JUSTICE WILLETT
delivered the opinion of the Court,
in which JUSTICE JOHNSON, JUSTICE GUZMAN, JUSTICE LEHRMANN, and JUSTICE BROWN joined.
We granted rehearing in this cause February 19, 2016, and now withdraw the opinion and judgment previously issued and substitute the following opinion.
This long-running dispute poses a question of constitutional law: whether governmental entities that engage in flood-control efforts are liable to homeowners who suffer flood damage, on the theory that the governments effected a taking of . the homeowners’ property by approving private development without fully implementing a previously approved flood-control plan. Under the circumstances presented, we answer no.
I. Background
A. Factual and Procedural Background
Plaintiffs (the homeowners) consist of about 400 homeowners whose homes were located in the upper White Oak Bayou watershed of Harris County. The homes suffered flood damage one or more times when flooding occurred during Tropical Storm Francis in 1998, Tropical Storm Allison in 2001, and another unnamed storm in 2002. The homeowners sued Harris County and the Harris County Flood Control District (collectively the County), asserting a takings cause of action.1 The homeowners sued other defendants as well, including the Texas Department of Transportation, municipal utility districts, engineering firms, and private developers; those claims were settled or dismissed and are not presented for review.
The District was created under Article XVI, section 59 of the Constitution, which authorizes the creation of conservation and reclamation districts. The District is charged with “the control ... of the storm and flood waters, and the waters of the rivers and streams in Harris County and their tributaries for ... flood control ... and other useful purposes.”2 The Defendants contend their conduct in this case with respect to flood control was coextensive, and the homeowners do not argue otherwise.3 The Harris County Commissioners Court is the governing body of the District.4
[796]*796• Most of the .homeowners’homes were built in the 1970s and early 1980s-. Prior to the, three flood events -in. issue, the homeowners’ properties had suffered little or no flood damage, although the area has a long history of flooding. In 1976 the U.S, Army Corps of Engineers prepared an “Interim Report on Upper White Oak Bayou.” The report was prepared for consideration by numerous federal and state entities including the District, the Cities of Houston and Jersey Village, and the Harris County Commissioners Court. The report noted recurring flooding in the upper White Oak Bayou drainage basin, an area covering 61 square miles, and described damaging flooding “occurring almost annually for the past several years.” It stated that the flooding was “caused primarily by inadequate channel capacities of the streams,” and that the problem was “compounded by continuing urbanization” of the fast-growing area. It predicted: “Additional residential development is expected to occur with or without an adequate plan for controlling the floods. Although current local regulations require that new structures be built above the level of the 100-year flood, damages will increase substantially in the future with increased rainfall runoff rates.” It proposed “enlargement, rectification, and partial paving” of the bayou and tributaries, together with other flood-control measures. The plan was to be funded primarily by the federal government.
The County concurred with Corps’ findings and agreed to act as a sponsor for the project, but federal funding was slow to materialize. The County approved new residential developments in the 1976-1984 period. The District began requiring new developments in the upper Bayou watershed to provide on-site detention ponds. The parties disagree on the extent to which the District deviated from this policy. The District eventually hired Pate Engineers to develop a flood-control plan, which was presented in a written report in 1984. The Pate Plan noted a “current policy requiring on-site stormwater detention on all new development projects in the Upper White Oak Bayou watershed,” and proposed channel improvements combined with detention basins, with the goal of eliminating “the [100-year] flood plain in the upper portion of the watershed.” The Plan stated that its implementation “should eliminate the existing flood plains through the existing developed portion of upper White Oak Bayou and provide for phased implementation of the ultimate plan to maintain 100-year flood protection on White Oak Bayou as future development occurs.” In 1984, the County approved the Pate Plan and authorized the District to implement it. The Plan was -to be funded through local taxes and impact fees, because federal funding was no longer available, and was to be implemented in phases. Developers who did not construct on-site detention facilities could pay an impact fee that would fund the construction of regional detention facilities.
The Pate Plan was never fully implemented, and flooding continued. In 1990 the District commissioned a new study by Klotz Associates to address flood concerns. The Klotz Plan called for measures that were different from the Pate Plan measures. ' The parties offer different characterizations of the shift from the Pate Plan to the Klotz Plan. The County contends that the Klotz Plan was necessary because assumptions in the Pate Plan proved wrong, and that the Klotz Plan was more ambitious than the Pate Plan. The homeowners contend that the Klotz Plan was less extensive than the Pate Plan for various reasons.
The homeowners claim that the flooding of their homes was caused by the County’s approval of “unmitigated” upstream development, combined with a failure to fully [797]*797implement the Pate Plan.5 Their expert, Larry Mays, relied on alleged unmitigated development occurring in the 1976-1990 time frame.6
The County filed a combined plea to the jurisdiction and motion for summary judgment, contending that no genuine issue of material fact had been raised on the elements of the takings claim. The trial court -grudgingly denied the motion,7 and the court of appeals affirmed.8
B. Contentions of the Parties on Appeal
The parties raise many arguments. Briefly, the County contends the homeowners failed to raise a fact issue on the issues of intent, causation, and public use.
On intent, the County argues that it never intended to cause flood damage to the homeowners’ properties. The County disputes that the evidence raised a fact [798]*798issue on whether the County was substantially certain that flooding would result from approval of development or failure to fully implement the Pate Plan. It argues that Mays’ opinion regarding intent and causation, is conclusory, suffers from analytical gaps, and therefore is not competent expert evidence.
On causation, the County contends that Mays opined that full implementation of the Pate Plan would have prevented flooding of the homeowners’ properties, because the three floods were all less than the 100-year flood and the Pate Plan would have prevented flooding up to the 100-year event. The County contends Mays made no attempt to show, with scientifically reliable analysis, that full implementation of the Plan would have met this goal. The County offered evidence that assumptions in the Pate Plan became outdated, and argues that the Plan’s goal of preventing flooding for anything less than a 100-year event was aspirational only. It points out that Mays, in his deposition, could not say that full implementation would have contained “the 100-year flood,” only that “the intent of [the Pate Plan] was to contain” such a flood.' The County also disputes that it ever authorized “unmitigated” development as described by Mays. It further complains of “Mays’ failure to connect any particular unmitigated development to any particular flooded property.”
On public use, the County argues that when, as here, property damage is merely the accidental result of the government’s acts, there is no public benefit and the property is not taken or damaged for public use under the Texas Constitution.- The County also argues there was no evidence the County consciously, sacrificed the homeowners’ properties for a perceived public use.
The homeowners respond to all of these arguments, contending that the trial court’s denial of summary judgment on the takings claim was proper. They argue that the County’s failure to fully implement the Pate Plan, combined with its approval of unmitigated development, resulted in flooding of their homes and amounted to a constitutional taking of their property. On the issue of intent, the homeowners contended in the trial court that Mays provided reliable expert testimony that unmitigated development “was substantially certain to result in increased flooding along the bayou in the vicinity of the Plaintiffs’ properties.” He claimed that, based on his analysis of the relevant data, the floods that damaged the homeowners’ homes were not 100-year floods, and that full implementation of the Pate Plan would have prevented flooding of all properties not in the 100-year .flood plain. He stated in the last of three expert reports: “ ‘But for’ the actions by the County of approving this unmitigated development, and/or by not fully implementing the Pate Plan (which would have eliminated the risk of flooding along the bayou for up to a 100-year event), the Plaintiffs’ properties would not have flooded during these three flood events.” The homeowners contend that the affidavit of District Director David Talbott, submitted to negate the intent element, was conclusory and self-serving.
II. Analysis
Where, as- here, evidence is presented with a plea to the jurisdiction, the court reviews the relevant evidence and may rule on the plea as a matter of law if the evidence does not raise a fact issue on the jurisdictional question, a standard that generally mirrors the summary-judgment standard.9
[799]*799Our decision does not turn on an analysis of the reliability of expert testimony. Assuming all disputed facts in favor of the homeowners, the record is clear -that the County never harbored a desire to cause flooding anywhere, Its motive was the opposite: it desired to prevent flooding and undertook extensive efforts over many years to accomplish this goal. Even under the most generous reading of the record, the County did, not know which particular properties would flood. At most it was aware of a risk that properties in the White Oak watershed somewhere, someday would flood. Assuming that a cause of the flooding was the affirmative act of approving private development, there indisputably were other causes: heavy rainfall, and, according to the. homeowners themselves, the failure to fully implement the flood-control measures of the Pate Plan. The confluence of these circumstances, in our view, does not give rise to a takings claim.
A. General Principles of Takings Law and the Intent Element
Article I, section 17 of our Constitution provides:
No person’s property shall be taken, damaged or destroyed for or applied to public use without adequate compensation being made, unless by the consent of such person.10 .
Generally, plaintiffs seeking recovery for a taking must prove the government “intentionally took or damaged their property for public use, or was substantially certain that would be the result.”11 Sovereign immunity does not shield the government from liability for compensation under the takings clause.12
Much of our takings jurisprudence focuses on mens rea. We have made clear that a taking cannot be established by proof of mere negligent conduct by the government.13 “[T]he requisite intent is present when a governmental entity knows that a specific act is causing identifiable harm or knows that the harm is substantially certain to result.”14
B. Other Elements of Takings Jurisprudence: Affirmative Conduct, Specificity, and Public Use
Much of the briefing focuses on the element of intent, but there are other elements of a taking that render the homeowners’ claim problematic.
1. Affirmative Conduct
Only affirmative conduct by the government will support a takings claim. We- have always characterized a takings claim as based on some affirmative “act” or “action” of the government. For example, in Gragg, we held “that the requisite intent is present when a governmental entity knows that a specific act is causing identifiable harm....”15 In Gragg, the [800]*800damage to the plaintiffs’ ranch was caused by the water district’s affirmative acts of constructing a reservoir and releasing water from the reservoir’s floodgates.16 Or as we stated in Jennings, “the parties agree that only an intentional act can give rise to such a taking.... There may well be times when a governmental entity is aware that its action will necessarily cause physical damage to certain private property, and yet determines that the benefit to the public outweighs the harm caused to that property.”17 A government cannot be liable for a taking if “it committed no intentional acts.”18 We have not recognized a takings claim for nonfeasance. The homeowners conceded this point in their briefing to the trial court, stating, “One of the elements of an ‘inverse condemnation’ case that a plaintiff must prove is that the government performed intentional act(s).” This element is important in today’s case because the homeowners have consistently based their claim on the failure of the County to fully implement the Pate Plan, combined with its alleged approval of “unmitigated” private development. Mays’ third expert report, for example, assigns two causes for the flooding: “[ujnmitigated land development, approved by the County,” and “[f]ailure of the [District] to fully implement the 1984 Pate Plan.” But the law does not recognize takings liability for a failure to complete the Pate Plan, despite the homeowners’ attempt to somehow bundle that inaction with the affirmative conduct of approving development. Assuming all other elements are met, liability for the taking of the homeowners’ properties can derive, if at all, from the County’s affirmative acts of approving development.
2. Specificity
In addition, a specificity element runs through our jurisprudence. The caselaw indicates that in order to form the requisite intent, the government ordinarily knows which property it is taking. For example, Jennings describes the intent requirement as covering situations where “a governmental entity is aware that its action will necessarily cause physical damage to certain private property.”19 The government must know that “a specific act is causing identifiable harm” or know that “specific property damage is substantially certain to result from an authorized government action.”20 We have not recognized liability where the government only knows that someday, somewhere, its performance of a general governmental function, such as granting permits or approving plats, will result in damage to some unspecified parcel of land within its jurisdiction.
3. Public Use in this Anti-Regulatory Takings Case
This most certainly is not an ordinary regulatory takings case, one where the [801]*801plaintiff complains that the government through regulation so burdened his property as to deny him its economic value or unreasonably interfere with its use and enjoyment.21 Today’s case does not fit this body of takings jurisprudence and is in a sense its antithesis. First, the homeowners are not complaining about regulation of their property but regulation of other private properties. Second, the complaint is not excessive regulation, but insufficient regulation via the alleged approval of “unmitigated development.” The homeowners similarly complain that in abandoning the Pate Plan the County did not regulate enough.
This uncharted theory should give us pause to ponder whether the claim, even if factually supported, is the stuff of a constitutional taking. If a private developer, after routine approval of its plat, uses its property in a manner causing damage to other properties, might the remecly lie against the developer rather than the county?22 One can certainly argue that if the government’s alleged affirmative conduct is nothing beyond allowing private developers to use their property as they wish, the more appropriate remedy is a claim against the private developers rather than a novel takings claim against the government. The homeowners in fact sued private developers and other private parties, and neither side contends that remedies against such parties do not exist.23
One way of analyzing this question is through the element of public use. Article I, section 17 of our Constitution provides for compensation where the property is “taken, damaged or destroyed for or applied to public use.” We have recognized that a taking may occur “if an injury results from either the construction of public works or their subsequent maintenance and operation,”24 but we have not held that the public-use element is met whére the government does nothing more than approve plats or building permits for private development.
The homeowners argue that the public-use element is met here because it was met in City of Keller v. Wilson. That case is factually distinguishable. In City of Keller, the city adopted a master drainage plan that called for a drainage easement on land belonging to the plaintiffs, the Wilsons.25 The easement was to contain a ditch, and the plan originally called for the city to condemn 2.8 acres of the Wilson property for construction of the ditch.26 The ditch would traverse the Wilson property and other properties and terminate in á creek.27 Developers were required to comply with the plan, and built a ditch on [802]*802an adjacent property, the Sebastian property, but made no provision for a drainage easement across the Wilson property.28 The city built a culvert to the creek, but it did not connect the ditch to the culvert , as planned because neither the city , nor the developers purchased an easement on the Wilson property, nor did they extend the ditch across the Wilson property as planned.29 The result was that the Wilson property would flood when it rained.30 Without analysis of the public-use element, we agreed with the court of appeals that this element had been satisfied,31 but we ultimately held that the intent element had not been established, due to lack of “proof that the City knew the plans it approved were substantially certain to increase flooding on the” Wilson property.32
City of Keller is factually distinguishable because the city there planned to condemn an easement and build a ditch across the plaintiffs’ property, knowing drainage across that specific property was part of the master plan. The jury found that the failure to build that leg of the ditch resulted in flooding on the Wilson property. As the court of appeals reasoned:
Clearly, had the City used its powers of eminent domain to condemn a portion of the Wilson property for an easement, that use would have been a “public use” to implement the City’s Master Drainage Plan. The fact that the City chose not to condemn any of the Wilson property and to instead, in violation of the Plan, allow the water flowing from the Sebastian easement to discharge, uncontrolled, across the Wilson property shows that the invasion of the Wilson property by the water flowing from the Sebastian easement was for “public use.”33
Hence, the public-use element as to 'the particular plaintiffs was not established merely by the approval of private development on other properties.34 In today’s case, there was no comparable proof that the County intended to purchase easements on the homeowners’ particular properties and construct drainage facilities as part of a master plan, such that with or without such easements and'drainage facilities runoff would under the plan traverse the homeowners’ properties. The public-use element was more apparent in City of Keller because with or without the easement and ditch the plaintiffs’ land would by city design be available to carry runoff to the creek. Similarly, in Gragg, the defendant water district denied that it had inversely condemned ranch property by releasing water through reservoir floodgates, but counterclaimed for a flowage easement if the court found that the property had been inversely condemned. The trial court granted such an easement on the portion of the ranch subject to flooding.35 And in Kopplow Development, Inc. v. City of San Antonio, when the City decided to build a detention facility it “knew the project would inundate part of [803]*803[plaintiff] Kopplow’s property before it ever began construction, prompting. the City to seek a drainage easement from Kopplow.”36 In today’s case, in contrast, whether -viewed through the lens of intent or public use, there was no evidence that the County ever had designs on the homeowners’ particular properties, and intended to use those properties to accomplish specific flood-control measures.
. We also find the United States Supreme Court’s landmark decision in Kelo v. City of New London37 factually and legally distinguishable.38 In Kelo, á city authorized a private nonprofit entity to condemn property as part of an economic revitalization plan.39 ' The city claimed and the Court accepted that the public-use requirement was met because the plan would enhance the overall economic health of the community.40 But Kelo was not an inverse condemnation case like today’s case. There, the government, through an agent, condemned plaintiffs’ properties for an alleged public purpose. Indisputably the plaintiffs’ properties were taken from them by government action. In today’s case, where the only affirmative conduct that allegedly damaged the homeowners’ properties was the approval of private development, there is in our view a real question whether a taking by the government even occurred, that is, whether the homeowners’ claim even belongs in the world of takings jurisprudence and is properly analyzed as a takings claim.41 In Kelo the plaintiffs argued that there was no public use within the ambit of the federal Takings Clause, in hopes of disallowing the taking of their land. Here the homeowners are in a completely different posture, arguing that the approval of private development was undertaken for the benefit of the public at large, with the result that the homeowners’ properties were damaged for a public use, so as to establish a taking within the ambit of the Texas Takings Clause. In light of these factual and legal distinctions, Kelo does not compel a result one way or the other in today’s case.
■ The Washington Supreme Court considered a case similar to today’s case where plaintiffs filed an inverse condemnation claim against a county, alleging that plaintiffs’ property flooded after the county approved a development plat for a neighboring property. The court held that mere approval of the plat' could not support a takings claim:
If all the County had done was to approve private development, then one of the elements of an inverse condemnation claim, that the government has damaged the [plaintiffs’] property for a public purpose, would be missing. There is no public aspect when the County’s only action is to approve a private development under then existing regulations. Furthermore, the effect of such automatic liability would have a completely [804]*804unfair result.42
The approval of private development in this case—doing nothing more than allowing private parties to use their properties as they wish—presents at best a highly attenuated basis for meeting the public-use element of a takings claim.
C. The Unavoidable Tension Between Takings Jurisprudence and Sovereign Immunity
While compensation to those whose property is taken for public use is an important and constitutionally imposed obligation of democratic government, governments must also be allowed to survive financially and carry out their public functions. They cannot be expected to insure against every misfortune occurring within their geographical boundaries, on the theory that they could have done more. No government..could afford such.obligations. Justices Jackson and Goldberg both recognized that the Bill of Rights is not a suicide pact.43
This Court has repeatedly, recently, and unanimously recognized that strong judicial protection for individual property rights is essential to “freedom itself.”44 Locke deemed the preservation of property rights “[t]he great and chief end” of government,45 a view we echoed almost 300 years later, calling it “one of the most important purposes of government.”46 Individual property rights are “a foundational liberty, not a contingent privilege.”47 They are, we reaffirm today, “fundamental, natural, inherent, inalienable, [and] not derived from the legislature,” and “preexist[ ] even constitutions.” 48
But there is always tension between the compensation obligation of the Takings Clause and the necessary doctrine of sovereign immunity, also a doctrine of constitutional significance.49 We long ago recognized that where government conduct caused damage to a plaintiffs property,
One’s normal reaction is that he should be compensated therefor. On the other hand, the doctrine of the non-suability of the state is grounded upon sound public policy. If the state were suable and liable for every tortious act of its agents, servants, and employees committed in the performance of their official duties, there would result a serious impairment of the public service and the necessary [805]*805administrative functions of government would be hampered.50
D. Application of Takings Law to the Facts Presented
Because inaction cannot give rise to a taking, we cannot consider any alleged failure to take further steps to control flooding, such as the failure to complete the Pate Plan. Because a taking cannot be premised on negligent conduct, we must limit our consideration to affirmative conduct the County was substantially certain would cause flooding to the homeowners’ properties and that would not have taken place otherwise. The only affirmative conduct on which the homeowners rely is the approval of private development. Further, the homeowners offered no proof that the County was substantially certain that the homeowners’ particular properties would flood if the County approved new housing developments. The homeowners did not even assert such a claim, alleging instead that the County was substantially certain that its actions in approving “unmitigated development” would result in flooding “in the vicinity of Plaintiffs’ properties.” They never explained whether the “vicinity” is a few square miles or hundreds of square miles. They never identified precisely or even approximately the area of unmitigated development. The Pate Plan described the White Oak Bayou watershed as 110 square miles in size. The homeowners’ expert Mays admitted that “I haven’t been asked to do anything concerning specific plaintiffs.” A County expert, Melvin Spinks, agreed that “Mays was unable to demonstrate that any actions of Defendants were the proximate cause of flooding of any particular Plaintiffs property,” nor did Mays “determine the causation of flooding subdivision by subdivision.” In the trial court, the County accurately described the homeowners’ particular parcels as “scattered in a checker board fashion in the upper White Oak Bayou watershed, stretching several miles.” •
Further, the homeowners offered no evidence that the County was consciously aware that approval of unmitigated development in one defined area, such as a specific block or neighborhood, was substantially likely to cause flooding in another specifically defined area of the White Oak Bayou watershed that included the homeowners’ properties. The County offered evidence to the contrary. District Director David Talbott attested that “[t]he District did not approve of land development knowing that there was inadequate stormwater runoff mitigation associated with a particular development,” and pointed out that the District was tasked with addressing severe flooding problems not only in the White Oak Bayou watershed where the homeowners resided, but also in the Clear Creek, Greens Bayou, Cypress Creek, and Brays Bayou watersheds. While the homeowners number 400 who suffered flood damage during three events, Talbott pointed out that Tropical Storm Allison alone flooded 73,000 residences. Talbott and the vice president of Klotz both attested that it is against District policy to “move a flood”—sparing one neighborhood that would otherwise flood by causing another neighborhood to flood. Talbott also attested: “Although White Oak Bayou was always a high priority, with limited District funding the District also had to consider other high priority projects throughout the County. District funds that were available were allocated to various projects around the County, with White Oak Bayou receiving an. appropriate share.”
[806]*806This case is qualitatively different from recent cases like Kopplow and Gragg where we recognized a taking. In Kop-plow, the city “knew the project would inundate part of [plaintiff] Kopplow’s property” and sought a drainage easement from Kopplow, the city’s project resulted in only one other property being placed below the 100-year flood plain, and the city obtained a drainage easement on that other property.51 In Gragg, one of the flood-control district’s experts acknowledged that his own modeling showed that higher than natural flooding would occur on the plaintiffs’ particular ranch in 10 out of 16 floods, the district’s records showed hundreds of releases by the district sufficient to cause flooding on the ranch, and there was evidence that the ranch had suffered “a large number of floods” after the district began the releases, whereas before the district’s actions the ranch had never suffered from extensive flood damage.52 In today’s case, in contrast, the record is devoid of evidence the County knew, at the time it allegedly approved of “unmitigated” development, that the homeowners’ particular properties would suffer from recurrent flooding. We recognized in Jennings that a taking occurs when property is “damaged for public use” in circumstances where “a governmental’ entity is aware that its action will necessarily cause physical damage to certain private property.”53 A conscious decision to damage certain private property for a public use is absent here.
The homeowners contend that Kopplow and Gragg are helpful to their case because both decisions recognized that the recurrence of flooding is probative on the issue of intent.54 But we also held, in City of San Antonio v. Pollock, that when deciding intent in the takings context, “[t]he government’s knowledge must be determined as of the time it acted, not with [the] benefit of hindsight.”55 This rule limits the persuasiveness of the homeowners’ argument. The homeowners alleged in 'their petition that “[m]ost, if not all of the plaintiffs herein, had never flooded before September, 1998,” the Tropical Storm Francis flooding and the first of three flood events about which they complain. Their recurrence argument to us is that “Plaintiffs’ homes flooded three times in five years in 1998, 2001, and 2002.” They contend these three floods are “probative evidence of intent under this Court’s holdings in Kopplow and Gragg.” But their expert, Mays, opined that the unmitigated development that caused the flooding of their homes ended no later than 1990, years before the three flooding events.56 The homeowners’ recurrence argument is made with the benefit of hindsight.
The determination of whether a taking has occurred is ultimately a question of law.57 The determination is not formulaic given the legal and factual complexities sometimes presented, and may [807]*807turn on the confluence of particular circumstances.58 We hold that the homeowners have not presented a cognizable takings claim where (1) the County never desired to cause .flooding, but desired only the opposite, (2) it undertook significant efforts to prevent flooding, spending tens of millions of dollars over many years, (3) the County never intended, as part of a flood-control plan, to use the homeowners’ particular properties for detention ponds, drainage easements, or the like, (4) the only affirmative conduct.allegedly causing the flooding was approval of private development, (5) the homeowners offered no proof that, the County was substantially certain its approval of development would result in the flooding of the homeowners’ particular lots, and (6) even by the homeowners’ reckoning the flooding resulted from multiple causes—Acts of God,59 the activities of other defendants,60 the alleged failure to complete the Pate Plan, and the approval of private development. We have never recognized a takings claim under such attenuated circumstances. This is not a case where the government made a conscious decision to subject particular properties to inundation so that other properties would be spared, as happens when a government builds- a flood-control dam knowing that certain properties will be flooded by the resulting reservoir. In such cases of course the government must compensate the owners who lose their land to the reservoir.
The homeowners’ theory of takings liability would vastly and unwisely expand the liability of governmental entities, a view shared by the many public and private amicus curiae who have urged rehearing of this cause.61 The theory lacks any discernible limiting principle and would ap[808]*808pear to cover many scenarios where the government has no designs on a particular plaintiffs property, but only knows that somewhere, someday, its routine governmental operations will likely cause damage to some as yet unidentified private property. The homeowners’ theory would apply to myriad property-damage claims and as to them might effectively abolish much traditional fault-based tort law, swallow much of sovereign immunity, and disrupt the carefully crafted waiver of immunity found in the Tort Claims Act. We have stated that sovereign immunity is universally recognized and fundamental to the nature and functioning of government, and that we leave it to the Legislature to make changes to that doctrine, as it has done in the Tort Claims Act.62 Therefore, we think so sweeping an expansion of takings jurisprudence should be made only for the soundest reasons—reasons that escape us here—given the corresponding and massive contraction in the scope of sovereign immunity that would follow. While the right to compensation for a taking is constitutionally mandated, sovereign immunity is also a matter of constitutional significance.
The homeowners’ notion of a taking is expansive indeed. Take, for example, a government such as the City of Austin that supplies electric utility service to its citizens.- It surely knows that in running power lines throughout the service area, fires or other damaging events will occasionally occur when acts of nature knock down lines or poles. Witness the recent devastating fires in the Bastrop area allegedly caused by power lines. The government also surely knows that some private properties—those adjacent to the lines, or, in the homeowners’ vernacular, “in the vicinity” of the lines, are especially vulnerable to such events. Under the homeowners’ theory, an Act of God, such as a bolt of lightning, that causes'a high-voltage line to topple or a transformer to blow, which in turn causes damage to a private property, is arguably a taking on the notion that properties near the grid have been sacrificed for the greater public good of providing electricity service to the whole community. Negligent maintenance of the utility grid is irrelevant, so long as the city is substantially certain that fires somewhere, someday, will occur along the grid, as surely is the case. It matters not that a storm contributed to the downed line, because in today’s case storms also played an essential role. Traditional fault-based tort law, sovereign immunity, and the Tort Claims Act are irrelevant if the claim is framed as a taking. The hypothetical is arguably a stronger case for a taking than today’s case, because providing an electric utility grid unquestionably involves a public use of property, whereas in today’s case the only affirmative conduct allegedly resulting in a taking was granting approval of private development, perhaps not a public use at all.
[809]*809Or take any large city, school district, or other governmental entity that has a fleet of vehicles. The government surely knows its vehicles regularly have accidents causing damage to private property. The government also, surely knows that collisions will occur more frequently in certain areas, such as along bus routes or near the garbage truck depot, school bus lot, etc., where the government’s vehicular traffic is concentrated. Under the homeowners’ theory, each accident resulting in damage to private property in a higher-risk area would appear to be a compensable taking, again on the theory that this property has been sacrificed for the greater good of providing city-wide public transportation. The claim is arguably stronger than the claim in today’s case, because (1) there is no Act of God that can be assigned at least part of the causation, (2) the city in the hypothetical is not just substantially certain but absolutely certain that accidents will occur, and (3) providing a city bus system is unquestionably a public use of property, unlike approval of private plats. No need to prove negligence on the city’s part; the intent element is met because the city is substantially certain that accidents happen. No need to predict exactly where the accidents will occur, because in today’s case the homeowners never contended or offered proof that the County formed any intent with respect to their particular properties. The homeowners’ expert on the intent element conceded, “I haven’t been asked to do anything concerning specific plaintiffs.”
Amicus curiae Texas Department of Transportation (TxDOT) • raises a similar hypothetical. It points out that it is required by law, in determining roadway access, to balance “the needs of mobility and safety on a highway system with the needs of access to adjacent land uses.”63 In its own access policy manual, it recognizes that numerous studies have shown that “[a]s access density increases, crash rates increase.” Thus, according to TxDOT, it “knows that the issuance of driveway access permits onto State roadways will result in more accidents, inevitably causing damage to as yet unidentified private property. .Arguably, under the homeowners’ theory, every accident resulting in damage to private property may be a compensable taking on the theory that the property has been sacrificed for the greater good, of providing property owners and their customers or residents access to roadways.”
Or take any large city with its contingent of high-rise buildings. A city may know that somewhere, someday, a fire will occur on a floor the city fire trucks cannot reach. Suppose a city has a study in hand recommending larger ladder trucks, but fails to purchase the trucks due to funding problems or other priorities. Under the homeowners’ approach, if a fire occurs on a higher floor of a building, and damages adjacent properties, the adjacent property owners have a takings claim if they can show that a larger ladder truck would have contained the fire. It does not matter whether the city’s conduct was reasonable given its tax base or funding priorities. In today’s case, too, the record shows that funding issues played a role in the County’s decisions. All that matters is that the city issued building permits, knowing that somewhere, someday, a fire would likely occur on an upper floor. This knowledge supplies the intent element, and the building permits supply the causation element, regardless of whether an act of nature started the fire. It does not matter whether the building in issue was privately owned, because according to the home[810]*810owners private ownership of the property-approved for development is no bar to recovery.
Amicus Curiae Harris County Metropolitan Transit Authority (METRO) imagines an even more disturbing hypothetical, asking whether hurricanes allegedly caused by global warming would be a compensa-ble taking under the homeowners’ reasoning. After all, METRO contends, “Experts can be hired who will testify that burning fossil fuels raises sea levels and makes storms more intense. Yet governments issue permits allowing exploration and production of fossil fuels, and construction and operation of the power plants that burn them.”
And, as several amici curiae fear in urging rehearing, the homeowners’ theory of takings would place, governments in a unending dilemma of requiring extreme flood-control measures and facing a regulatory takings claim from the owners directly subjected to such measures, or requiring less extensive measures and facing a takings claim when downstream property owners experience flooding. In the words of amicus curiae City of San Antonio, flood-control decisions will be reduced to “picking your plaintiff rather than responsible flood control management.”
We decline to extend takings liability vastly beyond the extant jurisprudence, in a manner that makes the government an insurer for all manner of natural disasters and inevitable man-made accidents.64 Holding otherwise would encourage governments to do nothing to prevent flooding, instead of studying and addressing the problem. The homeowners do not urge the existence of a general legal duty on the part of the County to prevent flooding, breach of which would give rise to a private cause of action. Their claim instead is that the County failed to complete the Pate Plan and approved private development, behavior allegedly resulting in a taking of their properties. If various state and federal governmental entities had not commissioned and conducted detailed studies of regional flooding, the homeowners would have no basis for contending that the County was substantially certain of the link between development and flooding, and would not be able to use that knowledge against the County. If the County had undertaken no efforts to control flooding, the homeowners could not assert the failure to complete the Pate Plan as a basis for liability. Further, by framing their claim as a constitutional taking, the homeowners have asserted a claim unbounded by any statutory caps on compensatory damages the Legislature might otherwise impose. Accepting such a capacious approach to takings would endanger the ability of governments to finance and carry out their necessary functions, the basis for sovereign immunity.
III. Conclusion
The plea to the jurisdiction was well-taken and should have been granted. We reverse the judgment of the. court of appeals and. render judgment dismissing the case.
JUSTICE LEHRMANN filed a concurring opinion. .
[811]*811JUSTICE DEVINE filed a dissenting opinion, in which CHIEF JUSTICE HECHT, JUSTICE GREEN, and JUSTICE BOYD joined.