Opinion

In a client’s action against an attorney for legal malpractice, the client must prove, among other things, that the attorney’s negligent acts or omissions caused the client to suffer some financial harm or loss. When the alleged malpractice occurred in the performance of transactional work (giving advice or preparing documents for a business transaction), must the client prove this causation element according to the “but for” test, meaning that the harm or loss would not have occurred without the attorney’s malpractice? The answer is yes.

I

In 1984, plaintiffs Michael Viner and his wife, Deborah Raffin Viner, founded Dove Audio, Inc. (Dove). The company produced audio versions of books read by the authors or by celebrities, and it did television and movie projects.

In 1994, Dove went public by issuing stock at $10 a share. In 1995, the Viners and Dove entered into long-term employment contracts guaranteeing the Viners, among other things, a certain level of salaries, and containing indemnification provisions favorable to the Viners. The Viners received a large share of Dove’s common stock and all of its preferred cumulative dividend series “A” stock.

Thereafter, Michael Viner discussed with longtime friend David Povich, a partner in defendant law firm Williams & Connolly in Washington, D.C., the possibility of selling the Viners’ interest in Dove. In the fall of 1996, Norton Herrick proposed buying the Viners’ entire interest in Dove. Attorney Povich assigned the matter to his partner, defendant Charles A. Sweet, a corporate transactional attorney. Sweet was not a member of the California Bar and was not familiar with California law. During the negotiations with Herrick, Sweet learned that under the Viners’ employment agreements with Dove, the latter owed the Viners a substantial amount of unpaid dividends on their preferred stock. Sweet also learned that the Viners wanted to preserve their right to engage in the television and movie businesses.

When the negotiations with Herrick were unsuccessful, Ronald Lightstone of Media Equities International (MEI) approached the Viners. Thereafter, in March 1997, the Viners and MEI entered into an agreement under which MEI was to invest $4 million, and the Viners $2 million, to buy Dove stock. By May 1997, disputes arose, and the parties to the agreement each threatened litigation. That same month, Ronald Lightstone of MEI and Michael Viner, without defendant attorney Sweet’s involvement, agreed that MEI would buy the Viners’ stock in Dove and the Viners would terminate their employment with Dove.

Defendant attorney Sweet and Lightstone of MEI negotiated the final agreement, which the parties signed on June 10, 1997. The deal consisted of a securities purchase agreement and an employment termination agreement. Under the former, MEI agreed to buy a significant portion of the Viners’ stock for more than $3 million. Under the latter agreement, the Viners’ employment with Dove was terminated, mutual general releases were given, and Dove was to pay the Viners a total of $1.5 million over five years in monthly payments, with Dove’s series “E” preferred stock to be held in escrow for distribution to the Viners if Dove defaulted on the monthly payments to them.

The employment termination agreement contained a noncompetition provision stating that the Viners would not “ ‘compete’ in any way, directly or indirectly, in the audio book business for a period of four years” in any state in which Dove was doing business. The agreement also had a nonsolicitation provision that the Viners would not “directly or indirectly contract with, hire, solicit, encourage the departure of or in any manner engage or seek to employ any author or, for purposes of audio books, reader, currently under contract or included in the Company’s book or audio catalogues for a period of four years.”

In addition, the employment termination agreement provided that Deborah Raffin Viner would receive “Producer Credit” on audiobook work initiated during her employment with Dove; that Dove would not amend documents to terminate or reduce its obligation to indemnify the Viners; and that disputes would be submitted to arbitration, whose costs were to be split equally between the parties, with attorney fees to the party seeking to enforce the arbitration in court.

Defendant attorney Sweet led the Viners to believe that the employment termination agreement gave them three years of monthly payments by Dove, retained the indemnity protection they had with Dove, and provided credit for work done before their departure from Dove. The Viners also thought that they could use their celebrity contacts for any work that did not compete with Dove’s audiobook business and involvement in film and television productions, and that if Dove defaulted on the agreed-upon monthly payments to them, the noncompetition clauses would be voided. The contracts did not so provide.

Later, several arbitration proceedings took place to resolve disputes between the Viners and MEI, including a claim by the Viners that the noncom-petition provision of the employment termination agreement violated Business and Professions Code section 16600’s restrictions on noncompetition agreements. The arbitrator rejected the claim, and the superior court confirmed the arbitrator’s decision.

On June 3, 1998, the Viners brought a malpractice action against Attorney Sweet and the law firm of Williams & Connolly. Presented at trial were these seven claims: (1) Sweet told the Viners that the nonsolicitation clause of the employment termination agreement prohibiting plaintiffs from using their contacts to obtain work in television and movie projects applied only to the book and audiobook parts of Dove’s business, but Dove, because the clause was ambiguous, asserted that the clause also encompassed Dove’s television and movie projects; (2) Sweet negligently agreed to the noncom-petition provision, which violated Business and Professions Code section 16600’s restrictions on such provisions; (3) the Viners had asked for an attorney fees provision, but the employment termination agreement disallowed attorney fees in any disputes, permitting them only in enforcing an arbitration award; (4) ambiguous language in the Producer Credit provision caused Dove not to give Deborah Raffin Viner credit as a producer; (5) the Viners lost rights to dividends on Dove’s series “A” preferred stock; (6) the employment termination agreement did not contain an indemnity provision providing the same level of protection as the Viners’ agreement with Dove; and (7) the series “E” stock afforded inadequate security to the Viners if Dove defaulted on the monthly payments due them under the employment termination agreement.

After deliberating five days, the jury found defendants liable on all seven claims of malpractice, awarding the Viners $13,291,532 in damages. Defendants moved for judgment notwithstanding the verdict or in the alternative for a new trial, arguing that the trial court erred in not instructing the jury that the Viners needed to prove they would have received a better deal “but for” defendant attorney Sweet’s negligence. The trial court denied both motions.

The Court of Appeal reduced the damage award to $8,085,732, but otherwise affirmed the judgment. The court first noted that it was undisputed that the Viners did not attempt to prove that without defendants’ alleged negligence MEI would have given them a better deal on the contract terms here in issue. The court determined that the case presented a pure question of law: whether plaintiffs in a transactional legal malpractice action must show that the harm would not have occurred but for the alleged negligence. It held that the “but for” test of causation did not apply, to transactional malpractice.

The Court of Appeal distinguished transactional malpractice from litigation malpractice, in which the plaintiff is required to prove the harm would not have occurred without the alleged negligence, and it offered three reasons for treating the two forms of malpractice differently. First, the court asserted that in litigation a gain for one side is always a loss for the other, whereas in transactional work a gain for one side could also be a gain for the other side. Second, the court observed that litigation malpractice involves past historical facts while transactional malpractice involves what parties would have been willing to accept for the future. Third, the court stated that “business transactions generally involve a much larger universe of variables than litigation matters.” According to the Court of Appeal, in “contract negotiations the number of possible terms and outcomes is virtually unlimited,” and therefore the “jury would have to evaluate a nearly infinite array ■of ‘what-ifs,’ to say nothing of ‘if that, then whats,’ in order to determine whether the plaintiff would have ended up with a better outcome ‘but for’ the malpractice.”

We granted defendants’ petition for review, and thereafter limited the issues to whether the plaintiff in a transactional legal malpractice action must prove that a more favorable result would have been obtained but for the alleged negligence.

II

Defendants contend that in a transactional malpractice action, the plaintiff must show that but for the alleged malpractice, a more favorable result would have been obtained. Thus, defendants argue, the Viners had to show that without defendants’ negligence (1) they would have had a more advantageous agreement (the “better deal” scenario), or (2) they would not have entered into the transaction with MEI and therefore would have been better off (the “no deal” scenario).

The Viners respond that in Mitchell v. Gonzales (1991) 54 Cal.3d 1041 [ 1 Cal.Rptr.2d 913 , 819 P.2d 872 ], this court repudiated the “but for” test of causation in tort cases alleging negligence. Not so.

In Mitchell , the parents of a boy who died while on a picnic with neighbors sued the neighbors for wrongful death. The child, who could not swim, was riding a paddleboard in a lake when the paddleboard capsized and he drowned. Addressing causation, a majority of this court held that, for use in jury instructions, the term “proximate cause” was “conceptually and grammatically deficient” because it could mislead jurors into focusing on the cause that as to time and space was nearest to the injury. (Mitchell v. Gonzales, supra, 54 Cal.3d at p. 1052 .)

In so holding, Mitchell did not abandon or repudiate the requirement that the plaintiff must prove that, but for the alleged negligence, the harm would not have happened. On the contrary, Mitchell stated that jury instructions on causation in negligence cases should use the “substantial factor” test articulated in the Restatement Second of Torts (Restatement), and Mitchell recognized that “the ‘substantial factor’ test subsumes the ‘but for’ test.” (Mitchell v. Gonzales, supra, 54 Cal.3d at p. 1052 , italics added.) Mitchell also stated that “nothing in this opinion should be read to discourage the Committee on Standard Jury Instructions from drafting a new and proper ‘but for’ instruction.” (Id. at p. 1054, fn. 10.) In Rutherford v. Owens-Illinois, Inc. (1997) 16 Cal.4th 953, 968 [ 67 Cal.Rptr.2d 16 , 941 P.2d 1203 ], this court affirmed that “California has definitively adopted the substantial factor test of the Restatement Second of Torts for cause-in-fact determinations.”

The text of Restatement section 432 demonstrates how the “substantial factor” test subsumes the traditional “but for” test of causation. Subsection (1) of section 432 provides: “Except as stated in Subsection (2), the actor’s negligent conduct is not a substantial factor in bringing about harm to another if the harm would have been sustained even if the actor had not been negligent.” (Italics added.) Subsection (2) states that if “two forces are actively operating . . . and each of itself is sufficient to bring about harm to another, the actor’s'negligence may be found to be a substantial factor in bringing it about.”

Thus, in Restatement section 432, subsection (1) adopts the “but for” test of causation, while subsection (2) provides for an exception to that test. The situation that the exception addresses has long been recognized, but it has been given various labels, including “concurrent independent causes” (Mitchell v. Gonzales, supra, 54 Cal.3d at pp. 1049, 1052), “combined force criteria” (Robertson, The Common Sense of Cause in Fact (1997) 75 Tex. L.Rev. 1765, 1778), and “multiple sufficient causes” (Rest.3d Torts, Liability for Physical Harm (Basic Principles) (Tent. Draft No. 2, Mar. 25, 2002) § 27, com. b, p. 70).

This case does not involve concurrent independent causes, which are multiple forces operating at the same time and independently, each of which would have been sufficient by itself to bring about the harm. Here, the Viners argued that their losses were caused by defendants’ negligence, the actions of MEI exploiting that negligence, the underlying economic situation, and “other factors.” Because these forces operated in combination, with none being sufficient in the absence of the others to bring about the harm, they are not concurrent independent causes. Accordingly, the exception stated in subsection (2) of Restatement section 432 does not apply, and this case is governed by the “but for” test stated in subsection (1) of Restatement section 432.

The Court of Appeal here held that a plaintiff suing an attorney for transactional malpractice need not show that the harm would not have occurred in the absence of the attorney’s negligence. We disagree. We see nothing distinctive about transactional malpractice that would justify a relaxation of, or departure from, the well-established requirement in negligence cases that the plaintiff establish causation by showing either (1) but for the negligence, the harm would not have occurred, or (2) the negligence was a concurrent independent cause of the harm.

“When a business transaction goes awry, a natural target of the disappointed principals is the attorneys who arranged or advised the deal. Clients predictably attempt to shift some part of the loss and disappointment of a deal that goes sour onto the shoulders of persons who were responsible for the underlying legal work. Before the loss can be shifted, however, the client has an initial hurdle to clear. It must be shown that the loss suffered was in fact caused by the alleged attorney malpractice. It is far too easy to make the legal advisor a scapegoat for a variety of business misjudgments unless the courts pay close attention to the cause in fact element, and deny recovery where the unfavorable outcome was likely to occur anyway, the client already knew the problems with the deal, or where the client’s own misconduct or misjudgment caused the problems. It is the failure of the client to establish the causal link that explains decisions where the loss is termed remote or speculative. Courts are properly cautious about making attorneys guarantors of their clients’ faulty business judgment.” (Bauman, Damages for Legal Malpractice: An Appraisal of the Crumbling Dike and Threatening Flood (1988) 61 Temp. L.Rev. 1127, 1154-1155, ins. omitted, italics added (hereafter Bauman, Damages for Legal Malpractice).)

In a litigation malpractice action, the plaintiff must establish that but for the alleged negligence of the defendant attorney, the plaintiff would have obtained a more favorable judgment or settlement in the action in which the malpractice allegedly occurred. The purpose of this requirement, which has been in use for more than 120 years, is to safeguard against speculative and conjectural claims. (Mattco Forge, Inc. v. Arthur Young & Co. (1997) 52 Cal.App.4th 820, 832-834 [ 60 Cal.Rptr.2d 780 ].) It serves the essential purpose of ensuring that damages awarded for the attorney’s malpractice actually have been caused by the malpractice. (Id. at p. 834.)

The Court of Appeal here attempted to distinguish litigation malpractice from transactional malpractice in order to justify a relaxation of the “but for” test of causation in transactional malpractice cases. One of the distinguishing features, according to the court, was that in litigation a gain for one side necessarily entails a corresponding loss for the other, whereas in transactional representation a gain for one side does not necessarily result in a loss for the other. We question both the accuracy and the relevance of this generalization. In litigation, as in transactional work, a gain for one side does not necessarily result in a loss for the other side. Litigation may involve multiple claims and issues arising from complaints and cross-complaints, and parties in such litigation may prevail on some issues and not others, so that in the end there is no clear winner or loser and no exact correlation between one side’s gains and the other side’s losses. In addition, an attorney’s representation of a client often combines litigation and transactional work, as when the attorney effects a settlement of pending litigation. The “but for” test of causation applies to a claim of legal malpractice in the settlement of litigation (Marshak v. Ballesteros (1999) 72 Cal.App.4th 1514, 1518-1519 [ 86 Cal.Rptr.2d 1 ]; Thompson v. Halvonik (1995) 36 Cal.App.4th 657, 661-663 [ 43 Cal.Rptr.2d 142 ]), even though the settlement is itself a form of business transaction.

Nor do we agree with the Court of Appeal that litigation is inherently or necessarily less complex than transactional work. Some litigation, such as many lawsuits involving car accidents, is relatively uncomplicated, but so too is much transactional work, such as the negotiation of a simple lease or a purchase and sale agreement. But some litigation, such as a beneficiary’s action against a trustee challenging the trustee’s management of trust property over a period of decades, is as complex as most transactional work.

It is true, as the Court of Appeal pointed out, that litigation generally involves an examination of past events whereas transactional work involves anticipating and guiding the course of future events. But this distinction makes little difference for purposes of selecting an appropriate test of causation. Determining causation always requires evaluation of hypothetical situations concerning what might have happened, but did not. In both litigation and transactional malpractice cases, the crucial causation inquiry is what would have happened if the defendant attorney had not been negligent. This is so because the very idea of causation necessarily involves comparing historical events to a hypothetical alternative. (E.g., 1 Dobbs, The Law of Torts (2000) § 169, p. 411; Robertson, The Common Sense of Cause in Fact, supra, 75 Tex. L.Rev. at p. 1770.)

The Viners also contend that the “but for” test of causation should not apply to transactional malpractice cases because it is too difficult to obtain the evidence needed to satisfy this standard of proof. In particular, they argue that proving causation under the “but for” test would require them to obtain the testimony of the other parties to the transaction, who have since become their adversaries, to the effect that they would have given the Viners more favorable terms had the Viners’ attorneys not performed negligently. Not so. In transactional malpractice cases, as in other cases, the plaintiff may use circumstantial evidence to satisfy his or her burden. An express concession by the other parties to the negotiation that they would have accepted other or additional terms is not necessary. And the plaintiff need not prove causation with absolute certainty. Rather, the plaintiff need only “ ‘introduce evidence which affords a reasonable basis for the conclusion that it is more likely than not that the conduct of the defendant was a cause in fact of the result.’ ” (Ortega v. Kmart Corp. (2001) 26 Cal.4th 1200, 1205 [ 114 Cal.Rptr.2d 470 , 36 P.3d 11 ], quoting Prosser & Keeton on Torts (5th ed. 1984) § 41, p. 269, ins. omitted.) In any event, difficulties of proof cannot justify imposing liability for injuries that the attorney could not have prevented by performing according to the required standard of care. (See Bauman, Damages for Legal Malpractice, supra, 61 Temp. L.Rev. at p. 1154.)

In urging us to exempt transactional malpractice from the “but for” test of causation, the Viners cite California State Auto. Assn. Inter-Ins. Bureau v. Parichan, Renberg, Crossman & Harvey (2000) 84 Cal.App.4th 702 [ 101 Cal.Rptr.2d 72 ] (CSAA). There, an attorney representing both the insured and the insurer in a personal injury action brought by a third party car accident victim negligently failed to forward to the insurer a medical report showing that the tort victim’s injuries were more serious than previously thought. Because of the attorney’s negligence, the insurance company rejected a $50,000 settlement offer but then, after learning the seriousness of the victim’s injuries, ultimately paid $850,000 to settle the litigation. In the insurer’s malpractice action against the attorney, the trial court refused to give a jury instruction requiring the insurer to prove that, but for the attorney’s negligence, the insurer would not have suffered harm. (Id. at p. 709, fn. 1.)

The Court of Appeal affirmed, concluding that the requested instruction was unnecessary because the issue was whether the insurance company’s settlement was reasonable in light of the facts and circumstances of the case. (CSAA, supra, 84 Cal.App.4th at p. 710 .) Referring to Professor Bauman’s law review article, which we discussed, ante, at page 1241, the court observed: “One commentator describes the difference between the proof of causation and damages in the ‘litigation’ and ‘transactional’ malpractice contexts in this way: ‘When legal malpractice takes place in a transactional setting—that is, in the advising and planning of business dealings—the courts take a much less structured approach to proof of damages. No longer wedded to a narrow interpretation of what can constitute adequate proof of the fact and amount of injury, the courts tend to treat such actions like ordinary business cases and allow considerably more flexibility to plaintiffs in proving their damages.’ ” (CSAA, at p. 711, quoting Bauman, Damages for Legal Malpractice, supra, 61 Temp. L.Rev. at p. 1150, italics added.)

Viner v. Sweet 30 Cal.4th 1232 ; 135 Cal.Rptr.2d 629 ; 70 P.3d 1046 [June 2003]

The CSAA court misunderstood the above quoted comment from Professor Bauman’s article as referring to both causation and damages, when it actually referred only to damages, specifically consequential damages. (Bauman, Damages for Legal Malpractice, supra, 61 Temp. L.Rev. at pp. 11SO-1153.) Professor Bauman thereafter observed, “Courts are properly cautious about making attorneys guarantors of their clients’ faulty business judgment”; hence, courts require that it be “shown that the loss suffered was in fact caused by the alleged attorney malpractice.” (Id. at pp. 1154-1155.)

For the reasons given above, we conclude that, just as in litigation malpractice actions, a plaintiff in a transactional malpractice action must show that but for the alleged malpractice, it is more likely than not that the plaintiff would have obtained a more favorable result.

Disposition

The judgment of the Court of Appeal is reversed, and the matter is remanded to the Court of Appeal for proceedings consistent with the views expressed here.

George, C. J., Baxter, J., Werdegar, J., Brown, J., Moreno, J., and Raye, J., concurred.

Causation analysis in tort law generally proceeds in two stages: determining cause in fact and considering various policy factors that may preclude imposition of liability. (Ferguson v. Lieff, Cabraser, Heimann & Bernstein (2003) 30 Cal.4th 1037, 1045 [ 135 Cal.Rptr.2d 46 , 69 P.3d 965 ]; PPG Industries, Inc. v. Transamerica Ins. Co. (1999) 20 Cal.4th 310, 315-316 [ 84 Cal.Rptr.2d 455 , 975 P.2d 652 ].) This case concerns only the element of cause in fact. The trial court refused defendants’ requested instruction on “but for” causation. The court did instruct the jury that a cause of an injury “is something that is a substantial factor in bringing about” the harm. Because the Court of Appeal addressed this case as presenting the “pure question of law” of whether the legal requirement of showing “but for” causation applies at all to transactional malpractice cases, and because we limited our review to that issue, we have not framed our discussion in terms of instructional error. “Concurrent independent causes” should not be confused with “concurrent causes.” The former refers to multiple forces operating at the same time and independently, each of which would have been sufficient by itself to bring about the harm. The latter refers simply to multiple forces operating at the same time. The requirement that the plaintiff prove causation should not be confused with the method or means of doing so. Phrases such as “trial within a trial,” “case within a case,” “no deal” scenario, and “better deal” scenario describe methods of proving causation, not the causation requirement itself or the test for determining whether causation has been established. California State Auto. Assn. Inter-Ins. Bureau v. Parichan, Renberg, Crossman & Harvey, supra, 84 Cal.App.4th 702 , is disapproved to the extent it is inconsistent with our decision in this case. Associate Justice of the Court of Appeal, Third Appellate District, assigned by the Chief Justice pursuant to article VI, section 6 of the California Constitution.