Mike Morath, in His Official Capacity as the Commissioner of Education v. Pecos-Barstow-Toyah Independent School District

Reversed and Rendered and Memorandum Opinion and Concurring and Dissenting Opinion filed July 3, 2025

In The Fifteenth Court of Appeals NO. 15-24-00101-CV MIKE MORATH, IN HIS OFFICIAL CAPACITY AS THE COMMISSIONER OF EDUCATION, Appellant V. PECOS-BARSTOW-TOYAH INDEPENDENT SCHOOL DISTRICT, ATHENS INDEPENDENT SCHOOL DISTRICT, BEEVILLE INDEPENDENT SCHOOL DISTRICT, BEN BOLT INDEPENDENT SCHOOL DISTRICT, BROWNSVILLE INDEPENDENT SCHOOL DISTRICT, ET AL., Appellees On Appeal from the 201st District Court Travis County, Texas Trial Court Cause No. D-1-GN-24-005018 MEMORANDUM OPINION In Morath v. Kingsville Independent School District, we held that the failure by the Commissioner of Education to issue school ratings, standards, and related documents by the statutory deadlines for the 2023 school year1 was not ultra vires conduct that rendered him subject to suit, nor did it support enjoining A to F school ratings for 2023 because “the clear legislative intent in Chapter 39 is to publish school ratings, not suppress them.”2 In this case, the Commissioner did meet the statutory deadlines for issuing standards in 2024, and could have done so for the 2024 ratings themselves. But he was sued again by 5 of the state’s 1,200 school districts (“the Districts”),3 and a Travis County district court again enjoined him, this time not from issuing school ratings late but from issuing them on time.

The Districts alleged that four grounds supported waiver of the Commissioner’s immunity from suit and a temporary injunction. We rejected three of them in Kingsville; the fourth asserts that “unexpected” and “unexplained” results on the 2024 STAAR tests “raised questions” about the automated grading software used to score some of them. But while selected results lifted out of three million STAAR tests may appear anomalous to lay witnesses, even those with long experience in public education, that does not make their doubts competent expert evidence that the tests were unreliable. We reverse the trial court’s judgment, dissolve the temporary injunction blocking publication of the 2024 ratings, grant the Commissioner’s jurisdictional plea, and dismiss him and the claims against him from this case.

Background Mike Morath, in his official capacity as the Texas Commissioner of School years typically extend from one calendar year into the next, so to avoid confusion we refer to school years by the calendar year in which they end. 710 S.W.3d 918, 926 (Tex. App. [15th Dist.] 2025, no pet.) (hereinafter “Kingsville”).

An additional 28 school districts intervened as plaintiffs on the morning of the injunction hearing. As the Plaintiffs and Intervenors filed a joint brief in this Court, we refer to them collectively as the “Districts.”

Education, is the executive officer of the Texas Education Agency, a state agency comprised of the Commissioner and agency staff. 4 Chapter 39 of the Texas Education Code gives the Commissioner broad legal authority “to adopt rules to evaluate school district and campus performance,” and to assign each district and campus an annual performance rating of A (exemplary), B (recognized), C (acceptable), D (needs improvement), or F (unacceptable).5 The Chapter requires the Commissioner to solicit input from school boards, administrators, teachers, and parents in establishing and implementing this system. 6 But it also gives him broad discretion that, along with the general immunity from suit provided to state officials, was intended to keep academic ratings “out of the courts.” 7 Yet no ratings were issued for five years—none in 2020 and 2021 due to COVID-19, none as to low-performing schools in 2022 for the same reason, and none in 2023 or 2024 due to temporary injunctions issued by two Travis County district courts. 8 In the first of those cases, 121 of roughly 1,200 Texas school districts sued the Commissioner in August of 2023, and successfully embargoed the 2023 school year ratings for 19 months.9 The Commissioner’s appeal in that case was transferred to this Court in late August of 2024, and we reversed and dismissed it on immunity grounds on April 3, 2025.10 In the current suit to bar the 2024 school ratings, things moved more quickly

See TEX. EDUC. CODE §§ 7.002(a); 7.055(b)(2).

Id. § 39.054(a). All statutory citations in this opinion refer to Chapter 39 of the Texas Education unless otherwise noted.

§ 39.001(b).

Honors Acad., Inc. v. Tex. Educ. Agency, 555 S.W.3d 54, 71 (Tex. 2018).

See Kingsville, 710 S.W.3d at 912.

See id. at 922.

See id. at 929. as the parties had pleaded and briefed many of the same issues the year before. Three days before the August 15th statutory deadline for issuing the 2024 ratings, 11 five of the Districts sued the Commissioner in Travis County seeking temporary and permanent injunctions prohibiting publication of the 2024 ratings. The trial court granted a temporary restraining order the day suit was filed. The Commissioner filed a plea to the jurisdiction asserting immunity with supporting affidavits 9 days later.

Within a month, the trial court held a two-day evidentiary hearing on September 16– 17, 2024, on the Districts’ claim for a temporary injunction and the Commissioner’s plea to the jurisdiction. Shortly before the hearing began, 28 new schools districts intervened as plaintiffs. The following day, the court granted the temporary injunction and denied the jurisdictional plea. The Commissioner filed his notice of this accelerated appeal the next day.

Normally, the Commissioner’s appeal would supersede the temporary injunction under Texas law. 12 To thwart that effect, the trial court added a clause purporting to enjoin supersedeas until the court of appeals could consider an emergency stay, which the Districts immediately requested from this Court. We deferred ruling on the motion and requested an agreed briefing schedule from the parties. Because of the similarity of the issues with the 2023 appeal in Kingsville, we set this appeal for submission without oral argument on March 19, 2025.13 Discussion As an official of state government, the Commissioner is immune from suit

§ 39.054(a-3).

See TEX. R. APP. P. 24.2(a)(3), 25.1(h), 29.1; TEX. CIV. PRAC. & REM. CODE § 6.001.

We dismiss the motion as moot. We express no opinion on the efficacy of the clause in the injunction purporting to delay supersedeas. absent waiver or exception.14 The Districts bear the burden of affirmatively showing immunity does not apply. 15 They sought to avoid immunity here under the ultra vires exception, alleging four reasons why issuing school ratings for 2024 would exceed the bounds of the Commissioner’s legal authority or conflict with the law itself. 16 I. The Ultra Vires Claims Based on Fair Notice The Districts assert two ultra vires claims based on fair notice, both of which their trial attorney (who represented them also in Kingsville) conceded “are the same this year” as those addressed in Kingsville. They alleged it would be ultra vires for the Commissioner to publish 2024 school ratings based on performance “measures, methods, and procedures” that were adopted (A) at the end of the school year rather than at the beginning; and (B) too late to allow a “mathematical possibility” for all districts and schools to receive an A rating. We reach the same conclusions here as we did in Kingsville.

A. The Commissioner can adopt standards at any time in a school year Chapter 39 grants the Commissioner legal authority to “adopt indicators and standards . . . at any time during a school year.” 17 Unlike the facts in Kingsville, in this case the Commissioner timely adopted the 2024 standards during the 2024 school year—issuing a short five-page overview of the accountability system in March of 2024, and publishing a 307-page Accountability Manual by rule on May See, e.g., Honors Acad., 555 S.W.3d at 68.

City of Austin v. Powell, 704 S.W.3d 437, 447 (Tex. 2024).

See Houston Belt & Terminal Ry. Co. v. City of Houston, 487 S.W.3d 154, 157–58 (Tex. 2016). The Districts do not allege immunity was waived for their claims under the Uniform Declaratory Judgment Act, see TEX. CIV. PRAC. & REM. CODE § 37.002(b), which waives immunity only for challenges to the validity of a law; claims like those here that an official failed to comply with the law must be brought under the ultra vires exception. See State v. Zurawski, 690 S.W.3d 644, 660–61 & n.33 (Tex. 2024).

§ 39.0541 (emphasis added).

10, 2024. 18 The Commissioner did nothing ultra vires by publishing documents within the statutory deadlines.

Yet the Districts complain that the Commissioner violated an implied deadline in § 39.0542, which requires that he “provide each school district a document . . . that explains the accountability performance measures, methods, and procedures that will be applied for that school year.”19 They say the phrase “will be applied” requires a definitive adoption of standards at the beginning of the school year. But § 39.0542 addresses when standards will be applied, not when they will be adopted—and they are not actually applied until the ratings are issued at the end of the school year.20 As we said in Kingsville, the statute allowing adoption of standards “at any time during a school year” plainly includes “up to the very last day of a school year.”21 Failing to adopt them earlier cannot be ultra vires.22 “We cannot write into Chapter 39 an earlier or more specific deadline” than Legislature did, nor can we “automatically cancel A to F ratings” for any delay when the Legislature did not.23 “[T]he clear legislative intent in Chapter 39 is to publish

49 TEX. REG. 3199, 3280–85, (effective May 14, 2024).

§ 39.0542(a) (emphasis added).

See § 39.054(a-3) (requiring ratings to be made publicly available by August 15th).

Kingsville, 710 S.W.3d at 926.

The Districts complain that the § 39.0542 document was not “definitive,” that it described only “proposed” changes, and that it was not formally adopted by rule. This mixes apples and oranges. The only document that Chapter 39 says must contain the “measures, methods, and procedures” of the ratings system (as pleaded by the Districts, specifically provides it was not intended for school administrators; it was designed to be a “simple, accessible” document that districts could “easily distribute to parents of students” to explain the ratings system. § 39.0542(b).

This overview did not have to be adopted by rule because it was an explanation of the rules rather the rules themselves. Id. By contrast, the manual for school professionals was the Accountability Manual, which was adopted by rule. But Chapter 39 does not specify what it must contain, and the Districts do not complain that its 300+ pages were insufficiently detailed.

Kingsville, 710 S.W.3d at 926–27. school ratings, not suppress them.”24 There is no evidence the Commissioner acted ultra vires in publishing the “measures, methods, and procedures” of the ratings system when he did.

B. The Commissioner can raise CCMR cutoffs at any time in a school year Chapter 39 also requires that the method used to evaluate school performance must be implemented in a manner that provides “the mathematical possibility that all districts and campuses receive an A rating” in both overall performance and in three statutory “domains” (school achievement, school progress, and closing the gaps).25 The Districts argue the Commissioner acted ultra vires by raising the cutoff scores for each A to F grade for college, career, and military readiness (“CCMR”) late in the school year, rendering it “mathematically impossible for many high schools to receive an A rating.” It was undisputed that CCMR ratings are “lagging indicators” based on statistics relating to students who left school the year before, since graduation rates, military enlistments, and completion of dual credit or industry courses cannot be quickly determined, reported, or addressed at the end of a school year.

The Districts do not argue that raising the cutoff late in the school year put an A rating permanently out of their reach; they argue only that adopting it late in 2024 made it impossible for them to do anything about it in the same year. But as our opinion in Kingsville reflects, this is simply one application of their previous argument that the Commissioner cannot adopt new standards unless schools have

Id. at 926.

§ 39.054(b). The parties agree the Legislature intended this 2017 amendment to end grading schools on a curve, where the bottom 5% received an F no matter how well they performed or improved. enough time to respond to them. 26 Since the Districts do not argue that reaching the higher cutoff was permanently impossible, their complaint is only about its timing.

But again, Chapter 39 allows the Commissioner to change standards “at any time during a school year,” 27 and the standards necessarily include cutoff scores.

We understand the disappointment and frustration of districts and schools, teachers and students who anticipate higher grades for improving their academic achievement but do not get them because the cutoff score for an A was changed too late to do much about it. Yet increasing the level of expected performance from schools and their students is “a feature rather than a flaw in the statute,”28 which requires the Commissioner to “modify standards to continuously improve student performance,” 29 and authorizes him to do so “at any time during a school year.”30 In applying the “mathematical possibility” requirement in § 39.054(b), not knowing in advance what higher score would be needed to receive an A does not mean an A was mathematically impossible.31 To the contrary, the Districts’ own argument that they could have gotten higher scores had they had earlier notice (and evidence from some districts that they did get higher scores in later years shows higher scores were mathematically possible. Late adoption may have made it physically impossible to get an A in the same year the new cutoffs were announced (a timing component that § 39.054(b) does not address). But that did not make it

See Kingsville, 710 S.W.3d at 926–27.

§ 39.0541 (emphasis added).

See Kingsville, 710 S.W.3d at 927.

§ 39.053(f) (emphasis added).

§ 39.0541 (emphasis added).

See Possibility, BLACK’S LAW DICTIONARY 1410 (12th ed. 2024) (“The quality, state, or condition of being conceivable in theory or in practice; . . . the word often (but not always) conveys a sense of uncertainty or improbability.”) (emphasis added)). mathematically impossible. The Districts’ argument stands or falls on the assumption that schools have a legal right to cancel school ratings for 2024 whenever they cannot immediately meet new standards that are adopted late in the school year.

We hold that is not ultra vires conduct.

II. The Ultra Vires Claims based on STAAR Tests The Districts assert two additional ultra vires claims, one old and one new.

They argue the Commissioner had no legal authority to publish 2024 ratings based on STAAR tests that (A) were not determined to be valid and reliable by an “independent” entity; and (B) were scored by automated machine grading they allege was unreliable. We rejected the first in Kingsville and reject the second below.

A. The STAAR tests were independently validated The Districts assert that issuing 2024 school ratings would be ultra vires if based on 2024 STAAR tests that were not determined to be “valid and reliable” by an “independent” entity as required by the statute.32 Chapter 39 requires the Commissioner to appoint a Technical Advisory Committee (TAC) of national experts on educational assessments and psychometrics to make that determination.33 In Kingsville, we applied the ordinary meaning of independent (“not subject to control by others”34), and held that relying on the TAC’s validation was not ultra vires without evidence that the Commissioner or TEA controlled its conclusions or work. That evidence is again missing here.

See § 39.023(a-11) (“Before an assessment instrument . . . may be administered . . . , the assessment instrument must, on the basis of empirical evidence, be determined to be valid and reliable by an entity that is independent of the agency and of any other entity that developed the assessment instrument.”).

See § 39.02302(a).

See Kingsville, 710 S.W.3d at 928 (citing Independent, MERRIAM-WEBSTER’S COLLEGIATE DICTIONARY 633 (11th ed. 2020)).

We also held in Kingsville that Chapter 39 draws a “bright statutory line between those who develop tests and those who advise the Commissioner on validity.” 35 The statute explicitly requires the TAC “to advise the commissioner and the agency regarding the development of valid and reliable assessment instruments.” 36 The Districts say that means TAC’s role is “to help develop” STAAR tests, putting it on the wrong side of the advise/develop divide. But advising the Commissioner regarding a test developed by a vendor is different from developing the test itself. And giving advice “regarding the development” of a standardized test implies that the advice will occur during development rather than after it. Erasing the line separating “advice regarding development” from “helping to develop” would mean the Commissioner could not get advice from experts during development without automatically disqualifying those experts as not “independent.” We hold the Commissioner did not commit an ultra vires act by getting advice from the independent panel of experts that Chapter 39 created for just that purpose.

B. “Unexplained” STAAR results are no evidence of invalidity The Districts complain that the 2024 ratings cannot be based on the 2024 STAAR tests administered a year ago due to the introduction of automated machine scoring for “constructed response” questions. 37 In 2024, an automated process graded 75% of the tests, and humans graded the remaining 25%. If a parent or district believed a machine-graded score was incorrect, they could submit it for regrading by a human grader.

See id. at 929.

§ 39.02302(a) (emphasis added).

In constructed-response questions, students write their own answer, instead of choosing one from a list (“selected response”).

Automated grading was not an experiment dreamed up by bureaucrats but a necessity due to legislative changes. Beginning in 2023, the Legislature required that constructed response questions constitute at least 25% of STAAR tests. 38 As the name implies, each student constructs their own response to those questions rather than pick one answer from a list, with the possible result that no two answers are the exactly same. The 2024 STAAR tests all told contained about 15.8 million answers to constructed response questions— which Chapter 39 requires to be graded and reported to districts within 21 days of the test.39 TEA estimated it would cost an additional $15–$20 million to grade these all by hand. At a legislative hearing, the Commissioner testified that “transitioning to automated scoring would be necessary unless we had additional funding from the state.” That additional funding was apparently not forthcoming.

The Commissioner offered testimony from several members of the TAC about the validity and reliability of automated grading. Dr. Andrew Ho is a professor at Harvard’s Graduate School of Education, a psychometrician,40 president of the National Council on Measurement in Education,41 and a member of advisory committees in Texas, California, New York, Massachusetts, Rhode Island, and Maryland. He testified that automated grading predicts what human graders would

§ 39.023(c–8) (“Beginning with the 2022-2023 school year, not more than 75 percent of the available points on an assessment instrument developed under Subsection (a) or (c) may be attributable to questions presented in a multiple choice format.”).

§ 39.023(h).

See Psychometrician, MERRIAM-WEBSTER’S COLLEGIATE DICTIONARY 1004 (11th ed. 2020) (“A person (such as a clinical psychologist) who is skilled in the administration and interpretation of objective psychological tests.”).

“The National Council on Measurement in Education (NCME) is a U.S. based professional organization for assessment, evaluation, testing, and other aspects of educational measurement.”

National Council on Measurement in Education, https://en.wikipedia.org/wiki/National_ Council_on_Measurement_in_Education (last visited June 20, 2025). do by analyzing hundreds of actual examples (far more that any one person could do), a process that “saves time, saves money, does better than what you would have gotten had you had another human look at it.” A second member of the TAC also testified, Dr. James Pellegrino, former professor of psychology and education at universities including Illinois–Chicago, Pittsburgh, U.C. Santa Barbara, and Vanderbilt. When questioned about the Districts’ assertion that automated scoring was not valid and reliable, Dr. Pellegrino stated: “I do not know of any evidence that really—that supports that assertion.” These experts and others supported the validity and reliability of automated scoring in great detail throughout 160 pages of testimony. We need not summarize it all here because the Districts offered no expert testimony whatsoever to contradict it.42 The Districts instead offered anecdotal evidence from school administrators with “concerns” and “questions” about some of the results. But the validity and reliability of standardized tests and automated scoring them is not a matter of common knowledge; expert testimony is required.43 The Districts offered none.

“When expert testimony is required, lay evidence supporting liability is legally insufficient.” 44 The Districts’ witnesses were experts in education, but none claimed to be psychometricians or experts in standardized tests. They pointed to several examples of “dramatic,” “very unexpected,” and “unexplained” results from the new STAAR tests that they believed “raise questions.” But questions that might occur to laypersons are not some evidence to contradict competent expert evidence

See, e.g., TEX. R. EVID. 705(a) (providing “an expert may state an opinion . . . without first testifying to the underlying facts or data”).

See, e.g., Helena Chem. Co. v. Cox, 664 S.W.3d 66, 75 (Tex. 2023) (“Expert testimony is required when an issue involves matters beyond jurors’ common understanding.” (quoting Mack Trucks, Inc. v. Tamez, 206 S.W.3d 572, 583 (Tex. 2006))).

City of Keller v. Wilson, 168 S.W.3d 802, 812 (Tex. 2005). when the latter is required.

For example, one of the Districts’ witnesses testified she could not “understand and explain an eight percentage point drop” in fifth-grade science scores for 2024. But in this case as in many others, “there were simply too many potential causes to assume from the one that the other must have been the culprit.”45 As TEA’s director of assessment testified, an alternate explanation is that fifth graders in the spring of 2024 were also first graders in the spring of 2020—when schools physically closed and moved online for lengthy periods during the pandemic. As Dr. Ho opined, “all the evidence that we reviewed points to the fact that it’s not the scoring system” that is at fault, but “a real decline in achievement that we should be concerned about and not try to sweep under the rug.”

Other District witnesses testified to individual concerns they had over particular questions or in isolated areas. But over three million STAAR tests were administered in 2024, averaging 30 to 60 questions per test. With this universe of over 100 million questions, the so-called “Law of Truly Large Numbers” recognizes that: “With a large enough number of opportunities, any outrageous thing is likely to happen.”46 Very large samples will always display coincidental “patterns” that seem highly unlikely to those of us who are not statisticians; 47 for example, a coin flip will likely produce 10 heads in a row if repeated a thousand times.48 Without

See Nissan Motor Co. Ltd. v. Armstrong, 145 S.W.3d 131, 137 (Tex. 2004) (requiring for causation of engine fire, “a specific defect must be identified by competent evidence and other possible causes must be ruled out”).

See Law of truly large numbers, THE CAMBRIDGE DICTIONARY OF STATISTICS 244 (B.S.

Everitt & A. Skrondal eds, 4th ed. 2010).

See Coincidences, id. at 90–91 (“the one chance in a million will undoubtedly occur, with no less and no more than its appropriate frequency, however surprised we may be that it should occur to us”).

See, e.g., United States v. Ford, 683 F.3d 761, 768 (7th Cir. 2012). competent expert testimony, such lay testimony is no evidence of invalidity.

The Districts’ witnesses also testified they had considerable success when selected tests they questioned were submitted for rescoring by humans and came back with higher scores. But constructed response questions do not lend themselves to precise scores; without knowing the questions and the range of answers, one can conclude from rescoring whether the automated system and a human grader agreed, but not necessarily which one was “right.”

Finally, the Districts’ brief repeatedly claimed that an early study of the Commissioner’s automated scoring system found it was “not successful.” But that comment was taken out of context from an early field test; the experts all testified they “never rely just on the field-test data” because it is merely a first-run model from which answers are collected and scored by humans, and then used to improve the automated system to “make it more accurate and more consistent with the human scoring.” After further reprograming, “[t]he automated scoring engine met the performance criteria for all items.”

Like any other plea to the jurisdiction, the trial court had to consider evidence adduced by the parties in connection with the Commissioner’s jurisdictional plea before confirming its jurisdiction.49 Absent competent expert testimony supporting the Districts’ claims challenging the automated scoring system, there is no evidence to support an ultra vires claim on this ground.

C. A response to the dissent The dissent argues we should give each of the parties half a loaf—reversing the temporary injunction so the 2024 ratings are published after a 10-month delay, but not the Commissioner’s plea to the jurisdiction, which would be left hanging for Fraley v. Tex. A & M Univ. Sys., 664 S.W.3d 91, 97 (Tex. 2023). “targeted discovery” of an unspecified nature and duration. For several reasons, we disagree.

First, whether automated scoring is valid and reliable is an important and highly technical issue—but it is not a “jurisdictional fact.” The Commissioner is immune from suit unless his actions were ultra vires, i.e., “conduct that is unlawful.”50 A valid ultra vires claim “does not alter government policy but enforces existing law”;51 the Districts thus had the burden to raise a fact question on whether the Commissioner’s acts “are unlawful under current law.” 52 Automated grading cannot meet that standard because state law says nothing about grading STAAR tests by hand or by machine. Chapter 39 requires that STAAR tests “provide reliable information relating to a student’s satisfactory performance,”53 but also expressly provides how that should be determined: (1) the Commissioner appoints “experts on educational assessments and psychometrics” to make that determination 54 (which he did); and (2) the TAC determines whether the test is “valid and reliable”55 (which it did). The Districts’ concede this in their brief: [T]he Legislature did not give [the Commissioner] the discretion to determine whether the STAAR test is reliable. Instead, in Section 39.023(a-11), the Legislature gave the Commissioner discretion to develop the STAAR test and imposed a mandatory requirement that . . .[he] must first have an independent third party determine that the test is valid and reliable based on empirical evidence.

We affirm again today that the TAC is independent. Since the Commissioner strictly Tex. Educ. Agency v. Houston Indep. Sch. Dist., 660 S.W.3d 108, 116 (Tex. 2023). Id. Id. § 39.023(a-11).

§ 39.02302(a).

§ 39.023(a). complied with Chapter 39, his actions cannot be ultra vires without re-writing the statute.

Second, even if we were to imply an order by the trial court denying targeted discovery (as the dissent suggests), we cannot also imply that it was an abuse of discretion. The trial court’s temporary restraining order barring publication of the 2024 ratings was set to expire the day after the temporary injunction hearing below was concluded. The Districts urged the trial court to extend it for an additional 14 days for the discovery they wanted. 56 But the trial court exercised its discretion by declining to do so. Rule 192.4 of the Texas Rules of Civil Procedure provides that discovery “should be limited by the court if it determines . . . the burden or expense of the proposed discovery outweighs its likely benefit.” 57 The trial court must have concluded either that the discovery sought would take far more than 14 days, and/or that the benefits would be illusory if the 2024 ratings were issued before it was concluded. We cannot imply that neither consideration played a role in the implied denial, nor that either was an abuse of discretion.

Third, “targeted discovery” is a euphemism the dissent employs 12 times in as many pages. But that is not what the Districts requested. The “targeted discovery” here involves fishing in a pond containing millions of test answers. A month before the temporary injunction hearing, the Districts’ counsel acknowledged that some documents they requested did not exist, some had to obtained from the vendor that developed the tests, and the rest required TEA to create spreadsheets detailing rescoring requests (2021–24), 5th-grade science scores (2023–24), extended constructed response scores (2023–24), and Domain 2A growth results (2022–24) broken down by school, district, number of students, percentages in each ratings TEX. R. CIV. P. 680.

Id. 192.4(b). category, and by student circumstances (economically disadvantaged, emergent bilingual, and special needs). To ensure the requested scores were not anomalous or taken out of context, the Commissioner would naturally need to gather millions of other scores broken down in these ways to represent the complete picture.

We disagree with the dissent that the timing and availability of discovery in jurisdictional pleas is the same as in summary judgment practice. While no-evidence review of the jurisdictional pleas “generally mirrors that of a summary judgment,”58 the operative word is “mirrors”—not “replicates.” Mirrors reflect a real thing; they are not the thing itself.59 The no-evidence review for legal sufficiency is the same for both,60 but the timing is not. The Rules of Civil Procedure provide that no- evidence summary judgments may be considered only “[a]fter adequate time for discovery,”61 but pleas to the jurisdiction are not governed by any rule in the rules book, but by caselaw providing that a trial court must consider them “at its earliest opportunity,” always being “mindful that this determination must be made as soon as practicable.”62 This is required because immunity “protect[s] the public fisc by shielding tax resources from being diverted to pay litigation costs and money judgments.”63 One reason the Texas Supreme Court did not apply the summary judgment rules to pleas to the jurisdiction may have been that “[g]overnmental units City of Austin v. Powell, 704 S.W.3d 437, 446 (Tex. 2024) (quoting Tex. Dep’t of Parks & Wildlife v. Miranda, 133 S.W.3d 217, 228 (Tex. 2004)).

MERRIAM-WEBSTER’S COLLEGIATE DICTIONARY 793 (11th ed. 2020) (“To reflect in or as if in a mirror; resemble”).

See City of Keller v. Wilson, 168 S.W.3d 802, 823 (Tex. 2005) (“[T]he test for legal sufficiency should be the same for summary judgments, directed verdicts, judgments notwithstanding the verdict, and appellate no-evidence review.”).

Tex. R. Civ. P. 166a(a), (b), (i).

Miranda, 133 S.W.3d at 226.

Hidalgo Cnty. Water Improvement Dist. No. 3 v. Hidalgo Cnty. Irrigation Dist. No. 1, 669 S.W.3d 178, 183 (Tex. 2023) (emphasis added). may incur unnecessary discovery costs and delays unless judges agree to hear summary judgment motions on jurisdictional matters as early in the case as they might hear a plea to the jurisdiction.”64 We do not hold (as the dissent says) that trial courts “do not need to allow for any discovery before ruling” on jurisdictional pleas. Courts have occasionally reversed dismissal on jurisdictional grounds for a little discovery, depending on how little that discovery is likely to be. In Hall v. City of Jersey Village (the case on which the dissent relies), a restaurant manager at a city’s golf course sued when she was hit on the head by a cart attendant’s errant golf ball. 65 This was targeted discovery (in several ways), primarily because it was limited: did the attendant act in the course and scope of employment with the city or in his private capacity? But the immunity question here (“and potentially other related evidence” says the dissent) presents millions of potential targets to explore.

We do not hold that governmental immunity can be extended simply by stonewalling discovery; we hold only that jurisdictional pleas cannot be postponed indefinitely without more limited requests than the Districts proposed here.

Conclusion For three years, A to F school ratings were not released because of the pandemic, but for the last two years they have not been released because of the courts. Millions of dollars and thousands of hours of work by teachers, administrators, and experts have been invested in creating the A to F ratings system; courts can decide only whether it is legal, not whether it is wise or fair, much less commandeer the job of running it. No one disputes that the Commissioner has

Miranda, 133 S.W.3dat 244 (Brister, J., dissenting). 2023 WL 3873351, at *1 (Tex. App.—Houston [1st Dist.] June 8, 2023, no pet.). repeatedly invited, encouraged, and considered complaints from stakeholders about the system, but a shrinking group of school administrators and school attorneys have simply thwarted it from taking effect.

In an amicus brief, an impressive group of education experts and activists, businesses, parents, students, and taxpayers point out that A to F accountability ratings “are the only meaningful tool that ordinary Texans have for assessing public school performance . . . . there is simply no substitute for them.” They argue that it is difficult for businesses to attract employees to communities without reliable information on local schools, or to hire well-prepared employees locally if there are too few. Nor can taxpayers tell whether the State’s comparatively high property taxes are being spent efficiently or effectively by the schools that consume the major part of them. The policy branches of government decided to mandate an A to F performance rating for schools to address these problems, and no one claims doing so was unconstitutional. It is time for local courts to stop obstructing those policies.

The record here contains no evidence to support the Districts’ claims that issuing A to F ratings for the 2024 school year would be an ultra vires act beyond the Commissioner’s legal authority. The trial court thus erred by denying the Commissioner’s plea to the jurisdiction. And because a court lacking jurisdiction cannot award injunctive relief, not “even temporarily,” 66 the trial court also erred by granting the Districts’ temporary injunction. We vacate both orders, dissolve the temporary injunction, and render judgment dismissing the Commissioner and the claims made against him for want of jurisdiction.67

In re Abbott, 601 S.W.3d 802, 805 (Tex. 2020) (orig. proceeding).

The trial court also denied petitions in intervention filed by three school-board members, IDEA Public Schools, and the Brewer Foundation. The intervenors alleged adverse effects from the Districts’ temporary injunction and sought declaratory relief and attorneys’ fees and costs _____________ Scott Brister Chief Justice Before Chief Justice Brister and Justices Field and Farris.

against the Districts for filing the 2024 suit. Those parties filed separate petitions for writ of mandamus or appeals, which remain pending.