Chief Justice HECHT delivered the opinion of the Court.
In actions for unintentional torts, the common law has long restricted recovery of purely economic damages unaccompanied by injury to the plaintiff or his property1 — a doctrine we have referred to as the economic loss rule.2 The rule serves to provide a more definite limitation on liability than foreseeability can and reflects a preference for allocating some economic risks by contract rather than by law.3 But the rule is not generally applicable in every situation; it allows recovery of economic damages in tort, or not, according to its underlying principles.4 The issue in this case is whether the rule permits a general contractor to recover the increased costs of performing its construction contract with the owner in a tort action against the project architect for negligent misrepresentations — errors — in the plans and specifications. We conclude that the economic loss rule does not allow recovery and accordingly reverse the judgment of the court of appeals5 and render judgment for the architect.
I
The Dallas Area Rapid Transportation Authority ("DART") contracted with LAN/STV to prepare plans, drawings, and specifications for the construction of a light rail transit line from Dallas's downtown West End to the American Airlines Center about a mile away. LAN/STV agreed to "be responsible for the professional quality, technical accuracy, and ... coordination of all designs, drawings, specifications, and other services furnished", and to be "liable to the Authority ... for all damages to the Authority caused by [LAN/STV's] negligent performance of any of the services furnished". DART incorporated LAN/STV's plans into a solicitation for competitive bids to construct the project. Martin K. Eby Construction Company, which had built two other DART light rail projects, one of which was designed by LAN/STV, submitted the low bid on this project, just under $25 million, and was awarded the contract. The contract provided an administrative procedure for Eby to assert contract disputes with DART, including complaints about design problems. Eby and LAN/STV had no contract with each other. Thus, LAN/STV was contractually responsible to DART for the accuracy of the plans, as was DART to Eby, but LAN/STV owed Eby no contractual obligation.6
Days after beginning construction, Eby discovered that LAN/STV's plans were full of errors — about bridge structures, manhole and utility line locations, subsurface soil conditions, an existing retaining wall, and many other aspects of the proposed construction. While Eby expected that, as on any project, 10% of the plans would be changed, it found that 80% of LAN/STV's drawings had to be changed. This disrupted Eby's construction schedule and required additional labor and materials. In all, Eby now calculates it lost nearly $14 million on the project.
Only seven months into what would turn out to be a 25-month job, Eby sued DART for breach of contract in the United States District Court.7 The court dismissed the case because Eby had not exhausted its administrative remedies against DART under their contract and Texas law.8 Eby then invoked DART's contract dispute procedures, claiming $21 million. The hearing officer not only rejected Eby's claim in its entirety, he concluded that DART was entitled to $2.4 million in liquidated damages from Eby. Eby filed an administrative appeal, but, before it was resolved, settled with DART for $4.7 million.
Meanwhile, Eby filed this tort suit against LAN/STV, asserting causes of action for negligence and negligent misrepresentation. After Eby and DART settled, this case proceeded to trial,9 but only on Eby's claim that LAN/STV negligently misrepresented the work to be done in its error-ridden plans.10 The jury agreed and assessed Eby's damages for its losses on the project at $5 million, but they also found that the damages were caused by Eby's and DART's negligence as well, and apportioned responsibility 45% to LAN/STV, 40% to DART, and 15% to Eby. The trial court concluded that Eby's $4.7 million settlement with DART should not be credited against the damages found by the jury, but that LAN/STV should be liable only for its apportioned share of the damages. Accordingly, the trial court rendered judgment for Eby for $2.25 million plus interest.
Both LAN/STV and Eby appealed, and following the court of appeals' affirmance,11 both petitioned for review. We granted both petitions,12 but as we view the case, we need only address LAN/STV's argument that Eby's recovery for negligent misrepresentation is barred by the economic loss rule.13 We begin by surveying the development of the rule in American law and its status in Texas. We then turn to its application in this case.
II
A
The law has long limited the recovery of purely economic damages in an action for negligence. An early example, oft-cited, is Justice Holmes's opinion in Robins Dry Dock & Repair Co. v. Flint,14 a suit by the charterers of a steamship against a dry dock for damages for loss of the use of the vessel from a delay in repairs due to the dry dock's negligence. The Supreme Court held that the charterers could not recover their economic damages from the dry dock, either as third-party beneficiaries of the contract between the owners and the dry dock,15 or for the dry dock's negligence. Justice Holmes explained:
Of course the contract of the [dry dock] with the owners imposed no immediate obligation upon the [dry dock] to third persons [the charterers] as we already have said, and whether the [dry dock] performed it promptly or with negligent delay was the business of the owners and of nobody else.... [The charterers'] loss arose only through their contract with the owners.... [N]o authority need be cited to show that, as a general rule, at least, a tort to the person or property of one man does not make the tort-feasor liable to another merely because the injured person was under a contract with that other unknown to the doer of the wrong.... The law does not spread its protection so far.16
Nearly sixty years later, Judge Higginbotham observed in State of Louisiana v. M/V Testbank that "Robins broke no new ground.... [T]he prevailing rule [in the United States and England] denied a plaintiff recovery for economic loss if that loss resulted from physical damage to property in which he had no proprietary interest."17 Judge Higginbotham cited Professor James's 1972 article, Limitations on Liability for Economic Loss Caused by Negligence: A Pragmatic Appraisal:
Under the prevailing rule in America, a plaintiff may not recover for his economic loss resulting from bodily harm to another or from physical damage to property in which he has no proprietary interest. Similarly, a plaintiff may not recover for economic loss caused by his reliance on a negligent misrepresentation that was not made directly to him or specifically on his behalf.18
"The reasons for this difference in treatment of indirect economic loss and physical damage," Professor James continued, "do not derive from the theory or the logic of tort law".19 Economic loss may be no less real than physical injury and just as foreseeable. In Robins, for example, the charterers' loss of business from the dry dock's negligent delay in repairing the steamship was readily foreseeable, but so would have been the charterers' clients' loss of business, and so on. Justice Holmes' abrupt curtailment of this rippling liability — "[t]he law does not spread its protection so far"20 — could have been achieved by taking a more restrictive view of foreseeability. But, wrote Professor James,
judges who have been unwilling to accept narrow and unrealistic views of what is foreseeable — or of what a jury may find to be unforeseeable — remain generally unwilling to allow recovery for indirect economic loss. The explanation for this reluctance, repeated in decisions over the years, is a pragmatic one: the physical consequences of negligence usually have been limited, but the indirect economic repercussions of negligence may be far wider, indeed virtually open-ended. As Cardozo put it in a passage often quoted, liability for these consequences would be "liability in an indeterminate amount for an indeterminate time to an indeterminate class."21
Liability for economic loss directly resulting from physical injury to the claimant or his property — such as lost wages or medical bills — is limited by the scope of the injury. Liability for a stand-alone economic loss is not.22
Often, a more appropriate remedy for the victim is to allocate the risk of loss by contract or to cover it through insurance.23 In Judge Posner's view:
This is simply generalizing to tort law the contract-law rule of Hadley v. Baxendale.... The point in Hadley ... was that the carrier could not estimate the loss that the customer would incur from a delay in the delivery of the repaired mill shaft to the customer, but the customer could estimate this cost and, therefore, was in a better position to avoid the loss by taking appropriate precautions or by buying insurance.24
Thus, for example, "when a defective product purchased in a commercial transaction malfunctions, injuring only the product itself and causing purely economic loss", protection from that kind of harm, the United States Supreme Court has held (in an admiralty case), should be "left entirely to the law of contracts" because "the parties may set the terms of their own agreements."25 Determining whether a provision. for recovery of economic loss is better left to contract helps delineate between tort and contract claims. As one commentator has explained:
If there is a convincing rationale for the economic loss rule, it is that the rule performs a critical boundary-line function, separating the law of torts from the law of contracts. More specifically, "[t]he underlying purpose of the economic loss rule is to preserve the distinction between contract and tort theories in circumstances where both theories could apply."26
Since Professor James's seminal article, much has been written on the development of the rule limiting recovery of economic damages in tort actions.27 From our review of the cases and commentary on the subject, we think the principal rationales for the rule are well-summarized by Dean Farnsworth in the recently approved Restatement (Third) of Torts: Liability for Economic Harm, which we quote at length:
Economic injuries may be no less important than injuries of other kinds; a pure but severe economic loss might well be worse for a plaintiff than a more modest personal injury, and the difference between economic loss in itself and economic loss resulting from property damage may be negligible from the victim's standpoint. For several reasons, however, courts impose tort liability for economic loss more selectively than liability for other types of harms.
(1). Indeterminate and disproportionate liability. Economic losses proliferate more easily than losses of other kinds. Physical forces that cause injury ordinarily spend themselves in predictable ways; their exact courses may be hard to predict, but their lifespan and power to harm are limited. A badly driven car threatens physical harm only to others nearby. Economic harm is not self-limiting in this way. A single negligent utterance can cause economic loss to thousands of people who rely on it, those losses may produce additional losses to those who were relying on the first round of victims, and so on. Consequences of this sort may be at least generally foreseeable to the person who commits the negligent act. Defendants in such cases thus might face liabilities that are indeterminate and out of proportion to their culpability. Those liabilities may in turn create an exaggerated pressure to avoid an activity altogether.
(2). Deference to contract. Risks of economic loss tend to be especially well suited to allocation by contract. First, economic injuries caused by negligence often result from a decision by the victim to rely on a defendant's words or acts when entering some sort of transaction — an investment in a company, the purchase of a house, and so forth. A potential plaintiff making such a decision has a full chance to consider how to manage the risks involved, whether by inspecting the item or investment, obtaining insurance against the risk of disappointment, or making a contract that assigns the risk of loss to someone else. Second, money is a complete remedy for an economic injury. Insurance benefits, indemnification by agreement, or other replacements of money payments are just as good as the money lost in a transaction that turns out badly. This fungibility makes those other ways of managing risk — insurance, indemnity, and the like — more attractive than they might be to a party facing a prospect of personal injury.
Those same points often will make it hard for a court to know what allocation of responsibility for economic loss would best serve the interests of the parties to a risky situation. A contract that settles responsibility for such a risk will therefore be preferable in most cases to a judicial assignment of liability after harm is done. The contract will better reflect the preferences of the parties and help prevent the need for speculation and litigation later. Contracts also are governed by a body of commercial law that has been developed to address economic loss, and thus will often be better suited for that task than the law of torts. In short, contracts to manage the risk of economic loss are more often possible, and more often desirable, than contracts to manage risks of other types of injury. As a result, courts generally do not recognize tort liability for economic losses caused by the breach of a contract between the parties, and often restrict the role of tort law in other circumstances in which protection by contract is available.28
Thus, the Restatement concludes, while there is "no general duty to avoid the unintentional infliction of economic loss",29 the duty may exist when the rationales just stated for limiting recovery are "weak or absent"30 — cessante ratione legis cessat et ipsa lex.31
B
The absence of a bright-line rule, and the failure to analyze whether denying tort recovery for an economic loss in a particular kind of situation is justified by the rationales for limiting recovery of such losses, has led to some confusion. In a 1992 article, then-Professor Powers called Texas law on the subject "murky".32 One thing certain was that the damage caused by a defective product to itself cannot be recovered in an action for strict products liability,33 even if there is also personal injury or injury to other property.34 Recovery of such damages must be for breach of contract or warranty. It was also fairly clear that one party to a contract cannot recover from another party, in an action for negligence, an economic loss to the subject of the contract.35
The Restatement now concludes generally that "there is no liability in tort for economic loss caused by negligence in the performance or negotiation of a contract between the parties."36 It was less clear twenty years ago, and still is today, the extent to which Texas precludes recovery of economic damages in a negligence suit between contractual strangers, notwithstanding the rule's genesis in such cases, like Robins. As Professor Powers observed, "[a]lthough cases between contractual strangers are the paradigm of the traditional `economic loss' rule, no Texas case involving `strangers' expressly addresses the economic loss rule."37 Professor Powers noted that this Court had suggested in dicta that purely economic damages are recoverable in a negligence action between contractual strangers but later appeared to have rejected that possibility.38
Since then, Texas courts of appeals have uniformly applied the economic loss rule to deny recovery of purely economic losses in actions for negligent performance of services.39 Professional malpractice cases are an exception. A client can recover purely economic losses from a negligent lawyer, regardless of whether the lawyer and client have a contract.40 Lawyer malpractice is actionable as negligence no doubt because agreements regarding legal representation are not required in Texas, except for contingent fees,41 and until relatively recently have not been the norm. Also, the standards governing legal representation are deeply developed and their application uniform and well-settled. These factors also support negligence actions against other professionals.42
Although Texas courts have repeatedly invoked the economic loss rule to disallow recovery of purely economic losses in actions for negligent services not involving professionals, this Court, without citing the rule, has allowed recovery of such losses in an action for negligent misrepresentation, the cause of action in the present case. We first recognized the action, defined by section 552 of the Restatement (Second) of Torts,43 in Federal Land Bank Association of Tyler v. Sloane, where we held that prospective borrowers could recover the costs they incurred (but not lost profits) in relying on their lender's negligent misrepresentation to them that their loan application would be approved.44 Later, in McCamish, Martin, Brown & Loeffler v. F.E. Appling Interests, we held that while a non-client cannot recover against a lawyer for negligence,45 a lawyer may be liable for negligent misrepresentation to a non-client, but only in narrow circumstances, "when information is transferred by an attorney to a known party for a known purpose", liability is not expressly limited or disclaimed but invited, and the claimant has "justifiably rel[ied] on a lawyer's representation of material fact", which cannot ordinarily occur in an adversarial context.46 Most recently, in Grant Thornton LLP v. Prospect High Income Fund, Ltd., we held that an accountant may be liable to a strictly limited group of investors who justifiably rely on negligent misrepresentations in a corporate audit report.47 But we denied the claims in that case because the plaintiffs were merely potential investors with no special relationship to the audited corporation, and given their knowledge of the corporation and the marketplace, their reliance was not justified.48
These cases should not be read to suggest that recovery of economic loss is broader for negligent misrepresentation than for negligent performance of services. We agree with the Restatement that "[t]he general theory of liability is the same" for both torts, which is that
[a] plaintiff's reliance alone, even if foreseeable, is not a sufficient basis for recovery; under either [tort] a defendant generally must act with the apparent purpose of providing a basis for the reliance. It may be useful to say that a defendant held liable under either [tort] must "invite reliance" by the plaintiff, so long as the expression is understood to refer to the defendant's apparent purpose and not to a temptation incidentally created by the defendant's words or acts.49
And for both torts, whether and how to apply the economic loss rule "does not lend itself to easy answers or broad pronouncements."50 Rather, as we have already observed, the application of the rule depends on an analysis of its rationales in a particular situation.
III
Eby argues that the economic rule should not apply in this case when it did not bar recovery in our other negligent misrepresentation cases, Sloane, McCamish, and Grant Thornton. LAN/STV counters that to allow such recovery on construction projects, where relationships are contractual and certainty and predictability in risk allocation are crucial, would be disruptive.
Construction projects operate by agreements among the participants. Typically, those agreements are vertical: the owner contracts with an architect and with a general contractor, the general contractor contracts with subcontractors, a subcontractor may contract with a sub-subcontractor, and so on. The architect does not contract with the general contractor, and the subcontractors do not contract with the architect, the owner, or each other.
We think it beyond argument that one participant on a construction project cannot recover from another — setting aside the architect for the moment — for economic loss caused by negligence. If the roofing subcontractor could recover from the foundation subcontractor damages for extra costs incurred or business lost due to the latter's negligent delay of construction, the risk of liability to everyone on the project would be magnified and indeterminate — the same result Justice Holmes rejected in Robins. As the Restatement explains:
There is no liability in tort ... when the owner of a construction project sues a subcontractor for negligence resulting in economic loss; nor is liability found when one subcontractor is sued by another because the negligence of the first drives up the costs of the second. A subcontractor's negligence in either case is viewed just as a failure in the performance of its obligations to its contractual partner, not as the breach of a duty in tort to other subcontractors on the same job, or to the owner of the project. This way of describing the subcontractor's role is not inevitable in all cases. General rules are favored in this area of the law, however, because their clarity allows parties to do business on a surer footing. In this setting, a rule of no liability is made especially attractive by the number and intricacy of the contracts that define the responsibilities of subcontractors on many construction projects. That web of contracts would be disrupted by tort suits between subcontractors or suits brought against them by a project's owner.51
The issues are whether to treat the architect differently and whether to distinguish between an action for negligent performance of services and an action for negligent misrepresentations. On the latter issue, we agree with the Restatement: "[b]oth [torts] are based on the [same] logic" and "[t]he general theory of liability is the same".52 The economic loss rule should not apply differently to these two tort theories in the same situation.
On the former issue, we diverge from the Restatement. We agree that
[t]he plans drawn by the architect are intended to serve as a basis for reliance by the contractor who forms a bid on the basis of them and is then hired to carry them out. The architect's plans are analogous to the audit report that an accountant supplies to a client for distribution to potential investors — a standard case of liability [for negligent misrepresentation].53
But we think the contractor's principal reliance must be on the presentation of the plans by the owner, with whom the contractor is to reach an agreement, not the architect, a contractual stranger. The contractor does not choose the architect, or instruct it, or pay it. Under McCamish, the contractor could not recover economic damages from the owner's lawyer's negligent drafting of the construction contract. And while there is some analogy between the architect's plans and an accountant's audit report, under Grant Thornton, the latter is not an invitation to all investors to rely, but only those to whom it is more specifically directed. Here, the architect's plans are no more an invitation to all potential bidders to rely.
The Restatement adds that if allowing recovery against the architect in negligence "is not congenial to the parties, they are free to change it in the contracts that link them."54 But the parties are just as free to provide for liability by contract that the law does not allow in tort. The Restatement acknowledges this, noting that if the architect is contractually liable to the owner for defects in the plans, and the owner in turn has the same liability to the contractor, the contractor is protected.55 But the Restatement concludes that while this assignment of risk by contract should be encouraged, it jeopardizes unsophisticated parties:
Forbidding tort claims between parties who are indirectly linked by contract would put pressure on them to specify their rights carefully in advance, thus sparing courts the need to inquire into them later. But that incentive is most likely to be noticed by sophisticated parties negotiating large projects, and for them the rule is unlikely to be of great importance. They will negotiate allocations of risk that look similar in the end notwithstanding the rule of tort law in the background. Meanwhile, less sophisticated parties would stand a good chance of being tripped up by a broad rule, as when they fail to provide for indemnification in some direction and inadvertently leave a party who has been wronged with no remedy.56
We think it more probable that a contractor will assume it must look to its agreement with the owner for damages if the project is not as represented or for any other breach.
Though there remains the possibility that a contractor may not do so, we think the availability of contractual remedies must preclude tort recovery in the situation generally because, as stated above, "clarity allows parties to do business on a surer footing".57 "Where contracts might readily have been used to allocate the risk of a loss," the Restatement observes, "a duty to avoid the loss is unlikely to be recognized in tort — not because the economic loss rule applies, but simply because courts prefer, in general, that economic losses be allocated by contract where feasible."58 We see no reason not to apply the economic loss rule to achieve this end.
Analyzing the economics of the construction site, Professor Powers proposed this result more than twenty years ago, and we quote his analysis at length:
In fact, construction disputes ... are good candidates for precluding recovery under the "economic loss" rule, because the parties are in a position to protect themselves through bargaining. Though the parties do not necessarily have contracts with each other, they typically all have contracts with the owner, or subcontracts with someone who does have a contract with the owner. If contractors want to be protected, they can insist on that protection from the owner who will get protection from the architect. The contractors can take less compensation from the owner, so that the owner can in turn compensate the architect for the added risk.
The issue is who will buy business protection insurance. It makes sense to let the parties bargain about this rather than impose a "legal" solution....
There are two additional reasons to decline imposing a general tort duty on architects and engineers. First, imposing the risk of economic loss on the architect requires the architect to pass the cost along to the owner. The owner will then pass the cost along to the various contractors and subcontractors. Different contractors and subcontractors have different susceptibilities to economic loss, but the owner has no way of distinguishing among the various contractors and subcontractors. Some contractors and subcontractors will benefit greatly, some will not. Yet all will pay the price for this protection, not in proportion to their benefit from the protection, but roughly in proportion to the dollar value of their services. This will lead to a cross-subsidization. Contractors and subcontractors who are not subject to losses from delays effectively "pay" for protection that they do not need. In effect, they subsidize other contractors and subcontractors who are more susceptible to this type of loss.
This inequity could be remedied if the owner could determine which contractors and subcontractors benefit most and then charge them more by paying them less. But this would require the owner to be in the business of evaluating contractors' susceptibility to economic loss, which would effectively put the owner in the insurance evaluation business. Individual contractors and subcontractors are in a better position to evaluate their own susceptibility to economic loss and determine whether to buy insurance. Thus, fairness and efficiency support leaving these losses on the contractors and subcontractors, who can decide for themselves whether and for how much to insure. I assume this is part of the explanation why current contractual practice does not shift these obligations to the architect.
Second, ... contracts between owners and supervising architects can vary. Sometimes the supervising architect might be hired for the benefit of the contractors and subcontractors. However, in most cases, the architect is hired either as a neutral arbitrator or, most often, as the agent of the owner.... If the architect is supposed to be neutral or to operate as the agent of the owner, negligence principles — which would be decided by the jury after the fact — would create a chilling effect on the architect's neutrality or fiduciary duty to the owner.
This analysis suggests that each situation is different and that courts should use contract principles[,] not tort principles, to determine whether the architect has "contractual" obligations to the contractors and subcontractors.59
Finally, the courts are fairly evenly divided over whether to apply the economic loss rule in this situation.60 We side with those who do.
DART was contractually responsible to Eby for providing accurate plans for the job. Eby agreed to specified remedies for disputes, pursued those remedies (when the federal court would not allow it to sue), and settled its claims for $4.7 million. Had DART chosen to do so, it could have sued LAN/STV for breach of their contract to provide accurate plans. But Eby had no agreement with LAN/STV and was not party to LAN/STV's agreement with DART. Clearly, the economic loss rule barred Eby's subcontractors from recovering their own delay damages in negligence claims against LAN/STV. We think Eby should not be treated differently.
* * * * * *
The reasons for the economic loss rule support its application in this case to preclude a general contractor from recovering delay damages from the owner's architect. Accordingly, we reverse the judgment of the court of appeals and render judgment that Eby take nothing from LAN/STV.
