5 S.W.3d 654
Tex.
1999

Lead Opinion

Justice ABBOTT

delivered a plurality opinion,

in which Justice HECHT and Justice OWEN join.

While driving late one night in the downtown Houston area, Angela Holder was stopped for an alleged traffic violation by Calvin Potter, an on-duty Houston police officer. Potter took Holder’s insurance and identification cards and told her to follow his squad car. Holder followed Potter several blocks to a parking garage owned by Mellon Mortgage Company. Once inside the garage, Potter sexually assaulted Holder in his squad car.

Holder sued Mellon and the City of Houston but did not sue her attacker. The trial court granted summary judgment for Mellon and the City on all of Holder’s claims. The court of appeals affirmed the summary judgment in favor of the City on the basis of sovereign immunity. With regard to Holder’s claims against Mellon, the court of appeals affirmed the summary judgment on Holder’s negligence per se claim, but reversed on the negligence, gross negligence, and loss of consortium1 claims. On petition for review to this Court, Mellon claims, among other things, that it owed no legal duty to Holder. Because we hold that it was not foreseeable to Mellon that a person would be accosted several blocks from Mellon’s garage and forced to drive to that garage where she would be sexually assaulted, Mellon owed no duty to Holder to prevent the attack. Accordingly, we reverse the court of appeals’ judgment and render judgment that Holder take nothing.2

I

With regard to criminal acts of third parties, property owners owe a duty to those who may be harmed by the criminal acts only when the risk of criminal conduct is so great that it is both unreasonable and foreseeable. See Timberwalk Apartments, Partners, Inc. v. Cain, 972 S.W.2d 749, 756 (Tex.1998). We focus our attention in this case on “foreseeability.” For most premises liability cases, the foreseeability analysis will be shaped by determining whether the plaintiff was an invitee, a licensee, or a trespasser. Because Holder was an unforeseeable victim regardless of her status, it is unnecessary to determine into which of the three categories she falls. Instead, we focus on general foreseeability principles that limit the scope of the defendant’s duty in this case.3

We have repeatedly stated that “[f]oreseeability requires only that the general danger, not the exact sequence of events that produced the harm, be foreseeable.” Walker v. Harris, 924 S.W.2d 875, 377 (Tex.1996); see also Nixon v. Mr. Property Management Co., 690 S.W.2d 546, 551 (Tex.1985). We have also frequently stated a two-prong test for foreseeability:

[I]t is not required that the particular accident complained of should have been foreseen. All that is required is [1] “that the injury be of such a general character as might reasonably have been anticipated; and [2] that the injured party should be so situated with relation to the wrongful act that injury to him or to one similarly situated might reasonably have been foreseen.”

Id. at 551 (citations and emphasis omitted); see also Texas Cities Gas Co. v. Dickens, 140 Tex. 433, 168 S.W.2d 208, 212 (1943); San Antonio & A.P. Ry. Co. v. Behne, 231 S.W. 354, 356 (Tex. Comm’n App.1921, judgm’t adopted). Thus, we consider not only the foreseeability of the general criminal act but also the foreseeability that the victim might be injured by the act. Stated more broadly, we determine both the foreseeability of the general danger and the foreseeability that a particular plaintiff— or one similarly situated — would be harmed by that danger.

This duty analysis has been widely embraced since Chief Judge Cardozo penned the seminal Palsgraf opinion. See Palsgraf v. Long Island R.R., 248 N.Y. 339, 162 N.E. 99 (1928). Palsgraf teaches that the duty question properly considers the foreseeability of the injured party. Mrs. Palsgraf was standing on a platform at the defendant’s railroad waiting for a train. Some distance away, porters tried to help a passenger board a train. As they assisted him, they dislodged a package of fireworks he was carrying. The package fell to the rails and exploded, knocking over scales and injuring Mrs. Palsgraf. See id. at 99.

The court held that, regardless of whether the railroad might have acted in a generally wrongful manner, it was not negligent with regard to Mrs. Palsgraf. See id. As Chief Judge Cardozo explained, ‘What the plaintiff must show is ‘a wrong’ to herself; i.e., a violation of her own right, and not merely a wrong to some one else....” Id. at 100. Because the plaintiff was not so situated to the wrongful act that her injury might reasonably have been foreseen, the defendant did not owe a duty to protect her from the resulting injury. “ ‘Proof of negligence in the air, so to speak, 'will not do.’ ... The risk reasonably to be perceived defines the duty to be obeyed, and risk imports relation; it is risk to another or to others within the range of apprehension.” Id. at 99, 100 (emphasis added). Because the railroad owed no duty to Mrs. Palsgraf, it was unnecessary to consider any question of proximate cause.

The Palsgraf dissent, however, illustrates the counter view that duty is owed generally and any limitations on liability should be through “proximate cause,” in which “foreseeability” must necessarily play a greater role than in the duty analysis. Writing for the dissent, Judge Andrews rejected the court’s view that the duties owed by a defendant were the particularized product of a relationship determined in part by foreseeability. “Every one owes to the world at large the duty of refraining from those acts that may unreasonably threaten the safety of others. Such an act occurs. Not only is he wronged to whom harm might reasonably be expected to result, but he also who is in fact injured, even if he be outside what would generally be thought the danger zone.” Id. at 103 (Andrews, J., dissenting). The Palsgraf dissent, like the dissent in this case, appears to contend that consideration of a particular plaintiffs relation to an alleged wrongful act is better considered under the guise of proximate cause.

Although judges and scholars have long debated the relative merits of the two views, the gist of Chief Judge Cardozo’s duty analysis has been widely embraced. Compare 3 Harper et al., The Law of Torts § 18.2, at 654-55 (2d ed.1986); Restatement (Second) of ToRts § 281 cmt. c (1965); Zipursky, Rights, Wrongs, and Recourse in the Law of Torts, 51 Vand. L. Rev. 1, 3-5 (1998); and Green, Proximate Cause in Texas Negligence Law, 28 Tex. L.Rev. 471, 472 (1950); with Keeton et al., Prosser and Keeton on the Law of Torts § 43, at 287 (5th ed.1984). The Restatement (Second) of Torts states:

In order for the actor to be negligent with respect to the other, his conduct must create a recognizable risk of harm to the other individually, or to a class of persons — as, for example, all persons within a given area of danger — of which the other is a member. If the actor’s conduct creates such a recognizable risk of harm only to a particular class of persons, the fact that it in fact causes harm to a person of a different class, to whom the actor could not reasonably have anticipated injury, does not make the actor liable to the persons so injured.

Restatement (Second) of Torts § 281 cmt. c (1965); see also 4 Harper et al., supra, § 20.5, at 138 (the scope of a duty is limited to “(1) those persons that are likely to be endangered by the act or omission, and (2) harm (to such person or interest) from a risk the likelihood of which made the act or omission negligent”). The result of this analysis is that “[a] plaintiff has no right of action unless there was a wrong relative to her or a violation of her right, and there is no such relational wrong or personal-rights violation in a negligence case where the duty to avoid foreseeable risk to the plaintiff has not been breached.” Zipursky, supra, at 15; see also Nixon, 690 S.W.2d at 551. A wrong in general is not enough; the plaintiff herself must be wronged. See Zipursky, supra, at 12.

When we consider whether a particular criminal act was so foreseeable and unreasonable as to impose a duty upon a landowner, we first examine the particular criminal conduct that occurred in light of “specific previous crimes on or near the premises.” Walker, 924 S.W.2d at 377. If, after applying the Timberwalk factors of similarity, recency, frequency, and publicity, see Timberwalk, 972 S.W.2d at 756-57, we determine that the general danger of the criminal act was foreseeable, we then apply the second prong of the foreseeability analysis and determine whether it was foreseeable that the injured party, or one similarly situated, would be the victim of the criminal act. In essence, we consider whether the plaintiff was within the range of the defendant’s apprehension such that her injury was foreseeable. See Palsgraf, 162 N.E. at 99-100. Only when we have analyzed the criminal act within the context in which it occurred can we determine whether the landowner owed a duty to the injured party. See, e.g., Centeq Realty, Inc. v. Siegler, 899 S.W.2d 195, 197 (Tex.1995) (when determining whether a duty lies, we must consider all “the facts surrounding the occurrence in question”).

Applying the Timberwalk factors, it was not unforeseeable as a matter of law that a rape might occur in the parking garage. Although no similar violent crimes had occurred in the parking garage before the attack on Holder, the summary judgment evidence shows that in the two years preceding the incident, 190 violent crimes, including rape and murder, were reported near the garage. This equates to a frequency of roughly one violent crime every four days.

While there is no evidence that any of these crimes received publicity and Mellon was not required to inspect police records to determine whether its garage was in a high crime area, the summary judgment evidence establishes that Mellon was aware that property crimes had occurred, including the theft of a Mellon employee’s car. Another Mellon employee complained to the garage manager “about the virtually non-existent security” in the garage, which compelled the employee to seek an escort to her car when she worked late. Furthermore, Mellon knew that vagrants frequented the garage and sometimes drank there.

Together, these facts constitute some evidence that violent criminal conduct was foreseeable. But while it may have been foreseeable that a violent crime such as rape might occur, this does not end our analysis. We must also consider whether Holder was situated such that Mellon could foresee that she would be the victim of this third-party criminal act. See Carey v. Pure Distrib. Corp., 133 Tex. 31, 124 S.W.2d 847, 849 (1939); Restatement (Second) of ToRts § 281 emt. c (1965). The facts of this case fall squarely within the second prong of the foreseeability analysis and show that Mellon could not have reasonably foreseen that its failure to secure the garage would lead to Holder’s injuries.

Certainly, Mellon expected that its employees would use the garage, often at times when it would be relatively vacant and thus more dangerous. It is not unreasonable to conclude that Mellon could foresee that an employee or some other person who frequents the garage could be the victim of a violent crime in the garage. To protect these garage users, Mellon provided armed security patrols weekdays from 5:45 a.m. to 11:30 p.m., in addition to random patrols by off-duty police officers during business hours. Holder, however, was not a member of this class nor any other that Mellon could have reasonably foreseen would be the victim of a criminal act in its garage.

Unlike any foreseeable victim, Holder was pulled over in her car at 3:30 a.m. by a third party over whom Mellon had no control, and she was led from several blocks away to the actual crime scene. Not only did Mellon have no control over the criminal, Potter, it had no knowledge of him nor any reason to know that he would pick the garage as the scene of his reprehensible crime. Moreover, Mellon had no knowledge of Holder nor any reason to believe that she, or a person similarly situated, could be subject to a crime on Mellon’s property. It simply was not foreseeable, beyond a remote philosophic sense, that this tragic event would occur to Holder on Mellon’s property. With relation to Mellon’s allegedly wrongful act of not securing its garage at three in the morning, Holder was not so situated that injury to her might reasonably have been foreseen. She was, in short, beyond Mellon’s reasonable apprehension.

Holder argues that Mellon- knew that the condition of its garage created an unreasonable and extreme degree of risk that an attack such as this would occur. However, nothing in Holder’s summary judgment evidence suggests that Mellon could have reasonably foreseen that its garage would be picked by Potter as the scene of his crime if it did not secure its garage. The mere fact that crimes are prevalent in downtown Houston is not enough. See Timberwalk, 972 S.W.2d at 756. Examining the evidence, it is true that Mellon was aware that a car had been stolen from its garage, but this does not indicate that the garage would be used as a place to bring Holder. It is also true that Mellon was aware that vagrants frequented the garage, but this does not suggest that it was a place that invited criminals to transport victims there. Holder’s summary judgment evidence provides little more than “proof of negligence in the air.” Palsgraf, 162 N.E. at 99. She provides no evidence of a foreseeable risk in relation to her.

In the end, Holder points again and again to the fact that Mellon was aware that cars could enter its garage without authorization. But to base foreseeability on this fact, without more, would effectively place a universal duty on any landowner with secluded property to prevent that property from becoming the scene of a crime. Whether it be a farmer’s field, an industrial park, or a twenty-four-hour laundromat, placing a duty on landowners to prevent criminal acts on their property simply because criminals could gain access to their land would make landowners the insurers of crime victims, regardless of the lack of connection between the landowner and either the victim or the perpetrator. “Courts across the country agree that an owner or possessor of property is not an insurer of the safety of those on the premises.” Lefmark Management Co. v. Old, 946 S.W.2d 52, 59 (Tex.1997) (Owen, J., concurring) (citing Exxon Corp. v. Tidwell, 867 S.W.2d 19, 21 (Tex.1998); Ann M. v. Pacific Plaza Shopping Ctr., 6 Cal.4th 666, 25 Cal.Rptr.2d 137, 863 P.2d 207, 215-16 (1993); and Kline v. 1500 Massachusetts Ave. Apartment Corp., 439 F.2d 477, 487 (D.C.Cir.1970)); see also Restatement (Second) of ToRts § 344 cmt. f (1965) (“[T]he possessor is not an insurer of the visitor’s safety....”).

Accordingly, Mellon owed no legal duty to Holder. To the extent that Mellon’s conduct may have created a risk of harm, it did not breach a duty to Holder because she was not so situated with relation to the wrongful aet such that her injury might have been foreseen.

II

The dissent implies that this analysis is inconsistent with Nixon. In Nixon, however, the Court did not discuss or analyze the common law aspects of duty. Instead, the Court held that the duty owed by the defendant was governed by an applicable ordinance. In doing so, the Court stated:

An ordinance requiring apartment owners to do their part in deterring crime is designed to prevent injury to the general public. R.M.V. falls within this class. Since the ordinance was meant to protect a larger class than invitees and licensees, and since R.M.V. committed no wrong in coming onto the property, these premises liability distinctions are irrelevant to our analysis.

Nixon, 690 S.W.2d at 549. Thus, the ordinance defined the scope of the second prong of foreseeability: “that the injured party should be so situated with relation to the wrongful act that injury to him or to one similarly situated might reasonably have been foreseen.” Id. at 551 (citations omitted).

Moreover, in considering the foreseeability aspect of proximate cause in Nixon, the Court’s discussion and its use of italics make clear that it focused solely on the first prong of foreseeability: that “[i ]t is not required that the particular accident complained of should have been foreseen. All that is required is that the injury be of such a general character as might reason ably have been anticipated .... ” Id. (citations omitted). In its proximate cause analysis, the Nixon Court did not discuss, italicize, or otherwise analyze the second prong of foreseeability. Thus, Nixon is inapposite to the analysis of this case.

The dissent also takes issue with this analysis of Mellon’s duty to Holder by claiming that it “improperly bootstraps proximate cause foreseeability into the threshold duty question.” 5 S.W.3d at 666 (O’Neill, J., dissenting). The dissent does not explain, however, how the foreseeability analysis under “proximate cause” differs from the foreseeability analysis under “duty.” Additionally, the dissent does not explain why it was not similarly improper for this Court, and other courts, to frequently use a singular foreseeability analysis interchangeably between duty and proximate cause. Furthermore, the dissent does not explain why the second prong of the foreseeability analysis — that the injured party should be so situated with relation to the wrongful act that injury to her or to one similarly situated might reasonably have been foreseen — applies only to proximate cause foreseeability and not to duty foreseeability. The dissent cannot be faulted, however, for failing to answer these questions because Texas jurisprudence on these issues has been unclear. But the answer is simple: The “foreseeability” analysis is the same for both duty and proximate cause.

The questions of duty and proximate cause “are often used in a confused and overlapping way” because both rest on a determination of “foreseeability.” 3 Harper et al., supra, § 18.1, at 650; see also Travis v. City of Mesquite, 830 S.W.2d 94, 98 (Tex.1992) (proximate cause consists of cause-in-fact and foreseeability); Greater Houston Transp. Co. v. Phillips, 801 S.W.2d 523, 525 (Tex.1990) (main determinant of duty is foreseeability). The confusion can be found, for example, in Nixon. There, foreseeability was analyzed only under the heading of “proximate cause” because the Court determined at the outset that the defendant owed the plaintiff a duty imposed by statute. See Nixon, 690 S.W.2d at 549. Yet, in defining “foreseeability” as applied to the case, the Court cited a case dealing exclusively with proximate cause, Missouri Pacific Railroad v. American Statesman, 552 S.W.2d 99, 103 (Tex.1977); a case dealing only with duty, Castillo v. Sears, Roebuck & Co., 663 S.W.2d 60, 64 (Tex.App.—San. Antonio 1983, writ ref'd n.r.e.); and another dealing with both, Walkoviak v. Hilton Hotels Corp., 580 S.W.2d 623, 625, 628 (Tex.Civ.App.—Houston [14th Dist.] 1979, writ ref'd n.r.e.). See Nixon, 690 S.W.2d at 550.

The confusion has been perpetuated since Nixon. In Walker v. Harris, 924 S.W.2d 375, 377 (Tex.1996), and Exxon Corp. v. Tidimll, 867 S.W.2d 19, 21 (Tex. 1993), this Court considered foreseeability as it relates to duty. In doing so, the Court cited to the Nixon foreseeability analysis, which, as has been noted, applied to proximate cause.

Interestingly, the court of appeals’ opinion in this case relies on Nixon’s discussion of foreseeability, as it was applied to causation, for support of its discussion of foreseeability as it applies to duty. 954 S.W.2d 786, 795. The court concluded that a duty was owed to Holder because, in part, her injury was foreseeable. Id. at 795. Turning to “proximate cause,” the court again considered whether Holder’s injury was foreseeable, but rather than repeat its analysis verbatim, the court simply refers to its previous discussion of foreseeability under “duty.” Id. at 801. The court relies upon a single discussion of foreseeability to establish foreseeability’s requirements for both duty and proximate cause. Neither the court of appeals in this case nor this Court in Nixon, Walker v. Harris, and Exxon Corp. v. Tidwell, were wrong for relying upon law that establishes a foreseeability standard that applies to both duty and proximate cause because the standard is the same. Consistent with that approach, it is entirely proper for the Court to apply the foreseeability standard stated in Nixon to the duty analysis in this case.

Justice ENOCH filed a concurring opinion. Justice BAKER filed a concurring opinion. Justice O’NEILL filed a dissenting opinion, in which Chief Justice PHILLIPS and Justice HANKINSON join. Justice GONZALES did not participate in the decision.

Notes

. Holder sued for loss of consortium as next friend for her minor son.

. Holder did not request this Court to review the part of the court of appeals’ judgment that was adverse to her. As a result, that portion of the court of appeals’ judgment is undisturbed.

. This analysis is complementary, not contradictory, to the traditional premises liability categories. Therefore, this opinion should not be construed as supplanting the traditional premises liability analysis as it relates to a plaintiff's status.



Concurrence Opinion

Justice ENOCH

concurring.

I join the Court’s judgment. I can join neither the plurality opinion nor Justice Baker’s writing because I believe those opinions skip a critical step that could lead some to assume the Court has adopted a new common law duty — that a landowner has a general duty to not be negligent. That is not the law in Texas, and is not after today. Because I am concerned that this omission might mislead, I write separately.

This case presents a simple question: Whether a landowner may be held liable for injuries caused to a stranger who was brought to the premises against her will by the criminal attack of another stranger.1 To begin answering this question, I note that a landowner has no general duty to not be negligent toward those entering the land. The extent of a landowner’s liability for injuries caused by a condition existing on the land depends on the status of the injured person. Thus, the scope of a landowner’s duty depends on whether, at the time of the injury, the person on the land was an invitee, a licensee, or a trespasser.2

To invitees, the landowner owes a duty to exercise reasonable care to keep the premises in a reasonably safe condition for use by the invitee.3 To licensees, the landowner owes a duty to warn of or to make safe hidden dangers known to the landowner and a duty not to intentionally, wil-fully, or through gross negligence cause injury.4 And to trespassers, a landowner owes only a duty not to intentionally, wil-fully, or through gross negligence cause injury.5

While this traditional classification system has been subject to debate, it remains the law in Texas. Thus, I believe it must be applied in this case.

Because this case is strikingly similar to Nixon v. Mr. Property Management Co.,6 I consider that case instructive. There, ten-year-old R.M.V. was dragged into an apartment complex that she didn’t reside in by an unknown assailant and was sexually assaulted. Her next friend sued Mr. Property, the manager of the apartment complex, alleging that it breached a duty of care to R.M.V. The trial court granted summary judgment for Mr. Property. Holding that R.M.V. was a “trespasser,” and that Mr. Property’s duty was not to injure her wilfully, wantonly, or through gross negligence, the court of appeals affirmed.7

We reversed and remanded on the ground that a Dallas city ordinance requiring property owners to “keep the doors and windows of a vacant structure or va cant portion of a structure securely dosed to prevent unauthorized entry” imposed a standard of care on Mr. Property without regard to R.M.V.’s classification. We said:

[T]he question of what duty Mr. Property owed to R.M.V. is answered by the ordinance. This ordinance legislatively imposes a standard of conduct which we adopt to define the conduct of a reasonably prudent person- The unex-eused violation of a statute or ordinance constitutes negligence as a matter of law if such a statute or ordinance was designed to prevent injury to the class of persons to which the injured person belongs .... A reasonable interpretation of this ordinance is that it was designed to deter criminal activity by reducing the conspicuous opportunities for criminal conduct.... An ordinance requiring apartment owners to do their part in deterring crime is designed to prevent injurg to the general public. R.M.V. falls within this class. Since the ordinance was meant to protect a larger class than invitees and licensees, and since R.M.V. committed no wrong in coming onto the property, these premises liability distinctions are irrelevant to our analysis.8

The facts of this case are virtually indistinguishable from Nixon — we have an innocent victim taken against her will into a vacant area and sexually assaulted, followed by tort claims against the landowner for not taking steps to prevent the assault. But unlike the plaintiff in Nixon, Holder does not claim in this Court that an ordinance makes the traditional classification system “irrelevant.” Thus, we are left with the traditional premises liability classifications to determine Mellon’s duty.

Addressing these classifications, I note that no one asserts that Holder was an invitee. At the other end, Holder argues that because she didn’t enter Mellon’s property for her own purposes, she was not a trespasser. But the court of appeals in the Nixon case rightfully explained that the classification of visitors on one’s land “does not depend upon ... volition but upon knowledge and consent of [the landowner].”9 And that “[i]n the absence of knowledge and consent [the landowner’s] duty ... was no greater than not to ... [be wilful, wanton or grossly negligent].”10 Thus I agree with Justice O’Neill that for purposes of determining Holder’s status on Mellon’s property, the relevant question is not whether Holder meant to be in the garage, but “whether Mellon expressly or impliedly consented to [Holder’s] entry.”11 Where Justice O’Neill and I part ways is in answering this question.

Justice O’Neill concludes that there is a fact question about whether Mellon, by its conduct, impliedly granted Holder license to come into its garage.12 I disagree. First, the cases Justice O’Neill cites don’t support this conclusion. Each of these cases demonstrate a nexus between the activity during which the injury occurred and the implied license.13 ⅛ And none hold, as Justice O’Neill would, that a license implied for some is a license implied for all.14 Evidence that Mellon was aware of vagrants in the garage in no way implies that Mellon opened the garage to vehicular traffic at all hours of the day or night. And while my colleagues struggle to avoid calling Holder a “trespasser,” the summary judgment evidence establishes that that was her status under the nomenclature of the traditional premises liability categories. Rather than struggling with the terminology, the Court could more easily establish another less harsh-sounding term. Regardless, and accepting Holder’s blamelessness, this does not affect the legal analysis of Mellon’s duty.

As part of her argument, Holder cites section 197(1) of the Restatement (Second) of Torts.15 She contends that she had a “privilege” to enter Mellon’s property because she was in fear for her safety, and therefore, she was not a “trespasser” for purposes of determining the scope of Mellon’s duty. This argument is incomplete. I may agree with Holder that, guided by section 197(1) of the Restatement, she was privileged to go on to Mellon’s property. But I read that section to mean only that she is relieved of liability to Mellon for having done so. Should this Court adopt section 197(1), Holder could not, as a matter of law, be liable to Mellon for entering Mellon’s garage.

But whether Holder had a privilege to be in Mellon’s garage has nothing to do with the scope of Mellon’s duty to Holder. While section 345(1) of the Restatement (Second) of Torts declares that a landowner owes the same duty to a privileged trespasser that the landowner owes a licensee,16 I would be reluctant to adopt that section. Mellon’s duty is determined by Holder’s status. And Holder’s status is determined by whether Mellon consented to her presence in the garage. Mellon’s duty to Holder can’t change simply because Holder went on the property involuntarily.

Mellon owned a parking garage in downtown Houston. The garage was not open for public use and was not used at night. Mellon’s duty to those who were on the premises without Mellon’s consent was only to not intentionally, wilfully, or through gross negligence cause them injury.

Having determined that this was the duty Mellon owed to Holder, the next inquiry would be whether Mellon met its summary judgment burden to conclusively prove that it did not intentionally, wilfully, or through gross negligence injure Holder. Mellon met that burden. Consequently, it was up to Holder to present summary judgment evidence that raised a fact issue on consent. The evidence presented by Holder does not. Thus I concur in the Court’s judgment.

. See Totten v. More Oakland Residential Housing, Inc., 63 Cal.App.3d 538, 134 Cal.Rptr. 29, 32 (1976).

. See, e.g., Carlisle v. J. Weingarten, Inc., 137 Tex. 220, 152 S.W.2d 1073, 1074 (1941); Galveston Oil Co. v. Morton, 70 Tex. 400, 7 S.W. 756, 757-58 (1888).

. See, e.g., Carlisle, 152 S.W.2d at 1075.

. See, e.g., Texas-Louisiana Power Co. v. Webster, 127 Tex. 126, 91 S.W.2d 302, 306 (1936).

. See, e.g., State v. Williams, 940 S.W.2d 583, 584 (Tex.1996); Burton Constr. & Shipbuilding Co. v. Broussard, 154 Tex. 50, 273 S.W.2d 598, 603 (1954).

. 690 S.W.2d 546 (Tex.1985).

. See 675 S.W.2d 585, 587 (Tex.App.—Dallas 1984), rev’d, 690 S.W.2d 546 (Tex.1985).

. Nixon, 690 S.W.2d at 549 (emphasis added).

. Nixon v. Mr. Property Management Co., Inc., 675 S.W.2d 585, 586 (Tex.App.—Dallas 1984).

. Id.

. 5 S.W.3d at 672 (O’Neill, J., dissenting).

. See id. at 672.

.See id. at 672 (citing two Texas court of appeals cases that found that boys who swam frequently on property owned by governmental units were "gratuitous licensees” because the governmental units knew that boys were using the property for that purpose and took no steps to prevent it).

.See id.

. Restatement (Second) of Torts § 197(1) (1965).

. Restatement (Second) of Torts § 345(1).



Concurrence Opinion

Justice BAKER,

concurring.

As a general rule, a landowner has no legal duty to protect another from the criminal acts of a third party who is not under the landowner’s control or supervision. See Timberwalk Apartments, Partners, Inc. v. Cain, 972 S.W.2d 749, 756 (Tex.1998); Walker v. Harris, 924 S.W.2d 375, 377 (Tex.1996). To the extent that the law does impose a duty, the threshold issue is whether the risk of harm was foreseeable. See Timbenualk, 972 S.W.2d at 756; Walker, 924 S.W.2d at 377. I conclude as a matter of law that, under the record here, Mellon could not foresee the risk that a sexual assault would occur in its employee parking garage. Therefore, I agree with the plurality’s conclusion that Mellon did not owe Holder a duty. I cannot agree, however, with the plurality’s duty analysis. Therefore, I concur in the judgment and write separately.

I. THE PLURALITY

The plurality relies on Palsgraf v. Long Island Railroad, for its two-prong foreseeability test for duty. See Palsgraf v. Long Island R.R., 248 N.Y. 339, 162 N.E. 99 (1928). But even the plurality’s cited authorities recognize that, contrary to the opinion’s claim, Palsgraf’s two-prong duty analysis has not been “widely embraced.” 5 S.W.3d 654, 655; see Restatement (Second) of Torts § 281 Reporter’s Notes (1966) (noting that Palsgraf is “controversial” and that, as late as 1966, the decisions on facts that are at all analogous to Pals-graf s facts are “few and divided.”); Zipursky, Rights, Wrongs, and Recourse in the Law of Torts, 51 Vand. L.Rev. 1, 3 (1998)(“Leading scholars treat Palsgraf as a proximate cause case.... Cordozo’s own reasoning in Palsgraf is typically ignored or derided.”); see also Powers, Judge and Jury in the Texas Supreme Court, 75 Tex. L.Rev. 1699, 1702-03 (1997) (explaining that Dean Keeton’s approach to duty and proximate cause, in which questions about whether a defendant’s liability extends to a particular type of plaintiff are questions of proximate cause and not duty, has prevailed in Texas). Further, as the plurality concedes, the Texas cases it cites for the two-prong foreseeability analysis discuss foreseeability only in the context of proximate cause, not duty. See Nixon v. Mr. Property Management Co., 690 S.W.2d 546, 549-50 (Tex.1985); Texas Cities Gas Co. v. Dickens, 140 Tex. 433, 168 S.W.2d 208, 212 (1943); Carey v. Pure Distrib. Corp., 133 Tex. 31, 124 S.W.2d 847, 849-50 (1939); San Antonio & A.P. Ry. v. Behne, 231 S.W. 354, 356 (Tex.1921). Moving the determination of whether harm to a certain class of potential plaintiffs is foreseeable from the proximate cause analysis to the duty analysis changes Texas law in this type of case. It also changes the law in every negligence case that requires a duty analysis as a threshold issue. More importantly, it shifts the allocation of power in such cases. See Powers, Judge and Jury in the Texas Supreme Court, 75 Tex. L.Rev. at 1703. Traditionally, duty is a threshold legal issue the court properly decides. See Walker, 924 S.W.2d at 377; Powers, Judge and Jury in the Texas Supreme Court, 75 Tex. L.Rev. at 1703. Proximate cause is usually a jury issue. See Clark v. Waggoner, 452 S.W.2d 437, 440 (Tex.1970); City of Houston v. Jean, 517 S.W.2d 596, 599 (Tex.Civ.App.— Houston [1st Dist.] 1974, writ ref'd n.r.e.); see also Flores v. Sullivan, 112 S.W.2d 321, 323 (Tex.Civ.App.— San Antonio 1937), rev’d on other grounds, 134 Tex. 55, 132 S.W.2d 110 (Tex.1939); Powers, Judge and Jury in the Texas Supreme Court, 75 Tex. L.Rev. at 1703. Consequently, changing the duty analysis to include the traditional proximate cause foreseeability test allocates more power to trial judges, as well as appellate judges, to decide questions traditionally and properly reserved for the jury.

Rather than change the law of duty to add a second-prong foreseeability analysis, we need only consider the Timberwalk factors — similarity, proximity, recency, frequency, and publicity — to analyze foreseeability within the duty context as it arises here. See Timberwalk, 972 S.W.2d at 759.

II. FORESEEABILITY

Common-law negligence consists of these elements: (1) a legal duty; (2) a breach of that duty; and (3) damages proximately resulting from the breach. See El Chico Corp. v. Poole, 732 S.W.2d 306, 311 (Tex.1987). Duty is the threshold inquiry, which is a question of law for the court to decide. See Greater Houston Transp. Co. v. Phillips, 801 S.W.2d 523, 525 (Tex.1990). As a general rule, a landowner has no duty to prevent criminal acts of third parties who are not under the landowner’s control or supervision. See Timberwalk, 972 S.W.2d at 756; Walker, 924 S.W.2d at 377; Exxon Corp. v. Tidwell, 867 S.W.2d 19, 21 (Tex.1993). To the extent that the law does impose a duty, foreseeability is the initial analysis. See Timberwalk, 972 S.W.2d at 756; Walker, 924 S.W.2d at 377. Only after foreseeability is established must we determine the parameters of the duty. See Timberwalk, 972 S.W.2d at 757.

In Timberwalk, this Court stated the factors courts should consider in determining if criminal conduct on a landowner’s property is foreseeable: (1) whether any criminal conduct previously occurred on or near the property; (2) how recently it occurred; (3) how often it occurred; (4) how similar the conduct was to the conduct on the property; and (5) what publicity the occurrences received to show that the landowner knew or should have known about them. See Timberwalk, 972 S.W.2d at 757-58. We summarize these foreseeability factors as similarity, proximity, recency, frequency, and publicity of previous criminal conduct. See Timberwalk, 972 S.W.2d at 759. Courts must consider all the factors together. See Timberwalk, 972 S.W.2d at 759.

Past crimes must be sufficiently similar, though not identical, to the crime at issue to put the landowner on notice of the specific danger. See Timberwalk, 972 S.W.2d at 758. For example, automobile vandalism in an apartment complex does not put the landowner on notice of the likelihood of a sexual assault. See Timberwalk, 972 S.W.2d at 758.

Proximity requires evidence of other crimes on or in the property’s immediate vicinity. See Timberwalk, 972 S.W.2d at 757. Evidence of criminal activity occurring farther from the landowner’s property is less relevant than past criminal activity in the specific area at issue. See Timberwalk, 972 S.W.2d at 757.

Foreseeability also depends on the re-cency of past criminal conduct. See Timberwalk, 972 S.W.2d at 757-58. A significant number of crimes occurring in a short time period on or near the property makes the crime in question more foreseeable. See Timberwalk, 972 S.W.2d at 758.

Publicity of prior crimes strengthens the claim that a particular crime was foreseeable because a property owner can be expected to have knowledge of such criminal activity. See Timberwalk, 972 S.W.2d at 758. Landowners, however, have no duty to inspect criminal records to determine the risk of crime in the area. See Timberwalk, 972 S.W.2d at 759.

III. ANALYSIS

Mellon is entitled to summary judgment if it can establish as a matter of law that the sexual assault in Mellon’s parking garage was not foreseeable. Forseeablilty requires an analysis of frequency, recency, publicity, and similarity of previous criminal activity. See Timberwalk, 972 S.W.2d at 759. In reviewing a summary judgment, we assume all evidence favorable to the nonmovant to be true. See Nixon, 690 S.W.2d at 548-49.

Mellon’s garage is in downtown Houston. In the twenty-two months before Holder’s assault, 190 violent crimes had occurred within a one-quarter mile radius of the garage. The year that Holder was sexually assaulted, 88 violent crimes occurred in the area surrounding the garage: 4 sexual assaults, 57' robberies, and 27 aggravated assaults. Indeed, Holder’s expert, relying on police reports, testified that there were high crime rates in the area surrounding Mellon’s garage. But “[t]he frequent occurrence of property crimes in the vicinity is not as indicative of foreseeability as the less frequent occurrence of personal crimes on the landowner’s property itself.” Timberwalk, 972 S.W.2d at 759. The only evidence of criminal activity in Mellon’s garage is evidence of vagrancy and automobile theft. There is no evidence of personal crimes occurring in the garage.

On the publicity of criminal activity in the area, Holder complains that Mellon did not regularly check Houston police records. But landowners have no duty to regularly inspect criminal records to determine the risk of crime in the area. Nevertheless, two Mellon employees had written memos to Mellon in response to auto thefts occurring when the garage was occupied by employees’ vehicles. One of the memos discussed a crime increase in the area surrounding the garage. But its author testified that he based his information on rumors he had heard from other Mellon employees. Mellon responded to these memos by employing armed security guards during hours that Mellon employees would be using the garage. Mellon also provided security escorts for Mellon employees going to and from the garage.

The fact that there may have been frequent and recent criminal activity in the area surrounding the garage and that Mellon knew about certain criminal activities occurring in its garage does not alone mean that a sexual assault in the garage was foreseeable. We have stated that the frequency of previous crimes necessary to show foreseeability lessens as the similarity of the previous crimes to the incident at issue increases. See Timberwalk, 972 S.W.2d at 759. The converse is also true — the less similar previous crimes are to the one at issue, the frequency necessary to show foreseeability increases. Thus, we must consider whether such criminal activity was similar to the crime at issue. There is no summary judgment evidence that violent or personal crimes had occurred in Mellon’s garage. The evidence only shows that automobile thefts during business hours and vagrancy had occurred in the garage. Automobile thefts and vagrancy do not suggest the likelihood of sexual assault. See Timberwalk, 972 S.W.2d at 758. Nor is there summary judgment evidence that any of the four reported sexual assaults in the area surrounding the garage occurred in either a public or private parking garage or were otherwise similar to Holder’s.

Considering the summary judgment evidence here and all the Timberwalk factors, I conclude that although there is evidence of frequent and recent criminal activity in the area surrounding Mellon’s garage, and evidence that Mellon knew of vagrancy and automobile thefts in the garage itself, it was not foreseeable to Mellon that a sexual assault would occur in its garage.

IV. THE DISSENT

The dissent misstates our view when it claims we discount the two employee memos. To the contrary, the memos are relevant to show that the nature of the crimes reported in Mellon’s garage were auto thefts and vagrancy, not violent crimes against persons. The dissent also argues that we completely disregard the nature and character of the premises at issue. Although the Timberwalk factors are not exclusive, nothing in Timberwalk suggests that a court must take into account the nature and character of the premises at issue. By citing Gomez v. Ticor, the dissent argues that all parking garage owners should inherently foresee rapists lying in wait for unsuspecting victims at all hours of the day and night. 5 S.W.3d 654, 669 (citing Gomez v. Ticor, 145 Cal.App.3d 622, 628, 193 Cal.Rptr. 600 (1983)). In effect, the dissent would make all property owners insurers of the general public. This is not the rule in Texas. See Lefmark Management Co. v. Old, 946 S.W.2d 52, 59 (Owen, J., concurring); see also Timberwalk, 972 S.W.2d at 756; Walker, 924 S.W.2d at 377. The flaw in the dissent’s analysis is that the dissent fails to properly consider all the Timberwalk factors together. See Timberwalk, 972 S.W.2d at 759.

V. CONCLUSION

Because I would hold that Mellon could not foresee a sexual assault in its garage, and therefore, did not owe Holder a duty as a matter of law, I concur in the judgment.





Dissenting Opinion

Justice O’NEILL,

dissenting, joined by Chief Justice PHILLIPS and Justice HANKINSON.

In three opinions applying three different rationales, a divided Court concludes that Mellon is entitled to summary judgment. These opinions, none of which carries a majority, alternately conclude that (1) the crime victim was not foreseeable, (2) the crime committed was not foreseeable, and (3) Holder was a trespasser toward whom Mellon fulfilled its duty. I cannot agree, in light of the summary judgment evidence, that any of these fac tors was established as a matter of law. Accordingly, I respectfully dissent.

I

Foreseeability of Plaintiff

Applying the Timberwalk factors, the plurality concludes, as I do, that there is some evidence to show that violent criminal conduct in Mellon’s garage was foreseeable. See Timberwalk Apartments, Partners, Inc. v. Cain, 972 S.W.2d 749 (Tex.1998). They proceed, however, to employ a so-called “second prong” foreseeability analysis that focuses on the class of victim to determine the existence of a duty. Although this approach produces a seemingly desired result, it improperly bootstraps proximate cause foreseeability into the threshold duty question, thereby usurping the function of the traditional premises liability classifications. Whether or not the foreseeability analysis is the same for both duty and proximate cause purposes, as the plurality posits, the concept of foreseeability in the context of premises liability is embodied in the classifications that have defined a landowner’s duty for over one hundred years.

It is true that in Nixon v. Mr. Property Management Co., 690 S.W.2d 546 (Tex. 1985), and in two other cases cited in the plurality opinion, we stated that foreseeability requires “ ‘that the injured party should be so situated with relation to the wrongful act that injury to him or to one similarly situated might reasonably have been foreseen.’” Id. at 551 (Tex.1985) (quoting Carey v. Pure Distrib. Corp., 133 Tex. 31, 124 S.W.2d 847, 849 (1939)). But that analysis applied to the determination of proximate cause, which is typically an issue for the jury, not duty, which is typically a question for the court. Id. And if the foreseeability analysis is the same, as the plurality reasons, it is difficult to reconcile their conclusion with that reached in Nixon. On almost identical facts—a young girl was abducted from another location and sexually assaulted in an abandoned apartment — we held that a fact issue existed on whether the criminal assault was foreseeable to the management company that had left the apartment unlocked:

With a litany of prior crimes ... and with deposition testimony that vagrants frequented the area, a material fact question exists on the foreseeability of this crime as it relates to the proximate cause issue.

Id. (emphasis added). To hold now, on nearly identical facts, that foreseeability is lacking as a matter of law for duty rather than proximate cause purposes defies logic and ignores a primary function of the traditional premises liability classifications. Were we to abandon the traditional classification system and impose upon landowners a generalized duty to exercise reasonable care toward all entrants, as Holder urges, there might be a place for the “foreseeable plaintiff’ approach.1 The requirement that injury to the plaintiffs “class” be foreseeable, however, is inherent in the premises liability distinctions between “invitee,” “licensee,” and “trespasser.” Like Justice Enoch, I believe that these classifications govern our analysis. The inquiry should be whether Mellon established as a matter of law that it acted within the scope of any duty that it owed to Holder. The nature of that duty depends upon the status of the person entering the property.

Both Justices. Baker and Enoch agree that the second-prong foreseeability analysis is flawed, and decline to join the plurality opinion. Justice Baker applies the Timberwalk factors and concludes that, as a matter of law, a sexual assault in Mellon’s garage was not foreseeable. Like Justice Enoch and the plurality, I cannot agree that such a conclusion may be drawn from this summary judgment record.

II

Foreseeability of Crime

The summary judgment evidence shows that, from January 1, 1990, through the date of the incident, 190 violent crimes, including murders, rapes, robberies, and aggravated assaults, were reported within a quarter-mile radius of Mellon’s garage. This amounted to one reported violent crime every five days, and was enough to support a “High Crime” designation for the area in 1991 and an “Above Average” designation in 1992.

John Hilliard, a Mellon employee, testified by deposition that his Jeep was stolen out of the garage in October 1992. Hilli-ard sent a memo to the garage manager, Curtis Oblinger, among others, expressing Ms concern about a “drastic increase in crime in the surrounding area” in the previous six months. Hilliard had heard rumors of increased criminal activity from other Mellon employees, including reports of violent crime in the surrounding area. Hilliard proposed a plan for increased garage security, but Oblinger never responded to his memo.

Cathleen Hackward, another Mellon employee, sent an e-mail to Oblinger and others to “lodge a formal complaint about the virtually non-existent security for our parking garage.” She wrote that “people are free to roam through there, obviously committing crimes,” and stated that she was concerned for her personal safety. Hackward testified by deposition that she had Mellon’s security guard escort her to her car when she worked late because she did not consider it safe to go to the garage alone.

According to Hilliard, it was obvious that people were sleeping in the garage. There were blankets and newspapers rolled up “like someone was sleeping in the stairwell.” Oblinger knew that vagrants were going into the garage, and that they were drinking beer there. He did nothing, however, to prevent their entry.

Reviewing this evidence, Justice Baker concludes that “the risk that someone would be sexually assaulted in Mellon’s garage was not foreseeable to Mellon as a matter of law.” Such a conclusion drawn from this summary judgment record, in my opinion, blinks reality and strains the Timberwalk factors beyond their logical or intended reach.

Justice Baker draws a bright line between property crimes occurring inside Mellon’s garage and personal crimes occurring outside. He thus discounts the employee memos identifying property crime within the garage, and dismisses their reference to violent crime in the vicinity as “rumors.” It is clear, however, that the employees’ memos were written out of concern for their own personal safety, not just the security of their cars. The Hackward memo explicitly states, “not only am I worried about my car, but I fear for my personal safety as well.” And Hilli-ard testified in Ms deposition that the “drastic increase in crime” in the surrounding area to which Ms memo referred included reports of violent crimes, including an armed robbery. Hilliard’s memo to Oblinger suggested that the garage should be patrolled “to prohibit automobile theft and potential danger to employees.” In Timberwalk we held that, for a risk to be foreseeable, evidence of crimmal activity “either on the landowner’s property or closely nearby” may be considered. See Timberwalk, 972 S.W.2d at 757. Considering the crime that had occurred in the garage and the abundance of violent crime in the immediate area, it was entirely foreseeable that a sexual assault might occur in Mellon’s open and abandoned garage.

Justice Baker also discounts Holder’s evidence of prior violent crimes because there is no evidence “that any of the four reported sexual assaults in the area surrounding the garage occurred in either a public or private parking garage or were otherwise similar to Holder’s.” It is true that neither party presented evidence detailing the circumstances of the sexual assaults or other 190 violent crimes committed in the vicinity. But Timberwalk does not require such a heightened degree of similarity for purposes of determining foreseeability. See Timberwalk, 972 S.W.2d at 758. As we recognized in Timberwalk, it is difficult to compartmentalize criminal activity, and “[pjroperty crimes may expose a dangerous condition that could facilitate personal crimes.” Id. at 758. See also Galloway v. Bankers Trust Co., 420 N.W.2d 437, 439 (Iowa 1988) (stating “[w]e do not believe, however, that crimes initially directed toward property are without any probative value on the question of foreseeability of injury.”); Aaron v. Havens, 758 S.W.2d 446, 447-48 (Mo.1988) (stating “[i]t is not necessary to allege that past crimes involving entry into unauthorized places are of the same general nature as the one which gave rise to the claim.... If a burglar may enter, so may a rapist.”).

To the extent Justice Baker bases his “similarity” distinction upon the manner in which Holder was assaulted, i.e., that she was lured into the garage from another location, it is immaterial, for we have long recognized that what must be foreseeable is not the exact sequence of events that produces the criminal conduct, but only the general danger. See Walker v. Harris, 924 S.W.2d 375, 377 (Tex.1996); Lofton v. Texas Brine Corp., 777 S.W.2d 384, 387 (Tex.1989). And to the extent his distinction is based upon the differing nature of other crimes in the area, I fail to see it. In the year Holder was assaulted, four sexual assaults, fifty-seven robberies, and twenty-seven aggravated assaults occurred in close proximity to the garage. Any distinction that might be drawn between Holder’s assault and these prior violent crimes is inconsequential at best. Moreover, we stated in Timberwalk that “the frequency of previous crimes necessary to show foreseeability lessens as the similarity of the previous crimes to the incident at issue increases.” See Timberwalk, 912 S.W.2d at 759. Conversely, the similarity of previous crimes necessary to show foreseeability should lessen, to a certain extent, as the frequency of the previous crimes increases. See Jardel Co., Inc. v. Hughes, 523 A.2d 518, 525 (Del.1987) (stating “the repetition of criminal activity, regardless of its mix, may be sufficient to place the property owners on notice of the likelihood that personal injury, not merely property loss, will result.”). Here, any distinction that might be drawn between Holder’s assault and the other violent crimes diminishes in light of their sheer number.

Holder presented additional foreseeability evidence that accounts for the nature and character of the premises in issue, a parking garage, which Justice Baker’s opinion altogether disregards. While it is true that our decision in Timberwalk articulated similarity, proximity, recency, frequency, and publicity of previous criminal conduct as factors relevant to determine foreseeability, there is nothing to suggest that these factors are meant to be exclusive.2 Oblinger admitted in his deposition that he knew parking garages in downtown Houston are inherently susceptible to criminal activity. And the report' of Holder’s security expert, Horace Loomis, refers to “the inherently dangerous nature of unattended and unprotected parking garages.” Justice Baker’s opinion gives no consideration to the fact that the particular premises at issue may, under certain circumstances, pose a peculiar attraction for criminal misconduct. See Gomez v. Ticor, 145 Cal.App.3d 622, 628, 193 Cal.Rptr. 600 (1983) (stating that “the deserted ... nature of these structures, especially at night, makes them likely places for robbers and rapists to lie in wait”).

I agree with my fellow justices that “it was not unforeseeable as a matter of law that a rape might occur in the parking garage,” and therefore cannot join Justice Baker’s opinion. And I agree with Justice Enoch that the plurality’s analysis comes dangerously close to imposing upon landowners a general common law duty not to be negligent. Like Justice Enoch, I believe that the traditional premises liability distinctions govern our analysis. The inquiry should be whether Mellon established as a matter of law that it acted within the scope of any duty that it owed to Holder. The nature of that duty depends upon the status of the person entering the property.

Ill

Holder’s Status

At the outset, Holder urges us to abolish the traditional premises liability classifications applied by Texas courts for well over a century to determine a landowner’s duty to persons coming onto the property. That duty is defined by the entrant’s status as an invitee, licensee, or trespasser to the premises. See Rosas v. Buddies Food Store, 518 S.W.2d 534, 536 (Tex.1975); Carlisle v. J. Weingarten, Inc., 137 Tex. 220, 152 S.W.2d 1073, 1074-75 (1941); Galveston Oil Co. v. Morton, 70 Tex. 400, 7 S.W. 756, 757-58 (1888). According to Holder, we should follow the lead of those jurisdictions that have abrogated the traditional classification scheme, and define Mellon’s duty under ordinary negligence principles.

It is true that some jurisdictions have abolished the traditional classification scheme, regarding it as “unjust, unworkable and unpredictable.”3 See, e.g., Mi chael Sears, Abrogation of the Traditional Common Law of Premises Liability, 44 U. Kan. L. Rev. 175, 184 (1995). Those courts now define a landowner’s duty not in terms of the plaintiffs status, but in terms of foreseeable risk and reasonable care. See id. The California Supreme Court first articulated the rationale for doing so:

A man’s life or limb does not become less worthy of protection by the law nor a loss less worthy of compensation under the law because he has come upon the land of another without permission or with permission but without a business purpose. Reasonable people do not ordinarily vary their conduct depending upon such matters, and to focus upon the status of the injured party as a trespasser, licensee, or invitee in order to determine the question whether the landowner has a duty of care, is contrary to our modern social mores and humanitarian values. The common law rules obscure rather than illuminate the proper considerations which should govern determination of the question of duty.

Rowland, 70 Cal.Rptr. 97, 443 P.2d at 568. Thus, the traditional classifications have been criticized as occasioning inequitable results.

It has been noted, however, that while the movement to abolish the traditional scheme gathered momentum through the mid-1970s, it has since come to “a screeching halt.” Prosser & Keeton on the Law of Torts § 62, at 433. In the last decade, only Nevada has abolished all entrant classifications. See Moody v. Manny’s Auto Repair, 110 Nev. 320, 871 P.2d 935, 942-43 (1994). Most other jurisdictions have decided to retain the traditional classifications in some form, recognizing that their abrogation in favor of what has been criticized as “a standard with no contours” would create corresponding problems. Younce v. Ferguson, 106 Wash.2d 658, 724 P.2d 991, 995 (1986).

The premises liability classifications reflect policy judgments carefully developed over time to balance the landowner’s interest in the free use and enjoyment of his land against the interests of persons injured by the land’s condition. The categories and their corresponding duties place rational limits on the liability of landowners, assuring that property owners do not become absolute insurers against all risk of injuries that others might sustain on their property. These distinctions afford a degree of certainty to what would otherwise be an amorphous standard of liability, and provide relatively predictable rules by which landowners and entrants may assess the propriety of their conduct. As recently stated by the Supreme Court of Missouri in deciding to retain the traditional categories: “To abandon the careful work of generations for an amorphous ‘reasonable care under the circumstances’ standard seems — to put it kindly — improvident.” Carter v. Kinney, 896 S.W.2d 926, 930 (Mo.1995).

It is not surprising, then, that most jurisdictions continue to apply the traditional premises liability classifications.4 And several jurisdictions have attempted to reach a middle ground by abolishing the distinction between licensees and invitees, but retaining limited duty rules toward trespassers.5 While I agree that this middle road is far more compelling than the wholesale abandonment of the traditional classifications, we are not faced with that issue in this case. Because the traditional classifications are supported by many years of carefully developed law and public policy and afford relative certainty to an otherwise nebulous premises liability standard, I would decline to abandon them now. Far from “mak[ing] all property owners insurers of the general public,” as Justice Baker charges, I rely on well-established precedent in defining the duty owed to Holder by determining her status as an invitee, a licensee, or a trespasser to Mellon’s garage.

An invitee enters onto another’s land with the owner’s knowledge and for the mutual benefit of both parties. See Ro-sas, 518 S.W.2d at 536. The owner owes an invitee a duty of reasonable care to protect her from foreseeable injuries. Id. It is undisputed that Holder was not an invitee; her presence in the garage was neither for Mellon’s benefit nor with its knowledge.

The closer question is whether Holder was a licensee or a trespasser. A trespasser enters another’s property without express or implied permission. See Texas-Louisiana Power Co. v. Webster, 127 Tex. 126, 91 S.W.2d 302, 306 (1936); Weaver v. KFC Management, Inc., 750 S.W.2d 24, 26 (Tex.App.—Dallas 1988, writ denied). A licensee, by comparison, is a person who is privileged to enter on land only by virtue of the owner’s consent and “under such circumstances that he is not a trespasser.” Rowland v. City of Corpus Christi, 620 S.W.2d 930, 933 (Tex.Civ.App.—Corpus Christi 1981, writ ref'd n.r.e.); see also Dominguez v. Garcia, 746 S.W.2d 865, 866-67 (Tex.App.—San Antonio 1988, writ denied); Restatement (Second) of ToRts § 330 (1965). Licensees have been found to include:

those taking short cuts across the property ...; loafers, loiterers and people who come in only to get out of the weather; those in search of their children; servants or other third persons; spectators and sightseers not in any way invited to come; those who enter for social visits or personal business dealings with employees of the possessor of the land; tourists visiting a plant at their own request; those who come to borrow tools or to pick up and remove refuse or chattels for their own benefit; salesmen calling at the door of private homes, and those soliciting money for charity; and a stranger entering an office building to post a letter in a mailbox provided for the use of tenant^ only.

ProsseR and Keeton on the Law of Torts § 60, at 413 (citations omitted).

For purposes of distinguishing an invitee from a licensee, courts have often looked to the entrant’s purpose in coming onto the property. Thus, it has been said that a licensee’s presence on the premises is “for his own purposes, benefits, convenience or pleasure.” Rowland, 620 S.W.2d at 933; Smith v. Andrews, 832 S.W.2d 395, 397 (Tex.App.—Fort Worth 1992, writ denied). However, the traditional premises liability classifications have also been retained, in large part, to afford owners an element of certainty regarding their duty to entrants upon the property. In the present case, where it cannot be said that Holder entered the garage “for her own purposes, benefits, convenience or pleasure,” the more appropriate inquiry is whether Mellon expressly or impliedly consented to the entry. See Webster, 91 S.W.2d at 306; Rowland, 620 S.W.2d at 933; see also Restatement (Second) of Torts § 330 (1965).

It is undisputed that Holder did not have Mellon’s express consent to enter the garage. But consent to enter property may be manifested by the owner’s conduct or by the condition of the land itself. See Prosser and Keeton on the Law of Torts § 60, at 413. Situations clearly exist “where a trespass has been tolerated for such a sufficient period of time that the public believes it has the ‘permission’ of the possessor to use the property.” Murphy v. Lower Neches Valley Auth., 529 S.W.2d 816, 820 (Tex.Civ.App.—Beaumont 1975), rev’d on other grounds, 536 S.W.2d 561 (Tex.1976); see also Boydston v. Norfolk S. Corp., 73 Ohio App.3d 727, 598 N.E.2d 171 (Ohio Ct.App.1991)(stating that “[consent] can be implied from acquiescence to continued use of the property by the public”).

In Murphy v. Lower Neches Valley Authority, for example, a teenage swimmer was injured when he jumped into a canal and struck his head on a lump of clay. 529 S.W.2d at 817. The summary judgment evidence showed that boys swam in the canal every day, the defendant knew that boys swam in the canal yet never asked them to leave, and no signs prohibited their activity. Id. at 820. The court concluded that the defendant did not prove, as a matter of law, that the injured boy was a trespasser and not a licensee. Id.

Likewise, in City of El Paso v. Zarate, the plaintiff sued the City of El Paso after her two sons drowned in a muddy city pond. 917 S.W.2d 326, 329 (Tex.App.—El Paso 1996, no writ). The City claimed that the evidence was legally and factually insufficient to support the jury’s finding that the boys were licensees and not trespassers. Id. at 330. The court of appeals disagreed, holding that the City gave its implied permission to use the premises because it failed to fence the area, put up barricades, or post warning signs, even though it knew people often entered the area to remove dirt and knew that four years earlier a child almost drowned in the pond. Id. at 381. Conversely, in Smither v. Texas Utilities Electric Company, the court classified the injured party as a trespasser, rather than a licensee, when the evidence showed that efforts were made to prevent access to the premises. 824 S.W.2d 693, 694-95 (Tex.App.—El Paso 1992, writ dism’d by agt.).

That is not to say that every tolerance of an intrusion will imply an owner’s consent to enter the land. Instead, courts have articulated sound principles to determine the conditions under which consent may be inferred from the owner’s tolerance of continued trespass. First, consent to enter is not implied unless the owner has actual knowledge that people have been entering the land. Cf. Hall v. Holton, 330 So.2d 81, 83 (Fla.Dist.Ct.App.1976); Gonzalez v. Broussard, 274 S.W.2d 737, 738 (Tex.App.—San Antonio 1954, writ ref'd n.r.e.). And implied consent may only be found when an owner with actual knowledge fails to take reasonable steps to prevent or discourage those persons from entering the land. Compare Zarate, 917 S.W.2d at 331-32 (upholding trial court’s finding that plaintiff was a licensee and not a trespass er when defendant knew people used land but made no attempt to keep them out) with Longbottom v. Sim-Kar Lighting Fixture Co., 651 A.2d 621, 622-28 (Pa.Commw.Ct.1994) (holding that defendant school conclusively proved it did not consent to people climbing on roof when evidence showed school undertook various measures to prevent access). Finally, an owner need not take steps to evict known trespassers when doing so would be unduly burdensome or futile. See Boydston, 598 N.E.2d at 174 (quoting PROSSER and Keeton on The Law of ToRts § 60, at 414: “[T]he mere toleration of continued intrusion where objection or interference would be burdensome or likely to be futile ... is not in itself and without more a manifestation of consent”).

In the present case, the summary judgment evidence shows that Mellon knew people were using the garage on nights and weekends for drinking alcohol and sleeping, yet took no action to keep them away. There is some evidence that Mellon impliedly consented to public entry by failing to make any attempt to impede access to the garage or post no trespassing signs when it knew the public was in fact entering the garage and sleeping there. Mellon presented nothing to indicate that it would have been unduly burdensome or futile to attempt to keep the public from the garage, but rather stated only that the problem “wasn’t noteworthy of any corrective action being taken.” Based on this summary judgment record, I cannot conclude as a matter of law that Holder was a trespasser, rather than a licensee, on Mellon’s premises. See Wiley v. National Garages, Inc., 22 Ohio App.3d 57, 488 N.E.2d 915, 928 (Ohio Ct.App.1984) (conferring licensee status on plaintiff who was assaulted after parking in defendant’s parking garage on Sunday during “off hours” with owner’s implied permission). Nor do I find any support for Justice Enoch’s position that a license for the public to enter the garage on foot does not imply a license to enter by car.

When the plaintiff is a licensee, the owner is negligent with respect to the condition of the premises if

a. the condition posed an unreasonable risk of harm;
b. defendant had actual knowledge of the danger;
c. plaintiff did not have actual knowledge of the danger; and
d. defendant failed to exercise ordinary care to protect plaintiff from danger, by both failing to adequately warn plaintiff of the condition and failing to make that condition reasonably safe.

State v. Williams, 940 S.W.2d 583, 584 (Tex.1996) (per curiam opinion denying application for writ of error). Mellon’s motion for summary judgment did not address its potential liability if Holder were found to be a licensee, nor do we.

In sum, after properly placing the summary judgment burden on Mellon and resolving all inferences from the facts in Holder’s favor, I conclude that fact issues exist as to the foreseeability of the risk of criminal conduct in the garage and Mellon’s actual knowledge of that risk. Because the Court concludes otherwise, I respectfully dissent.

. Although disclaiming an intent to supplant the traditional premises liability classifications, the plurality does just that by analyzing the case as one involving a negligent activity, as in Palsgraf, rather than a premises defect. Such an approach comes dangerously close to imposing a general negligence duty on landowners for premises defects. Far from espousing the dissent’s position in Palsgraf, as the plurality charges, I follow well-established precedent that defines the duty of a landowner in the premises liability context.

. Substantial authority supports consideration of the nature and character of the premises as a factor in the foreseeability analysis. See Kendrick v. Allright Parking, 846 S.W.2d 453, 458 (Tex.App.—San Antonio 1992, writ denied) (recognizing the distinction between premises that are prone to attract criminal activity and those that are not); Castillo v. Sears, Roebuck & Co., 663 S.W.2d 60, 66 (Tex.App.—San Antonio 1983, writ ref’d n.r.e.) (recognizing that to leave a washateria open and unattended all night may impose a duty on the business to provide some sort of security, a duty that may not apply to a department store in a mall with employees present); see also Isaacs v. Huntington Mem'l Hosp., 38 Cal.3d 112, 211 Cal.Rptr. 356, 695 P.2d 653, 661 (1985) (stating that the "nature, condition and location of the defendant’s premises" should be considered in the duty analysis); W. Page Keeton et al., Prosser and Keeton on the Law of Torts § 33, at 201 (5th ed.1984) (stating that the defendant has a heightened duty to protect the plaintiff from third party crimes when "an especial temptation and opportunity for criminal misconduct” exists); Restatement (Second) of Torts § 344 cmt. f (1965) (stating that “[i]f the place or character of [a] business ... is such that [the landowner] should reasonably anticipate careless or criminal conduct on the part of third persons,” the landowner may have a duty to guard against it); id. § 302B, cmt. e, subcmt. G (1965) (noting that, when the defendant’s property affords "a peculiar temptation or opportunity for intentional interference likely to cause harm,” the defendant is required to guard against the intentional, or even criminal, conduct of others).

. In 1968, the Supreme Court of California abolished the traditional classifications and declared the ordinary negligence principles of foreseeable risk and reasonable care to be the standard for premises liability in California. See Rowland v. Christian, 69 Cal.2d 108, 70 Cal.Rptr. 97, 443 P.2d 561 (1968), superseded in part by statute as explained in Calvillo-Silva v. Home Grocery, 19 Cal.4th 714, 80 Cal.Rptr.2d 506, 968 P.2d 65, 71-72 (1998). Courts in a number of jurisdictions later followed California in abandoning all classifications, including that of trespasser. See, e.g., Webb v. City and Borough of Sitka, 561 P.2d 731, 732-33 (Alaska 1977), superseded in part by statute as explained in Alaska v. Shanti, 835 P.2d 1225, 1227 (Alaska 1992); Mile High Fence Co. v. Radovich, 175 Colo. 537, 489 P.2d 308, 311-15 (1971), superseded by statute as explained in Lakeview Assoc., Ltd. v. Maes, 907 P.2d 580, 582-83 (Colo.1995); Smith v. Arbaugh’s Restaurant, Inc., 469 F.2d 97, 100 (D.C.Cir.1972), cert. denied, 412 U.S. 939, 93 S.Ct. 2774, 37 L.Ed.2d 399 (1973); Pickard v. City & County of Honolulu, 51 Haw. 134, 452 P.2d 445, 446 (1969); Keller v. Mols, 129 Ill.App.3d 208, 84 Ill.Dec. 411, 472 N.E.2d 161, 163 (Ill.1984) (abolishing distinctions only with regard to child entrants); Cates v. Beauregard Elec. Coop., Inc., 328 So.2d 367, 370-71 (La.1976), cert. denied, 429 U.S. 833, 97 S.Ct. 97, 50 L.Ed.2d 98 (1976); Limberhand v. Big Ditch Co., 218 Mont. 132, 706 P.2d 491, 496 (1985) (construing statute to require duty of ordinary care to all); Moody v. Manny’s Auto Repair, 110 Nev. 320, 871 P.2d 935, 942-43 (1994); Ouellette v. Blanchard, 116 N.H. 552, 364 A.2d 631, 633-34 (1976); Basso v. Miller, 40 N.Y.2d 233, 386 N.Y.S.2d 564, 352 N.E.2d 868, 871-72 (1976); Mariorenzi v. Joseph DiPonte, Inc., 114 R.I. 294, 333 A.2d 127, 130-33 (1975) (but see Tantimonico v. Allendale Mut. Ins. Co., 637 A.2d 1056, 1062 (R.I.1994) (restoring trespasser status)).

. By 1996, twenty-three jurisdictions had abolished some or all of the premises liability categories. See Heins v. Webster County, 250 Neb. 750, 552 N.W.2d 51, 54-55 (1996) (providing comprehensive analysis of how other jurisdictions have dealt with question of whether to abrogate traditional classifications). However, fourteen jurisdictions had expressly retained the categories, and another fourteen had continued to apply the common-law classifications without specifically addressing their continued validity. See id. at 55.

. At least fifteen jurisdictions have repudiated the licensee-invitee distinction while maintaining the limited-duty rule for trespassers. See Nelson v. Freeland, 349 N.C. 615, 507 S.E.2d 882, 886-87 (1998) (abolishing licensee-invitee distinction but maintaining categories and citing the following cases as doing the same: Wood v. Camp, 284 So.2d 691 (Fla.1973); Jones v. Hansen, 254 Kan. 499, 867 P.2d 303 (1994); Poulin v. Colby College, 402 A.2d 846 (Me.1979); Baltimore Gas & Elec. Co. v. Flippo, 348 Md. 680, 705 A.2d 1144 (1998); Mounsey v. Ellard, 363 Mass. 693, 297 N.E.2d 43 (1973); Peterson v. Balach, 294 Minn. 161, 199 N.W.2d 639 (1972); Heins, 250 Neb. 750, 552 N.W.2d 51; Ford v. Bd. of County Comm'rs, 118 N.M. 134, 879 P.2d 766 (1994); O'Leary v. Coenen, 251 N.W.2d 746 (N.D.1977); Ragnone v. Portland Sch. Dist. No. 1J, 291 Or. 617, 633 P.2d 1287 (1981); Tantimonico v. Allendale Mut. Ins. Co., 637 A.2d 1056 (R.I.1994); Hudson v. Gaitan, 675 S.W.2d 699 (Tenn.1984); Antoniewicz v. Reszcynski, 70 Wis.2d 836, 236 N.W.2d 1 (1975); Clarke v. Beckwith, 858 P.2d 293 (Wyo.1993)). Six others have modified the common-law categories without abolishing them outright. Missouri and Kentucky, for example, recognize a duty of care to all entrants equal to that owed to invitees once the landowner is aware of the entrant’s presence. See Heins, 552 N.W.2d at 54—55. Connecticut passed legislation modifying the common law status of a social guest from licensee to invitee. Id. at 55. Illinois eliminated the classifications by statute in 1984. Id. at 55 Indiana and Maine judicially altered the status of social guest from licensee to invitee. Id.