Wesley Rattray, Marco Nunez, Martha Saavedra, Antonio Vindell, Carmen Pashos, Steve Tullos, Cesario Pedraza, Minerva Pedraza, Roger Luly, Nora Gonzalez, and Rosalinda Castillo v. City of Brownsville, Texas

Supreme Court of Texas

══════════

No. 20-0975

══════════

Wesley Rattray, Marco Nunez, Martha Saavedra, Antonio

Vindell, Carmen Pashos, Steve Tullos, Cesario Pedraza, Minerva

Pedraza, Roger Luly, Nora Gonzalez, and Rosalinda Castillo,

Petitioners,

v.

City of Brownsville, Texas,

Respondent

═══════════════════════════════════════

On Petition for Review from the

Court of Appeals for the Thirteenth District of Texas

═══════════════════════════════════════

Argued October 27, 2022

JUSTICE YOUNG delivered the opinion of the Court.

The Texas Tort Claims Act waives governmental immunity for

“property damage” that “arises from the operation or use of . . . motor-

driven equipment.” Tex. Civ. Prac. & Rem. Code § 101.021(1)(A). We

must decide whether a city’s closure of a stormwater gate during a

rainstorm, which immediately preceded the flooding of a neighborhood,

falls under this waiver of immunity. We hold that petitioners have

1

successfully invoked the statutory waiver, at least at this stage of the

case. We accordingly reverse the judgment of the court of appeals and

remand to the trial court for further proceedings.

I

The plaintiffs below, and petitioners in this Court, are eleven

homeowners who live in the Quail Hollow subdivision of Brownsville.

The homeowners allege that the overflow of a nearby resaca 1 flooded

their homes and caused extensive property damage. According to their

seventh amended petition, the accumulation of water in the resaca—and

its resulting overflow—would not have occurred but for the City’s decision

to close a stormwater gate during a severe rainstorm.

The rainstorm rolled into the City around noon on a warm day in

late August 2015. Jose Figueroa, the City’s stormwater manager, was

enjoying lunch at a local Chick-fil-A at the time. When Figueroa noticed

the rain’s intensity, he left the restaurant and directed his crew to report

to their assigned posts and monitor the situation. He also instructed

Leo Saldivar, the City’s raw-water technician, to meet him near the

Quail Hollow subdivision. Figueroa wanted to keep an eye on the

stormwater flowing from the Resaca de la Guerra—a resaca running

through Quail Hollow that serves as part of the City’s water-drainage

system and which the homeowners describe as a “large waterway.”

As in other neighborhoods in the City, Quail Hollow’s stormwater

1A “resaca” is a “former course or channel of a stream.” Webster’s New

International Dictionary 2117 (2d ed. 1934). As the court of appeals noted, the

resacas here are “former channels of the Rio Grande found in the southern half

of Cameron County.” 647 S.W.3d 710, 714 n.2 (Tex. App.—Corpus Christi–

Edinburg 2020) (internal quotations omitted).

2

drains into that resaca, which generally flows from west (upstream) to

east (downstream). To control the flow of water, the resaca has five

sluice gates that open and close. One of those gates—the North Laredo

Gate—is immediately downstream of Quail Hollow. The North Laredo

Gate typically is kept open. According to the homeowners, motor-driven

actuators installed sometime before the rainstorm can remotely open

and close the North Laredo Gate.

Petitioners’ Oral Argument Exhibit

When Figueroa and Saldivar arrived in Quail Hollow, they

observed that the North Laredo Gate was open, as usual; that water in

the resaca was flowing normally, from upstream to downstream; and

that the resaca’s water was at a normal level, about thirty feet above sea

level. After making these assessments, Figueroa and Saldivar proceeded

to check other locations in the City. Except for “high water” in some

3

areas, they observed similar conditions. The South Laredo Gate, for

instance, was also open and had normal waterflow.

Figueroa and Saldivar returned to the North Laredo Gate about

an hour later. Conditions had changed. They noticed that the resaca’s

waterflow had reversed course and was now flowing from downstream

to upstream (what the parties refer to as “negative waterflow”). In

response, Figueroa closed the North Laredo Gate. He did that, he says,

to prevent the resaca’s water from overflowing into Quail Hollow.

After Figueroa closed the gate, he and Saldivar continued their

watch in the unrelenting rain. They returned to various areas in the

western part of the City and decided to place portable water pumps near

some of the other gates to push water downstream and away from Quail

Hollow. They also observed, again, that the water near the South Laredo

Gate had normal waterflow. But because of high water levels in other

areas, they carried on with the task of pumping water out of the resaca.

After Figueroa and Saldivar began pumping water out of various

parts of the resaca, they returned to the North Laredo Gate, which was

still closed. Once again, conditions had worsened. Amidst the heavy

rainfall and negative waterflow, the water had risen to about knee-deep

over Laredo Road and begun spilling over into Quail Hollow. Both

Figueroa and Saldivar witnessed the water flow over Laredo Road; there

was nothing they could do to stop it, they said. All told, the severe

“supercell” rainstorm dropped about four to six inches of water onto the

City within approximately three hours.

The homeowners allege that about two feet of water from the

resaca spilled over its banks and into their homes. To recover for their

4

property damage, the homeowners sued the City for negligence under

the Tort Claims Act. They alleged that the City and its employees

should have known that abnormal waterflow at the North Laredo Gate

was only “temporary,” that closing the gate would trap the water, and

that its resulting accumulation would cause the resaca to overflow and

flood their neighborhood. Based on those allegations, the homeowners

invoked § 101.021(1)(A) of the Act,2 which waives immunity for

“property damage” that “arises from the operation or use of . . . motor-

driven equipment.”3 In response, the City filed a plea to the jurisdiction.

It asserted, among other things, that the homeowners’ allegations

concerned the nonuse, rather than the use, of motor-driven equipment,

and that there was no evidence that the North Laredo Gate’s closure

caused the homeowners’ property damage.

The trial court denied the City’s plea. A divided court of appeals

reversed. 647 S.W.3d 710 (Tex. App.—Corpus Christi–Edinburg 2020).

The majority sided with the City on both points. As for the first, it held

that “the gravamen of the [homeowners’] complaint is based on the

City’s nonuse of the North Laredo Gate,” id. at 718, underscoring the

homeowners’ various allegations about how the gates were not opened

and how the pumps were not activated, which allegedly led to the

overflow, id. at 718–19. As for the second, the majority held that “the

2 Unless otherwise stated, references to and citations of statutory

provisions are to the Texas Tort Claims Act as codified in the Texas Civil

Practice and Remedies Code.

3 In addition to alleging misuse of the North Laredo Gate, the

homeowners also pleaded that their damages arose from misuse of the motor-

driven water pumps. Whether those pumps would also qualify under the statute

is not before us and we express no opinion on it.

5

mere act of closing or failing to open a gate does not cause flooding,”

because “[i]f no rainstorm had occurred . . . and the City closed the North

Laredo Gate, the homeowners would not have suffered property damage.”

Id. at 720. Accordingly, the court of appeals directed the trial court to

dismiss the homeowners’ suit for lack of jurisdiction. Id. at 722. The

homeowners then filed a petition for review, which we granted.

II

We begin our review of the court of appeals’ judgment by addressing

the jurisdictional nature of the homeowners’ suit.

A

The homeowners assert tort claims against the City, a subdivision

of the State of Texas. They therefore must overcome the City’s

governmental immunity. Travis Cent. Appraisal Dist. v. Norman, 342

S.W.3d 54, 57–58 (Tex. 2011). They can do so only by demonstrating

that the legislature, as the branch of government constitutionally

empowered to manage the State’s financial affairs, has waived

immunity by statute. Fed. Sign v. Tex. S. Univ., 951 S.W.2d 401, 405

(Tex. 1997). Without such a waiver, the court would lack jurisdiction to

proceed. Accordingly, if a Tort Claims Act plaintiff cannot satisfy “the

burden to affirmatively demonstrate the trial court’s jurisdiction” by

showing that the claim falls within a statutory waiver of immunity, the

court must dismiss the suit. Town of Shady Shores v. Swanson, 590

S.W.3d 544, 550 (Tex. 2019); see also Tex. Dep’t of Parks & Wildlife v.

Miranda, 133 S.W.3d 217, 226 (Tex. 2004).

The requirement to show that a claim falls within a statutory

6

waiver of immunity sounds simple enough, but the volume of our Tort

Claims Act cases illustrates how complex it often turns out to be. That

complexity is at least mitigated by focusing on the foundational principle

that the Act’s text and structure guide every determination of whether

a plaintiff has met the burden to show a waiver. The starting point is

always the status quo: a presumption against any waiver until the

plaintiff establishes otherwise. To do so, a plaintiff may invoke various

provisions in Subchapter B of the Act that affirmatively describe when

immunity is waived. Without such a provision, no court is empowered

to hear tort cases against and impose liability on “governmental unit[s].”

The Act’s detailed descriptions of the contours of the waiver of immunity

delineate the extent of the judicial authority. A plaintiff must begin,

therefore, by alleging circumstances that fit within a provision of the Act

that authorizes a waiver, such as the homeowners’ assertion here that

the flooding of their homes followed from circumstances described in

§ 101.021(1)(A).

But a plaintiff cannot stop there. The Act also provides various

exceptions or caveats that function as a withdrawal of the waiver, and

thus of the court’s jurisdiction to proceed, under certain conditions.

Some exceptions turn on the kind of defendant who is named. Even

when allegations would otherwise describe circumstances within the

Act’s waiver, for example, the Act withdraws that waiver for any “claim

arising from the activities of the state military forces when on active

duty under the lawful orders of competent authority.” § 101.054; see

also, e.g., §§ 101.052–.053 (similar limitation for suits against legislators

and judges). Other exceptions to the waiver may turn on the legal theory

7

that a plaintiff deploys. If, for example, a claim against certain

governmental defendants depends on “[t]he common law doctrine of

vicarious liability because of participation in a joint enterprise,”

§ 101.0211, then any waiver of immunity is withdrawn as to that claim

even if the claim otherwise would be within the waiver. A third type of

exception eliminates the waiver for certain kinds of actions. For example,

the waiver of immunity does not apply to “a claim arising . . . from the

action of an employee while responding to an emergency call or reacting

to an emergency situation if the action” either complied with laws

governing emergency situations or, “in the absence of such a law . . . , if

the action is not taken with conscious indifference or reckless disregard

for the safety of others.” § 101.055(2). Nor is immunity waived for

claims challenging the failure to undertake a discretionary act.

§ 101.056 (titled “Discretionary Powers”).4

Finally, assuming that a plaintiff successfully establishes a waiver

and negates any relevant expressed withdrawal of the waiver, the Act

also contains some provisions that operate to limit recovery. For

example, § 101.024 eliminates the availability of punitive damages and

§ 101.023 “limits the amount of the government’s liability . . .

depend[ing] on the type of governmental unit being sued[.]” Gulf Coast

Ctr. v. Curry, 658 S.W.3d 281, 285 (Tex. 2022). As we recently said in

Curry, “a trial court must ascertain, as part of determining its

4 “This chapter does not apply to a claim based on: (1) the failure of a

governmental unit to perform an act that the unit is not required by law to

perform; or (2) a governmental unit’s decision not to perform an act or on its

failure to make a decision on the performance or nonperformance of an act if

the law leaves the performance or nonperformance of the act to the discretion

of the governmental unit.”

8

jurisdiction, whether and to what extent the Tort Claims Act waives

immunity from suit.” Id. at 286.

B

Consideration of each of those defining features of immunity is

essential to the jurisdictional inquiry that lies at the heart of litigation

under the Texas Tort Claims Act. Taken together, these features

implicate the judicial branch’s very authority to act—a limitation on

jurisdiction that protects the separation of powers inherent in our

constitutional order. In private litigation, parties risk no larger

constitutional values if they waive or forfeit claims and defenses that

are personal to themselves; such matters lack jurisdictional significance

in any sense. The Tort Claims Act, though, represents a delicate balance

that the legislature alone can strike. Even if a governmental unit would

be happy to waive “its” immunity, it is not the governmental unit’s

immunity to waive. More to the point, the courts cannot be part of any

such transaction. One key reason that we use the “jurisdictional” label

is because it reflects the judicial obligation to adjudicate only those

claims that are authorized without trespassing on the authority of the

political branches.

This concern for ensuring the integrity of the judicial process—and

the entirety of the separation of powers—is why courts that detect a non-

waivable jurisdictional problem may at any time demand that the parties

resolve the resulting doubts about whether the case may proceed. A court

that raises a jurisdictional issue does not aim to serve the interests of

one side or the other. Instead, such a court discharges its duty to ensure

that the court itself is functioning in an authorized and properly judicial

9

capacity. For the same reason, the parties likewise remain duty-bound

for as long as they invoke or submit to a court’s authority to confirm the

presence of jurisdiction and to raise jurisdictional defects.

As a general matter, a plaintiff initially discharges this burden by

alleging facts that bring a claim within the waiver. But being “within”

the waiver entails both key parts described above: satisfying the

provisions that clearly and affirmatively waive immunity and negating

any provisions that create exceptions to, and thus withdraw, that

waiver. The Act may waive immunity in one breath and in the next take

back part of the waiver. Both parts of the analysis are needed to answer

the common issue that underlies every Tort Claims Act case: whether

immunity is waived. Imagine a building with this sign: NO VISITORS

WELCOME (akin to the default status of “no suits permitted” because of

immunity), followed by EXCEPT IN THE SUMMER (like the affirmative

statutory waivers of immunity). A visitor on the first Saturday in

August, though, would still be a trespasser if the sign continued BUT

NEVER ON WEEKENDS (playing the role of the Act’s various withdrawals

of parts of the waiver of immunity). The sign could have said VISITORS

WELCOME ON SUMMER WEEKDAYS ONLY. And the Act could have

phrased its own waivers in highly detailed terms that do not rely on

separate exceptions. The consequence of how the legislature actually

framed the Act, however, is that plaintiffs can affirmatively establish

jurisdiction only if they show that their claims are not within the

withdrawn part of the waiver.

To be sure, no great effort is needed to negate many or indeed

most statutory exceptions to the general waiver of immunity. A plaintiff

10

has never had to march through the Act provision by provision. Pleadings

whose affirmative allegations functionally exclude the exceptions are

typically sufficient. Plaintiffs thus need only expressly negate those

exceptions that their allegations plausibly implicate, which will depend

on the nature of the dispute. It will be readily apparent in most cases,

for example, that the defendant is not a state legislator and is not being

sued for an “act or omission . . . in his official capacity” as a legislator,

§ 101.052, so plaintiffs will not need to affirmatively or explicitly negate

that exception (or any other plainly inapplicable one) in their pleadings.

A governmental defendant, in turn, plays its role by identifying

where jurisdiction might be lacking and raising any such deficiencies,

most commonly in a plea to the jurisdiction, as the City deployed below.

If a plaintiff has not shown that a claim affirmatively falls within a

statutory waiver, the defendant should say so. Likewise, if the plaintiff

omits or otherwise fails to negate a relevant exception to the waiver, the

defendant should speak up. Both sides may develop and introduce

evidence to support their contentions. This iterative process will ensure

that both sides present their jurisdictional arguments to the court, which

can play its continuing role of assessing its own jurisdiction.

The rigor of proof required to satisfy a court that jurisdiction is

present increases at each stage of litigation, as with disputes over

traditional subject-matter jurisdiction. See, e.g., Lujan v. Defs. of

Wildlife, 504 U.S. 555, 561 (1992) (noting that jurisdictional “facts (if

controverted) must be supported adequately by the evidence adduced at

trial” (internal quotations omitted)). And as our decision in Curry

reflects, this jurisdictional analysis continues all the way past a verdict:

11

“When . . . the plaintiff establishes that the Act applies but fails to

establish which cap applies, the plaintiff has failed to demonstrate that

the trial court has jurisdiction to render a judgment exceeding the

minimum statutory cap.” 658 S.W.3d at 283.

Just one valid jurisdictional obstacle is enough for the court to halt

further proceedings. The fundamental rule is that the court may not

reach the merits if it finds a single valid basis to defeat jurisdiction. When

one such ground exists, it is not necessary that every other potential

jurisdictional defect be raised, fleshed out, or resolved at the outset.

When defendants challenge jurisdiction on multiple grounds,5 courts are

therefore not duty-bound to address them all if any one of them warrants

dismissal (although addressing multiple grounds, as the court of appeals

helpfully did in this case,6 may be more efficient and may prevent further

proceedings after an appeal). See, e.g., Sinochem Int’l Co. v. Malaysia

Int’l Shipping Corp., 549 U.S. 422, 431 (2007) (describing courts’ “leeway

to choose among threshold grounds for denying audience to a case on the

5Raising all reasonably available objections early is best practice and

most efficient even if it is not absolutely required. A defendant that sports

with the court by making seriatim, piecemeal objections to jurisdiction that

could have been raised up front may lose credibility, of course. Plaintiffs bear

the burden to establish jurisdiction, so they subject themselves to some risk by

waiting for a plea to the jurisdiction (or other appropriate vehicle) to establish

a waiver of immunity or to attempt to negate clearly relevant exceptions to the

waiver rather than doing so from the start. But we recognize that, in some

challenging cases, the parties may not anticipate their opponents’ positions,

which is why our process permits the development of the argument through

discovery and competing motions.

6 See 647 S.W.3d at 719 (observing, after rejecting jurisdiction on one

dispositive ground, that “even if the homeowners’ complaint stemmed from the

use of equipment, their damages did not arise from that use” and proceeding

to resolve the other issue presented, thus facilitating our review).

12

merits” (internal quotations omitted)). As a corollary, the court may not

move to the merits if even one jurisdictional argument remains unresolved.

The statement that courts have the authority and indeed the duty

to resolve any jurisdictional doubts that arise before proceeding to the

merits does not mean that we expect courts to become Inspector Javert,

hunting for defects that the parties do not see or raise. Courts are

empowered to note potential jurisdictional defects sua sponte, but the

adversary process remains the touchstone of litigation even in this

context. See Rusk State Hosp. v. Black, 392 S.W.3d 88, 102 (Tex. 2012)

(Hecht, J., concurring). True, we have often said that sovereign immunity

“implicates” a court’s subject-matter jurisdiction, e.g., Hous. Belt &

Terminal Ry. Co. v. City of Houston, 487 S.W.3d 154, 160 (Tex. 2016),

such that an opinion in the face of a valid assertion of immunity may

correctly be called “advisory,” Rusk, 392 S.W.3d at 95 (maj. op.). But, by

the same token, we have also said that immunity does not equate to

subject-matter jurisdiction. Engelman Irrigation Dist. v. Shields Bros.,

514 S.W.3d 746, 755 (Tex. 2017). So while the “implication” of jurisdiction

may allow a governmental entity to raise immunity arguments later

than at the outset of litigation—even for the first time on appeal, Rusk,

392 S.W.3d at 94–96—such issues do not allow that entity, for instance,

to collaterally attack a final judgment, Engelman, 514 S.W.3d at 751.

In summary, no court should proceed to the merits of a case

brought under the Act if that requires turning a blind eye to jurisdictional

concerns that the court itself perceives or that the parties have raised.

And in the event a court does perceive a potential problem that counsel

have not raised, the court retains discretion to reserve that issue if it is

13

already prepared to grant a plea to the jurisdiction for a different reason

or to otherwise dismiss a case.

C

As this case comes to us, the parties essentially dispute the

applicability of only one provision in the Tort Claims Act—§ 101.021(1)(A),

which waives immunity for property damage arising from the “operation

or use” of “motor-driven equipment.” We have just described the

plaintiff’s obligation to negate any potentially relevant exception at the

outset by, at a minimum, filing pleadings containing allegations that, if

true, would displace the applicability of such exceptions. In their briefing

to us, the parties have not disputed whether the City would otherwise

retain immunity even if the two grounds presented for our review do not

justify dismissal. Given that the court of appeals believed that it could

resolve the case on those two jurisdictional grounds (and either one would

suffice), that court had no obligation to consider or raise any further

jurisdictional issues, either.7

Regardless of whether one or more exceptions to the waiver of

immunity may be relevant, none is before us now. We could direct

supplemental briefing on any other jurisdictional question and proceed to

resolve that question ourselves in the first instance, but as a prudential

matter, the law is typically better served when the lower courts review

7 Cases involving statutory exceptions to the waiver of immunity in the

context of floods are legion. See, e.g., City of San Antonio v. Hartman, 201

S.W.3d 667, 671–73 (Tex. 2006) (considering the applicability of the emergency-

situation exception to a flood case); see also id. at 673 n.23 (noting other flood

cases). Our decision today should not be read to express any view regarding

whether those exceptions would apply here.

14

a legal issue before this Court does. “Ours is a court of final review and

not first view. Ordinarily, we do not decide in the first instance issues

not decided below.” Zivotofsky v. Clinton, 566 U.S. 189, 201 (2012)

(internal quotations omitted). Nor is there any need for us to do so. Our

decision today only reverses the court of appeals’ judgment on the

grounds considered by that court. We do not affirmatively declare that

there is jurisdiction. We instead remand for further proceedings, during

which the parties and lower courts may determine how best to comply

with the Act’s jurisdictional requirements.

III

We accordingly turn to the applicability of § 101.021(1)(A), which

provides as follows:

A governmental unit in the state is liable for:

(1) property damage, personal injury, and death

proximately caused by the wrongful act or omission

or the negligence of an employee acting within his

scope of employment if:

(A) the property damage, personal injury, or

death arises from the operation or use of a motor-

driven vehicle or motor-driven equipment.

The parties do not dispute that the North Laredo Gate has a motor.

Nor do they dispute that Figueroa, a City employee, closed it.8 They

8 The parties submitted conflicting evidence, however, as to whether

Figueroa closed the gate manually or by motor. For whatever reason, that

distinction is important to the waiver of immunity in the context of

§ 101.021(1)(A). Like the court of appeals, we view the evidence in the light

most favorable to the homeowners and presume, for the purpose of answering

the questions presented, that Figueroa closed the gate with its motor-driven

actuators. See Town of Shady Shores v. Swanson, 590 S.W.3d 544, 550 (Tex.

15

instead disagree as to whether the statutory phrase “operation or use”

encompasses the allegations and record facts established here and, if so,

whether the homeowners’ property damages “arose from” that operation

or use.9

A

We turn first to whether the City’s “operation or use” of the

equipment is even at issue here. The homeowners argue that the City

operated or used the gate by closing it to block water, which is the gate’s

intended purpose. This “use,” they claim, caused the flooding because

closing the gate allowed the water to accumulate and overflow into their

2019).

9The parties dispute, and the court of appeals addressed, only the

“arises from” standard in § 101.021(1)(A), and not the “proximately caused by”

standard in § 101.021(1). As shown by the statutory text, and as we have

previously noted, these are two “separate and independent requirements,” so

“[s]atisfying the ‘arises from’ requirement does not excuse a plaintiff from

demonstrating proximate cause.” City of San Antonio v. Riojas, 640 S.W.3d

534, 537 n.13 (Tex. 2022). On the other hand, while both requirements achieve

distinct goals, they do so in a related way. The “proximately caused”

requirement of § 101.021(1) simply ensures that the familiar and ordinary

proximate-cause standard in tort litigation (and not some lesser standard) be

the outer bound of liability here, too, and that it be tethered to a governmental

employee acting within the course and scope of his job. The “arises from”

requirement in § 101.021(1)(A), in turn, narrows the universe of actions that

fall within the statutory waiver—here, by including only certain actions with

a link (or “nexus”) to a motor-driven vehicle or equipment within the relevant

waiver. As we discuss below, see infra Part III.B, this “arises from” requirement

mandates an additional showing, but it is not a different kind of causation—it

is still a proximate-cause requirement, just one that focuses on motor-driven

equipment. In any event, like the applicability of certain exceptions to the Tort

Claims Act’s waiver of immunity addressed in Part II, we leave it to the parties

and lower courts on remand to address § 101.021(1)’s separate causation

requirement in the first instance, if need be.

16

homes. The City, on the other hand, argues that what the homeowners

truly allege is nonuse: the City’s failure to later open the floodgate and

thus relieve the accumulated overflow. We think that, however one

examines the essentially undisputed facts on the day in question, “the

operation or use” of the gate is inseparable from the homeowners’ tort

allegations and the resulting property damage that they claim.

Decades ago, we expounded on the meaning of “operation or use”

in the Tort Claims Act. We defined “use” as “to put or bring into action

or service; to employ for or apply to a given purpose,” Mount Pleasant

Indep. Sch. Dist. v. Estate of Lindburg, 766 S.W.2d 208, 211 (Tex. 1989),

and “operation” as “a doing or performing of a practical work,” id.

(internal quotations and citations omitted). These definitions,

admittedly, are “not particularly enlightening,” PHI, Inc. v. Tex. Juv.

Just. Dep’t, 593 S.W.3d 296, 303 (Tex. 2019), but they reflect our effort

to apply them in a way that conforms to common sense.

To that end, our cases require more than merely stating that

government property plays a role in an alleged injury. Instead, they

have been careful to distinguish between genuine use and nonuse. In

LeLeaux v. Hamshire-Fannett Independent School District, for example,

the plaintiff sued a school district for injuries she sustained when she

hit her head on the rear door of a school bus. 835 S.W.2d 49, 50–51 (Tex.

1992). We held that because the bus was “parked, empty, with the motor

off,” the plaintiff was unable to show “operation or use” of a motor-driven

vehicle and was thus unable to invoke the statutory waiver in § 101.021.

Id. at 51–52. Similarly, in Texas Natural Resource Conservation

Commission v. White, the plaintiff alleged that her store was destroyed

17

by fire after the Commission removed a motor-driven pump on her

property that dissipated gas fumes from a nearby gasoline tank. 46

S.W.3d 864, 866 (Tex. 2001). We held that those allegations did not

concern the “pump’s operation or use” because the pump was not even

“on [the plaintiff’s] property when the fire began.” Id. at 870 (internal

quotations omitted). In so holding, we declined to equate “use” with

“failure to use,” for doing so “would be tantamount to abolishing

governmental immunity” altogether. Id. at 870–71 (quoting Kerrville

State Hosp. v. Clark, 923 S.W.2d 582, 585 (Tex. 1996)). Hence our

admonition in City of North Richland Hills v. Friend: that plaintiffs

cannot “enlarge the scope of the waiver” by conflating use and nonuse

through “artful pleading.” 370 S.W.3d 369, 373 (Tex. 2012).

We remain heedful of this difference because, as we have often

noted and reiterate above in Part II, § 101.021 is “a limited waiver of

governmental immunity,” Alexander v. Walker, 435 S.W.3d 789, 790

(Tex. 2014), and its precise scope must be clearly expressed. At the same

time, however, we aim not to be unduly restrictive or to engage in

sophistry in our understanding of “operation or use.” They are, after all,

“nothing if not common, everyday words” that “should be given their

everyday meaning.” PHI, 593 S.W.3d at 303. We have therefore

declined to read the words of § 101.021 as “terms of art intelligible only

to experts in the case law,” id., or as language burdened by “nit-picking

technicalities” that do not “accompany other causes of action,” Antonin

Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts

285 (2012). As always, we construe the statutory text reasonably—and

this approach, happily and perhaps unsurprisingly, often yields

18

reasonable answers.

The Southwestern Reporter has no shortage of examples. Our

decisions in PHI and Ryder Integrated Logistics, Inc. v. Fayette County,

453 S.W.3d 922 (Tex. 2015), are two among many others. In Ryder, a

deputy sheriff pulled off on the side of a road and positioned his cruiser

to face oncoming traffic—in the middle of the night with his headlights on.

Id. at 926. The lights blinded a driver and a deadly car collision resulted.

Id. We held that these allegations involved the “operation or use” of a

“motor-driven vehicle” because the deputy was “not just operating the

headlights—he was driving the car” at the time of the accident. Id. at

928. Our most recent § 101.021(1)(A) decision, PHI, also involved the

disputed use of a vehicle. There, a state employee parked a van on an

incline near a helipad, failed to engage the emergency brake, walked

away, and allowed the van to roll down the incline and into a helicopter.

PHI, 593 S.W.3d at 300. True, the employee was not in the van at the

time it crashed, but these allegations still successfully invoked the

statutory waiver because engaging the emergency brake was “an integral

part of the ‘operation or use’ of a vehicle.” Id. at 303–04. That the employee

was not “actively operating” the van “at the time of the incident,” as the

officer was in Ryder, did not change our analysis. Id. at 305. “The

statute,” we noted, “does not explicitly require that the operation or use

be ‘active’ or that it be ongoing ‘at the time of the incident.’ ” Id.

Many of these cases, of course, have fact patterns that differ in

various ways from the one now before us. Taken together, though, they

illustrate how we have come to understand the scope of § 101.021(1)(A)’s

potentially confusing “operation or use” requirement. As we observed in

19

PHI, “the multiplicity of possible fact-patterns and the vagaries of

litigation can create complexity,” which is precisely why we strive for “a

simple construction” that will usefully guide the resolution of most cases

without requiring expensive and time-consuming litigation. Id. at 303.

Applying that guidance to the facts here requires us to respectfully

disagree with the conclusion reached by the court of appeals. The North

Laredo Gate has a motor; city employees closed the gate during a

rainstorm; and the flooding of the homeowners’ properties happened soon

after—within about an hour of—the closure. We think this is enough to

conclude that the North Laredo Gate was put to “operation or use” as we

have come to understand that phrase in § 101.021(1)(A). Closing the gate

put it to its intended purpose: blocking water. And the act of closing the

gate, along with the homeowners’ property damage and the rainstorm

itself, all occurred together as part of a single episode, giving them the

“close temporal proximity” we spoke of in PHI, 593 S.W.3d at 305.

The City’s response—that the homeowners actually complain

about negligent failure to open the North Laredo Gate and thus the

nonactionable nonuse of motor-driven equipment—proves too much.

The only reason that the North Laredo Gate would have had to be opened

was because, during the same storm, the City closed it. What matters

here is that, as all parties agree, the North Laredo Gate is used to control

waterflow in the resaca, the City closed the gate, and it was that use of

the gate (the attempt to control waterflow) that immediately preceded

and allegedly caused the flooding of the homeowners’ neighborhood.

These events, as alleged, fit comfortably within the scope of the

statutory waiver and our precedent interpreting that waiver. Counsel

20

for the homeowners properly acknowledged that the analysis would be

quite different if the gate had been closed long before—even a day

before—the storm. Then the failure to open the gate would have been a

matter of nonuse. But here, all the operative facts were part of the same

larger narrative without any logical disconnect that could justify the

City’s theory.

Under this record, therefore, we see no meaningful way to separate

the alleged “nonuse” (the failure to close the gate) from the immediately

preceding “use” (the opening of the gate). In short, we conclude that, at

least at this stage, the homeowners’ allegations concern the “operation

or use” of “motor-driven equipment” under § 101.021(1)(A) of the Texas

Tort Claims Act.

B

Our inquiry into the applicability of § 101.021 does not end there,

however. The City also challenges the homeowners’ tort claims under

what we have called the statute’s nexus requirement, which allows us

to determine if the property damage has “arise[n] from” the operation or

use of the North Laredo Gate. § 101.021(1)(A). The City argues, and a

majority of the court of appeals agreed, that it did not. We hold that at

this stage of the proceedings, the homeowners’ allegations are sufficient

to meet this causation requirement.

Much like “operation or use,” the statutory phrase “arises from”

does not explain itself, so in prior cases we have attempted to provide

more guidance. In LeLeaux, we explained that it “requires a nexus

between the injury negligently caused . . . and the operation or use of a

motor-driven vehicle or piece of equipment.” 835 S.W.2d at 51. “This

21

nexus,” we have added, “requires more than mere involvement of

property,” and the property must do more than merely “furnish the

condition that makes the injury possible.” Dall. Area Rapid Transit v.

Whitley, 104 S.W.3d 540, 543 (Tex. 2003). Rather, as we have stated

before, “the equipment’s use must have actually caused the injury.”

White, 46 S.W.3d at 869 (emphasis added).

We are aware, however, that we have not always been so plain in

our descriptions of the “arises from” standard. The largest source of

confusion in this area seems to derive from dicta in Ryder, in which we

cited an insurance-policy dispute for the proposition that the statutory

phrase “arises from” meant “something more than actual cause but less

than proximate cause.” 453 S.W.3d at 929 (citing Utica Nat’l Ins. Co. of

Tex. v. Am. Indem. Co., 141 S.W.3d 198, 203 (Tex. 2004)). In some

contexts, that in-between articulation of “arises from” can be accurate

and helpful, but in this context it is not. See PHI, 593 S.W.3d at 302

(describing Ryder as “perhaps unhelpfully” articulating the standard);

see also Riojas, 640 S.W.3d at 537 n.13 (rejecting the idea that claims

under § 101.021(1)(A) require something less than proximate cause).

That formulation is not susceptible to definite or useful guidance.

Indeed, in Ryder itself, we quickly walked back this in-between standard,

turning to the traditional proximate-cause standard in the very next

sentence: “Accordingly, a plaintiff can satisfy the ‘arising from’ standard

by demonstrating proximate cause.” 453 S.W.3d at 929.10

10 We held in Ryder that the defendant’s conduct proximately caused

the plaintiff’s injuries and accordingly had no occasion to apply, much less

further opine on, any other causation standard. Thus, we may depart from the

in-between standard without being bound by the ordinary concerns of stare

22

Accordingly, we return, as we did in Ryder, to the familiar tort-

law standard of proximate cause. Notably, this is also the very standard

that § 101.021(1) announces in its text. To satisfy § 101.021(1)(A)’s

nexus requirement, therefore, plaintiffs must show that the

governmental employee’s use or operation of the vehicle or equipment

proximately caused the relevant injury.

“Proximate cause has two elements: cause in fact and

foreseeability.” W. Invs., Inc. v. Urena, 162 S.W.3d 547, 551 (Tex. 2005).

Cause in fact is established when “the act or omission was a substantial

factor in bringing about the injury” and, without it, the harm would not

have occurred. Nixon v. Mr. Prop. Mgmt. Co., 690 S.W.2d 546, 549 (Tex.

1985). Foreseeability, on the other hand, requires “the actor [to] have

reasonably anticipated the dangers that his negligent conduct created

for others.” Univ. of Tex. M.D. Anderson Cancer Ctr. v. McKenzie, 578

S.W.3d 506, 519 (Tex. 2019). The danger that must be reasonably

anticipated is only the “general danger, not the exact sequence of events

that produced the harm[.]” Id. In making this causal assessment,

particularly in the Tort Claims Act context, we look to the record and

pleadings to determine if the alleged cause is too geographically or

temporally attenuated from the alleged effect. Ryder, 453 S.W.3d at

929–30.

Based on this standard, we hold that the homeowners, at least at

decisis. See, e.g., Bryan A. Garner, et al., The Law of Judicial Precedent 44–45

(2016). Because it falls to us and not the lower courts to determine when

statements in our decisions warrant revision, and because the consequences of

the in-between standard are visited primarily upon the lower courts, we think

it particularly important that we expressly disclaim that standard now.

23

this stage, have met their burden to create a fact issue on whether their

property damage arose from the closure of the North Laredo Gate. After

all, the temporal and geographic links are both tight. As we have

concluded already, a close temporal proximity existed between the

closing of the gate and the flooding of the homeowners’ properties. The

rainstorm, the gate’s closure, and the flooding all happened within the

same episode of events—one closely following the occurrence of the

other. Likewise, the City does not suggest that there is any significant

geographical attenuation between the gate and the homeowners’

properties. Indeed, according to one of the homeowners’ unrebutted

allegations, the North Laredo Gate is the closest of the five gates in the

resaca and is “immediately downstream” of Quail Hollow.11 In this sense,

the use/nonuse analysis and the nexus requirement understandably

inform each other. Vast time gaps or vast distances could defeat either

showing under these circumstances. But the existence of a logically

defined single episode within a small spatial area, where the challenged

governmental actions were undertaken both soon after and because of

the downpour, makes it more likely (but certainly not guaranteed) that

both the temporal and geographic showings can be made.

Despite these close geographical and temporal proximities, the

City contends that the homeowners’ property damage cannot be traced

to the gate’s closure. We agree with the City that it is the homeowners’

11We can also make a rough approximation of the physical distance

between the North Laredo Gate and the homeowners’ properties based on the

deposition of Figueroa’s supervisor, Santana Torres, who said that the distance

between the North Laredo Gate and the South Laredo Gate is about five

hundred yards.

24

burden to establish causation, not the City’s burden to disprove it, and

we subject the homeowners to the proper level of scrutiny. The City

urges us to do that by viewing the homeowners’ evidence and allegations

as insufficient for two reasons. First, according to the City, it was the

rainstorm, not the mere act of opening and closing the gate, that caused

the damage. Second, the City argues, the flooding of Quail Hollow would

have occurred regardless of whether the North Laredo Gate had been

closed. The first argument presents primarily a legal question about the

nature of the homeowners’ claim; the second challenges the adequacy of

the record to survive an attack on causation.

The City’s first argument—that the rainstorm was the only but-

for cause of the property damage—would render the homeowners’

evidence essentially useless, for it rejects the idea that two events can

work together to bring about a consequence. As the dissent below

explained, the City’s position is essentially predicated on a sole-factor-

causation standard that our Tort Claims Act cases have not embraced.

647 S.W.3d at 724–25 (Hinojosa, J., dissenting) (citing cases). And as we

have stated multiple times before, the tortious act can be a substantial

factor in causing the injury, even if it is not the sole factor. E.g., City of

Dallas v. Sanchez, 494 S.W.3d 722, 726 (Tex. 2016); Ryder, 453 S.W.3d

at 929. The homeowners can still establish causation if the rainstorm

was necessary but not sufficient for the flooding of Quail Hollow. Put

another way, if the same storm could have happened without Quail

Hollow flooding, and it was the use of the gate that made the difference,

the homeowners would satisfy a minimal requirement of causation.

The City’s second argument—that the rainstorm made the

25

property damage inevitable regardless of the use of the North Laredo

Gate—would defeat the homeowners’ claim if the homeowners could not

produce evidence that shows otherwise. But we think that they have

met their burden of putting forth sufficient evidence to show that the

closure of the gate proximately caused their property damage. For

example, the homeowners allege, and the City does not dispute, that the

water at the South Laredo Gate, in contrast to the water at the North

Laredo Gate, had maintained a positive waterflow throughout the day.

This discrepancy, the homeowners submit, indicated that the negative

waterflow at the North Laredo Gate was only temporary. The

homeowners also cite testimony from the City’s Public Works Director,

Santana Torres, and their expert witness, Lawrence Dunbar, who both

opined that closing the North Laredo Gate had the effect of trapping

water, which is what allowed the water to accumulate and overflow into

the homeowners’ properties. According to Dunbar, closing the gate

“effectively prevent[ed] any water from flowing . . . in the Resaca.”

Torres similarly agreed that closing the gate meant that the water

pouring into the resaca “just built up” and “had no place to go.”

This evidence is far from dispositive, but it is enough for the

homeowners to meet their burden of showing that the City’s theory of

causation (or the lack thereof ) cannot yet be deemed established as a

matter of law. Accordingly, “[w]e conclude that a fact issue remains as

to whether the [property damage] ‘arose from’ the [City’s operation of

the gate]. At this stage of the proceedings, no more is required to satisfy

section 101.021(1)(A)’s ‘arises from’ requirement.” PHI, 593 S.W.3d at

303. And while we agree that the resolution of this issue implicates

26

jurisdiction, we have observed before that, “in some cases, disputed

evidence of jurisdictional facts that also implicate the merits of the case

may require resolution by the finder of fact.” Miranda, 133 S.W.3d at

226. So too here.

IV

The homeowners’ allegations concern the “operation or use” of the

North Laredo Gate and there is sufficient evidence at this stage for a

factfinder to infer that the property damage arose from the gate’s

closure. The judgment of the court of appeals is therefore reversed and

the case is remanded to the trial court for further proceedings.

Evan A. Young

Justice

OPINION DELIVERED: March 10, 2023

27