive information.

American Oversight also sent four Public Information Act requests to the Attorney General's office. The first sought all emails sent by the Attorney General or his solicitor general between January 6 and January 8, 2021. The second sought official communications involving any non-governmental email address used by the Attorney General. The third sought all text messages sent or received by the Attorney General regarding official business. The fourth sought emails between the AG's office and the National Rifle Association, the Texas State Rifle Association, and other similar groups.

For the first and third requests, the AG's office produced some responsive documents but sought to withhold others. It requested an AG's office ruling,¹ and the ruling confirmed its authority to withhold the remaining information. For the second request, the AG's office sought to withhold all responsive information, and the resulting ruling confirmed that approach. For the fourth request, the AG's office identified no responsive information.

American Oversight was unsatisfied with both offices’ responses. It filed a petition for writ of mandamus in Travis County district court against the Governor and the Attorney General. It brought the mandamus petition under section 552.321 of the PIA, which provides that “[a] requestor or the attorney general may file suit for a writ of mandamus compelling a governmental body to make information available for public inspection if the governmental body ․ refuses to supply public information.” Id. § 552.321(a). Section 552.321 further provides that the suits authorized by subsection (a) “must be filed in a district court for the county in which the main offices of the governmental body are located.” Id. § 552.321(b).

The Governor and the Attorney General (collectively, “the State”) challenged the district court's jurisdiction. They argued that sovereign immunity bars the mandamus petition because American Oversight did not plead a viable claim of the State's refusal to provide public information. They contended they had properly followed the PIA's procedures for withholding the documents not produced, and they provided affidavits indicating they had conducted a “diligent and good faith search” for responsive information. The district court denied the State's pleas to the jurisdiction, and the State filed an interlocutory appeal. See Tex. Civ. Prac. & Rem. Code § 51.014(a)(8).

On appeal, the State re-urged its initial position but also pointed out an additional potential defect in the district court's jurisdiction. See Tex. Dep't of Transp. v. Self, 690 S.W.3d 12, 20 (Tex. 2024) (questions of subject-matter jurisdiction “cannot be waived” and “can be raised for the first time on appeal”). The State contended that the Legislature has empowered only the Supreme Court to issue writs of mandamus against constitutional executive officers, see Tex. Gov't Code § 22.002(c), and that the PIA does not authorize district courts to do so. American Oversight disagreed, arguing that section 552.321 grants district courts mandamus jurisdiction over any “governmental body,” including governmental bodies headed by constitutional executive officers.

The court of appeals affirmed. 683 S.W.3d 873, 884 (Tex. App.—Austin 2024). It held that despite the general statutory reservation of mandamus authority over constitutional executive officers to the Supreme Court, section 552.321(b) authorizes district courts to issue mandamus relief when any governmental body, including one headed by a constitutional executive officer, “refuses to supply public information.” Id. The court of appeals further held that American Oversight had alleged a viable claim that the State “refuse[d] to supply public information.” Id. at 885; see Tex. Gov't Code § 552.321(a). The State petitioned for review, and we granted the petition.

II.

A.

The Legislature has authorized PIA requestors like American Oversight to “file suit for a writ of mandamus compelling a governmental body to make information available.” Tex. Gov't Code § 552.321(a) (emphasis added). Section 552.321 could certainly have authorized a procedural vehicle other than a mandamus petition, with all of its attendant substantive and procedural quirks. For instance, the very next section authorizes suits by district and county attorneys for declaratory and injunctive relief, either on their own or at the urging of a dissatisfied requestor. Id. § 552.3215(c), (e). But rather than give requestors a unilateral right to sue for declaratory relief, injunctive relief, or some other conventional remedy available in run-of-the-mill civil litigation, the Legislature chose to authorize requestors to pursue petitions for writ of mandamus. And a common-law term like “writ of mandamus,” when transplanted into the statutory law, “brings the old soil with it.” Hall v. Hall, 584 U.S. 59, 73, 138 S.Ct. 1118, 200 L.Ed.2d 399 (2018) (internal quotation omitted). The outcome of this appeal turns on the Legislature's choice to provide mandamus relief as the sole remedy available to requestors under section 552.321(a). That choice, which we must assume was deliberate, naturally incorporates Texas law's pre-existing limitations on mandamus relief, including limitations on jurisdiction.

The history of the writ of mandamus in Texas dates (at least) to the 1836 Constitution of the Republic of Texas, which charged the new Congress with adopting the English common law, from which the writ derives. Repub. Tex. Const. of 1836, art. IV, § 13, reprinted in 1 H.P.N. Gammel, The Laws of Texas 1822–1897, at 1074 (Austin, Gammel Book Co. 1898); Act approved Jan. 20, 1840, 4th Cong., R.S., § 1, 1840 Repub. Tex. Laws 3, 4, reprinted in 2 H.P.N. Gammel, The Laws of Texas 1822–1897, at 177, 178 (Austin, Gammel Book Co. 1898) (adopting the common law of England).² The writ itself long pre-existed the Republic, of course. See Thomas Tapping, The Law and Practice of the High Prerogative Writ of Mandamus 56 (London, Wm. Benning & Co. 1848) (tracing the origins of the writ of mandamus perhaps to the fourteenth-century reign of King Edward III, but at the latest to Bagg's Case from the King's Bench in 1615).

This Court often issues the writ of mandamus—or does so provisionally—to compel a lower court to perform a duty that is “simply ministerial and involves no judicial discretion.” Lloyd v. Brinck, 35 Tex. 1, 10 (1871). But the writ's historical uses extend beyond supervision of lower courts and include the possibility that writs of mandamus may be issued against officials in other branches of government, in appropriate cases where the issuing court is lawfully authorized to do so. The writ is an order from a court “whereby a person or officer is required to do something which he wrongfully declines to do,” and “in exceptional cases it may properly be given a restraining effect” to “revers[e] or amend[ ] ․ a previous act.” Seagraves v. Green, 116 Tex. 220, 288 S.W. 417, 425 (Tex. [Comm'n Op.] 1926); cf. Lewright v. Bell, 94 Tex. 556, 63 S.W. 623, 623 (Tex. 1901) (“The courts cannot, by the writ in question, compel an officer to perform an official duty, where that duty involves ․ discretion on his part.”).

In mandamus actions against non-judicial officials, we have stated that “[m]andamus may issue to compel public officials to perform ministerial acts, as well as ‘to correct a clear abuse of discretion by a public official.’ ” In re Williams, 470 S.W.3d 819, 821 (Tex. 2015) (quoting Anderson v. City of Seven Points, 806 S.W.2d 791, 793 (Tex. 1991)). “An act is ministerial when the law clearly spells out the duty to be performed by the official with sufficient certainty that nothing is left to the exercise of discretion.” Id. (quoting same).

With these general observations about the nature of the writ of mandamus in mind, we turn to the question of the power of the district court in this case, under our Constitution and statutes, to issue the writ requested by American Oversight.

B.

The question is whether section 552.321 of the Government Code authorizes the district court to issue a writ of mandamus against two constitutional executive officers, the Governor and the Attorney General. To answer that question, we must first understand the constitutional and statutory landscape in which section 552.321 operates.

District court jurisdiction is generally defined by Article V, section 8 of the Constitution:

District Court jurisdiction consists of exclusive, appellate, and original jurisdiction of all actions, proceedings, and remedies, except in cases where exclusive, appellate, or original jurisdiction may be conferred by this Constitution or other law on some other court, tribunal, or administrative body. District Court judges shall have the power to issue writs necessary to enforce their jurisdiction.

Tex. Const. art. V, § 8. This provision does not mention writs of mandamus, but the Constitution is not unconcerned with the subject. With respect to the Supreme Court, Article V twice mentions the writ of mandamus. Section 3(a) provides that the “Supreme Court and the Justices thereof ․ may issue ․ writs of mandamus ․ as may be necessary to enforce its jurisdiction,” subject to “such regulations as may be prescribed by law.” Id. § 3(a). That same section further provides: “The Legislature may confer original jurisdiction on the Supreme Court to issue writs of quo warranto and mandamus in such cases as may be specified, except as against the Governor of the State.” Id.³ The latter provision reflects our Constitution's acknowledgement of the obvious separation-of-powers concerns implicated by judicial authority to issue writs of mandamus against high-ranking executive officials. See Tex. Const. art. II, § 1.⁴ The Legislature has acknowledged those concerns as well, as reflected in the limitations it has placed on the judiciary's authority to issue writs of mandamus against constitutional executive officers. See Tex. Gov't Code § 22.002(c). We turn now to those limitations.

A district court's constitutional grant of jurisdiction consists of “exclusive, appellate, and original jurisdiction of all actions, proceedings, and remedies,” which the parties agree includes mandamus actions. Tex. Const. art. V, § 8.⁵ An important caveat follows, however. The apparently broad grant of jurisdiction does not extend to “cases where exclusive, appellate, or original jurisdiction may be conferred by this Constitution or other law on some other court, tribunal, or administrative body.” Id. (emphasis added). The Legislature is thus authorized to withdraw jurisdiction from district courts and confer it upon other courts. One way in which the Legislature has done so is section 22.002 of the Government Code:

(a) The supreme court or a justice of the supreme court may issue ․ all writs of quo warranto and mandamus agreeable to the principles of law regulating those writs, against a statutory county court judge, a statutory probate court judge, a district judge, a court of appeals or a justice of a court of appeals, or any officer of state government except the governor, the court of criminal appeals, or a judge of the court of criminal appeals.

․

(c) Only the supreme court has the authority to issue a writ of mandamus or injunction, or any other mandatory or compulsory writ or process, against any of the officers of the executive departments of the government of this state to order or compel the performance of a judicial, ministerial, or discretionary act or duty that, by state law, the officer or officers are authorized to perform.

Tex. Gov't Code § 22.002 (emphases added).

Setting aside the constitutional question of whether any court can issue a writ of mandamus against the Governor, see post at 25–37 (Young, J., concurring), section 22.002(c) states that only this Court has authority to do so “against any of the officers of the executive departments of the government of this state,” Tex. Gov't Code § 22.002(c). In other words, no court, except for the Supreme Court, has that authority. As we have held before, the phrase “officers of the executive departments of the government of this state” includes the officers of the Executive Department enumerated in the Constitution:

The Executive Department of the State shall consist of a Governor, who shall be the Chief Executive Officer of the State, a Lieutenant Governor, Secretary of State, Comptroller of Public Accounts, Commissioner of the General Land Office, and Attorney General.

Tex. Const. art. IV, § 1; see Sharp, 904 S.W.2d at 672 (“[T]he [C]onstitution identifies seven officials as executive officers” and section 22.002(c) “confer[s] exclusive original jurisdiction on this Court over mandamus proceedings against executive officers, except for the [G]overnor.”).⁶

C.

To summarize, district courts generally have broad original jurisdiction unless another law provides otherwise. Section 22.002(c) is another law providing otherwise. It says no court but the Supreme Court may issue a writ of mandamus against constitutional executive officers, which includes the two parties at issue here, the Governor and the Attorney General. District courts therefore lack jurisdiction to issue a writ of mandamus against the Governor or the Attorney General, unless some other statute overrides section 22.002(c) and empowers them to do so.

American Oversight contends that section 552.321 is just such a statute. It states:

(a) A requestor or the attorney general may file suit for a writ of mandamus compelling a governmental body to make information available for public inspection if the governmental body refuses to request an attorney general's decision as provided by Subchapter G or refuses to supply public information or information that the attorney general has determined is public information that is not excepted from disclosure under Subchapter C.

(b) A suit filed by a requestor under this section must be filed in a district court for the county in which the main offices of the governmental body are located ․

Tex. Gov't Code § 552.321. This is not the first time we have considered the interplay between the PIA's mandamus remedy and section 22.002’s restrictions on mandamus jurisdiction against executive officers. In 1995, the statute authorizing PIA mandamus actions was, in relevant respects, identical to today's subsection (a), quoted above. Subsection (b) did not yet exist. The question arose of a district court's authority to issue a writ of mandamus under the PIA against the Comptroller. Relying on section 22.002(c)’s clear reservation to the Supreme Court of exclusive mandamus jurisdiction over constitutional executive officers, we held:

[D]istrict courts generally have no jurisdiction over executive officer respondents. Any exception to this rule would require express statutory authorization by the legislature naming district courts as the proper fora. See id. § 552.353(b)(3). Under the current statutory scheme, when a relator seeks to compel an executive officer to perform duties imposed by law, generally this Court alone is the proper forum.

Sharp, 904 S.W.2d at 672. Further, although the statute authorized mandamus relief against a “governmental body,” we held that the constitutional executive officer himself, in that case the Comptroller, was “the proper party,” because he had “the legal obligation to perform the duties under [the PIA].” Id. at 673. Because the constitutional executive officer himself is the proper respondent in a mandamus action to enforce the PIA against a constitutional office, the law's restrictions on mandamus relief against such officers were implicated. See id. The Legislature had not avoided this quandary by authorizing mandamus relief against the “governmental body,” as opposed to its officer. See id. at 681.⁷

In 1999, the Legislature added subsection (b) to section 552.321, which now states that “[a] suit filed by a requestor under this section must be filed in a district court for the county in which the main offices of the governmental body are located.” Tex. Gov't Code § 552.321(b). Determining this provision's meaning determines the outcome of this appeal.

American Oversight argues that subsection (b) is the Legislature's effort, in response to Sharp, to empower district courts to issue writs of mandamus against constitutional executive officers. It contends that, in addition to being later in time, subsection (b) is more specific than section 22.002(c) in that it specifically governs PIA mandamus actions, whereas section 22.002(c) governs mandamus actions generally.

The State responds that section 552.321(b) merely specifies where a mandamus action under the PIA must be filed, rather than expanding the authority of a district court over an executive officer respondent who is not otherwise within the court's jurisdiction. In the State's view, subsection (b) does not mention constitutional executive officers or jurisdiction, and it should be understood as a general venue provision for PIA mandamus actions, which remains subject to section 22.002(c)’s specific limitation on a district court's mandamus authority over constitutional executive officers.

We conclude that section 552.321(b) did not expand the jurisdiction of district courts over constitutional executive officers. The Supreme Court's exclusive jurisdiction under section 22.002(c) of the Government Code to issue writs of mandamus against constitutional executive officers has been a settled feature of Texas law for at least a century, and section 552.321(b) contains none of the indicia required to overcome it.

To begin with, section 552.321(b) does not mention jurisdiction, authority, power, or anything of the sort. The key holding of Sharp, which American Oversight contends has been abrogated by subsection (b), is that district courts lack the jurisdiction—that is, the power—to issue writs of mandamus against constitutional executive officers under the PIA. But nothing in section 552.321(b) speaks to the question of the district court's power. The provision reads:

(b) A suit filed by a requestor under this section must be filed in a district court for the county in which the main offices of the governmental body are located ․

Id. By its own terms, subsection (b) merely establishes a mandatory venue for the suits authorized by subsection (a). That venue is the local district court, which in the overwhelming majority of PIA disputes across all levels of state and local government throughout Texas will have the power to entertain such a suit. In the small fraction of PIA disputes involving constitutional executive officers, however, Sharp instructs that district courts lack jurisdiction. American Oversight asks us to read into subsection (b)’s venue requirement an unspoken legislative intent to abrogate Sharp and to expand the power of district courts to include all PIA mandamus actions, including those against constitutional executive officers that would otherwise run afoul of section 22.002(c) of the Government Code. The problem is that subsection (b) does not say that. It reads as a garden-variety venue provision, which resembles many other venue provisions,⁸ none of which are thought to alter the jurisdiction of any court. See Tex. Mut. Ins. Co. v. Chicas, 593 S.W.3d 284, 289 (Tex. 2019) (statute providing for venue transfers “clearly did not [implicate] subject-matter jurisdiction” (citing Morales v. Liberty Mut. Ins. Co., 241 S.W.3d 514, 516 n.1 (Tex. 2007))).

A key pillar of American Oversight's position is the notion that subsection (b) was the Legislature's answer to Sharp. Whether or not that is true as a matter of legislative history, the operative force of subsection (b) derives from its text, not from the purpose it was thought to serve at the time it was enacted. Yet even if we assume subsection (b) was indeed the Legislature's “answer” to Sharp, that gets us nowhere, because the question remains: What was the answer? American Oversight suggests that the answer must have been to empower district courts with mandamus authority over constitutional executive officers. But we must discern the Legislature's answer from the text it chose, and nothing in the text of subsection (b) accomplishes such a result.

Subsection (b) simply does not pick up the glove thrown down by Sharp, which holds that “an action against” the “executive officer respondents” “would require express statutory authorization by the legislature naming district courts as the proper fora.” 904 S.W.2d at 672 (emphasis added). The proper fora for what? For actions against executive officer respondents. Section 552.321(b) does not mention constitutional executive officers at all. It therefore does nothing to expressly authorize district courts as the proper fora for mandamus suits against those officers.

Immediately following Sharp’s indication that only “express statutory authorization” would do the trick, the Court cited an example, from within the PIA, of the kind of express language it had in mind. See id. The example the Court provided, section 552.353(b)(3), at the time referred to:

[A] petition for a declaratory judgment, a writ of mandamus, or both, against the attorney general in a Travis County district court seeking relief from compliance with the decision of the attorney general ․

Act of May 4, 1993, 73d Leg., R.S., ch. 268, § 1, 1993 Tex. Gen. Laws 583, 607 (amended 2009) (current version at Tex. Gov't Code § 552.353(b)(3)) (emphasis added). Thus, had the Legislature wanted to respond to Sharp by further empowering district courts, the Court provided line-by-line instructions. It did so by pointing to a provision specifically contemplating a mandamus petition against the attorney general in a district court. But subsection (b)’s venue requirement looks nothing like the statute the Court pointed to as a rubric for accomplishing the result American Oversight seeks. It looks, instead, like many other venue provisions throughout the statute books, to which courts do not typically assign jurisdictional significance. See supra note 8.

Two basic rules arose from Sharp: (1) district courts lack the power to issue PIA mandamus relief against constitutional executive officers; and (2) the Supreme Court generally has the power to issue PIA mandamus relief against constitutional executive officers (setting aside the Governor). 904 S.W.2d at 672–73. To these two rules, the Legislature added another: (3) all PIA mandamus actions must be brought in district court. Tex. Gov't Code § 552.321(b). American Oversight would like the Legislature's answer to Sharp to have been that the third rule overrides the first rule. But subsection (b) does not say that. Indeed, by restricting venue in PIA mandamus actions to district court without mentioning constitutional executive officers or the jurisdiction of any court, the Legislature may very well have been responding to Sharp by preventing parties from filing mandamus actions against constitutional executive officers in the Supreme Court. In other words, the Legislature may have intended the third rule mentioned above to override the second rule, not the first.

We need not decide today whether an original mandamus action in this Court under the PIA against a constitutional executive officer could proceed in the face of subsection (b)’s mandatory venue requirement. It is enough, for present purposes, to hold that the text of subsection (b) of section 552.321 contains no indication that its aim was to expand the authority of district courts over constitutional executive officers. The text of subsection (b) fixes venue in the local district court for the mandamus actions authorized by subsection (a). It does so without changing the nature of the actions authorized by subsection (a) or the power of various courts over those actions. Because that is all the text says, that is all the text does.

D.

For American Oversight to prevail, it would have to be the case that a statute authorizing lower courts to issue writs of mandamus against a general category of respondent that includes constitutional executive officers is enough to override section 22.002(c)’s grant to this Court of exclusive mandamus jurisdiction against constitutional executive officers. Yet if this were true, section 22.002(c) would always give way to general statutory grants of mandamus authority to lower courts. We have never taken such a limited view of section 22.002(c). To the contrary, as we held in Sharp, overcoming section 22.002(c)’s prohibition on lower-court mandamus authority over constitutional executive officers will typically require “express statutory authorization” of the sort we pointed to in that case. See Sharp, 904 S.W.2d at 672 (pointing to a provision specifically contemplating mandamus relief in district court against the attorney general).

American Oversight argues that if subsection (b) is not a jurisdictional grant, then it must be superfluous. See Hunter v. Fort Worth Cap. Corp., 620 S.W.2d 547, 551 (Tex. 1981) (“[T]he Legislature is never presumed to do a useless act.”). That is not the case. As already observed, subsection (b) could be read to foreclose the direct mandamus actions in this Court against constitutional executive officers contemplated by Sharp. Whether or not it goes that far, there is no question that subsection (b) provides clarity on an important and muddled question the statute previously left unaddressed—where petitioners should file the PIA mandamus actions authorized by subsection (a). Unlike most lawsuits, mandamus petitions may originate at all three levels of our court system. A mandamus petition initiates an original action in the court in which it is filed, and as a colloquial matter mandamus petitions are often more associated with appellate courts rather than with district courts. It would therefore come as no surprise—and would certainly not be superfluous or redundant—for the Legislature to specify that these mandamus petitions, authorized by section 552.321(a), must originate in district court, rather than leaving it to litigants and courts to divine such a rule from the pre-existing legal landscape.

The Legislature's provision of clarity and specificity is generally a welcome feature of a user-friendly statutory scheme. Legislative clarity and specificity should not lightly be confused for superfluity or redundancy. When the Legislature spells out the rules governing a cause of action, such as by providing a venue rule, the canon against superfluity does not require us to ask whether, in the absence of the legislative venue rule, the pre-existing law would have required the same rule—and then to assign an alternative meaning to the statutory venue rule, beyond the natural meaning of its text, in a quest to avoid superfluity. Like any canon of construction, the presumption against “useless acts” is merely one tool, among many, to be used in service of the overriding goal, which is to understand the meaning of the statutory text, within its context, as would an ordinary reader of English. See Malouf v. State ex rels. Ellis, 694 S.W.3d 712, 718 (Tex. 2024) (“We ․ apply the common, ordinary meaning” of a “statute's text,” “constru[ing] the words in light of their statutory context ․”). Stretching the text beyond its natural meaning to avoid perceived superfluity or redundancy is just as misguided as stretching the text beyond its natural meaning for any other reason.

Finally, American Oversight suggests that if the State's arguments prevail, then constitutional executive officers and their state agencies will be able to violate the PIA with impunity. That is not so. The PIA contains abundant criminal penalties. To take one capacious example, “[a]n officer for public information, or the officer's agent, commits an offense if, with criminal negligence, the officer or the officer's agent fails or refuses to give access to, or to permit or provide copying of, public information to a requestor as provided by this chapter.” Tex. Gov't Code § 552.353(a). There are also civil remedies beyond the mandamus action at issue here. Section 552.3215 authorizes parties like American Oversight to “file a complaint alleging a violation of this chapter.” Id. § 552.3215(e). If the governmental body is a state agency, the complaint may be filed with the Travis County district attorney, who then has thirty days to decide whether a violation of the PIA was committed and to decide whether to initiate an action for declaratory or injunctive relief. Id. § 552.3215(e), (g)(1). The district attorney must inform the complainant in writing of these determinations. Id. § 552.3215(g)(2).

American Oversight objects that this pathway makes civil litigation of its allegations contingent on the agreement of the Travis County district attorney to take up the mantle. It does. But that requirement hardly diminishes the incentive of state officials, who face the prospect of both criminal and civil liability at the hands of the Travis County district attorney, to comply in good faith with the PIA. American Oversight is correct, of course, that its reading of the statutory scheme would maximize the ability of private litigants to unilaterally enforce the PIA against constitutional executive officers through civil litigation. We are under no obligation, of course, to maximize the scope of the PIA's remedies. Our job is to understand what those remedies are, as the Legislature has written them. “Liberally construed” or otherwise, the words of section 552.321 do not support the remedy American Oversight seeks. See id. § 552.001(a). If that result is thought to be undesirable, the Legislature may of course provide otherwise.

III.

For these reasons, the district court lacks jurisdiction over the petitions for writ of mandamus against the Governor and the Attorney General. Without reaching any other issues, we reverse the judgment of the court of appeals and remand the case to the district court for dismissal.

The judgment below depends on two mistaken holdings. The first is statutory: that the legislature attempted to authorize district courts to enforce the Public Information Act against the State's executive officers by writ of mandamus. The second, which is applicable only to the governor, is constitutional: that the legislature could authorize a district court to issue a writ of mandamus against the governor. I join the opinion of the Court, which resolves the case on purely statutory grounds and thus properly declines to address the second holding below. I write separately to address that holding, however, because it reflects an error of constitutional dimension that should not pass unnoticed.

The Constitution describes the governor as “the Chief Executive Officer of the State.” Tex. Const. art. IV, § 1. It requires him, and him alone, to “cause the laws to be faithfully executed.” Id. § 10. It excludes him, and him alone, from this Court's constitutional mandamus jurisdiction by allowing the legislature to vest in this Court the power “to issue writs of quo warranto and mandamus in such cases as may be specified, except as against the Governor of the State.” Id. art. V, § 3(a).

In isolation, it is linguistically possible to read § 3(a) as the court of appeals necessarily did: that the framers and ratifiers under no circumstances wanted the State's highest court to review the governor's actions or inactions by mandamus, but they were perfectly content for any district court in Texas to do so. But another possible reading is that if this Court lacks mandamus authority over the governor, then no court has such authority. Text must always be read in its context, and particularly for a constitutional provision, courts must strive to give the text its original public meaning. Under a proper analysis, it is highly likely that no court in this State is now or ever can be authorized to issue a writ of mandamus against the governor.

Bound up in the immediate question—the governor's constitutional amenability to mandamus in a district court—is something larger: the Constitution's vision of the governor's role in our constitutional order. My tentative view is that the court of appeals too easily disregarded the unique constitutional status of the governor, at least in part due to the conventional wisdom holding that Texas lacks an institutionally “strong” governor and that our Constitution does not erect a “unitary executive.” The conventional wisdom may not be entirely wrong, but it seems to be far from entirely right.

True, unlike the federal chief executive, the governor may not choose the other executive officers except the secretary of state (or if there is a vacancy, and then, only until the next election). That limitation does not say very much about what the governor can do. After all, the constitutional provisions I quote above—making the governor the chief executive officer and charging him with ensuring the faithful execution of the laws—establish him not just as an executive official, but as the superior executive official. Other provisions reinforce that understanding. Without quibbling over terminology—“unitary,” “modular,” “directed,” “coordinated,” or some other adjective—the governor clearly has constitutional authority that transcends the rest of the executive branch. He is not just first among equals.

Many governors since 1876 may well have preferred to choose the other executive officers themselves. But the popular election of those officers represents the People's desire to ensure that at least some executive-branch decisions begin with a diversity of perspectives brought from a group of leaders who have every incentive to pursue what they regard as the best public policy of the State. Because the governor may not hire or fire them, one would expect them to voice their views candidly and independently. In many instances, the governor may conclude that this “plural” executive well serves the interests of the State, requiring no further intervention from him beyond using his political tools (which, of course, are considerable).

But ultimately it is the governor's constitutional duty to ensure the proper functioning of the entire executive branch. In part, that is what it means to be a “chief executive officer.” At a minimum, therefore, being the CEO likely means that there are certain circumstances in which the governor may or even must settle the executive branch's policy even despite disagreement from other executive officers. If the governor invokes that authority, he will then assume direct responsibility for the relevant decisions and results.

Under this view, grounded in its text, the Constitution seems to construct a nuanced and sophisticated mechanism in which Texas gets the best of both worlds. It begins with a distribution of executive authority that facilitates the airing-out of competing perspectives and creates incentives to refine differences. It avoids the problem of echo chambers that can follow when officers are primarily beholden to an individual rather than to the People. But the Constitution's elevation of the governor as chief executive officer also adds the possibility of a uniform executive-branch position when such uniformity is necessary for him to discharge his constitutional duties. After all, “[i]n the construction of Constitutions, as well as of statutes, it has been often held that the powers necessary to the exercise of a power clearly granted will be implied,” Imperial Irrigation Co. v. Jayne, 104 Tex. 395, 138 S.W. 575, 586 (Tex. 1911), and it would be strange indeed if this principle applied to everyone except the governor.

In this duality, therefore, the Constitution simultaneously encourages a distribution of authority but allows for uniformity. Exactly how the Constitution authorizes the governor to be “the chief executive officer” in the face of potential disagreement largely exceeds the scope of my analysis today. It is enough for now to recognize that the governor's CEO title and his structural role and responsibilities mean that his authority is sometimes far greater than the conventional wisdom might suggest. That authority, in turn, is relevant to today's case. It shows why the governor alone, as the head of the co-equal executive branch, is no more subject to mandamus in any court than the legislature as a body could be. If mandamus could issue from any court, it would probably have to be from the Supreme Court, the head of this branch. But even that would be doubtful. The concept of mandamus—certainly in 1876, if to a lesser degree today—conveys a sense of command that uneasily fits when directed from the head of one branch to the head of another.

I reiterate that the Court properly resolves this particular dispute as to both the governor and the attorney general on purely statutory grounds. The Court itself should not reach out to decide constitutional matters unless unavoidable. The luxury of a concurring opinion is the opportunity to identify and begin to sketch the contours of complex and important issues that arise within a case without binding even its author, much less the Court as a whole, to any position. The very fact that this Court has not yet given the chief-executive-officer clause any real legal significance further justifies both the Court's not addressing that matter now and my flagging it for future purposes.

I proceed as follows. Part I briefly recounts how the constitutional issue arose in this case. It then addresses the apparent textual ambiguity within Article V, § 3(a) and § 8—whether any court may exercise mandamus jurisdiction over the governor—and suggests that, at least in part, reading them in light of Article IV's text and history may resolve it.

In Part II.A, therefore, I invoke various tools to begin ascertaining the original public meaning of the chief-executive-officer clause. Although I emphasize that my views remain tentative, and I intend humility in their expression, it seems clear from this analysis that the governor's CEO title is no mere honorific. It instead reflects actual authority that the governor, when acting as chief executive officer, may invoke to supervise and manage the executive branch. This Court's cases that are close in time to the Constitution's enactment reflect that understanding. From there, I more fully address this case's central question in Part II.B and explain why it is likely that no court could subject the governor to mandamus review by turning to, among other things, the history of the legislature's conferral of authority on Texas courts to issue the writ and this Court's early discomfort with doing so against any executive officer. I briefly examine federal and early state court practices, which lend additional support to this conclusion.

Finally, in Part III, I identify two potential consequences of a proper understanding of the governor's role as chief executive officer. For one, my analysis casts some of this Court's precedents that might seem to suggest an undue constriction of gubernatorial authority in a different light, and it explains why those cases need not be read to do so. For another, it raises the question of what legal tools the governor may have to ensure that the executive branch faithfully executes the law—a question that only future cases can fully resolve.

I

At the heart of this case is American Oversight's invocation of the Public Information Act to demand various communications to and by the governor and the attorney general. See Tex. Gov't Code § 552.001(a) (providing a general right “to complete information about the affairs of government and the official acts of public officials and employees”). American Oversight alleges that the governor and the attorney general withheld public information, leading it to seek writs of mandamus against both officials in a Travis County district court. Government Code § 552.321(b) provides that a writ of mandamus to enforce the Act must be filed in that court, whereas Government Code § 22.002(c) provides that this Court alone may issue writs of mandamus “against any of the officers of the executive department.” The court of appeals regarded § 552.321(b) as an exception to § 22.002(c) and ultimately determined that American Oversight should have the opportunity to secure the writs in the district court. See 683 S.W.3d 873, 882–83, 889 (Tex. App.—Austin 2024). Stated more simply, the court concluded that the Act vests the district courts with jurisdiction to issue writs that were otherwise exclusively issuable from the State's highest court. Id.

Behind the court's construction of the Government Code was its construction of Article V, § 3(a) and § 8 of the Constitution. The latter provides that the district courts have original jurisdiction over “all actions ․ except in cases where exclusive, appellate, or original jurisdiction may be conferred by this Constitution or other law on some other court.” Tex. Const. art. V, § 8. The former, Article V, § 3(a), vests this Court with appellate jurisdiction “over all cases except in criminal law matters and as otherwise provided in this Constitution or by law.” And it further authorizes the legislature to “confer original jurisdiction on the Supreme Court to issue writs of quo warranto and mandamus in such cases as may be specified, except as against the Governor of the State.” Id. In isolation, that text is susceptible of two readings:

1. the Constitution ensures that the State's highest court may never issue a writ of mandamus against the governor, but it is wholly unconcerned about allowing any lower court to do so; or

2. the textual restriction of this Court's mandamus authority means that the only court that might have such authority still may not subject the governor to its mandamus jurisdiction.

The fact that the court of appeals chose the former reading underscores that interpreting a legal text requires a full appreciation of its context. Cf. Malouf v. State ex rels. Ellis, 694 S.W.3d 712, 736 (Tex. 2024) (Young, J., dissenting) (quoting Pulsifer v. United States, 601 U.S. 124, 141, 144 S.Ct. 718, 218 L.Ed.2d 77 (2024)). The court of appeals’ reading is linguistically possible but yields a result that—to put it mildly—should have struck the court of appeals as odd: that the Constitution contemplates jurisdiction for a district court anywhere in the State to issue a writ of mandamus against the governor while forbidding this Court from ever doing so.

Especially when a court reads a significant legal provision in a way that generates a strange outcome, the court should pause, think again, and perhaps even a third time. Constitutional text requires a special commitment to understanding its original public meaning. Should the judiciary err on that front, the sole way to correct that error outside the judiciary itself is through the extremely onerous exercise of the sovereign prerogative of the People to amend the Constitution. Courts’ commitment to understanding constitutional text with as much precision as possible is, of course, part of their job. But it is an especially delicate part because it is also the one and only way to ensure that when the People do amend their Constitution, it is to change its meaning to something they now want it to mean, rather than to restore the meaning that it always had but that the judiciary wrongly abrogated.

That the court of appeals’ foundational premise touches on the governor's authority only casts the point in sharper relief. The courts must neither expand nor erode the constitutional role or authority of another branch, and a decision that would have either effect should, again, be one that courts consider and reconsider with caution and self-doubt. Here, the court reached a startling result by impliedly twisting the negative-implication (also known as the expressio unius) canon, as it found that even though this Court is prohibited from issuing such writs, the authority could be properly placed elsewhere: “The expression of one thing”—that this Court cannot issue the writ against the governor—was “an expression of all that share[d] in the ․ prohibition involved.” Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 107 (2012).

This analysis ignored the rest of the Constitution's text, as well as the history and tradition animating both that text and the legislature's conferral of mandamus authority on the judiciary. Beyond all that, the court of appeals never grappled with the most severe consequence of its holding—namely, that a trial court from anywhere in Texas could commandeer the highest authority in the executive department.

The opposite construction—that no court can issue a writ of mandamus against the governor—is the one with intuitive appeal. There is no precedent to the contrary. And given that this Court has always considered mandamus to be an extraordinary remedy, it makes sense that, if any court could do so, this Court alone could have jurisdiction to issue such a writ against the executive officers. Disputes giving rise to such mandamus requests are likely to involve statewide issues of immense public importance. And that no court has mandamus jurisdiction over the governor in particular would reflect the constitutional separation of powers between the three coordinate branches of government. Still, weighing the relative stakes is not a satisfactory answer to the legal question of whether any court may issue a writ of mandamus against the governor.

The Court today properly avoids all this, thereby “[s]etting aside the constitutional question.” Ante at ––––. Yet we cannot hope to ever answer it without looking at other constitutional text, context, history, and tradition.

Take how Article V, § 3(a) of the Constitution immunizes only the governor, and not the other executive officials, from becoming targets of this Court's writs of mandamus. And like Article V, § 3(a), Article IV, § 1 distinguishes the governor from the rest of the constitutional executive officers by providing that

[t]he Executive Department of the State shall consist of a Governor, who shall be the Chief Executive Officer of the State, a Lieutenant Governor, Secretary of State, Comptroller of Public Accounts, Commissioner of the General Land Office, and Attorney General.

(Emphasis added.) Apart from denoting his role as chief executive officer, the Constitution provides that the governor is duty-bound to “cause the laws to be faithfully executed.” Tex. Const. art. IV, § 10. It provides that he alone may demand from the rest of the executive branch “information in writing” about “any subject relating to the duties, condition, management and expenses of their respective offices,” subject to criminal “punish[ment]” and “remov[al] from office.” Id. § 24.

The combined force of all these—or at least these—constitutional provisions is what commands my attention. Specifically, in giving the chief-executive-officer clause its original public meaning, we can understand whether, given that meaning, any court could issue a writ of mandamus against him. His authority as chief executive officer, after all, elevates him above the other constitutional and statutory inferior officers who exercise executive authority. Charged with causing the laws to be faithfully executed, the governor's role underscores his immunity from our mandamus authority. I explore all this, and more, below.

II

This Part proceeds in two subparts. First, I recount how the governor came to be called chief executive officer in our constitutional system and identify clues that might give that clause its original public meaning. Second, I summarize the historical understanding of the writ of mandamus, explore the legislature's prior conferrals of authority on the courts to issue the writ, highlight some early opinions from this Court, and comment on how the federal courts and our sister States approached “mandamusing” the chief executive officer.

A

The Chief Executive Officer. Ascertaining the original public meaning of the chief-executive-officer clause benefits from examining a variety of textual sources from the nineteenth century, this Court's opinions that reference the clause, and records of the constitutional conventions of 1845 and 1875. Like “the federal Constitution, we have records of the debates from the convention[s],” but “[t]hese records may be even more useful than the federal versions,” as “[t]hey were not drafted in secret, and they are not overshadowed by interpretations offered at subsequent ratifying conventions.” Holden T. Tanner, Lone Star Originalism, 27 Tex. Rev. L. & Pol. 25, 83–84 (2023).

While my views remain tentative and my research incomplete, I hope to spur interested parties, amici, scholars, the profession, and fellow judges to supplement it, regardless of whether their contributions confirm or rebut the views that I express here. Thus far, however, I find that notwithstanding his inability to choose certain high members of the executive department, the governor—when acting as chief executive officer—has constitutionally conferred authority to supervise and manage the affairs of the executive department. The constitutional executive officers owe their positions not to the will of the governor but of the People, of course, and that independence reflects the genius of our constitutional design, which encourages the governor to lead what could become a “team of rivals” without resorting to the political process or the courts.

Consider first, however, that a leading Texas constitutional scholar has said that the clause “apparently has no legal significance.” George D. Braden et al., The Constitution of the State of Texas: An Annotated and Comparative Analysis 301 (1977). This conclusory statement strikes me as highly doubtful. In expounding our Constitution, no less than its federal analogue, courts consider it axiomatic that “every word must have its due force, and appropriate meaning.” Williams v. United States, 289 U.S. 553, 572–73, 53 S.Ct. 751, 77 L.Ed. 1372 (1933) (quoting Holmes v. Jennison, 39 U.S. 540, 570–71, 14 Pet. 540, 10 L.Ed. 579 (1840)). “[N]o word was unnecessarily used, or needlessly added,” meaning that we must treat each as if it was “weighed with the utmost deliberation, and its force and effect to have been fully understood.” Id. In doing so, we dignify “the high talent, the caution, and the foresight of the illustrious men who framed” our Constitution. Id.

It was not just the framers but the People, after all, who designated the governor as the chief executive officer of the State. By removing the chief executive's ability to name for himself his whole cabinet, the People departed markedly and purposefully from the federal model. Texans, seasoned by years under the federal Constitution, distributed some executive power among officials of their, and not the governor's, choice. In the wake of Reconstruction, “the proper limitations to place on an executive branch” that some had “perceived as tyrannical” was an obvious focus at the framing. William J. Chriss, Six Constitutions Over Texas: Texas’ Political Identity, 1830–1900, at 130 (2024) (noting a sentiment among some 1875 constitutional convention delegates “who more than anything wanted homestead protection, railroad regulation, and emasculation of the spendthrift Reconstruction government that had raised their taxes, worsened their lot, and diluted their political ability to protect themselves”). And still, they put the governor above the rest of the executive branch, conveying added responsibility and the power incident to it.

The “chief executive officer” title was new in the 1876 Constitution. Given this change's potential significance, I trace its use from the 1836 Constitution of the Republic of Texas to the Texas Constitution of 1869. During those years, “the executive authority” or the “supreme executive power” was “vested in a Chief Magistrate.” The Constitution of the Republic of Texas of 1836 provided that

[t]he executive authority of this government shall be vested in a chief magistrate, who shall be styled the president of the republic of Texas.

Repub. Tex. Const. of 1836, art. III, § 1, reprinted in 1 H.P.N. Gammel, The Laws of Texas 1822–1897, at 13, 13 (Austin, Gammel Book Co. 1898). The style changed slightly in 1845:

The supreme executive power of this State shall be vested in a Chief Magistrate, who shall be styled the Governor of the State of Texas.

Tex. Const. of 1845, art. V, § 1. Meanwhile, the modern cabinet was beginning to take shape with the executive article providing for a lieutenant governor, secretary of state, treasurer, and comptroller. Id. §§ 12, 16, 23.

Article V, § 1 did not change whatsoever in the 1861 or 1866 constitutions, each replicating the 1845 Constitution. Compare id. art. V, § 1, with Tex. Const. of 1861, art. V, § 1, and Tex. Const. of 1866, art. V, § 1. It was not until 1869 that Article V, § 1 became Article IV, § 1 (with the executive and judiciary articles flipping places). New Article IV looked a bit different from its predecessor but still maintained the governor's designation as chief magistrate:

The executive department of the State shall consist of a Chief Magistrate, who shall be styled the Governor, a Lieutenant Governor, Secretary of State, Comptroller of Public Accounts, Treasurer, Commissioner of the General Land Office, Attorney General and Superintendent of Public Instruction.

Tex. Const. of 1869, art. IV, § 1. The 1869 version was the first “to expressly define the different executive offices which constitute the executive department.” Bledsoe v. Int'l R.R. Co., 40 Tex. 537, 565–66 (1874). By listing these executive offices, the People vested the executive power “in the entire magistracy composing the executive department, with the powers of each separately defined,” thereby “more clearly defin[ing] the boundaries of power” among “the different offices composing the executive department.” Id. at 566.

In the 1876 Constitution, the People finally provided that the governor shall be the “chief executive officer” of the State. When they adopted Article IV, § 1 in 1876, it provided that

[t]he executive department of the State shall consist of a governor, who shall be the chief executive officer of the State, a lieutenant governor, secretary of State, comptroller of public accounts, treasurer, commissioner of the general land office and attorney general.

Tex. Const. art. IV, § 1 (amended 1995, eliminating the office of the treasurer). The shift from “chief magistrate” to “chief executive officer,” like any overt shift in constitutional or even statutory text, is presumptively significant. For modern ears, the former term may sound somewhat antiquated, and the latter may sound somewhat corporate. As I describe below, I find substantial overlap between the two—some potential distinctions, but certainly no dramatic shift. Both terms, meanwhile, implicate significant oversight responsibilities.

“Chief,”¹ of course, first modified “magistrate,” which was defined throughout that period as “[a] public civil officer.” Magistrate, Webster's 1845; Magistrate, Webster's 1868 (“A person clothed with power as a public civil power.”); Magistrate, Webster's 1897 (defining the term as a “public civil officer”). The presidents of the United States were often called—and called themselves—the “chief magistrate,” as President Washington and President Lincoln both did even in their inaugural addresses.² “Chief” then modified the compound noun “executive officer.” “Executive” was defined as an administrator or superintendent of government. Executive, Webster's 1845; Executive, Webster's 1868; Executive, Webster's 1897. Like the definitions for “magistrate,” those for “executive” often included the word “officer”—someone who held public office or was authorized to perform a public duty. Officer, Webster's 1845; Officer, Webster's 1868; Officer, Webster's 1897.

The dictionary distinctions between “chief magistrate” and “chief executive officer,” therefore, seem rather slight.³ “Chief magistrate” described the highest or head individual who held and could exercise civil governmental authority; “chief executive officer” described the highest or head administrator or superintendent who held and was authorized to perform the responsibilities of a public office.

Confirming a substantial degree of overlap, at least in this context, is corpus linguistics, which helps verify the contexts in which terms—like “chief magistrate” and “chief executive officer”—were used throughout different historical periods.⁴ I find (or, in fairness, my law clerk who in fact can run the searches has shown me) that the common usage of “chief magistrate” between 1860 and 1879 referred most often to the president, then to governors, and sometimes to mayors.⁵ When compared to “chief magistrate,” the term “chief executive officer” was not nearly as commonly used.⁶ Still, “chief executive officer” referred at times to a king, a president, a palatine, a mayor, and, of course (and unlike “chief magistrate”), a leader of a private company. Both terms—chief executive officer and chief magistrate—were frequently juxtaposed against text providing that individuals with these titles bore heavy responsibilities in managing the affairs of state.

From all this, we can begin to define the governor as the head superintendent of state government, distinguished above those other constitutional executive officers in Article IV, § 1. To illustrate, we can refer to the governor as the “chief magistrate” and as the “chief executive officer,” whereas we would not (at least ordinarily) refer to the leader of a company as its “chief magistrate.” At least more so than “chief magistrate,” therefore, “chief executive officer” connotes a latent authority to compel obedience from an entire organization, which in this context would include constitutional and statutory executive officers. The obvious overlap between the two terms is analytically significant, because it further links the framers’ (and the People's) thoughts on executive power between both the 1845 and 1875 constitutional conventions beyond even what is evident from the debates.

The framers of the 1845 Constitution, of course, sought to depart from the federal model and tendency among other States “to create a cabinet for the Governor, over whom he may exercise absolute control.” 1845 Debates, supra, at 119 (emphasis added). The election of executive officers by persons other than the governor ensured that he could not pack the executive department with yes-men or sycophants; rather, the People (or their representatives) packed it with those whom they trusted.

The governor's inability to choose these individuals, however, was separate and apart from his ability to supervise and manage the affairs of the executive department. For example, one delegate at the 1845 convention recognized the governor as “the Executive head of the nation, chief of that department, as it were the head and sign board of government.” Id. at 129–30. One imagined that he would be “plain, honest, sensible, and good” and urged the convention, therefore, to vest him “with certain powers and privileges” so that he could “perform the duties pertaining to his office.” Id. at 121–22, 129. Similarly, at the 1875 convention, one delegate observed that the office of the governor “requir[ed] the highest order of talent, probity, and integrity, and the very credit of the State.” 1875 Debates, supra, at 153. It was their duty, said another, “to give the chief executive authority to protect” those in the State who were “endangered in their rights, liberty, and property.” Id. at 157.

In giving the chief executive officer that authority, one delegate in 1845 noted that to fulfill his duty to “see the laws faithfully executed,” the governor must have “certain aids, adjuncts[,] and assistants,” for otherwise, “he as one individual c[ould not] attend to the due execution of the laws.” Id. at 129. Top of mind in 1875 was, in one delegate's words, “filling all the offices of the State with persons of [the People's] own selection.” 1875 Debates, supra, at 257 (emphasis added). For example, like the “independen[ce]” some at the 1845 convention would have afforded the secretary of state, see 1845 Debates, supra, at 118, still others at the 1875 convention thought that the secretary “could act independently and fearlessly when he knew that he owed his election to the people,” not “the mere breath of the Governor.” 1875 Debates, supra, at 257.

From the debates, the “independence” apparently conferred to elected executive officers consisted of their owing their offices to the People and not to the governor. But that fact did not vitiate those officers’ duty to report to, and to follow, the governor, who remained responsible for directing their affairs. For example, while discussing the gubernatorial salary in 1875, one delegate successfully helped defeat a proposal to “reduc[e] the salary of the chief executive below that of any other State officer.” Id. at 153. He observed that doing so would run counter to the prevailing sentiment that the governor “was made responsible” for the “good conduct” of all other executive officers. See id. Another delegate later parroted that the governor “had been made responsible for the good conduct and honest management of all other State officers,” and still another delegate was “answered in the affirmative” after asking whether the Governor was “responsible for all of the acts of the subordinate officers of the Executive Department.” Id. at 163.

To be sure, the delegates knew that officers who were elected would be endowed with a confidence to resist the governor's demands. Notably, some delegates at the 1845 convention contemplated the possibility of a fractured executive department, one in which an elected official might “be brought into collision with the Governor.” See 1845 Debates, supra, at 119. Unlike appointed officials, who at all times “ought to be in harmony with the Executive,” any single elected official might be a “spy,” or worse, “an opposition power ․ inimical to the Governor,” one “disposed to break him down in order to take his place.” Id. at 120, 122. And yet, the ultimate status of these officers “most effectually ensure[d] the due execution of the duties assigned [to the governor]”—namely, his duty “to see the laws faithfully executed.” Id. at 129. The governor was trusted to lead the department through its inherent potential for disharmony, division, and discord.

Displaying that leadership at the 1875 convention, then-Governor Coke delivered a message to the delegates, stating:

I have the honor, in behalf of the Executive Department of the State Government, to tender the earnest co-operation of all the officers of that department, as far as their aid may be desired in forwarding the labors and advancing the purposes of the Convention․ Any information to be found in any of the offices of the Executive Department, not embraced in th[e] reports [that were already made by executive-department officials], will be cheerfully and promptly furnished, upon request from yourself or the honorable Convention.

1875 Journal, supra, at 12 (emphasis added). Of course, it is easy to read too much into Governor Coke's message—he and the rest of the executive department may well have agreed, before he sent his missive, to cooperate with the convention in such ways. Still, the delegates’ response was telling—“thank[ing] the Governor for his polite and patriotic communications” and promising to “avail themselves of such information he has offered, as occasion may require.” Id. at 13 (emphasis added). The debates that day underscored the delegates’ thinking. Though they had no real “use for the message,” they understood that “[i]f they required any further information they would ask it respectfully and the Governor would give it cheerfully.” 1875 Debates, supra, at 2 (emphasis added). The governor himself was thus perceived as responsible for seeing that the executive department offered the convention all it needed, despite his inability to personally choose certain persons who would constitute his “aids, adjuncts[,] and assistants.” Cf. 1845 Debates, supra, at 129.

Today, Article IV, § 1 embodies the nuanced, tenuous, and potentially complicated arrangement that the framers gave us. And yet, from this Court's early days, we have appreciated the power of the chief magistrate and chief executive officer to command his executive department, treating his distinct authority with special solicitude, and recognizing it as superior to other officers of the executive branch, including with respect to the particular work that they were chosen to do.

The governor, we said, “is the head of the executive department of the state, and it is made his duty, by the constitution, to ‘take care that the laws be faithfully executed.’ ” Hous. Tap & Brazoria Ry. Co. v. C. H. Randolph, 24 Tex. 317, 343 (1859). By his status within the structure of the Constitution, “[i]t is evidently contemplated[ ] that he shall give direction to the management of affairs, in all the branches of the executive department.” Id. (emphasis added). “Otherwise,” of course, “he has very little to do.” Id.

The governor's direction transcends the boundaries of his subordinates’ otherwise-siloed authority. For example, we have suggested that where the attorney general lacks constitutional or statutory authority to institute a suit, the governor, “as [the State's] chief executive officer,” may “have the power to require the attorney general to institute, or to cause to be instituted, a suit of [such] character, when in [the governor's] judgment the welfare of the state required it, even though the legislature had not so directed.” Day Land & Cattle Co. v. State, 68 Tex. 526, 4 S.W. 865, 867 (1887). And we have legitimized the governor's conduct by virtue of his status as “chief executive officer.” E.g., Arnold v. State, 71 Tex. 239, 9 S.W. 120, 120 (1888) (holding that an act “creat[ing] new offices, to be held by the heads of the executive department,” was constitutional given, among other things, the governor's role as “the chief executive officer”); see also Jones v. Alexander, 122 Tex. 328, 59 S.W.2d 1080, 1082 (Tex. [Comm'n Op.] 1933) (noting that in Arnold, the “Court held that [the] appointments of the Governor were not prohibited by the Constitution, because he was the chief executive officer of the state”).

Of course, when compared to the federal government, in which the president enjoys “the entire unity of [his] executive department” via his absolute “authoritative control” over who fills its divisions, the governor of Texas enjoys comparatively less authority. Hous. Tap, 24 Tex. at 343. “The absence of that absolute power of the chief executive in this state[ ] must occasionally produce a want of harmony in the executive administration, by the inferior officers of that department, declining to comply with the wishes, or to follow the judgment[,] of the governor.” Id. It is through his role as chief executive officer that he commands the executive department, and in circumstances—hopefully rare—in which he concludes that it is essential to have a unity of policy, it is that supervisory and managerial authority that he must invoke. “[T]he power to control the decisions of subordinate levels of government is not extraordinary at all.” Abbott v. Harris County, 672 S.W.3d 1, 18 (Tex. 2023). As the “Chief Executive Officer of the State,” who is “obligate[d] [ ] to ‘cause the laws to be faithfully executed,’ ” it is “little surprise” and “nothing extraordinary” that the governor may “control the executive branch of government.” Id. at 18–19.

* * *

To sum up, the text, history, and tradition of the chief-executive-officer clause suggests that the governor is vested with meaningful authority to manage, supervise, and direct the affairs of all those who reside in the executive department. Both “chief magistrate” and “chief executive officer” reflect a meaning that could not escape ordinary speakers of English—that, as the saying goes, the buck stops with him. True, this State may not have a “unitary executive” in the sense that the members of the executive department are all removable at will by the governor. Cf. Trump v. Wilcox, ––– U.S. ––––, 145 S. Ct. 1415, 1415, ––– L.Ed.2d –––– (2025). In its place, rather, is a chief executive officer who must manage elected officials whom he might not have appointed but with whom he must work to cause the laws to be faithfully executed.

This is part of the genius of the Constitution—vesting certain executive officers with a measure of independence that was more likely to deliver on the policies of the electorate who voted for them. President Lincoln, both a “chief magistrate” and “chief executive officer,” famously chose for himself this very benefit—a cabinet made up of ambitious and fractious leaders, potentially divisive, but when channeled, of enormous benefit to the country, as told in Doris Kearns Goodwin's famous Team of Rivals: The Political Genius of Abraham Lincoln (2005). The People of our State chose to impose an analogous model of leadership on future Texas governors.

Division within the executive department is to be expected, but nothing can occur within or without it that does not at least potentially implicate the governor's authority and discretion. The framers believed that the governor would be wise enough to recognize that it was often not in his or the State's best interest to override other executive officers, but they did not deprive him of his ultimate authority to act in the best interest of the State when necessary or when division arose within his branch. And as I discuss below, it is his special distinction as chief executive officer, and the authority incident to it, that likely bars the courts from issuing writs of mandamus against him.

B

The Chief Executive Officer and the writ of mandamus. I first briefly recount the history of the writ of mandamus. I then follow the legislature's history of conferring mandamus authority on Texas courts and describe this Court's early discomfort with issuing such writs against executive officers. All this, plus the approach of the federal courts and of early courts in sister states, combined with my discussion above about the role of the governor, make it likely that no court may constitutionally exercise mandamus jurisdiction over him.

To begin, as the Court today notes, the writ of mandamus “long pre-existed the Republic.” Ante at ––––. The writ is, at its most basic, a tool—one “whereby various public duties and powers [are] commanded and enforced.” Thomas Tapping, The Law and Practice of the High Prerogative Writ of Mandamus 56 (London, Wm. Benning & Co. 1848). Courts historically dispensed the writ “in all cases where there was a legal right to justice, but for which right the law had not provided any specific legal remedy.” Id. at 57. Or put differently, a party successfully prosecuted the writ “where [he] ha[d] a legal power consequent upon the violation of some legal right or duty, for which the law ha[d] not established any specific or adequate legal remedy, and where, in justice and good government, there ought to be one.” Id. at 58–59.

Mandamus today is primarily a tool used by superior courts to exercise control within the judicial branch itself, but modern practice also includes this Court's issuance of a writ of mandamus against an executive official to compel him “to perform ministerial acts” or “to correct a clear abuse of discretion.” In re Williams, 470 S.W.3d 819, 821 (Tex. 2015) (citation omitted). If courts are not cautious, that standard can be slippery. “How easily the doctrine may be pushed and widened to any extent[.]” Decatur v. Paulding, 39 U.S. (14 Pet.) 497, 518, 10 L.Ed. 559 (1840) (opinion of Catron, J.) (stating that entertaining the writ in that case created a “dangerous” conflict between the executive and judicial departments).

Understanding for what purpose the writ was traditionally deployed is just as important as understanding to whom the writ was appropriately directed. The writ of mandamus, we often have said, is “extraordinary.” E.g., Ferguson v. Huggins, 122 Tex. 95, 52 S.W.2d 904, 907 (1932). And given its extraordinary nature, the writ was only properly used to supervise inferior officers. See 2 Isaac ’Espinasse, A Digest of the Law of Actions and Trials at Nisi Prius 302 (Gould, Banks & Gould 1811) (1793) (“The Writ of Mandamus is a prerogative writ, issuing out of the Court of King's Bench, by virtue of that general superintendency which that court possesses over all inferior jurisdictions and persons.” (emphasis added)). The distinction between inferior officers’ vulnerability to writs of mandamus and their superior officer's immunity from them “is a fundamental principle of law,” for the writ should “never be granted in cases where, if issued, it would prove unavailing.” Edward J. Myers, Mandamus Against a Governor, 3 Mich. L. Rev. 634, 648 (1905).

Consistent with this inferior–superior dichotomy, the People had vested Texas district courts with jurisdiction to issue writs of mandamus beginning early in the State's history. For example, the 1845 Constitution provided that district courts had “power to issue all writs necessary to enforce their own jurisdiction, and give them a general superintendence and control over inferior jurisdictions.” Tex. Const. 1845 art. IV, § 10; see also Tex. Const. 1861 art. IV, § 10 (same); Tex. Const. 1866 art. IV, § 6 (similar); Tex. Const. 1869 art. V, § 7 (also similar). Relatedly, in 1846, a statute provided that the district courts had authority to grant petitions for writs of mandamus and that

all writs of mandamus, sued out against the heads of any of the departments or bureaux of government, shall be returnable before the district court of the county in which the seat of government may be.

Act approved May 11, 1846, 1st Leg., R.S., 1846 Tex. Gen. Laws 200, 201, reprinted in 2 Gammel, supra, at 1507. Discussing this statute and the 1845 Constitution's Article IV, § 10, we found it “evident” that a district court could issue the “extraordinary writ” of mandamus. Jones v. T.H. McMahan & Gilbert, 30 Tex. 719, 728, 730 (1868) (cautioning that the writ “should not be used when an ordinary writ or suit will be as effectual”).

Yet in Houston Tap, we questioned the authority of our district courts to issue writs of mandamus against executive officers (i.e., the “heads of any of the departments or bureaux of government”) at all. See 24 Tex. at 342–43. Justice Roberts—eventually Chief Justice Roberts and then Governor Roberts—wrote for the Court, noting that the 1846 statute “obviate[d] the difficulty, which had grown out of the practice of seeking the writ of mandamus against the commissioner of the general land office ․ in counties all over the state.” Id. Consolidating and streamlining mandamus petitions in one geographic location was admirable, but still, the Court found that the statute “t[ook] for granted the existence of the power” to “grant the writ” against “executive officers” in the first place. See id. at 343.

Houston Tap underscores the discomfort among the early justices of this Court in issuing “the extraordinary remedy of mandamus” and pitting the judiciary against the executive branch. See id. at 329. The case involved a railroad company's attempt to obtain a writ of mandamus against the state treasurer from a Travis County district court. Id. We ultimately affirmed the district court's dismissal of the suit, holding that the company had failed to sufficiently plead its right to mandamus and that the company was seeking to “force the treasurer of the state to perform” an act that was “not only official, but [ ] require[d] the exercise of his judgment, as an officer.” Id. at 333, 338. Throughout, however, the Court emphasized that the Constitution's separation-of-powers provision “contemplate[d] that the persons employed in each department, will be wise enough, and honest enough, to discharge the duties intrusted to them, without the aid or interference of the others.” Id. at 336.

The writ of mandamus, the Court said, ran headlong into these constitutional principles, flouting the People's “right to expect[ ] that the respective duties allotted to each department shall be performed by those they have chosen to perform them.” Id. at 336. Justice Roberts engaged in a thought experiment, sketching out a seeming dystopia in which even the governor could be subject to mandamus in a district court. He wrote that the People “would be not a little surprised to find” that “the governor, and heads of departments, elected by the whole people of the state, were summoned before the district court of Travis [C]ounty, and required there to contest the propriety of any of their official acts, done within the scope of their authority.” Id. at 336–37. Then, “after a tedious struggle, the facts in issue being tried and determined by a jury of twelve men,” the People would be equally surprised to find that these officials would be “compelled, under the penalty of attachment and imprisonment for contempt, to do an act, which they had refused to do, acting under their oath of office, and under a sense of responsibility to their constituents.” Id. at 337. The Court then imagined the “consequence[s]” of issuing a writ of mandamus not just against an executive official but rather against the governor himself:

[Say, for example, that] [t]he governor is required to [perform] a mere ministerial act by writing his name; the right of the plaintiff has been made clear in the district court; and the reasons given by the governor for his refusal [to perform], are not deemed sufficient by the district judge. The governor, under a sense of duty, and to resist aggression upon his official rights, is obstinate, and will not obey the mandate of the court,—will not [perform] officially, as “governor of the state of Texas,” upon compulsion; the sheriff of Travis [C]ounty must enter the governor's mansion with his posse, and take possession of the governor, and put him in jail, and keep him there, until he will [perform his ministerial act].

Id. Worse, the Court said, would be the consequences of when such a governor “yield[s] [his] judgment” and “obey[s] the mandate”: “Who ‘takes care, that the laws are faithfully executed;’ the governor or the district judge? Surely not the governor, if he must obey the mandate of the court, in the performance of an official duty.” Id. (emphasis added).

Fifteen years later, in Bledsoe v. International Railroad Co., the Court walked back some of Houston Tap’s dicta, noting that it did not mean to “entertain[ ] the slightest intention of annulling” the courts’ ability to issue writs of mandamus against public officers “to perform a purely ministerial duty, positively required by law, and involving neither official judgment nor discretion.” 40 Tex. at 552. Still, the Court “admitted” that this rule was “difficult” to apply “to particular cases.” Id. at 557. The “independence of power in the different departments was intended to act as a check and restraint against usurped authority,” and so remedies by mandamus should be rarely, if ever, available. Id. at 566.

Reiterating Houston Tap’s concerns, however, the Court then noted that “[t]he word ‘ministerial’ ha[d] reference generally to an act done under authority of a superior; and in this sense it could never apply to the chief executive with respect to anything required by the legislative authority.” Id. at 557 (emphasis added). The Court then concluded by underscoring that the 1846 statute vesting district courts with jurisdiction to issue writs of mandamus against the “heads of departments” of the executive branch neither “enlarged [nor] extended the remedy by mandamus beyond the enforcement of a merely ministerial act by any of such officers.” Id. at 568–69.

It was the year following Bledsoe that the delegates met in Austin for the constitutional convention. The Constitution as adopted in 1876 provided that the district courts had “power to issue writs of habeas corpus in felony cases, mandamus, injunction, certiorari, and all writs necessary to enforce their jurisdiction.” Tex. Const. art. V, § 8 (amended 1891, 1973, 1985). And it specified that this Court had exclusively appellate jurisdiction and could only issue writs of mandamus for the purpose of enforcing that appellate jurisdiction. Id. § 3 (amended 1891, 1930, 1980, 2001); accord Wells v. Littlefield, 62 Tex. 28, 30 (1884). In 1881, however, the legislature provided as follows:

No court of this State shall have power, authority or jurisdiction to issue the writ of mandamus, or injunction, or any other mandatory or compulsory writ or process against any of the officers of the executive departments of the government of this State, to order or compel the performance of any act or duty, which, by the laws of this State, they or either of them are authorized to perform, whether such act or duty be judicial, ministerial or discretionary.

Act approved Feb. 15, 1881, 17th R.S., ch. 12, § 4, 1881 Tex. Gen. Laws 7, 7–8, reprinted in 9 Gammel, supra, at 99–100 (emphasis added). By this statute, the legislature apparently took the Court's evident discomfort about exercising mandamus jurisdiction over executive officers, as expressed in Houston Tap and Bledsoe, to a unique extreme—vitiating the authority of any court anywhere to issue writs of mandamus against “officers of the executive departments of the government.” See id.

That solution proved unworkable. In 1891, the legislature proposed a series of amendments to the Constitution to provide, among other things, that “[t]he Legislature may confer original jurisdiction on the Supreme Court to issue writs of quo warranto and mandamus in such cases as may be specified, except as against the Governor of the State.” Act approved Apr. 28, 1891, 22d Leg., R.S., S.J.R. 16, § 3, 1891 Tex. Gen. Laws 197, 197–98, reprinted in 10 Gammel, supra, at 199–200. The People adopted the amendments narrowly, 37,445 to 35,695. “In pursuance of the power granted by that [new constitutional] provision,” McKenzie v. Comm'r of Gen. Land Off., 88 Tex. 669, 32 S.W. 1038, 1039 (1895), the legislature provided the following:

The supreme court, or any justice thereof, shall have power to issue writs of habeas corpus as may be prescribed by law; and the said court, or the justices thereof, may issue writs of mandamus, procedendo, certiorari and all writs necessary to enforce the jurisdiction of said court; and in term time or vacation may issue writs of quo warranto or mandamus against any district judge or officer of the state government, except the governor of the state.

Act approved Apr. 13, 1892, 22d Leg., 1st C.S., ch. 14, § 1, art. 1012, 1892 Tex. Gen. Laws 19, 21, reprinted in 10 Gammel, supra, at 386 (emphasis added). The legislature accordingly vested this Court with jurisdiction to issue writs of mandamus against executive officers except the governor. Meanwhile, the 1881 statute prohibiting any court from exercising “power, authority or jurisdiction to issue the writ of mandamus ․ against any of the officers of the executive departments ․ to order or compel the performance of any act or duty ․ whether such act or duty be judicial, ministerial or discretionary” remained on the books.

This Court confronted the tension between the 1881 and 1892 statutes in McKenzie, 32 S.W. at 1039. There, the petitioner filed an original proceeding requesting that the Court issue a writ of mandamus against the commissioner of the general land office. Id. In finding that it had “power to grant the writ of mandamus against the head of a[n] [executive] department,” the Court observed that the later-enacted statute “restricted the operation of the former law in so far as it applied to the supreme court, and repealed it to that extent.” Id. The 1892 statute vesting this Court with mandamus jurisdiction “clearly manifest[ed] that it was the purpose of the legislature to pursue the policy for which a way had been laid out by” the 1891 amendments to the Texas Constitution, thereby “continu[ing] the power previously granted to issue the writ of mandamus to the heads of the departments of the state government.” Id. at 1040 (emphasis added). Thus, we construed the 1881 statute “to read: ‘No court of this state, except the supreme court, shall have power,’ etc.,” which “preserve[d] both [statutes].” Id.

We have not taken this responsibility lightly. In Betts v. Johnson, for example, we explained that the legislature “did not intend to confer original jurisdiction upon this court except in cases where there existed some special reason for its exercise.” 96 Tex. 360, 73 S.W. 4, 5 (1903). “A mandamus proceeding against the head of a department, as a rule, involves questions which are of general public interest and call for a speedy determination.” Id. And it was “obvious” that such cases “are of far more importance than those ordinarily arising in mandamus suits against other officers, whether of the state, or of a district, or [of] a county.” Id.

Then-Justice Hecht explained that following the 1925 statutory recodification, the 1881 statute disappeared, and in its place was “a provision giving the Supreme Court exclusive mandamus jurisdiction over ‘officers of the executive departments.’ ” In re TXU Elec. Co., 67 S.W.3d 130, 157 (Tex. 2001) (Hecht, J., dissenting). “In effect, the Legislature made its 1892 enactment an exception to the 1881 statute, consistent with our construction of the two” in McKenzie; “[t]he provision is now [§] 22.002(c) of the Government Code.” Id. Similarly, the 1892 statute vesting this Court with mandamus jurisdiction eventually became “[§] 22.002(a) of the Government Code.” Id. at 152, 157.

The writ of mandamus, in sum, is an extraordinary remedy that should be rarely issued against executive-branch officers. Today's decision, rendered entirely based on the Court's construction of the PIA and the Government Code § 22.002, is consistent with the history I have just described. Indeed, as late as 1874, “a petition for writ of mandamus ha[d] never been sustained in this state against the governor, secretary of state, comptroller, [or] treasurer ․ though many ha[d] been filed and prosecuted.” Bledsoe, 40 Tex. at 567; see also Kuechler v. Wright, 40 Tex. 600, 648, 667 (1874) (Roberts, C.J., concurring in the denial of the second motion for rehearing) (making a similar observation and responding to Justice Moore's concurring opinion).

Given its English origins, the writ was applicable only against “inferior” officers, and notwithstanding district courts’ broad statutory authority to subject constitutional executive officers to mandamus jurisdiction, this Court exhibited extreme discomfort with that practice in its early years. Hous. Tap, 24 Tex. at 337. We later noted, albeit in dicta, that as the titular “superior” executive officer, the governor had no “ministerial” duties to perform and so was never a proper target of mandamus. See Bledsoe, 40 Tex. at 557. After the legislature granted this Court alone jurisdiction to issue writs of mandamus against executive officers—but not the governor—we observed that its actions “continue[d] the power previously granted” to the courts. See McKenzie, 32 S.W. at 1040. Keeping with this history and tradition, therefore, it is almost certain that no court can issue a writ of mandamus against the governor.

This analysis corresponds to the approach of federal and other state courts. While it formally remains an open question within the federal system whether the courts can issue a writ of mandamus against the president, “[i]t is extremely doubtful” that courts can do so. Harris v. Bessent, No. 25-5057, 2025 WL 1021435, at *6 (D.C. Cir. Apr. 7, 2025) (Rao, J., dissenting) (collecting cases in which the D.C. Circuit “declined to issue the writ ‘in order to show the utmost respect to the office of the Presidency and to avoid ․ any clash between the judicial and executive branches of Government’ ” (first quoting Nat'l Treasury Emps. Union v. Nixon, 492 F.2d 587, 616 (D.C. Cir. 1974); and then citing Nat'l Wildlife Fed'n v. United States, 626 F.2d 917, 928 (D.C. Cir. 1980))). In Mississippi v. Johnson, the Supreme Court refused to “express[ ] any opinion” on “whether, in any case, the President of the United States may be required, by the process of this court, to perform a purely ministerial act under a positive law, or may be held amenable, in any case, otherwise than by impeachment for crime.” 71 U.S. (4 Wall.) 475, 498, 18 L.Ed. 437 (1866). Still, it noted that “[a]n attempt on the part of the judicial department of the government to enforce the performance” of the president's duties was “an absurd and excessive extravagance.” Id. at 499. “[I]mplicit in the separation of powers established by the Constitution” is that “the President and the Congress (as opposed to their agents)” cannot “be ordered to perform particular executive or legislative acts at the behest of the Judiciary.” Franklin v. Massachusetts, 505 U.S. 788, 827, 112 S.Ct. 2767, 120 L.Ed.2d 636 (1992) (Scalia, J., concurring). Consequently, it is likely that no court can issue the writ against the president, but such a remedial gap is nothing new. E.g., Marbury v. Madison, 5 U.S. (1 Cranch) 137, 179–80, 2 L.Ed. 60 (1803).

As for the state courts, after surveying “all the adjudicated cases upon th[e] question [of whether the courts could mandamus the governor], dating from the year 1839 up to and including cases decided as late as June 30, 1903,” one scholar identified “two irreconcilable lines of decision”:

1. the governor “is not answerable to the writ to compel the performance of his duty, be it either discretionary or ministerial”; or

2. the governor “is liable to the writ to compel the performance of duties purely ministerial in nature.”

Myers, supra, at 634, 647; see also R. E. Heinselman, Annotation, Mandamus to Governor, 105 A.L.R. 1124 (1936) (similarly noting “an irreconcilable conflict of authority”). It was the former line of decision, and not the latter, that was “based upon the more logical and cogent reasoning.” Myers, supra, at 647.

Compelling any executive officer to take an action by mandamus risks serious separation-of-powers concerns, see, e.g., In re Stetson Renewables Holdings, LLC, 658 S.W.3d 292, 296 (Tex. 2022) (refusing to compel an executive branch official to act because it would unnecessarily pit the judicial department against the prerogatives of the coordinate branches), and doing so against the governor would take that concern to the highest level, cf. Decatur, 39 U.S. at 519, 521 (opinion of Catron, J.) (noting that even when weighing whether mandamus should issue against a lower-level executive official, “the conflict between the executive and judiciary departments could not well be more direct, nor more dangerous” and that “entertaining such a cause [was] calculated to alarm all men who seriously th[ought] of the consequences”).

Of course, the governor “should follow the law because it is the law” and not because some court somewhere told him to do so. Stetson, 658 S.W.3d at 297. We presume that he does so at all times; at least as much as any other government official in a coordinate branch, the governor must benefit from the presumptions of regularity, good faith, and lawfulness. See, e.g., Webster v. Comm'n for Law. Discipline, 704 S.W.3d 478, 488, 496, 501, 505 (Tex. 2024). It is one thing for a court to rule against the governor if he is a party to ordinary litigation; it is quite another to purport to instruct him as to what actions of his are mandatory rather than discretionary, and then purport to order him to personally undertake an action that he regards as improper. Cf., e.g., id. at 487 (“[C]onstitutional problems arise when one branch pushes beyond the boundaries to interfere with another branch's exercise of its constitutional powers.”).

For a myriad of reasons, therefore, it is hard to imagine that the Constitution's text or structure would ever abide subjecting a governor of this State to a court's mandamus process—and certainly not that of a trial court.

III

Thus far, I have concluded—again, subject to reconsideration if I am shown to be wrong—that the chief-executive-officer clause vests the governor with at least some meaningful authority to direct and manage the executive branch of government. Beyond the consequence most directly relevant here—the governor's likely lack of amenability to mandamus—two further potential consequences merit additional discussion.

First, because the governor is charged with causing the laws of this State to be faithfully executed, Tex. Const. art. IV, § 10, any precedents of this Court that might seem at odds with that role—such as by suggesting that he cannot direct agency action or that he is not implicated when public officials exercise executive authority—deserve further explanation. Such cases can and should be harmonized with the principles that I have described. Second, given the heavy responsibilities in superintending the executive branch, the governor must have tools to ensure that other officers faithfully execute the laws. I discuss both points in turn.

A

To reiterate, in addition to and as part of his role as chief executive officer, the governor is duty-bound to “cause the laws to be faithfully executed.” Id. The People have given the governor preeminence over the other constitutional executive officers, and indeed the entirety of the executive branch of government. This understanding of gubernatorial power may, at least at first glance, be in some tension with two of this Court's recent precedents. See State v. Volkswagen Aktiengesellschaft, 692 S.W.3d 467 (Tex. 2022); In re Abbott, 645 S.W.3d 276 (Tex. 2022). Those decisions, however, were based on statutory construction rather than constitutional principles, and they say nothing about the governor's role as chief executive officer. That role was invoked in neither case, and perhaps purposefully so; as I see it, the governor may invoke the Constitution's chief-executive-officer authority, but he may also allow the process to unfold in the ordinary way. In other words, a future case that does implicate that authority will not necessarily be governed by precedents that turn on something else. Accordingly, I should not be mistaken as criticizing these decisions as wrongly decided.

Take, for example, In re Abbott, which involved the governor's directing the Department of Family and Protective Services to “follow the law as explained” by an attorney general opinion. 645 S.W.3d at 279. After the plaintiffs challenged DFPS's stated intention to follow the governor's directive, the court of appeals reinstated a temporary injunction restraining DFPS's ability to take any action “based on th[at] directive.” Id. Among other things, we granted mandamus relief “with respect to the order's injunction against the Governor, as there [was] no allegation that he [was] taking, or ha[d] authority to take, the enforcement actions the order enjoins.” Id. at 280.

We first noted that “[u]nlike the federal constitution, the Texas Constitution does not vest the executive power solely in one chief executive.” Id. “[T]he executive power,” we said, “is spread across several distinct elected offices, and the Legislature has over the years created a wide variety of state agencies—including DFPS—whose animating statutes do not subject their decisions to the Governor's direct control.” Id. (emphasis added). Crucially, “[t]he State d[id] not contend in this Court that the Governor's letter formally changed the legal obligations of DFPS.” Id. at 280–81. What is more, “the Governor's letter cite[d] no legal authority that would empower [him] to bind state agencies with the instruction contained in the letter's final sentence, and we [were] directed to none.” Id. at 281. Thus, we concluded that “neither the Governor nor the Attorney General ha[d] statutory authority to directly control DFPS's investigatory decisions.” Id. at 281 (emphasis added). “DFPS alone b[ore] legal responsibility for its decisions.” Id.

Similarly, in Volkswagen, we broadly observed that “Texas does not have a unitary executive.” 692 S.W.3d at 473. At issue was whether we could fairly “impute the status of party to the Governor himself,” and because the actions at issue “were not brought by the Governor, at his direction, or on his authority,” we concluded that he was not a true party to the suit. Id. at 474. Imputing party status to any potential litigant—public or private—carries a host of jurisprudential concerns that have nothing to do with a coordinate branch's authority. That party-status focus in Volkswagen provides the proper lens for viewing the Court's statements that “the Governor is not automatically implicated in every state action or even every executive-branch action” and that “nothing in [an agency's] enabling statute g[ave the governor] the authority to direct [its] actions.” Id. (emphasis added).

Significantly, at no point in Abbott or Volkswagen was the Court referred to the governor's authority as chief executive officer to direct and manage the affairs of the executive branch writ large. Nor did the governor invoke his status as chief executive officer to direct any action in either case. In this sense, it is correct that he had not assumed direct responsibility for the government's work. Yet as I have explained above, the framers, and the People, always understood that the governor is potentially implicated in and responsible for every action of an executive-branch officer—elected or otherwise. Again, whether he is implicated to the extent that he could be fairly considered a party to the litigation is a separate question. But it would go too far to suggest—and the Court did not so suggest—that he could not be implicated by executive action. Such an assertion would undermine his constitutional role. And while nothing in a statute might provide that the governor can direct an agency's action, the governor's preexisting authority over the executive branch fills that apparent void. I find it hard to imagine that a statute could create an executive agency over which the governor's authority as chief executive officer could be displaced without violating Article IV, § 1.

It is hard to overstate the importance of getting all this right. As the head of the executive branch, the governor has been charged by the People to cause the laws to be faithfully executed. He cannot do that task alone; instead, he must supervise the work of other elected and very many unelected officials who wield executive authority. When the People elect the governor, in other words, they do so with the expectation that he may prioritize and implement his policies that, at a minimum, supersede whatever the bureaucracy may generate. Recognizing this constitutional minimum avoids a host of serious constitutional questions, including the propriety of unelected government officials exercising executive authority. See, e.g., Webster, 704 S.W.3d at 494 (noting that the first assistant attorney general “operate[s] next to and in tandem with the constitutional source of power” and that when he “acts under the direction of the attorney general, he does so as if the attorney general himself had acted”). At times the governor may give suggestions; at times he may issue directives. And when he does the latter, at least if he does so by exercising his chief-executive-officer responsibility, the law cannot treat his commands as empty words. Rather, they are imbued with constitutional authority.

Thus, Abbott and Volkswagen may be harmonized with my understanding of executive authority. Again, in neither case did the governor purport to direct agency or executive action under his authority as chief executive officer, duty-bound to cause the laws to be faithfully executed. That choice may have been purposeful. And until it is otherwise—i.e., until he invokes his executive authority to superintend the actions of the executive branch and doing so leads to litigation—the internal machinations of that branch offer no opportunity for this Court or any other to opine on the boundaries of gubernatorial authority.

B

As chief executive officer and as the official charged with seeing that the laws are faithfully executed, the governor must have some tools to discharge these constitutional roles and duties. The framers feared that the office of the governor “w[ould] increase in strength and power as it progress[ed],” e.g., 1845 Debates, supra, at 116, but likely could not have anticipated that so much executive authority would be placed in people largely invisible to the public eye. Of course, given the structure of the executive branch of government, the governor sits atop constitutional executive officers who owe their offices to their own (albeit often overlapping) constituents. But he remains their constitutional superior, and we have acknowledged that his authority transcends that of his inferiors. Cf. Day Land & Cattle Co., 4 S.W. at 867.

Division and dissention, this Court has said, are to be expected. In Houston Tap, for example, we acknowledged “an inherent difficulty in the organization of th[e] [executive] department.” 24 Tex. at 343. When a “want of harmony” among executive officers results in an injury to a private plaintiff, mandamus may not, and often will not, lie; such an injury “cannot justify another department, to wit, the judiciary, in overstepping the boundary of its prescribed authority, for the purpose of furnishing a remedy.” Id.

But what about the injury to the governor—or, given the source of his authority, to the State as an institution? With respect to an executive officer who fails to perform an official action, the governor surely has a host of political options to impose discipline. With respect to lesser officials, the governor may be able to terminate their employment. There are undoubtedly many other tools at his disposal.

Beyond those, the courts can sometimes provide a mechanism for both public and private rights to be enforced. The Constitution, for example, provides that “the officers of the Executive Department” as well as “all officers and managers of State institutions” must provide, under oath, a semi-annual report to the governor. Tex. Const. art. IV, § 24. Under that provision, the governor can also, “at any time, require information in writing from any and all of [those] officers or managers, upon any subject relating to the duties, condition, management[,] and expenses of their respective offices and institutions.” Id. Should any officer “wilfully make a false report or give false information,” he can be “removed from office” and “punish[ed]” for “perjury” following conviction. Id. The Constitution, in other words, expressly contemplates judicial enforcement as one tool available to the governor to ensure compliance within his branch with his directives.

Likewise, this Court has never held that the governor cannot seek a writ of mandamus to compel an inferior officer to perform a purely ministerial duty, for example. Contra Braden, supra, at 319 (asserting that the governor “apparently lacks tools” to cause the laws to be faithfully executed and that “no governor appears to have asserted any authority that might be derived from” his chief-executive role or duty to cause the laws to be faithfully executed). Though we have acknowledged that “the conflicts arising out of” the executive department's “declining to comply with the wishes, or to follow the judgment of the governor”—conflicts which in that case only injured a private party—“cannot be adjudicated or settled by the judiciary,” Hous. Tap, 24 Tex. at 343, we have also suggested that the governor could, if necessary, seek redress in this Court against his inferiors via mandamus. In In re Texas House of Representatives, for example, we noted that should an executive agency “ignore” the governor's exercise of his constitutional authority, “it may be appropriate for the courts to order the [agency] to comply.” 702 S.W.3d 330, 347 (Tex. 2024). The particular power at issue there—a reprieve—is only one example. The governor's specific responsibility in Article V, § 10 must be construed as the People's intention to vest in him “whatever power is reasonably necessary to fulfill [that] function or perform [that] duty.” Cf. PUC v. GTE–Sw., Inc., 901 S.W.2d 401, 407 (Tex. 1995); see also Jayne, 138 S.W. at 586.

In a proper case, after all, mandamus can lie against any other executive officer at the behest of an injured party. If the governor is constitutionally empowered to direct an executive officer about an executive-branch policy matter, that task would no longer be discretionary; if the officer refused to comply, the governor may have a variety of tools at his disposal, including presumably having at least as much right to seek mandamus as a private citizen. Cf., e.g., Tex. Const. art. IV § 24. The governor has “legal power” as chief executive officer, “consequent upon the violation of [the] legal ․ duty” to cause the laws to be faithfully executed, to remedy his injury by mandamus, at least where the law “has not established any specific or adequate legal remedy.” Cf. Tapping, supra, at 58. Whether any particular order was proper under the chief-executive-officer (or some other) clause presents a distinct question—but the governor is surely empowered in various situations to issue such binding orders, and when he does, it is hard to see why mandamus could not lie.

Such an exercise of executive authority would not be novel, at least in other states. See, e.g., State ex rel. Withycombe v. Stannard, 84 Or. 450, 165 P. 566, 566 (1917) (“Where a public official charged with a duty to the whole state ․ refuses to execute the law and to perform his duty in that regard, we think the Governor is acting only in obedience to this requirement of the Constitution in appealing to the court to compel [via mandamus] that official to perform such legal duty.”); Chiles v. Milligan, 659 So.2d 1055, 1056 (Fla. 1995) (involving the conditional grant of a governor's petition for writ of mandamus against the secretary of state and statewide elected comptroller).

As I noted at the outset, however, this case—and this concurring opinion—need not explore all the ways in which the governor's authority over the executive branch is enforceable, whether through judicial means or otherwise. It is enough for today to recognize that the authority exists, and perhaps to a far greater degree than the conventional wisdom has assumed.

* * *

The Court today broke no new jurisprudential ground. As it observes, consistent with our prior decision in A & T Consultants, Inc. v. Sharp, 904 S.W.2d 668, 669 (Tex. 1995), the legislature has not sought to vest the district courts with the authority to issue writs of mandamus against any constitutional executive officer, let alone the governor. Even had the legislature done so, I strongly suspect no court could issue a writ of mandamus against the governor, though the courts could possibly issue writs of mandamus against his inferiors, including at his request, given his capacious constitutional authority as chief executive officer to manage and direct the executive department.

Above all, I hope that in the future, the lower courts, counsel, amici, and scholars will accept my invitation to analyze both the chief-executive-officer clause, including within the larger context of Article IV and the entire Constitution, and the effect it has on cases involving the governor or inferior executive officers. Hermeneutically synthesizing the text, history, and tradition of the Texas Constitution is no small feat—one that will be accomplished only through partnership between bench and bar. My opinion today purports to do nothing more than scratch the surface, and tentatively at that. I remain “open to any outcome that faithfully reflects the original meaning of our constitutional text.” Tex. Dep't of State Health Servs. v. Crown Distrib. LLC, 647 S.W.3d 648, 666 (Tex. 2022) (Young, J., concurring).

With these thoughts, therefore, I am pleased to concur.

FOOTNOTES

1.   When the AG's office requests a ruling under the PIA, an oddity arises in that it must submit the request to another section of its own office. In such circumstances, we understand that the section of the AG's office requesting the ruling and the section of the AG's office making the ruling are sealed off from one another and, at least in principle, operate independently. See generally Pub. Util. Comm'n of Tex. v. Cofer, 754 S.W.2d 121 (Tex. 1988). We are not asked to consider whether two sections of the same office, both of which answer to and exercise authority derived from the same constitutionally elected officer, can ever truly operate independently of one another. The PIA affords no alternative way of handling information requests submitted to the AG's office, which means the office is often obligated to “wear two hats” in this way and to navigate the consequent ethical challenges as best it can.

2.   See also Bradley v. McCrabb, Dallam 504, 506, 508 (Tex. 1843) (holding that the inadequate-legal-remedy requirement for writs of mandamus was part of the English common law adopted into the law of the Republic); Yett v. Cook, 115 Tex. 175, 268 S.W. 715, 718–19 (Tex. 1925) (observing that the writ of mandamus is “construed in light of the common law”).

3.   Article V elsewhere mentions the writ of mandamus in section 5(c), which authorizes the “Court of Criminal Appeals and the Judges thereof,” “[s]ubject to such regulations as may be prescribed by law,” to issue the writ of mandamus “in criminal law matters.” Tex. Const. art. V, § 5(c).

4.   “The powers of the Government of the State of Texas shall be divided into three distinct departments, each of which shall be confided to a separate body of magistracy ․ and no person, or collection of persons, being of one of these departments, shall exercise any power properly attached to either of the others ․”

5.   Until 1985, the Constitution specifically mentioned the writ of mandamus as within a district court's authority. See Tex. Const. art. V, § 8 (amended 1985) (providing that “[d]istrict [c]ourt[s] ․ have power to issue writs of ․ mandamus”). It no longer does so, instead providing only that “District Court judges shall have the power to issue writs necessary to enforce their jurisdiction.” Tex. Const. art. V, § 8. Section 24.011 of the Government Code authorizes a district judge to “grant writs of mandamus, injunction, sequestration, attachment, garnishment, certiorari, and supersedeas and all other writs necessary to the enforcement of the court's jurisdiction.” Tex. Gov't Code § 24.011. The parties do not question the general authority of district courts to issue writs of mandamus. They focus, as do we, only on whether that authority extends to constitutional executive officers in PIA cases.

6.   Sharp’s count of seven constitutional executive officers included the Treasurer, a position eliminated by a 1995 constitutional amendment. See Tex. S.J. Res. 1, 74th Leg., R.S., 1995 Tex. Gen. Laws 4980 (adopted at the general election on Nov. 7, 1995, by a vote of 495,181 to 218,473).

7.   The extent to which any legally relevant distinction may be drawn between a constitutional executive officer and the governmental body he heads is a complicated and interesting question. The parties join issue on it to some extent, but we need not endeavor to answer it here. It is enough, for present purposes, to follow Sharp’s holdings that (1) the proper respondent in a mandamus action under the PIA involving a constitutional executive office is the constitutional executive officer; and (2) the PIA's mandamus remedy against a “governmental body” necessarily implicates section 22.002(c)’s limitations on the judiciary's authority to issue mandamus relief against constitutional executive officers.

8.   See, e.g., Tex. Gov't Code § 2007.044(b) (“A suit under this section must be filed in a district court in the county in which the private real property owner's affected property is located.”); Tex. Gov't Code § 551.142(d) (“A suit filed by the attorney general under Subsection (c) must be filed in a district court of Travis County.”); Tex. Transp. Code § 342.302(c) (“A suit under this section must be filed in a justice court of the county [from which a license is required under this chapter].”); Tex. Transp. Code § 342.304(b) (“A suit under this section must be filed in a justice court of the county in which the license is issued.”); Tex. Gov't Code § 2007.021(a) (“A suit under this subchapter must be filed in a district court in the county in which the private real property owner's affected property is located.”).

1.   Between 1845 and 1897, “chief” referred to the “[h]ighest in office or rank,” the “most eminent,” and “the head” of a (typically military) group. Chief, Webster's An American Dictionary of the English Language (rev. ed. 1845) [hereinafter Webster's 1845]; Chief, Webster's A Dictionary of the English Language (acad. ed. 1868) [hereinafter Webster's 1868]; Chief, Webster's An American Dictionary of the English Language (1897) [hereinafter Webster's 1897].

2.   President George Washington, Second Inaugural Address (Mar. 4, 1793) (“I am again called upon by the voice of my country to execute the functions of its Chief Magistrate.”); President Abraham Lincoln, First Inaugural Address (Mar. 4, 1861) (“The Chief Magistrate derives all his authority from the people ․”).

3.   This is hardly surprising given that delegates at even the 1845 convention were already referring to the governor not only as the chief magistrate but also as the chief executive. E.g., Wm. F. Weeks, Debates of the Texas Convention 120 (Houston, J. W. Cruger 1846) [hereinafter 1845 Debates] (referring to the governor as “the Executive”); see also id. at 129 (referring to the governor in one breath as “the Executive,” “the chief,” and the “chief magistrate of the country, standing above all the officers of the State”). Delegates at the 1875 convention likewise frequently referred to him simply as “the executive.” E.g., Debates in the Texas Constitutional Convention of 1875, at 157 (Seth Shepard McKay ed., 1930) [hereinafter 1875 Debates] (referring to the governor as the “executive” and noting that “it [was] the duty of th[e] Convention to give the chief executive authority to protect [the People]”). When the 1875 “Committee on Executive Department” reported the initial executive article that the convention was to debate, it already styled the governor as the “chief executive officer” and no longer as the “chief magistrate.” See Journal of the Constitutional Convention of the State of Texas, Begun and Held at the City of Austin, September 6th, 1875, at 228 (Galveston, “News” Office 1875) [hereinafter 1875 Journal]. As far as I can tell, no delegate either suggested, objected to, or even discussed the change.

4.   Corpus linguistics “employs a massive database that enables date-specific searches for the possible, common, and most common uses of words or phrases as they were used in newspapers, books, magazines, and other popular publications.” Matthews v. Indus. Comm'n of Ariz., 254 Ariz. 157, 520 P.3d 168, 174 (2022) (Bolick, J.) (citing Thomas R. Lee & Stephen C. Mouritsen, Judging Ordinary Meaning, 127 Yale L.J. 788, 831–32 (2018)). Within the corpus-linguistics database are “large bodies of naturally occurring text,” called “corpora,” which are “drawn from a particular speech community” and “reflect (both in diversity and relative frequency) the language patterns within that community.” Thomas R. Lee & Stephen C. Mouritsen, The Corpus and the Critics, 88 U. Chi. L. Rev. 275, 277, 291 (2021). Of particular use to me are “concordance line[s],” or “sample sentence[s] from real-world language that show[ ] how” the terms “chief magistrate” and “chief executive officer” have “been used in the past.” Cf. id. at 292.

5.   Search of “Chief Magistrate” from 1860 to 1879, Corpus of Hist. Am. Eng., https:www.english-corpora.org/coha (last visited June 24, 2025) (populating 94 source references).

6.   Compare id., with Search of “Chief Executive Officer” from 1820 to 1899, Corpus of Hist. Am. Eng., https:/www.english-corpora.org/coha (last visited June 24, 2025) (populating 30 source references).

Chief Justice Blacklock delivered the opinion of the Court.

Justice Young filed a concurring opinion. Justice Sullivan did not participate in the decision.