Griffin Dewatering Corp. v. Northern Ins. Co.

Full title: GRIFFIN DEWATERING CORPORATION, Plaintiff and Respondent, v. NORTHERN…

Court: Court of Appeal of California, Fourth District, Division Three

Date published: Jul 31, 2009

176 Cal.App.4th 172 (Cal. Ct. App. 2009)

97 Cal. Rptr. 3d 568

Opinion

G036896.

July 31, 2009.

Appeal from the Superior Court of Orange County, No. 00CC04293, W. Michael Hayes, Judge.

Horvitz Levy, Mitchell C. Tilner, Curt Cutting, Felix Shafir; Williams Law Group, Kelly Lytton Williams, Richard D. Williams; Meckler Bulger Tilson, Meckler Bulger Tilson Marick Pearson, Scott Seaman and Karen M. Dixon for Defendant and Appellant.

Law Offices of Timothy C. Cronin, Timothy C. Cronin, Michael A. Mazzocone; Law Offices of Daniel U. Smith and Daniel U. Smith for Plaintiff and Respondent.

Dickstein Shapiro, Kirk A. Pasich, Cassandra S. Franklin and Stephanie A. Sullins as Amicus Curiae on behalf of Plaintiff and Respondent.

OPINION

SILLS, P.J.

I. INTRODUCTION

At first we did not know what to make of this case. There was a $10 million judgment obtained by a nationwide groundwater pumping and control company, against its liability insurer. The compensatory damages — all of which consisted of the attorney fees and costs incurred to sue the insurer for the $10 million — were about $1 million.¹ What foul deeds had the insurer committed, we wondered, that merited such punishment?

1 $681,840 in fees, plus $379,348.21 in costs, which equal $1,061,188.21, to be exact.

At first the answer seemed simple. At a meeting in Houston in May 1997, representatives of the insurer had, in order to induce the insured to renew its CGL² (that is, "third party" liability) policy, orally promised representatives of the insured that the insurer would cover any "future" liability claims based on the release of sewage, even though the insurer, to that point, had steadfastly maintained that any liability based on sewage releases was excluded under the insurer's "total pollution exclusion."³ The Houston meeting had arisen out of the insurer's disavowal of any coverage, or potential for coverage, for liability arising out of a certain sewage overflow in Laguna Beach. In particular, the insurer had disclaimed any coverage for the damages to the Laguna Beach home of Dr. and Mrs. Waters about a year and one-half before arising out of an allegedly faulty sewer bypass constructed by the insured for a water district.⁴ (In this opinion we will refer to that sewage backup as the "Waters claim" and the (alleged) promise made by the insurer in Houston as the "Houston Oral Promise."⁵)

2 CGL now stands for "commercial general liability" though it once stood for "comprehensive general liability." (See Buss v. Superior Court (1997) 16 Cal.4th 35, 39 [ 65 Cal.Rptr.2d 366, 939 P.2d 766].) The various editions of the CGL represent probably the most litigated contract in history.

3 To the degree that there may be any difference between the "total pollution exclusion" and the more common "absolute pollution exclusion," it appears that the "total" one excludes a little more than the "absolute" one, that is, less favorable to the insured. (See In re Idleaire Technologies Corp. (Bankr.D.Del. 2009) 2009 WL 413117, p. *6 ["The last five to ten years have seen yet another shift in the language of pollution exclusions. The newest generation, the Total Pollution Exclusion, was first introduced in 1988, and subsequently became the standard form of language. This variation is even broader than the Absolute Pollution Exclusion. Specifically, it further limits coverage by excluding releases from products and for certain off-site releases of pollutants."].)

4 The South Coast Water District, to be exact. From now on we will mostly use "the district" instead of "Water District" or "South Coast," so as to convey to readers the thought that (a) this was a public agency while (b) avoiding confusion with Dr. and Mrs. Waters, the damage to whose home started this whole case.

5 At the time of the Houston meeting the Waters claim was only a "claim" that had not, as of that moment, reified itself in the form of any lawsuit against the insured. In fact, Dr. and Mrs. Waters never did sue anybody for their damages. They were recompensed by the district for which the insured did the work, the district sued the insured, and eventually the insurer settled the district's claim against the insured. But that's getting ahead of ourselves for the moment.

When the district on whose behalf the insured had built the bypass settled the Waters claim, the district sued the insured to get its money back. But the insurer refused to defend this latter suit against the insured for about 11 months even though it had promised the insured that it would cover any "future" claims.

Hence, it initially looked to us like this case might indeed warrant punitive damages. After all, coming to us after a jury trial, the ambiguity in the word "future" (could it encompass the district's post-May 1997 suit against the insured? — or, because it originally surfaced in the form of an informal claim by the district sometime earlier, was it a "past" claim?) would have to be construed in favor of the prevailing party — the insured.

But then we started digging into the voluminous record with the help of able counsel on both sides who provided two rounds of supplemental briefing plus a second oral argument. And it was only after the second oral argument in April of this year that the case finally unfolded itself. The whole theory of liability based upon the Houston Oral Promise turned out to be an illusion that dissolved under scrutiny. Two items in particular made liability based on the Houston Oral Promise untenable:

First, the complaint never actually mentioned the oral promise made in Houston at all. Rather, the complaint was predicated on a straightforward coverage question based not on some oral promise made in Houston in May 1997, but on the written insurance policy as it stood in 1996. Essentially, the complaint said: We, the insured, had a sewage claim against us, and the insurance company denied our request for a defense of that claim because it interpreted the written insurance policy, with its total pollution exclusion, unreasonably.

Second, the complaint was never amended to include any cause of action based on the oral promise made in Houston in May 1997. In fact, before the trial, the insured's counsel expressly dropped an attempt to amend the complaint to state a claim based on that promise as a "stand alone" cause of action.

Why? The answer came out in the second oral argument. Rather than expose the merits of the issue to the jury, the insured's counsel wanted to rhetorically exploit the promise as a simple "concession" (his word at oral argument) by the insurer that its coverage position had been unreasonable all along.

But absent an amendment to the complaint, the Houston Oral Promise could not serve as a basis for recovery. It is elementary that a party cannot recover on a cause of action not in the complaint. (E.g., Mondran v. Goux (1875) 51 Cal. 151, 153 ["In other words, the cause of action, if any, established by the findings, is wholly different from that averred in the complaint, and is foreign to any issue raised by the pleadings. The rule is well settled that a plaintiff must recover, if at all, upon the cause of action set out in his complaint, and not upon some other which may be developed by the proofs."]; Walker v. Belvedere (1993) 16 Cal.App.4th 1663, 1670 [ 20 Cal.Rptr.2d 773] ["`It is a fundamental principle of pleading that "a plaintiff must recover, if at all, upon the cause of action set out in the complaint, and not upon some other which may be developed by the proofs."'"].)

That left the breach of the written contract (the insurance policy) which was, after all, the actual basis for the jury's punitive damage assessment. But that assessment turned out to be the result of an error of law in a motion in limine in favor of the insured. Specifically, the trial court had ruled, in an in limine motion, that, as a matter of law based on the written contract and totally independent of the Houston Oral Promise, the insurer had breached the contract unreasonably so as to expose the insurer to tort, and maybe even punitive damages.

However, that theory would not hold up either. The ruling on the in limine motion (as we explain in probably too much detail below) was clear error, as shown by this court's opinion in Morris v. Paul Revere Life Ins. Co. (2003) 109 Cal.App.4th 966 [ 135 Cal.Rptr.2d 718], a case which turns out to be directly on point. The trial court erroneously thought that because the case law was "unsettled" when the insurer first turned down the claim, that unsettledness created a potential for a covered claim. Morris, however, explained that if an insurance company's denial of coverage is reasonable, as shown by substantial case law in favor of its position, there can be no bad faith even though the insurance company's position is later rejected by our state Supreme Court.

Exactly that had happened in the case before us. Back in the late 1990's, at the time this insurer denied a request for a defense, there was ample case law and policy language to support the insurer's position. On top of that, this insurer changed its mind in favor of the insured more than six months before the California Supreme Court settled the question of correctness of the insurer's original position in MacKinnon v. Truck Ins. Exchange (2003) 31 Cal.4th 635 [ 3 Cal.Rptr.3d 228, 73 P.3d 1205].

And finally, there was the matter of damages for the insurer's initial and incorrect denial of coverage. It turned out, there weren't any. As to contract damages, the insurer had, long prior to the MacKinnon case, (a) settled all the litigation against the insured and (b) paid all the insured's attorney fees incurred in that litigation. As to tort damages, the insured's claims for attorney fees foundered on the reasonableness of the insurer's initial and incorrect denial. Because the denial was reasonable, no tort damages were available, including attorney fees — often called " Brandt fees." (See Brandt v. Superior Court (1985) 37 Cal.3d 813 [ 210 Cal.Rptr. 211, 693 P.2d 796]; Cassim v. Allstate Ins. Co. (2004) 33 Cal.4th 780, 808 [ 16 Cal.Rptr.3d 374, 94 P.3d 513] ["without a tort judgment, there could be no Brandt fees"].)

We therefore reverse the approximately $11 million judgment, with directions to enter a judgment in favor of the insurer.⁶

6 To aid readers we provide an outline of our opinion here:
I. INTRODUCTION
II. BACKGROUND
A. Events Prior to Litigation
B. Events After the Litigation Began
III. DISCUSSION
A. The Basic Three-part Paradigm for Analyzing Bad Faith Cases
B. The Reasonableness (or Lack Thereof) of the Insurer's Denial of a Defense to the Waters Claim
1. A Primer on the Potentiality Rule
2. The Question of an Objective Standard
3. The MacKinnon Case
4. " Congeries of Cerebrates": Why MacKinnon and Morris Show the Insurer's Decision Here Was Reasonable
5. The Insured's Attempt to Distinguish Morris Is Not Persuasive
a. Reasonability Is the Test in Both First and Third Party Cases
b. A Third Party Insurer Has the Right to Make a Reasonable Coverage Decision Even If It "Benefits Its Own Interests"
C. Damage Issues
1. The Failure to Amend the Complaint to Include a Cause of Action Based on the Houston Oral Promise Precludes Any Recovery Based on That Promise 2. The Reservation of Rights Issue
3. Miscellaneous Brandt Fees Issue
IV. CONCLUSION AND DISPOSITION

II. BACKGROUND

Because this is a bad faith case, perspective is best attained if the story is told in a timeline fashion. In particular, noting the time that specific events occurred is important here because it shows that:

(a) the insurer denied the requested defense long before the MacKinnon case came down;

(b) the insurer actually changed its mind and provided a defense relatively quickly after it had first denied it, and did so before the Supreme Court handed down the MacKinnon decision;

(c) the insurer changed its mind with sufficient speed that the insured never really paid anything in defense costs. Those costs were picked up by the insurer before the insured was ever billed for them.

Also, because this is a bad faith case, we quote the precise and full language of a number of important documents, so readers can easily see the "source materials" which reveal how the parties were dealing with each other.⁷

7 Which, alas, makes for a longer opinion. Brevity may be the soul of wit, but it is sometimes not conducive to legal proof. Law is, after all only language. (Exceptions to that last sentence can be debated in jurisprudence class.) If readers want to know whether a paraphrase or characterization is accurate, the best way is for writers to quote the source material, so readers can see for themselves the exact words behind the paraphrase or characterization. Too many legal writing instructors discourage quotation of source materials. That only makes more work for those readers who may not necessarily want to take the writer's word for it.

A. Events Prior to Litigation

1. November 1995 to February 1996: The insured's work and the origin of the Waters claim.

In early November 1995, Griffin Dewatering Corporation (usually referred to in this opinion as "the insured," sometimes as "Griffin"⁸) agreed to fix a 75-foot manhole feeding into the main sewer line for the South Coast Water District.⁹

8 Throughout this opinion we refer to plaintiff Griffin Dewatering as "the insured." When referring to an "insured" or "insureds" in a generic sense, or in describing the facts or thinking of another case, we will use the words "policyholder" or "policyholders" (except, of course, in a direct quotation).

9 The insured's primary work involves movement of water. Its president would later testify that "sewer bypass jobs are maybe five to ten percent" of the firm's "total business."

The insured worked on the job sometime between November 1995 and February 1996. On February 6 and February 21, sewage backed up into the Laguna Beach home of Dr. and Mrs. Ron Waters, obviously resulting in extensive damage.

In late February 1996, the Waters submitted a claim form to the district. The claim form submitted by the Waters said that the "sewer line under construction by South Coast Water District backed up causing massive flow of raw sewerage [ sic] into ground floor of our home on two occasions."

The Waters, however, never sued, and never would sue, on their claim.

2. March to April 1996: The insurer is requested to cover a claim not yet reduced to a lawsuit, and denies the request.

In March 1996 (about March 8), the insurer¹⁰ received notice of the Waters claim. It is not clear from the record precisely from whom the notice first came — from the district or from the insured — though it is clear that at least the insured had brought the matter to the insurer's attention in March 1996.

10 Throughout this opinion we will use "insurer" to refer to defendant Northern Insurance Company of New York, which, in various incarnations in quoted source material may also be referred to as "Maryland" or "Zurich." When referring to an "insurer" or "insurers" in a generic sense, or in describing the facts or thinking of another case, we will use the words "insurance company" or "insurance companies" (except, of course, in a direct quotation).

The next month, in April 1996, the insurer denied the claim in a letter dated April 11.¹¹ The insured's broker immediately disputed the conclusion in a letter sent on April 15, 1996.

11 The letter said: "Unfortunately, this Total Pollution Exclusion Endorsement, and specifically section f (1), bars coverage for this loss. Pursuant to this section of the exclusion, the claimant's property damage `. . . . would not have occurred in whole or part but for. . . .' the discharge of raw sewage. Raw sewage is waste and as such, constitutes a pollutant within the meaning of the above-quoted exclusion. Thus, there is no coverage for any of the alleged `property damage' to Mr. Waters' residence resulting in any way from the discharge of this pollutant."

3. May 1996 to May 1997: The Waters claim is still a possible suit against the insured, and a bone of contention between the insurer and the insured.

There is not much in the record about events in the May 1996 through May 1997 time period, except that the policy had come up for renewal in early February 1997 and the earlier denial of coverage for the Waters claim was clearly a sore spot for the insured.

4. May 1997: The Houston meeting.

In May 1997, there was a meeting between representatives of the insurer (including members of its environmental claims unit) and the insured (including its president, its in-house attorney, and its insurance broker) in Houston. This is the origin of the Houston Oral Promise mentioned above. The insured was unhappy with the denial of the Waters claim, and the insurer was apparently eager to retain the insured's business.

There is no dispute that there was some sort of oral promise made at the May 1997 meeting, but there is a dispute as to the precise nature of the promise. According to one of the insurer's underwriters who was at the meeting (as he would later testify at trial¹²), the promise was that the insurer would consider sewage claims under a reservation of rights "and in many instances" make payments as a "business consideration." This same underwriter would also later testify that by the time of the Houston meeting, the decision on the Waters claim could not be altered but "similar-type claims" would be covered in the future.

12 Again, we emphasize that the trial was not about a breach of the Houston Oral Promise; rather, the Houston Oral Promise was used by the insured as evidence of the insurer's bad faith dealing (a supposed "concession") in the context of the insurer's breach of the written insurance contract.

In contrast to the resolute quality of the insurer's decision that there was no coverage at all for the Waters claim as described by one of the insurer's underwriters, the insured's president would later testify that he asked the same underwriter "if anything comes from South Coast are we covered, and the answer was yes." The insured's in-house attorney also testified that he "understood" the underwriter to be promising that "if the Waters claim suddenly had come back to life" that the insurer would "pay" it. And the insured's broker testified that he understood Griffin would be receiving something in writing to confirm Northern's new coverage position. (There is no dispute that nothing was ever received afterwards except a change in the pollution exclusion in the renewed policy to remove the phrase "in whole or in part" from the pollution exclusion.)

5. May 1997 to May 1999: The district settles the Waters claim on its own and begins to think about looking to the insured for reimbursement.

Around May 1999, the district paid the Waterses nearly $417,000 to settle their claim against the district for the sewage backup. Now the district wanted to recoup that money from the insured.¹³

6. Late September 1999 to October 1999: The district sues the insured, the insured wants coverage from the insurer, the insurer denies the request.

13 The exact date of the district settlement with the Waterses is not given in the record (unless we missed something in the more than 8,000 pages of which it consists) or specified by the parties, but it appears that the settlement was on or just before May 6, 1999, which is the date of a letter from the district's attorney to the insured, the gravamen of which was: We want you, or your insurance company, to indemnify us for the $417,000. We quote from the letter of that date: "The Association of California Water Agencies/Joint Powers Insurance Authority has paid approximately $416,889.35 and now seeks to recover the costs incurred due to Griffin's failure to furnish adequate equipment and material pursuant to the agreement. [¶] Please forward this letter to your insurance carrier requesting it contact me to further discuss this matter. I would be pleased to answer any questions you may have."

In September 1999, the district sued both the insured and the insurer¹⁴ to recoup its expenditures to settle the Waters claim. The insured was served with the suit "within a day or two" of September 29, 1999, and, in a letter dated October 1, 1999, the insured's broker enclosed the lawsuit brought by the district against the insured (as well as the insurer itself), requesting coverage (presumably a defense or settlement, since no judgment had been entered against the insured at that point).¹⁵

14 The district obtained a certificate of insurance under Griffin's CGL policy with Northern, which explains the addition of the insurer as a defendant in the district's lawsuit. The district's status as an additional insured, however, is only of passing importance to the case at hand, because there never was any dispute over whether the district might not be covered when Griffin otherwise was.

15 Here is the first paragraph of the letter to the insurer: "Dear Mr. Lysaught: "Attached please find litigation documents regarding the above captioned. On April 11, 1996, at the time Maryland Insurance Company, denied coverage to the insured for this loss (Declination letter and ROR letter from Marcus Vergi attached). On May 14, 1996, you upheld the Maryland's decision denying coverage to the insured for this loss (your letter of 5/14/96 also attached). Please note, Plaintiff, South Coast Water District, has now filed a lawsuit against Defendants, Griffin Dewatering Corp. and Northern Insurance Company (The Insured's General Liability Carrier). Per the attached litigation, we now believe there is coverage for Griffin Dewatering Corporation for this loss. Please review all litigation documents carefully, taking the necessary action to protect the insured's interests."

A representative of the insurer telephoned with a "verbal response," to the effect that there would be no coverage for the suit. The insurer did not think a written response was necessary because there had been three prior denials.

As it turned out, though, the insured was never "out of pocket" for any defense costs during the pendency of the suit. The insured had an excess insurance policy with another insurance company (AIG — yes, that AIG), and AIG agreed to hire a defense firm to defend the insured against the district and "invoice" the bills to the insured. Essentially the excess insurer would be fronting money for the defense. The insured's in-house counsel would later admit on cross-examination that the district's case against it was settled (as we shall soon see, it was settled by the insurer) prior to the insured receiving any invoices from the excess insurer for the hiring of the defense firm.¹⁶

16 Here is the relevant text of the cross-examination:
"Q. OKAY. LET ME TRY AGAIN. MAYBE MY QUESTION WASN'T CLEAR. DID GRIFFIN EVER MAKE ANY PAYMENTS TO THE WALSWORTH FIRM FOR THEIR WORK IN DEFENDING THE SOUTH COAST LAWSUIT
"A. ACTUALLY, GRIFFIN ENDORSED A CHECK WRITTEN BY NORTHERN TO PAY THE WALSWORTH FEES. SO YES, I WOULD SAY THEY DID.
"Q. OKAY. IN TERMS OF — MAYBE MY QUESTION STILL WASN'T CLEAR. IN TERMS OF GRIFFIN ACTUALLY WRITING A CHECK, A GRIFFIN CHECK TO THE WALSWORTH FIRM, DID THAT EVER HAPPEN
"A. NO.
"Q. OKAY. GRIFFIN NEVER WENT OUT OF POCKET ANY MONEY TO THE WALSWORTH FIRM FOR ITS SERVICES IN DEFENDING GRIFFIN IN THE SOUTH COAST LAWSUIT
"A. FOR THE WALSWORTH FIRM FEES, NO, SIR." (Italics added.)

B. Events After the Litigation Began

1. October 1999 to April 2000: The insured prepares and files its bad faith complaint against the insurer.

About six months after the October verbal denial, in April 2000, the insured filed a complaint against the insurer for bad faith breach of insurance contract. (The other causes of action listed were breach of contract, fraud, intentional misrepresentation, and declaratory relief.) As we have mentioned, the complaint made no reference to the Houston meeting in May 1997. The insurer filed an answer in June 2000.

2. September 2000: The insurer changes its mind, settles the district suit against the insured, and unsuccessfully attempts to give up its right to seek reimbursement in exchange for the insured's dropping of the bad faith suit.

Less than three months after it had filed its answer, in a letter of September 8, 2000, the insurer agreed to defend the district's suit against the insured.¹⁷

17 Here is the relevant text:
"Dear Mr. McAnelly:
"We acknowledge receipt of the complaint filed in South Coast Water District v. Griffin Dewatering Corp., et al., Case No. 814290, pending in the Orange County Superior Court, State of California ("South Coast"), which you forwarded to us. As I advised during our telephone conversation today, please be advised that, based upon the information provided to us to date. Northern Insurance Company of New York ("Northern") will participate in Griffin Dewatering Corporation's ("Griffin") defense of the South Coast lawsuit, from the date you tendered the lawsuit to Northern, subject to a complete and full reservation of all of Northern's rights, remedies, and defenses as set forth more fully below." (Underscoring omitted.)

In a letter of September 28, 2000, to the insured's counsel, the insurer's counsel stated that it had settled all the liability of the insured asserted by the district. There was a confirmation of the settlement in a letter to the insurer from the insured's defense lawyers the next day, acknowledging that the insurer had provided $350,000 to settle the case.

Now we come to what appears to be a case of mistaken purposes. The letter of September 28, 2000, from the insurer's lawyer to the insured's lawyer also stated that the insurer had "agreed" to give up any right to reimbursement and further stated that the insurer was willing to pay the insured's fees incurred to sue the insurer in the bad faith case up to that point, which then were the relatively small sum of $8,921.40. On the other hand, the letter made an ambiguous reference to "mediation," which suggests that the insurer still entertained a wistful hope of getting some of its money back, at least in an alternative dispute resolution forum. We quote the totality of the letter in the margin.¹⁸

18 Here is the text:
"Now that Northern has resolved Griffin's exposure in the underlying case, it would like to turn its full attention to resolving the coverage case. In addition to having paid all of the underlying defense fees and costs, Northern agrees to waive any rights to seek reimbursement of those fees and costs. Northern is also offering to reimburse Griffin for its attorneys' fees and costs incurred in the coverage case.
"Northern has conveyed to you a $10,000 offer to cover Griffin's fees and costs in that regard.
"Northern previously invited you to the mediation scheduled for October 2, 2000. You declined this offer. It was — and remains — Northern's hope to explore settlement possibilities of the coverage action. Even though the underlying case has settled prior to the mediation, Northern remains willing to work with Griffin to resolve the coverage dispute as amicably and efficiently as possible. To this end, Northern again invites Griffin to mediate the coverage dispute before Justice Wiener (or any other mutually agreed upon mediator) at a convenient time.
"Despite Griffin's decision to withdraw from that voluntary mediation, we hope that Griffin will agree to reconsider engaging in negotiations to settle the coverage action."

The insured's counsel didn't like the offer one bit, and rejected the offer in a letter dated October 2, 2000. Basically, the insured's counsel did not perceive the insurer to be really giving up the chance to seek reimbursement, though it may be significant that the phrase "rights to seek reimbursement later" or something like it was not the phrase used in the rejection letter. Rather, counsel for the insured (also lead counsel in this appeal) used the phrase "withdrawal of all coverage defenses." We note this because there might be a substantive difference in the two thoughts: The phrase "withdrawal of all coverage defenses" suggests the possibility of the bad faith case continuing on, but in the posture of: "We the insurer now admit we were wrong in denying coverage — we now throw ourselves on the jury's mercy, punish us as you see fit." The thought of "dropping any claim to reimbursement" suggests: "We were right all along, but even so we won't seek our money back."

On the other hand, perhaps the insured wasn't demanding unconditional surrender at that point. We also note that the insured's in-house counsel, as he would testify later, perceived the September 28, 2000 letter as one simply leaving open the possibility of reimbursement in a way that money might still come out of the insured's pockets. The in-house counsel testified he would have been "interested" in settlement if the insurer had given up its right of recoupment.

The October 2, 2000 rejection letter made one arresting point, namely that the insured's counsel didn't like the fact that the insurer had paid $350,000 to settle the case on its behalf. He thought $350,000 paid on behalf of his client might have an adverse effect on future premiums. (No action would come of this thought, though, and the idea of increased future premiums as a result of the insurer's settling the case played no role in this suit.)

The insured's counsel also made it quite clear (as the italicized passages in the margin demonstrate) that he viewed the case as warranting punitive damages regardless of whether his client suffered any actual damages, including any attorney fees incurred to sue the insurer up to that point. As is the case with the September 28 letter from the insurer's counsel to the insured's counsel, we quote the entirety of the October 2 rejection letter from the insured's counsel to the insurer's in the margin.¹⁹

3. October 2000 to April 2002: Case law temporarily turns in favor of the insurer, and the insurer goes for a quick win in the bad faith case with a summary judgment motion.

19 Here is the text:
"Thank you for your letter dated September 28, 2000. In connection therewith, please note the following.
"First, we have forwarded a copy of your letter offering to settle the above-entitled bad faith matter for $10,000 to our client Griffin Dewatering Corporation. We note that this written settlement offer differs from that conveyed to us via telephone. Your oral settlement offer contained a withdrawal of all coverage defenses by your client Northern Insurance Company of New York (`Northern') in addition to a $10,000 payment. Robert Gokoo, Esq. has informed us that you have reached settlement in the underlying case and, it is for that reason, we assume that the written settlement offer does not contain a provision for, as we discussed, Northern's withdrawal of all coverage defenses. We understand that settlement with Mr. Gokoo's client, South Cost [ sic] Water District, was reached after considerable discussion of Northern's bad faith liability given South Coast's position as an additional insured on CGL Policy No. EC 86161692. In any event, Griffin hereby rejects Northern's settlement offer as contained in your September 28, 2000 letter.
"Second, this correspondence serves to put Northern on notice of Griffin's objection to Northern settling the underlying lawsuit without Griffin's approval and consent. While it is true that as a general proposition an insurer need not secure the consent of an insured to settle a case, there is an exception to this rule applicable in the case at bar. Where a settlement would be prejudicial to the insured's interest, the consent of the insured to the settlement terms is required. This rule is founded on the two-fold rationale that (1) an insurer may not use its discretionary power to settle within policy limits in a manner that impairs its insured's financial interests without the insured's consent; and (2) the covenant of good faith and dealing requires the insurer to give at least as much consideration to the welfare of its insured as it gives to its own interest. Indeed, where an insurer settles in a manner which is financially prejudicial to the insured without the insured's consent, that insurer is subject to bad faith liability.
"As you are aware, the underlying lawsuit involves two occurrences, or spills, implicating Griffin's potential liability. It is, and always has been, Griffin's position that Griffin's liability in connection with these two spills, collectively, is limited to a certain dollar amount. Evidence developed in the underlying lawsuit supports Griffin's position in this regard. Northern's settlement of the underlying lawsuit for an amount in excess of Griffin's actual liability will undoubtedly cause a negative financial impact upon Griffin vis a vis future insurance premiums. Such financial prejudice to Griffin is exactly what is at the heart of the exception to the general rule requiring that an insurer not settle without the consent of the insured. Accordingly, Griffin has no choice but to provide Northern with notice of its objection to Northern's settling of the underlying claim without Griffin's consent.
"Third, as previously advised in our correspondence dated September 19, 2000, Griffin is not in a position to accurately evaluate the settlement value of its bad faith case given Northern's failure to provide meaningful discovery responses. Accordingly, Griffin can neither provide Northern with a damages estimate nor agree to mediate the parties' dispute at this time. Should Northern agree to withdraw all of its objections to Griffin's interrogatories and inspection demand and agree to provide Griffin with all of the discovery it needs to meaningfully participate in mediation it will certainly do so.
"Parenthetically, we note that settlement of the underlying lawsuit will not affect the value of the instant bad faith case or the damages which Griffin is seeking pursuant thereto. It comes as a bit of a surprise, indeed, that Northern has not heretofore attempted to settle the underlying lawsuit so that Northern could proffer the argument that Griffin has not suffered significant damages. Without any damages, as the argument will undoubtedly be framed, Griffin should accept what amounts to a nominal settlement amount. While we understand the logic behind this strategy, it does not comport with the realities of this case or Northern's claims handling in similar cases. We are confident that a jury would not look favorably upon Northern's use of its discretionary power to settle the underlying claim as a basis for the contrived argument that absent damages it has no bad faith liability.
" From the start, Griffin has viewed the instant bad faith litigation as a punitive damages case notwithstanding that it has suffered actual damages. We need not, and will not, rehash here the theory of Griffin's case. Suffice it to say, we are confident that the evidence in this case will show that Northern has engaged in a company-wide practice of asserting the Total Pollution Exclusion to deny coverage when it was unreasonable to do so. Indeed, Griffin, as well as other entities, have previously fallen victim to such despicable claims handling practices under facts remarkably similar, if not identical, to those in the case at bar. We find it difficult to believe that a jury would simply find the timing of Northern's decision to provide a defense under a reservation of rights in this matter to be reasonable given its past application of the Total Pollution Exclusion to deny coverage and the existence of nothing other than the filing of Griffin's bad faith action between Northern's initial denial and its recent decision to provide a defense. Probatively, Griffin was first notified of Northern's reversal of its coverage decision in Northern's responses to Griffin's discovery requests. To date, no explanation, let alone a reasonable one, has been provided to Griffin regarding Northern's coverage determination.
"Griffin remains committed to amicably resolving the parties' dispute as expeditiously and equitably as practicable. To this end, Griffin reaffirms its commitment to entertain any and all settlement offers commensurate with Northern's liability as discussed herein." (Italics added.)

The case had not settled. Two events in the period October 2000 through summer 2002 are worth recounting:

If the insurer had been willing to drop its reservation of the right to seek recoupment in its September 2000 settlement offer, it apparently had not been willing to do so unilaterally. The insurer had not filed a cross-complaint for reimbursement when it filed its answer, but in a letter dated March 2, 2001, it floated the idea of obtaining leave to file such a cross-complaint, and asked opposing counsel if he would be amenable to stipulating to such a filing.²⁰

20 Here is the text of the letter:
"Northern intends to file a motion for leave to file a cross-complaint against Griffin seeking recoupment of the three hundred fifty thousand dollars ($350,000.00) it paid South Coast Water District to settle South Coast Water District v. Griffin Dewatering Corporation, et al. (Case No. 814290).
"Please advise if Griffin will stipulate to the filing of said cross-complaint. Given the fact that Griffin has thus far conducted very little discovery, and that the trial was continued, at your request, for that reason, Griffin will suffer no prejudice by virtue of the filing of the cross complaint. Thank you for your cooperation in this matter." (Underscoring omitted.)

Two decisions of the Court of Appeal came down in early 2002, both of which supported the insurer's position as to the scope of the total pollution exclusion; one case involved the Bechtel Petroleum Corporation, the other would become MacKinnon v. Truck Ins. Exchange, supra, 31 Cal.4th 635 ( MacKinnon) when it was taken by the Supreme Court.

4. May 2002 to October 2002: The insurer loses the first battle and unilaterally withdraws any right to seek reimbursement.

The insurer's two allies in the appellate case law proved false when, in May 2002, the California Supreme Court granted review in both of them. The Bechtel Petroleum case and MacKinnon were dead — at least as far as they might have helped the insurer's legal position. It is not surprising, then, that the insurer's motion for judgment or adjudication in its favor failed in a ruling of July 2002. The minute order gave no explanation.

On October 8, 2002, a few months after the insurer's failure to gain a quick victory, the insurer unilaterally withdrew any right to seek reimbursement of any of the funds expended to defend or settle the Waters claim.²¹

5. Summer 2002 to September 2004: A series of inconclusive battles and a big change in the case law in favor of the insured.

21 Here is the full text of the letter. Readers will note that the insurer asked for no quid pro quo.
"At the June 25, 2002 hearing on Northern's motion for summary adjudication, you argued that Northern was threatening to sue Griffin to get back the money Northern expended on Griffin's behalf to resolve the South Coast matter. You raised this same point again in Griffin's reply brief filed in connection with the motion heard today before Judge Brenner. Please be advised that while Northern contends and believes that the Total Pollution Exclusion did indeed bar coverage for the South Coast case, Northern nonetheless does not intend to seek reimbursement for any of those sums it paid for defense of Griffin or for settlement in that case.
"Your argument is apparently premised on the fact that Northern has previously reserved rights in connection with the South Coast matter. Northern hereby withdraws any reservation of its right to seek reimbursement or recoupment from Griffin." (Underscoring omitted.)

For the next two years, the parties battled in a series of discovery fights plus motions for summary judgment and summary adjudication, none of which are worth retelling now, except to make two points: While the insurer would, in these motions, press something called the "genuine dispute doctrine" as one of the reasons the insurer was not vulnerable to tort damages, there is also no question that the insurer also argued that its denials of coverage in the period 1996 through 2000 were objectively reasonable.

In this opinion, we do not address at all any arguable applicability of the "genuine dispute doctrine" to this case. We have nothing to say on that subject.²² We need only point out here that the insurer did not waive the right to prevail in this appeal on an objective reasonability theory, as distinct from the genuine dispute doctrine. There is no question that the insurer did raise the issue of the objective reasonability of its denial in the trial court, so the issue of objective reasonability was not waived.

22 The applicability of the "genuine dispute doctrine" was the only issue addressed by amicus curiae for the insured, arguing that the doctrine shouldn't apply in this case. As noted, we express no opinion on the point, one way or the other.

The record discloses that the Houston Oral Promise was not the basis of any motion by the insured. Nor was it the basis of any defense by the insured to any motion of the insurer.

The big event in the period was in August 2003, when our Supreme Court handed down MacKinnon, supra, 31 Cal.4th 635. We explore MacKinnon in detail in part III.B.3. of this opinion. For the moment, suffice to say that no longer could the insurer even hope for a decision that its initial coverage denial had been correct. Now all it could hope for was that that denial would be adjudged reasonable.

6. October 2004 to October 2005: In limine battles culminate in a big breakthrough for the insured when the trial court ruled in an in limine motion that the insurer's denials of coverage under the policy were unreasonable as a matter of law.

Trial was now looming, and beginning in October 2004 the parties began filing a series of in limine motions. The one we are concerned with was "plaintiff's number 5," heard on the afternoon of October 5, 2005. The trial court had just concluded that the insurer had breached the contract (which, given the Supreme Court's ruling in MacKinnon, was no great revelation). It then turned to the issue of the "reasonableness" of the breach.

The trial court specifically referenced CalFarm Ins. Co. v. Krusiewicz (2005) 131 Cal.App.4th 273 [ 31 Cal.Rptr.3d 619] ( CalFarm), which rejected a bad faith claim because the insurance company had been objectively reasonable in its denial. The CalFarm opinion arose out of a contractor's failure to properly seal certain retaining walls, which allowed water to permeate the walls and damage their exterior paint. The insurance company, said this court, "could make an objectively reasonable determination" that the costs of "removing and replacing" certain backfilled dirt and landscaping in the course of repairing the walls fell within a policy's "`own work'" exclusion; thus there could be no bad faith liability. ( Id. at pp. 291-292.)

The trial court and counsel thoroughly discussed the CalFarm opinion. Unfortunately (as shown anon) the trial court misapplied it. The trial court held that the insurer's denials had been unreasonable as a matter of law. Why? The trial judge said that the issue of the scope of the total pollution exclusion was "unsettled" when the insurer had disclaimed coverage.²³ The trial judge indicated that he thought CalFarm inapplicable because in the case before him, there had been no defense pursuant to a reservation of rights.²⁴

7. October 5, 2005: The insured unequivocally drops the attempt to predicate a separate claim based on the Houston Oral Promise.

23 Here is the relevant portion of the transcript:
"THE COURT: I WANT TO MAKE CRYSTAL CLEAR HERE WE HAVE — QUESTION. I REALLY SEE IT AS TWO PARTS. THE FIRST PART IS DID THE POLICY COVER THE LOSS? OKAY? WAS THIS POLLUTION WITHIN THE MEANING OF THE EXCLUSION. IF YOU WILL?
"AND ONE CAN VERY EASILY FIND IN CALIFORNIA — I COULD ADOPT YOUR [referring to the insurer's counsel] ARGUMENT THAT THERE WAS NO SUPREME COURT DECISION, THAT IT WAS — THERE WAS NO DECISION ONE WAY OR ANOTHER. I COULD ADOPT THAT. THE PROBLEM I HAVE IS THERE'S A SECOND ISSUE, THAT SINCE IT WAS UNSETTLED, AND FOLLOWING MY REASONING ALL ALONG, WAS IT UNREASONABLE NOT TO DEFEND WHILE THEY WENT AND GOT THE ANSWER." (Italics added.)

24 Here is the relevant portion of the transcript:
"THE COURT: . . . IN THAT CASE [referring to CalFarm] THEY [referring to the insurance company] PROVIDED A DEFENSE SO THEY [referring to the appellate court] SAID WAS THE CONDUCT OF PROVIDING A DEFENSE IN DISPUTING THE COVERAGE REASONABLE. ISN'T THAT WHAT THEY [not clear whether this refers to the insurance company or the appellate court] DID IN CALFARM?"

A little while thereafter during the same hearing, the trial court moved on to the problem of the Houston Oral Promise, which had been lurking in the background in the case, but had never been squarely presented as a cause of action.²⁵ At one point in the proceedings — and one must remember this point was after the trial court had ruled that the insurer had been unreasonable in denying coverage under the written contract — it appeared that the insured's counsel wanted to offer an amendment to the complaint based on the Houston Oral Promise (referred to by the parties at that point as the "Gentleman's Agreement") and the insurer's counsel was prepared not to oppose it.

25 There had been causes of action for fraud in the inducement of the written contract, but those were dropped.

The trial court readily grasped the essence of the problem for the insurer posed by the Houston Oral Promise. As the judge said at one point, it was like the insurer promising coverage and then pointing to an integration clause in the contract and saying "`neener, neener.'"²⁶

26 Here is the relevant portion of the transcript: "THE COURT: WHAT ABOUT IF PEOPLE WITH AUTHORITY FROM YOUR COMPANY COME IN AND SAY, OKAY, WE'RE GOING TO HAVE A GENTLEMEN'S AGREEMENT, AND LET'S GO FOR MY EXAMPLE THE WAY THAT THE PLAINTIFF WOULD LIKE TO HAVE IT, AND THAT IS, ALL FUTURE CLAIMS, AND WE DON'T GET INTO POLICY YEARS, AND THE WATER DISTRICT FILES THEIR LAWSUIT FOR INDEMNITY. CAN THEY THEN TURN AROUND AND GO, NEENER, NEENER, GOT YOU. FINGERS CROSSED BECAUSE WE HAVE AN INTEGRATION CLAUSE IN THE CONTRACT AND, YOU KNOW, YOU REALLY COULDN'T TRUST ANYTHING WE SAID AT THAT DARN MEETING BECAUSE UNLESS YOU GET IT IN WRITING, IT REALLY DOESN'T COUNT?"

After some more colloquy, the insured's counsel made it very clear that he was withdrawing any attempt to include the Houston Oral Promise as a "stand alone cause of action." Rather than do that, he said, he planned to use the Houston Oral Promise in the context of the causes of action that already were in the complaint.²⁷ And in supplemental briefing to this court, the insured has also made it abundantly clear that its case at trial was premised on the tortious breach of the 1996 written policy and not the Houston Oral Promise.²⁸

8. Late 2005 to March 2006: The insurer loses big at trial, brings the usual posttrial motions, then files this appeal.

27 "MR. WILLIAMS: AND THE RELATED MOTION WAS NORTHERN 32 THAT DEALT WITH THE GENTLEMEN'S AGREEMENT.
"THE COURT: LET ME RUN IN THE BACK REAL QUICK. I THOUGHT I HAD THEM RIGHT HERE. ALL RIGHT. DO YOU WANT TO GO THROUGH THEM INDIVIDUALLY OR IS AN ANSWER TO ONE AN ANSWER TO ALL?
"MR. CRONIN: YOUR HONOR, I THINK I CAN CLEAR IT UP PRETTY QUICKLY. WHERE WE LEFT OFF, TO REORIENT THE COURT, WAS WE WERE GOING TO AMEND OUR THIRD AND FOURTH CAUSES OF ACTION ON THE FRAUD. WHAT WE'VE DECIDED TO DO IS WE'RE JUST GOING TO WITHDRAW THOSE CAUSES OF ACTION IN LIGHT OF THE COURT'S RULING IN CONNECTION WITH BAD FAITH. SO THOSE ARE NO LONGER AN ISSUE. AND COUNSEL WAS ADVISED, WHATEVER IT WAS, TWO WEEKS AGO, YOUR HONOR, OF THIS WE STILL INTEND TO PRESENT TESTIMONY AND EVIDENCE RELATING TO THE GENTLEMEN'S AGREEMENT AND THE CIRCUMSTANCES SURROUNDING REPRESENTATIONS MADE AT THE MAY 20 97, MEETING. BUT INSOFAR AS THE FRAUD AND THE INTENTIONAL MISREPRESENTATION AS STAND-ALONE CAUSES OF ACTION, WE'RE WITHDRAWING THEM." (Italics added.)

28 To quote from a heading of the supplemental briefing filed by the insured March 10, 2009: "Northern's bad faith liability rests on its unreasonable refusal to defend the 1999 South Coast Lawsuit, which tortiously breached Griffin's 1996 policy, not the Houston Promise." (Original italics, boldface omitted.) To this heading the brief appended a footnote that drove home the absence of a claim based on the Houston Oral Promise: "Northern's bad faith could only rest on its refusal to defend under Griffin's 1996 policy. . . . Here, Northern's bad faith rested on the breach of its duty to defend under Griffin's 1996 policy." (Italics added.)

There are no issues in this appeal regarding the conduct of the trial (at least, given our determination of the objective reasonability of the insurer's denials of coverage as explained in pt. III.B.4. below). The major point about the trial for our purposes here was the insured's emphasis on the insurer's reservation of rights to recoup monies paid to settle the Waters claim. The reason is that much of the trial was devoted to establishing the rather large amount of attorney fees incurred by the insured to bring its bad faith case against the insurer to trial in the first place, so as to obtain fees under Brandt v. Superior Court, supra, 37 Cal.3d 813, 817 ("When an insurer's tortious conduct reasonably compels the insured to retain an attorney to obtain the benefits due under a policy, it follows that the insurer should be liable in a tort action for that expense. The attorney's fees are an economic loss — damages — proximately caused by the tort.").

In early January 2006, the jury came back with its verdicts. The jury found that the insured had paid about $1 million (all in legal fees and costs) to "collect the benefits due under the contract." It found that the insurer had acted with "oppression or fraud or malice in doing the acts that caused" the insured to incur those fees and costs, denominated "damages" in the special verdict form. It then proceeded to award $10 million in punitive damages.

After a futile round of postjudgment motions for new trial or judgment notwithstanding the verdict in late January, ²⁹ the insurer filed this appeal. As noted above, the case has taken two oral arguments and two rounds of additional briefing, largely because it has taken us some time to ascertain the true nature of the Houston Oral Promise, and how it fit into the rest of the case. Indeed, counsel for the insured did not make that clear until the second oral argument in April 2009, when he candidly explained that he saw the Houston Oral Promise as a "concession" by the insurer that its original coverage position was unreasonable. (The insurer disputed that characterization.)

29 These postjudgment motions served to extend the time to appeal the judgment, notice of entry of which was served January 13, 2006, so that the notice of appeal from the judgment and the denial of those motions, filed March 23, 2006, was timely.

III. DISCUSSION A. The Basic Three-part Paradigm for Analyzing Bad Faith Cases

By the time the trial court entertained the October 2005 motion in limine that essentially won the case for the insured, the MacKinnon case was already two years old. Accordingly, there was no question that the insurer had indeed breached the written contract of insurance when, in October 1999, it had denied the insured's request for a defense of the district's lawsuit stemming from the Waterses' sewage damage claim. The trial court's predicate ruling on which the in limine motion was based — that the insurer had breached the contract — was indeed correct.

But a breach of an insurance contract does not automatically subject an insurer to tort damages for bad faith breach. Bad faith cases are analyzed in a three-step process: First, was there a breach at all so as to warrant contract damages? Second, was the breach unreasonable so as to warrant tort damages? Third, was the breach so egregious that there is evidence of "oppression, fraud, or malice" under Civil Code section 3294, subdivision (a) so as to warrant punitive damages?³⁰

30 See Croskey et al., California Practice Guide: Insurance Litigation (The Rutter Group 2008) (hereinafter the "Croskey Rutter Insurance Treatise") paragraph 12:224, page 12B-6 (rev. # 1, 2008): "The first step is to determine whether the insurance company breached its contract with the insured (i.e., whether the claim was covered or potentially covered so that a defense was owed)" then "Next, determine whether the insurer's conduct breached its implied covenant of good faith and fair dealing with the insured (i.e., whether the insurer acted unreasonably)"; and "Finally, consider whether punitive damages are recoverable (i.e., whether the insurer acted with `oppression, fraud or malice' within the meaning of Civ.C. § 3292(a) and other elements of a punitive damages claim are satisfied)."
See also Croskey Rutter Insurance Treatise, supra, paragraphs 12:49 to 12:54, pages 12A-18 to 12A-19 (rev. # 1, 2008).

It is important to recognize the reason for the possibility of tort, and perhaps even punitive damages on top of regular tort damages, for an insurance company's unreasonable breach of an insurance contract. Insurance contracts are unique in that, if the insurance company breaches them, the policyholder suffers a loss (often a catastrophic loss) that cannot, by definition, be compensated by obtaining another contract. (E.g., Cates Construction, Inc. v. Talbot Partners (1999) 21 Cal.4th 28, 44 [ 86 Cal.Rptr.2d 855, 980 P.2d 407] ["Unlike other parties in contract who typically may seek recourse in the marketplace in the event of a breach, an insured will not be able to find another insurance company willing to pay for a loss already incurred."]; Medina v. Safe-Guard Products, Internat., Inc. (2008) 164 Cal.App.4th 105, 111 [ 78 Cal.Rptr.3d 672] ["The reason that insurance contracts may be enforced by tort (bad faith) as well as ordinary contract damages is that, unlike all other contracts, a policyholder who is wrongfully denied a claim cannot, by definition, obtain a substitute in the marketplace. Once it is known that the insurable loss has occurred, the insured will not be able to obtain insurance for that loss."].)

Thus, without the possibility of tort damages hanging over its head when it makes a claims decision, an insurance company may choose not to deal in good faith when a policyholder makes a claim. The insurance company could arbitrarily deny a claim, thus gambling with the policyholder's "benefits of the agreement." (E.g., Major v. Western Home Ins. Co. (2009) 169 Cal.App.4th 1197, 1209 [ 87 Cal.Rptr.3d 556] ["The fundamental purpose of the implied covenant of good faith and fair dealing is `"that neither party will do anything which will injure the right of the other to receive the benefits of the agreement."'"].)³¹ If the insurance company gambled wrong, it would be no worse off than it would have been if it had honored the claim in the first place. In effect, if the law confined the exposure of the insurance company under such circumstances to only contract damages, it would be pardoned and still retain the fruits of its offense.

31 To be sure, an insurance company that did that too often might quickly get a bad reputation for claims service, and the market might eventually force it to clean up its act, but only after leaving a trail of innocent victims.

On the other hand, if the coverage decision is reasonable, no lawyer has power to charm a jury into awarding any tort damages against the insurance company based on that coverage decision.³² As Justice Bedsworth wrote for this court in Morris v. Paul Revere Life Ins. Co., supra, 109 Cal.App.4th at page 977 ( Morris): "As long as the insurer's coverage decision was reasonable, it will have no liability for breach of the covenant of good faith and fair dealing. An insurer which denies benefits reasonably, but incorrectly, will be liable only for damages flowing from the breach of contract, i.e., the policy benefits." (Italics added.) We therefore now turn to the question of the reasonability of this insurer's coverage decision in October 1999.

32 There is one possible and relatively abstruse exception to this rule, which we note in footnote 38, post, and which most assuredly does not apply to this case.

B. The Reasonableness (or Lack Thereof) of the Insurer's Denial of a Defense to the Waters Claim

A CGL policy (remember that the L in CGL stands for "liability") such as the insured carried in this case typically provides two benefits: One, it will pay for liability a policyholder incurs by virtue of a lawsuit against the policyholder, at least up to a certain limit, and assuming that the terms of the policy (the risks covered) otherwise provide for such payment. The technical word for this benefit is called "indemnification." Two, the CGL contract will also pay for a defense of any lawsuit which even might potentially result in the insurance company indemnifying, in whole or in part, the policyholder. (See generally Croskey Rutter Insurance Treatise, supra, ¶ 7:4, pp. 7A-1 to 7A-2 (rev. # 1, 2008) ["A liability insurer generally has two basic obligations to its insured: [¶] • Indemnification: First, the insurer promises to pay on the insured's behalf (i.e., to indemnify the insured), up to the policy limits, the sums the insured shall become legally obligated to pay on account of specified risks, [¶] • Defense: Second, the insurer promises to defend any suit against the insured alleging damages payable under the terms of the policy."].)

Since insurance companies don't have law degrees, in practical effect the "duty to defend" means a duty to hire competent counsel to conduct the defense of a lawsuit against the policyholder. (E.g., Kroll Tract v. Paris Paris (1999) 72 Cal.App.4th 1537, 1542 [ 86 Cal.Rptr.2d 78] ["In the usual tripartite insurer-attorney-insured relationship, the insurer has a duty to defend the insured, and hires counsel to provide the defense."].)

The big question in this case, of course, was whether the insurer's denial of a defense to the district's lawsuit against the insured was reasonable. But to answer this question, we must address three interrelated problems: First, do we gauge the reasonableness of the insurer's denial on an objective standard as a matter of law (which is what the trial court here did, holding the decision unreasonable as a matter of law), or should the matter simply have been committed to the jury? Second, what is the impact of subsequent case law upon our evaluation of the insurer's denial? In the present case, for example, the insured pointed to the subsequent MacKinnon decision which, to say the least, invalidated the rationale behind the insurer's denial. Third, what is the nature of the "potentiality rule" that formed the basis of the trial court's reasoning that the insurer's denial was not reasonable?

1. A Primer on the Potentiality Rule

There is no better place to understand the potentiality rule than the nearly half-century-old fountainhead case that gave us the rule: Gray v. Zurich Insurance Co. (1966) 65 Cal.2d 263 [ 54 Cal.Rptr. 104, 419 P.2d 168]. There, a policyholder was sued for assault and battery — classic intentional torts not covered by insurance policies — after an altercation arising out of an auto accident. The policyholder claimed, however, that he acted in self-defense. The insurance company denied a defense of the lawsuit, reasoning that liability for intentional torts was not covered by the policy. But — insurance law's most cited opinion ever went on to explain — there was a potential that the policyholder might be found liable not for assault and battery, but merely for the negligent use of unreasonable force in the altercation. That potential liability thus created the possibility of a judgment for a negligent tort, not an intentional one, and if the judgment came down that way, the insurance company would have to pay for it. And because the insurance company might have to pay for such a judgment, it definitely had an obligation to defend.

The next logical question is: From what facts precisely is such a potential gauged? The answer is: (a) The facts alleged in the complaint, and (b) facts known to the insurance company at the time of the coverage decision (usually the inception of the suit). (See Scottsdale Ins. Co. v. MV Transportation (2005) 36 Cal.4th 643, 654 [ 31 Cal.Rptr.3d 147, 115 P.3d 460] ( Scottsdale).) It must be remembered in that regard, though, that the precise causes of action are not determinative of coverage. If, under the facts alleged, the complaint could be fairly amended to state a cause of action alleging a covered liability, there will be a duty to defend. (See ibid. ["Determination of the duty to defend depends, in the first instance, on a comparison between the allegations of the complaint and the terms of the policy. . . . But the duty also exists where extrinsic facts known to the insurer suggest that the claim may be covered. . . . Moreover, that the precise causes of action pled by the third party complaint may fall outside policy coverage does not excuse the duty to defend where, under the facts alleged, reasonably inferable, or otherwise known, the complaint could fairly be amended to state a covered liability." (Citations omitted.)].)

However, the facts do not include speculative facts not in the complaint, or otherwise unknown by the insurance company. (See Friedman Prof. Management Co., Inc. v. Norcal Mutual Ins. Co. (2004) 120 Cal.App.4th 17, 34-35 [ 15 Cal.Rptr.3d 359] ["the universe of facts bearing on whether a claim is potentially covered includes extrinsic facts known to the insurer at the inception of the suit as well as the facts in the complaint, it does not include made up facts, just because those facts might naturally be supposed to exist along with the known facts"]; Westoil Terminals Co., Inc. v. Industrial Indemnity Co. (2003) 110 Cal.App.4th 139, 154 [ 1 Cal.Rptr.3d 516] [refusing to consider scenario not in facts of complaint]; Low v. Golden Eagle Ins. Co. (2002) 99 Cal.App.4th 109, 114 [ 120 Cal.Rptr.2d 827] [refusing to consider possibility of amendment of third party's complaint against policyholder because it would speculate "about unpled causes of action"]; Hurley Construction Co. v. State Farm Fire Casualty Co. (1992) 10 Cal.App.4th 533, 538-539 [ 12 Cal.Rptr.2d 629] [refusal to consider hypothetical amendment to third party's complaint against policyholder]; Gunderson v. Fire Ins. Exchange (1995) 37 Cal.App.4th 1106, 1114 [ 44 Cal.Rptr.2d 272], quoting Hurley.)

An insurance company can thus get into trouble by refusing to consider facts which it knows, but which are extrinsic to the complaint and which show, ala Gray, a potential for coverage.

The insurance company committed precisely that sort of error in Amato v. Mercury Casualty Co. (1993) 18 Cal.App.4th 1784 [ 23 Cal.Rptr.2d 73] ( Amato I). In Amato I, the policyholder was sued by his mother-in-law, for injuries arising out of an auto accident in which the policyholder was driving and the mother-in-law was a passenger. Application of a certain exclusion depended on whether the policyholder resided with his mother-in-law. The Amato I court noted the insurance company "possessed information which, if true, indicated that [he] was residing at locations other than the home of" his mother-in-law. ( Id. at p. 1789.) Despite that knowledge, the insurance company refused to defend or, after a default judgment against the policyholder, indemnify. As it turned out, though, in a subsequent bad faith action, the jury found that the policyholder did, indeed, live with his mother-in-law. ( Ibid.) That fact, however, did not exonerate the insurance company. At the time of the coverage decision, the insurance company had facts in its possession, albeit extrinsic to the mother-in-law's complaint against the policyholder, which created a potential that the exclusion did not apply, ergo there was a potential for indemnification, ergo there was definitely a duty to defend. It made no difference that a jury ultimately found that those extrinsic facts were ultimately disproven at trial.

2. The Question of an Objective Standard

Unlike Amato I, the peculiar circumstances of the case before us do not implicate any "facts known to the insurer" extrinsic to the district's complaint against the insured which suggest the nonapplicability of the total pollution exclusion to sewage claims. Nor has there been any attempt by the insurer to disclaim coverage based on facts outside the complaint that might establish that the liability for the sewage backup was never potentially covered in the first place, as if, for example, the insured actually "expected" the damage. (E.g., Shell Oil Co. v. Winterthur Swiss Ins. Co. (1993) 12 Cal.App.4th 715, 772 [ 15 Cal.Rptr.2d 815] [insurance company lawyers argued policyholder "should have known it was polluting"].)

From the beginning, the insurer's coverage decision in this case, and the insured's quest for a defense, was a straightforward, binary matter based on a discrete, almost pristine question of law: Did the insurer's total pollution exclusion categorically eliminate any potential for coverage for sewage backups?

As such, we believe we are on safe ground to conclude, as indeed the trial court concluded, that the question of the reasonableness of the categorical application of the pollution exclusion to the sewage backup must be determined in this case using an objectively reasonable standard. After all, as shown especially in the MacKinnon opinion itself (see MacKinnon, supra, 31 Cal.4th at pp. 645-647 [noting multitude of court decisions in area]; cf. ACL Technologies, Inc. v. Northbrook Property Casualty Ins. Co. (1993) 17 Cal.App.4th 1773, 1779-1783 [ 22 Cal.Rptr.2d 206] [canvassing decisions across country on interpretation of prior version of pollution exclusion]), the determination of the scope of the pollution exclusion in the CGL is the quintessential legal question, on which many courts in many jurisdictions opine. (See CalFarm, supra, 131 Cal.App.4th 273, 287 ["When the issue of the insurer's objective reasonableness depends on an analysis of legal precedent, reasonableness is a legal issue reviewed de novo."]; Chateau Chamberay Homeowners Assn. v. Associated Internat. Ins. Co. (2001) 90 Cal.App.4th 335, 346 [ 108 Cal.Rptr.2d 776] [reasonableness of "claims-handling conduct . . . becomes a question of law where the evidence is undisputed and only one reasonable inference can be drawn from the evidence"].) This case is not like Amato I where coverage rested on a disputed issue of fact.

The tidiest statement in the area may be found in Morris, supra, 109 Cal.App.4th 966, 973, footnote 1: "We recognize there are numerous cases reciting that `reasonableness' of a defendant's conduct is a factual question for the jury. However, the reasonableness of the legal position taken by [the insurance company here] depends entirely on an analysis of legal precedent and statutory language. Those are matters of law, not facts which can effectively be ascertained by lay jurors."

Because it was an objective matter and thus an issue of law, the Morris court would ultimately be able to go on to affirm a summary judgment declaring that the insurance company had not acted in bad faith.³³

33 A corollary of this conclusion is that there is no issue in this case involving a failure-to-investigate for its own sake, that is, where coverage actually turned on facts the insurer should have unearthed, such as Egan v. Mutual of Omaha Ins. Co. (1979) 24 Cal.3d 809 [ 169 Cal.Rptr. 691, 620 P.2d 141], where a disability insurance company didn't even bother to ask the policyholder's own doctors about his condition before deciding that he was malingering. From the beginning, the case before us, in contrast to cases like Egan, has turned on a pure question of law.

3. The MacKinnon Case

We thus determine that an explication of the MacKinnon decision is the appropriate way to determine the reasonableness of the insurer's coverage determination under the facts of the present case.

Prior to the 2003 MacKinnon decision, there was no decision explaining the scope of either the "absolute" or the "total" pollution exclusion in California, though as we have noted in part II.B., ante, for a brief few weeks in early 2002 there were two published appellate decisions which appeared to go in the insurer's favor. The only decisions to that point had been in the context of "traditional" pollution, i.e., long-term gradual exposure. (See MacKinnon, supra, 31 Cal.4th at p. 641, fn. 1.) MacKinnon itself arose out of the application of insecticide by a landlord. Unfortunately, the pest control company he hired failed to warn his tenant of the spraying, and her reaction to the subsequent exposure killed her. Her heirs sued the landlord, but his insurance company denied coverage, specifically withdrawing an initially provided defense after deciding that the absolute pollution exclusion precluded coverage.³⁴ The withdrawal, in turn, provoked a coverage suit by the landlord. The trial court granted summary judgment to the insurance company on the strength of the absolute pollution exclusion, and the appellate court affirmed. ( MacKinnon, supra, 31 Cal.4th at pp. 640-641.)

34 The MacKinnon court pointed out that the term "absolute" pollution exclusion did not appear in the CGL policy before it. ( MacKinnon, supra, 31 Cal.4th at p. 644, fn. 3.) Like that court, we use the phrase "absolute pollution exclusion" as a handy way to refer to the post-1985 pollution exclusion in the CGL, which, at the very least, excluded more than the pre-1985 clause, or "qualified pollution exclusion."

But the Supreme Court reversed. The structure of the discussion portion of the opinion bears study: First, the high court set up the problem: The scope of the absolute pollution exclusion was a big, nationwide issue: Would courts take an expansive view or a narrow view of the new pollution exclusion? The issue had generated a lot of controversy on which courts around the country were split. (See MacKinnon, supra, 31 Cal.4th at pp. 641-642, and particularly p. 642, fn. 2, noting the scoreboard as it stood to date.) The MacKinnon court surveyed the score as it stood in 2003: If we read the court's recount correctly, it was nine jurisdictions (Illinois, Indiana, Kansas, Louisiana, Maryland, Massachusetts, New York, Ohio, Wyoming) for a narrow, i.e., propolicyholder reading of the absolute pollution exclusion, versus four jurisdictions for an expansive reading (Florida, Montana, Oklahoma, Texas) with Pennsylvania being in both camps.

The MacKinnon court then summarized the history of the CGL pollution exclusion as told by the Illinois Supreme Court in American States Ins. Co. v. Koloms (1997) 177 Ill.2d 473 [227 Ill.Dec. 149, 687 N.E.2d 72] ( Koloms), plus noted the positions taken by commentators to the effect that the exclusion was intended to be limited to "traditional environmental pollution rather than all injuries from toxic substances." ( MacKinnon, supra, 31 Cal.4th at pp. 643-645.)

Next, the MacKinnon court recapitulated the arguments for and against a narrow interpretation of the pollution exclusion. (See MacKinnon, supra, 31 Cal.4th at pp. 645-647.) Readers of that portion of the opinion will note a thorough discussion of both sides, particularly of the side it would ultimately reject. Just as the MacKinnon opinion featured the Illinois Supreme Court's decision in Koloms, by using language from that decision to demonstrate the historical case for a narrow view of the exclusion (see MacKinnon, supra, 31 Cal.4th at pp. 643-644) so did the MacKinnon court quote what the Wisconsin Supreme Court had said in Peace ex rel. Lerner v. Northwestern Nat'l Ins. Co. (1999) 228 Wis.2d 106 [ 596 N.W.2d 429] ( Peace) in favor of a broader view of the exclusion based on its "plain language." (See MacKinnon, supra, 31 Cal.4th at pp. 646-647.)

After restating some principles of insurance contract construction, the MacKinnon court next came to the nitty-gritty of its decision: The literal language of the absolute pollution exclusion, particularly the words "`irritant or contaminant,'" was just "overly broad," and thus yielded absurd results: "The application of iodine," wrote Justice Moreno for a unanimous court, "onto a cut through an eyedropper may be literally characterized as a discharge or release of an irritant." ( MacKinnon, supra, 31 Cal.4th at p. 650.) The MacKinnon court made it clear that "interpreting an exclusionary clause so broadly that it logically leads to absurd results," was not the course it would follow. ( Ibid.; see also id. at p. 652 ["In short, because Truck Insurance's broad interpretation of the pollution exclusion leads to absurd results and ignores the familiar connotations of the words used in the exclusion, we do not believe it is the interpretation that the ordinary layperson would adopt."].) The court then went on to demonstrate that the more reasonable interpretation of the exclusion was that it should be limited to injuries arising from "conventional environmental pollution." ( Id. at p. 655.)

4. " Congeries of Cerebrates": Why MacKinnon and Morris Show the Insurer's Decision Here Was Reasonable

The case before us is remarkably similar to Morris. Remember that in Morris an insurance company that issued a disability policy denied coverage based on a policy interpretation that had garnered considerable, if minority, support in various jurisdictions (there the score had been seven to nine in favor of policyholders). And, as here, the insurance company's interpretation was rejected, years later, by the California Supreme Court in a decision that settled the matter for this state. Recognizing that "so many courts" had agreed with insurance companies about what was a pure question of law, the Morris court held that the insurance company's position was reasonable. ( Morris, supra, 109 Cal.App.4th at p. 976.) In a line that recalls Brahms's comment on Strauss's "Blue Danube," Justice Bedsworth summed up its rationale: "We cannot just dismiss such a congerie of cerebrates as `unreasonable.'" ( Ibid.)

Another similarity was the temporary precedent favoring the insurance company's position. In Morris, two Court of Appeal decisions — as here — had gone with the position advocated by the insurance company. And, as here, one of those two cases would end up as the Supreme Court case on point, while the other disappeared in the wake of the Supreme Court's ultimate position. (See Morris, supra, 109 Cal.App.4th at p. 971.) That is, as in Morris, more than just one trial judge had ruled in favor of the insurance company's position; the majority on two appellate panels had as well, plus the majority of about seven state supreme courts outside of California. Here, as shown in MacKinnon, we also have the majority of two appellate courts, and (depending on how you count Pennsylvania) four or five state supreme courts outside of California.

Morris was not just remarkably similar to the present case. It is almost on all fours.

Ironically, whatever few significant differences that do exist between the case before us and Morris actually make the conclusion stronger that Morris requires us to conclude that the denial here was reasonable.

First, in Morris, the same disability insurance company before the court had itself been the company that had litigated the issue in jurisdictions "favoring its position." ( Morris, supra, 109 Cal.App.4th at p. 976.) The policyholder argued that the insurance company was just running around the country creating favorable precedent for itself, in effect conjuring outside precedent with a deceptively pleasing shape. The Morris court rejected that line of reasoning by pointing out that the insurance company there was "entitled to be an advocate for its own interests." ( Ibid.) Here, by contrast, we perceive no special role that this insurer has taken beyond any other CGL insurance company in establishing whatever favorable precedent there is for its position. (See MacKinnon, supra, 31 Cal.4th at p. 642, fn. 2.)

The second significant difference is that in Morris there actually was one, lone standing appellate precedent against the insurance company's position, McMackin v. Great American Reserve Ins. Co. (1971) 22 Cal.App.3d 428, 439-440 [ 99 Cal.Rptr. 227]. To be sure, that precedent didn't count for much because, as Morris noted that Galanty had noted, that lone precedent was "`without useful discussion.'" ( Morris, supra, 109 Cal.App.4th at p. 971, quoting Galanty v. Paul Revere Life Ins. Co. (2000) 23 Cal.4th 368, 377 [ 97 Cal.Rptr.2d 67, 1 P.3d 658].) Even so, it was there. The current might have been weak, but the insurance company in Morris was still swimming upstream. Here, by contrast, there was zero precedent on point against the insurer's position at the time it made its coverage decision.

A third difference is the degree to which one can perceive a difference in the relative strength of the insurance company's linguistic and analytical position in the two cases. The position of the insurance company was, if anything, weaker in Morris than the insurer's position here.

Remember that Morris was a case where a disease (multiple sclerosis) had manifested itself before the inception of the policy but the claim was made after the first two years of the policy, in light of an incontestability clause which said, essentially, that the insurance company would not dispute any claim presented after two years had passed.³⁵ Thus, in Morris the policyholder had the literal language of the policy's incontestability clause going in his favor. The insurance company could only prevail if a court were to give precedence to the contract's limitation of coverage to sicknesses first manifesting themselves after the inception of the policy over the clear implication of the incontestability clause that claims presented after two years would be honored without dispute.

35 The exact language, quoted in Morris, was: "The policies also included an incontestability clause, as required by California Insurance Code section 10350.2, which provided: `a. After Your Policy has been in force for 2 years, excluding any time You are Disabled, We cannot contest the statements in the application. [¶] b. No claim for loss incurred or Disability that starts after 2 years from the Date of Issue will be reduced or denied because a sickness or physical condition not excluded by name or specific description before the date of loss had existed before the Date of Issue.'" ( Morris, supra, 109 Cal.App.4th at p. 969.)

By contrast, in the case before us, the insurer's position was rooted in the literal language of the absolute pollution exclusion, so the insurer was going to win if a court were to apply the literal language of the exclusion. After all, pollution was defined to include "waste," which is often a euphemism for sewage.³⁶ Moreover, plain language remained (and remains today) an important principle of contract interpretation. Just two months prior to the MacKinnon decision, the California Supreme Court had enforced the plain language of a collapse provision in a homeowner's insurance policy ( Rosen v. State Farm General Ins. Co. (2003) 30 Cal.4th 1070, 1073 [ 135 Cal.Rptr.2d 361, 70 P.3d 351] ["By failing to apply the plain, unambiguous language of the policy, the Court of Appeal erred."]) and, of course, the high court had invoked plain language as the governing rule numerous times before. (E.g., Palmer v. Truck Ins. Exchange (1999) 21 Cal.4th 1109, 1115 [ 90 Cal.Rptr.2d 647, 988 P.2d 568] [when language "`is clear and explicit, it governs'"]; Bank of the West v. Superior Court (1992) 2 Cal.4th 1254, 1264 [ 10 Cal.Rptr.2d 538, 833 P.2d 545] [same].) MacKinnon itself had also noted that the Wisconsin Supreme Court in Peace had invoked the "`plain language'" of the policy to uphold the insurance company's position there. ( MacKinnon, supra, 31 Cal.4th at pp. 646-647.)

36 The total pollution exclusion in the 1996 policy said: "This insurance does not apply to: [¶] (1) . . . `property damage' which would not have occurred in whole or part but for the actual, alleged or threatened discharge, dispersal, seepage, migration, release or escape of pollutants at any time." Pollutants were defined to "mean[] any solid, liquid, gaseous, or thermal irritant or contaminant including smoke, vapor, soot, fumes, acid, alkalis, chemicals and waste. Waste includes material to be recycled, reconditioned or reclaimed."

To be sure, the MacKinnon court didn't apply the literal language of the exclusion, finding it "overly broad," and MacKinnon stands as a decision refusing to carry literal language to an absurd result. But surely the insurer here, facing no actual contrary language in the contract, was on firmer ground than the insurance company in Morris, which (in light of the incontestability clause) did.³⁷

37 One can often detect an understandable judicial reluctance on the part of courts to invoke the "absurd results" rule in the face of literal language. (E.g., Benson v. Kwikset Corp. (2007) 152 Cal.App.4th 1254, 1270 [ 62 Cal.Rptr.3d 284] [rejecting dissent's argument based on absurd results rule in statutory context].) As some courts have described the absurd results approach to language, one would think a sparrow dies every time it is invoked. (See Unzueta v. Ocean View School Dist. (1992) 6 Cal.App.4th 1689, 1699 [ 8 Cal.Rptr.2d 614] ["Each time the judiciary utilizes the `absurd result' rule, a little piece is stripped from the written rule of law and confidence in legislative enactments is lessened."].) Perhaps, then, there is a special provenance whenever a court departs from the plain language rule because of some absurdity that would thereby ensue, which is what happened in MacKinnon. But that sort of specialness only shows the insurance company's position in this case to have been all the more reasonable, even if not ultimately prevailing.

5. The Insured's Attempt to Distinguish Morris Is Not Persuasive

Readers may already have noted one other difference between Morris and the present case that did not make our list of "significant" differences: Morris was a "first party" case, where the policyholder's loss was directly compensable by the insurance company — the relationship is direct between insurance company and policyholder. The present case is a "third party" case where the insurance company is to defend and if necessary indemnify a "third party's" claim against the policyholder, so the compensation is indirect. There would be no problem if there weren't a third party out there who wanted the policyholder to pay it money.

The insured here attempts to distinguish Morris solely on that difference. According to the insured here, the fact that Morris was first party and the present case is third party means that in Morris the insurer had the right to (and we now quote directly from the respondent's brief): "`argue for whatever interpretation of the law and policy language most benefited its own interests,' so long as the interpretation was `objectively reasonable.' [ Morris, supra, 109 Cal.App.4th at p. 974.] That rule simply does not apply to third-party insurance coverage, where the insurer must construe the policy language favorably to the insured." (Original italics.)

The insured here seems to be intimating that, because Morris was a first party case, the insurance company there had the right to pick its side of the legal issue which had divided courts as noted in Galanty, but here, since we have a third party situation, the insurer was obligated to side with the anti-insurance company side of the division in case law as noted in MacKinnon.

We beg to differ. The passage from the respondent's brief betrays several points of confusion which need clearing up. Preliminarily, we should note that we are well aware of the differences between first and third party insurance. Readers who want to consult an extended discussion on the difference are referred to Garvey v. State Farm Fire Casualty Co. (1989) 48 Cal.3d 395, 405-408 [ 257 Cal.Rptr. 292, 770 P.2d 704], which was a case where the difference between first and third party insurance actually did make a difference in the outcome.

a. Reasonability Is the Test in Both First and Third Party Cases

First, as far as bad faith law is concerned, we can divine no difference in basic doctrine between first party cases and third party cases under the facts of this case. The basic rule of reasonability of coverage decision holds for both first and third party insurance policies. In either case, tort damages are predicated on the policyholder's being stuck with costs that an insurance company would have otherwise paid for, but which — after the event giving rise to those costs — now cannot be covered by another insurance contract. As we have already pointed out, imposition of a reasonability standard is required to prevent the insurance company from, in essence, denying its obligation under the contract risk-free. (See pt. III.A., ante.)³⁸

38 We can think of but one possible situation where a liability insurance company might be tagged for "tort" damages when it was reasonable, but incorrect. That is the situation where the insurance company had the chance to settle a case within policy limits, but passed up the opportunity because it incorrectly determined that there was not even any potential coverage. In that particular situation, even though the coverage decision may have been reasonable, but incorrect, the insurance company still might have to pay the excess of any ensuing judgment over the policy limits as bad faith damages. (See generally Comunale v. Traders General Ins. Co. (1958) 50 Cal.2d 654 [ 328 P.2d 198] [insurance company's "wrongful" refusal to defend and thereby settle resulted in liability for excess judgment over policy limits].) We express, however, no opinion on the scenario, other than to mention our awareness of the Comunale case and to note that here the insurer protected itself from any such exposure by settling the district's suit against the insured before that suit ever went to judgment.

It is illustrative of the settledness of the reasonability standard that, for example, Justices Moore and Fybel, who differed in CalFarm — a third party case — on the question of whether the issue of reasonability was a matter of law or, alternatively, should have been submitted to a jury in the case, both agreed with each other in R B Auto Center, Inc. v. Farmers Group, Inc. (2006) 140 Cal.App.4th 327 [ 44 Cal.Rptr.3d 426] — also a third party case that because the insurance company's position was reasonable, there was no tort liability. (Compare id. at p. 354 (maj. opn. of Moore, J.) ["At the time Truck Insurance evaluated the claim, it was reasonable to deny it."] with id. at p. 374 (dis. opn. of Fybel, J.) ["I concur in the majority's affirmance of the trial court's decisions in favor of defendants on the causes of action for breach of contract, breach of the implied covenant of good faith and fair dealing, and breach of fiduciary duty."].)

b. A Third Party Insurer Has the Right to Make a Reasonable Coverage Decision Even if It "Benefits Its Own Interests"

The insured's assertion that the insurer had no right to take a legal position on policy language that "benefited its own interests" is an attempt to set up a kind of mousetrap for liability insurance companies. The trap is that if a liability insurance company proffers a legal position regarding a matter of pure contract interpretation in its own "interest" — just like any other litigant might do in a contract — it ends up proclaiming its own bad faith.

No. As the court in Love v. Fire Ins. Exchange (1990) 221 Cal.App.3d 1136 [ 271 Cal.Rptr. 246] ( Love), pointed out, insurance companies do not stand in a fiduciary relationship with policyholders, though the unique nature of insurance contracts may make the relationship "akin" to a fiduciary one. Thus: "Unique obligations are imposed upon true fiduciaries which are not found in the insurance relationship. For example, a true fiduciary must first consider and always act in the best interests of its trust and not allow self-interest to overpower its duty to act in the trust's best interests." ( Id. at p. 1148, italics added.)

By contrast, said the Love court, an insurance company retains the power to give its interests "consideration equal to that it gives the interests of its insured" and "is not required to disregard the interests of its shareholders and other policyholders when evaluating claims." ( Love, supra, 221 Cal.App.3d at pp. 1148-1149.)

Most significantly, for our purposes, an insurance company " is not required to pay noncovered claims, even though payment would be in the best interests of its insured." (Love, supra, 221 Cal.App.3d at pp. 1148-1149, italics added.)

Since Love was decided, our Supreme Court has quoted with approval its distinction between a true fiduciary and the fiduciarylike duties of an insurance company. (See Vu v. Prudential Property Casualty Ins. Co. (2001) 26 Cal.4th 1142, 1151 [ 113 Cal.Rptr.2d 70, 33 P.3d 487] [quoting Love's language that while "`these "special" duties are akin to, and often resemble, duties which are also owed by fiduciaries, the fiduciary-like duties arise because of the unique nature of the insurance contract, not because the insurer is a fiduciary'"]; accord, Morris, supra, 109 Cal.App.4th at p. 973 ["An insurer is not a fiduciary, and owes no obligation to consider the interests of its insured above its own."]; Oakland Raiders v. National Football League (2005) 131 Cal.App.4th 621, 633, fn. 9 [ 32 Cal.Rptr.3d 266] ["Although an insurer has special obligations of good faith towards its insured, an insurer is not a `true fiduciary' with respect to its insured."].)

Now, we must point out that in asserting a contractual position, an insurance company must take a reasonable position under rules of contract interpretation, which rules generally favor policyholders. For example, if there is an ambiguity in an insurance policy provision, the insurance company must interpret the ambiguity in favor of the policyholder. (The words of the contract must be ambiguous in a given context, of course, see Bank of the West v. Superior Court, supra, 2 Cal.4th at p. 1265 [language must be construed in context and there is no ambiguity in the abstract] and that context must, of course, also take into account the rule of contract interpretation prescribed in § 1649 of the Civ. Code.)

However, the question of whether there is an ambiguity in the first place is different. As the Love court said, nothing obligates insurance companies to pay "noncovered" claims just because such a payment will (obviously) benefit the policyholder.

In sum, we divine the following rule: The question of whether an insurance contract is, in context, ambiguous at all, is one where the insurance company may, if reasonable, take a position that "benefits its own interest." Otherwise, a liability insurance company would lose every time the policyholder could find a court — anywhere, in any jurisdiction (even if that court was having a bad day³⁹) — that took the policyholder's side of a given question of law over the interpretation of contract language, and even if that court employed a rationale that would be generally at odds with California rules regarding interpretation of policy language. It would mean that an insurance company could never argue in favor of its side in a pure question of law if there were some court somewhere that had taken the opposite view. (Remember that we are not dealing in this case with any issues involving the reasonableness of the insured's coverage position.)

39 The Morris opinion perceptively suggested that even an appellate ruling might be "obviously flawed." ( Morris, supra, 109 Cal.App.4th at p. 976, fn. 2.) We assume that a court can issue an "obviously flawed" ruling in a policyholder's favor as well as one in an insurance company's favor.

This court specifically rejected such reasoning in ACL Technologies, Inc. v. Northbrook Property Casualty Ins. Co., supra, 17 Cal.App.4th at page 1787, footnote 39 ("The mere fact that judges of diverse jurisdictions disagree does not establish ambiguity under the particular principles which govern the interpretation of insurance contracts in California. . . .").

And, more significantly, the MacKinnon decision clearly bypassed the opportunity to say that the mere fact that jurisdictions were split showed the absolute pollution exclusion to be ambiguous. Rather, our high court grounded its conclusion on classic rules of California policy interpretation, most importantly the rule that policy terms should be interpreted as an ordinary person would understand those terms.

Indeed, in its summary of the respective arguments for and against the insurance company's application of the absolute pollution exclusion beyond traditional environmental pollution, the MacKinnon court quoted a passage from the Wisconsin Peace court to the effect that the mere fact that courts disagree does not make a term ambiguous. ( MacKinnon, supra, 31 Cal.4th at p. 647.) To be sure, MacKinnon would ultimately reject the outcome in Peace. But, it bears noting, the MacKinnon court did not use an if-courts-disagree-ergo-the-term-must-be-ambiguous rationale. Rather, after quoting the passage from Peace, the MacKinnon court then went on to list principles for construing insurance contracts in California, none of which included the idea that because courts disagree a clause is therefore ambiguous. (See MacKinnon, supra, 31 Cal.4th at pp. 647-648.) After recitation of those principles, the MacKinnon court rested its rationale on the overbreadth of the insurance company's attempt to apply the absolute pollution exclusion beyond traditional environmental pollution.

The essential flaw in the insured's argument is thus that it conflates two different sets of ideas: (a) the basic rules of insurance contract interpretation on the one hand, and (b) the question of the reasonability of a given position taken by an insurance company in light of the basic rules of contract interpretation.

This brings us face to face with precisely where the trial court went wrong. As our Supreme Court emphasized in Scottsdale, the mere existence of a legal dispute does not create a potential for coverage: "However, we have made clear that where the third party suit never presented any potential for policy coverage, the duty to defend does not arise in the first instance, and the insurer may properly deny a defense. Moreover, the law governing the insurer's duty to defend need not be settled at the time the insurer makes its decision." ( Scottsdale, supra, 36 Cal.4th at p. 657, italics added.)

C. Damage Issues 1. The Failure to Amend the Complaint to Include a Cause of Action Based on the Houston Oral Promise Precludes Any Recovery Based on That Promise As the trial judge rightly intuited, an insurance company cannot first promise coverage, then barefaced repudiate that promise by pointing to an integration clause in the written contract, and saying in effect, hey "neener, neener." Indeed, it was the Houston Oral Promise that has posed the biggest problem in the case for us: Surely, we first thought, at least some punitive damages were merited given such a barefaced repudiation.

Except it didn't happen that way. A cause of action based on the repudiation of the alleged oral promise made in Houston was never pled, and the complaint was never amended to add such a cause of action, up through and including trial. The insured's counsel told us at the second oral argument, rather, that the Houston Oral Promise was effectively a concession that the insured's position was right all along.

In a word, the insured had its chance to amend its complaint to allege such a cause of action — indeed, apparently gave some deliberation to the idea — and ultimately bypassed the opportunity. Indeed, even as recently as the supplemental briefing preparatory to the second oral argument, the insured's counsel made it clear that the insured was not seeking recovery based on a breach of the Houston Oral Promise.

Thus, the judgment cannot be affirmed based on breach of the Houston Oral Promise. No cause of action based on it was ever alleged.

We have already noted the governing rule: A plaintiff must recover, if at all, upon a cause of action set out in the complaint, and not on some other cause of action which may be developed by the proofs. The Supreme Court authority stating the rule is abundant. ( Lavely v. Nonemaker (1931) 212 Cal. 380, 385 [ 298 P. 976] ["It is a fundamental principle of pleading that `a plaintiff must recover, if at all, upon the cause of action set out in the complaint, and not upon some other which may be developed by the proofs.'"]; Schirmer v. Drexler (1901) 134 Cal. 134, 139 [ 66 P. 180] ["`A plaintiff must recover, if at all, upon the cause of action set out in the complaint, and not upon some other which may be developed by the proofs.'"]; Reed v. Norton (1893) 99 Cal. 617, 619 [ 34 P. 333] [source of quote in Schirmer]; Mondran v. Goux, supra, 51 Cal. at p. 153 [source of quote in Reed].) We may assume that the relative age of the Supreme Court cases stating the rule stems from its obviousness. One of the reasons courts have developed a policy of liberality in allowing amendments to pleadings, even up to and including trial, is precisely to avoid the loss of a meritorious cause of action (or defense) because counsel simply forgot to include it in an earlier pleading. (Cf. City of Stanton v. Cox (1989) 207 Cal.App.3d 1557, 1563 [ 255 Cal.Rptr. 682] [noting ability of trial judge to permit amendments to pleadings during trial].)

Here, however, the insured never asked, even during the course of the trial, to be allowed to amend its complaint to state a cause of action based on breach of the Houston Oral Promise. Whether it would have been within the trial court's proper exercise of its discretion to have allowed such an amendment is now academic.

2. The Reservation of Rights Issue

To recap the evidence concerning the insured's damages:

The insured never actually paid any defense costs to defend the lawsuit brought against it by the district because of a lucky happenstance of circumstances: AIG, the excess insurance company, fronted the money for lawyers to defend the district's lawsuit in the period summer 1999 through September 2000 with the understanding that the insured would be billed for the lawyers then being paid by AIG, but, before the insured was ever billed, the insurer changed its mind and defended the district's lawsuit against the insured, i.e., paid the costs of defense that had been incurred.

The insured never incurred any liability to the district by way of a judgment in the district's lawsuit against the insured, because the insurer settled the case below policy limits prior to any judgment.

By late September 2000, the insured had incurred less than $10,000 in fees in the pending bad faith suit ($8,921.40 to be exact). And in the letter of September 2000, the insurer offered to pay those fees as well. But — and this is stressed by the insured — when the insurer offered to pay those fees, it reserved its right to seek recoupment of what it had paid in the case so far.

So we come to the next question in the case: Could not the fees and costs incurred in the bad faith suit after September 2000 — about $1 million worth — still be justified as " Brandt fees," that is, fees incurred by the insured to obtain what it was owed under the contract, on the theory that the insured never really got what it was owed under the contract until the insurer had relinquished all possible recoupment rights? The insured's theory in this regard was (and is on appeal) that, by reserving its rights, the insurer had deprived it of the full "benefits" it was owed under the policy.⁴⁰

40 We quote the insured's brief directly on the reservation of rights issue to give readers a better sense of the argument:
— Characterizing the reservation as a "demand" that the insured pay the $350,000 settlement which the insurer had paid to settle the district's claim, "turned Griffin, in effect, into Northern's insurer."
— "What Northern said by retaining its right to pursue Griffin is that we're using Griffin's money to settle with South Coast."
— "`[A] reservation of rights letter' creates `a string,' so that the insured does not have the benefits of the policy until after that string is cut." (Original italics, as to what one of the insurer's expert's "admitted.")
— "The $350,000 payment in 2000 was not unconditional because Northern never withdrew its reserved rights and denied coverage even at trial."

There are two parts of the theory. The first part is simple: The insurer, says the insured, had no right to reserve its right to seek recoupment from the time it first asserted a reservation of rights in September 2000 until that reservation was formally withdrawn in October 2002.

The second part is a little more esoteric, and is used to bridge the gap between October 2002 and the fees incurred up to and including the trial. The insured points out that there was no explicit withdrawal of any reservation of rights to seek recoupment from the excess insurance company, AIG, even in the October 2002 letter (which otherwise unilaterally withdrew any right to seek recoupment). The AIG theory goes like this: Because the insurer still clung to the possibility of a recoupment action against AIG, there remained the theoretical possibility that the insured might ultimately get stuck with the $50,000 deductible in the AIG policy in the wake of a successful recoupment action against AIG (apparently assuming that AIG would turn around and sue the insured). Thus, the reasoning goes, in order to obtain "full" policy benefits, the insured had to incur all the fees it did, not only until October 2002, but until trial in 2005 as well.⁴¹

41 To quote again from the insured's respondent's brief: — "Northern is wrong to claim that the Williams letter of October 8, 2002, provided Griffin all policy benefits . . . because that letter did not withdraw Northern's right of contribution from AIG [the excess insurance company], which exposed Griffin to pay AIG's $50,000 deductible."
— The trial judge "ruled correctly in allowing the jury to find that Northern's reservation of rights to pursue AIG made policy benefits a legitimate object of Griffin's suit through October 2005."
— "[I]f Northern had gotten contribution from AIG, triggering the deductible, Griffin would have been harmed."

The insured's reservation of rights theory — both parts — betrays a profound misunderstanding — based more on linguistic art than legal matter — of an insurance company's reservation of rights to obtain recoupment, as outlined in the seminal case on that topic, Buss v. Superior Court, supra, 16 Cal.4th 35 ( Buss). The misunderstanding is this: The argument fails to recognize that the right of recoupment, under Buss, is necessarily predicated on the proposition that the insurance company has expended funds on the policyholder's behalf that were never "even potentially covered" under the insurance contract. ( Buss, supra, 16 Cal.4th at p. 50 ["As to the claims that are not even potentially covered, however, the insurer may indeed seek reimbursement for defense costs."].) Thus, by the same token, an insurance company "may not seek reimbursement for defense costs" as to "claims that are at least potentially covered" because a defense of those claims is owed under the contract. ( Id. at p. 49.)

That is, by definition and as a matter of law, an insurance company's only hope of reimbursement is if the policyholder has received a benefit that the policyholder never paid for in the first place. ⁴² As Scottsdale put it in the course of holding that an insurance company could indeed obtain reimbursement, in the circumstances of that case, of its expenses in defending a policyholder: "[I]f, as a matter of law, neither the complaint nor the known extrinsic facts indicate any basis for potential coverage, the duty to defend does not arise in the first instance." ( Scottsdale, supra, 36 Cal.4th at p. 655.)

42 Here is the entirety of the salient language from the two paragraphs that span pages 48 and 49 of Buss, supra, 16 Cal.4th 35 opinion in the official reporter. Readers should bear in mind that the high court had defined "`mixed' action" two paragraphs before on page 48 as an action in which "some of the claims are at least potentially covered and the others are not. . . ." Readers should take special note in this passage that the Buss court was careful to make a point as co what the insurance company owed contractually and what was required as a matter of common law. (To demonstrate that point, we have italicized that language and deleted original italics.)
"We cannot justify the insurer's duty to defend the entire `mixed' action contractually, as an obligation arising out of the policy, and have never even attempted to do so. To purport to make such a justification would be to hold what we cannot — that the duty to defend exists, as it were, in the air, without regard to whether or not the claims are at least potentially covered. (See Waller v. Truck Ins. Exchange, Inc. [(1995) 11 Cal.4th 1,] 19 [ 44 Cal.Rptr.2d 370, 900 P.2d 619.1; Gray v. Zurich Insurance Co., supra, 65 Cal.2d at p. 275.) As stated, the duty to defend goes to any action seeking damages for any covered claim. If it went to an action simpliciter, it could perhaps be taken to reach the action in its entirety. But it does not. Rather, it goes to an action seeking damages for a covered claim. It must therefore be read to embrace the action to the extent that it seeks such damages. So read, it accords with the general rule, set out above, that the insurer has a duty to defend as to the claims that are at least potentially covered, but not as to those that are not. (See 14 Couch on Insurance [(2d ed. 1982)] § 51:47, p. 482; Annot. [(1955)] 41 A.L.R.2d 434,] 435.) Even if the policy's language were unclear, the hypothetical insured could not have an objectively reasonable expectation otherwise.
"That being said, we can, and do, justify the insurer's duty to defend the entire `mixed' action prophylactically, as an obligation imposed by law in support of the policy. To defend meaningfully, the insurer must defend immediately. ( Montrose Chemical Corp. v. Superior Court [(1993)] 6 Cal.4th [287,] 295 [ 24 Cal.Rptr.2d 467, 861 P.2d 1153].) To defend immediately, it must defend entirely. It cannot parse the claims, dividing those that are at least potentially covered from those that are not. To do so would be time consuming. It might also be futile: The `plasticity of modern pleading' ( Gray v. Zurich Insurance Co., supra, 65 Cal.2d at p. 276) allows the transformation of claims that are at least potentially covered into claims that are not, and vice versa. The fact remains: As to the claims that are at least potentially covered, the insurer gives, and the insured gets, just what they bargained for, namely, the mounting and funding of a defense. But as to the claims that are not, the insurer may give, and the insured may get, more than they agreed, depending on whether defense of these claims necessitates any additional costs." ( Buss, supra, 16 Cal.4th at pp. 48-49, fn. italics omitted, italics added.)

In Prichard v. Liberty Mutual Ins. Co. (2000) 84 Cal.App.4th 890 [ 101 Cal.Rptr.2d 298] ( Prichard), this court confronted a variation of the argument the insured makes here. And rejected it. In Prichard, the policyholder reneged on a deal to bail out a yogurt company, and was sued on a variety of contract-based causes of action, with one exception: a cause of action for defamation based on the policyholder's alleged statement that the head of the yogurt company was stealing from it. Because the cause of action for defamation at first appeared potentially covered, the insurance company paid the costs of the entire action (including the contract claims that were not potentially covered). The judgment was for over $200,000 on the defamation claim and $1.35 million on nondefamation claims. The judgment was quickly appealed. ( Id. at pp. 896-898.)

After the trial and prior to the appeal, however, the insurance company's counsel wrote the policyholder a letter about a contemplated appeal. In that letter, counsel took the position that the insurance company did not owe any further defense of the suit, because the evidence at trial showed the first publication of the defamation was prior to the inception of the insurance company's policy — hence, no coverage potentiality at all. However, the insurance company also said it would pay a "`proportionate share'" of the costs of appellate counsel (there were other insurance companies sharing the misery) but — and this is why Prichard is important to the case before us — the payment of appellate defense costs would be subject to a "`strict reservation of rights.'" ( Prichard, supra, 84 Cal.App.4th at p. 899.) And that strict reservation of rights included "`the right to seek reimbursement from the insureds for all of the defense fees paid on behalf of the insured from the date upon which the undisputed evidence adduced at the trial . . . showed that there was no potential for coverage. . . .'" ( Ibid.)

Afterwards, the case was settled without the insurance company's consent and, to make a long story less long, the policyholder sued the insurance company (and others) for bad faith and related causes of action. The trial court would soon conclude that the reservation of rights letter had improperly tried to "`condition the retention'" of appellate counsel on its right of reimbursement, which the trial court further concluded was a breach of the insurance company's "`contractual duty to defend'" the policyholder in the appeal. ( Prichard, supra, 84 Cal.App.4th at pp. 899-900.) Based on that conclusion, the trial court concluded that the insurance company had forfeited its right in the bad faith cause of action to even argue that the defamation claim was based on a publication made before the inception of the policy. Consequently, the trial judge gave judgment to the policyholder for the amount of the defamation award. But the trial judge also determined that the insurance company had acted reasonably and thus was not subject to bad faith tort damages. ( Ibid.)

In light of the trial judge's determination that the insurance company had acted reasonably in reserving its rights, the policyholder, still seeking bad faith damages, sought a new trial, and, remarkably, the trial court judge granted it. The case came to us on the insurance company's appeal from the new trial order. ( Prichard, supra, 84 Cal.App.4th at pp. 899-900.)

We first determined that the insurance company was in the wrong as to the termination of its duty to defend the appeal.⁴³ However, even though the insurance company's reservation of rights was effectively in vain (it was reserving rights to recoup that didn't exist because, under the contract, it still had to keep defending), we rejected the argument that the insurance company had acted in bad faith in reserving its rights.

43 Our reasoning was a corollary of the potentiality rule: Since it was still possible that facts after remand after an appeal would show that the first publication of the defamation was made inside the policy period, the defense obligation continued. (See Prichard, supra, 84 Cal.App.4th at pp. 903-904.)

Why? Because the right of an insurance company to obtain reimbursement is not contractual. Here are the relevant passages from Prichard: "The trial court bought Prichard's argument that the right of an insurer to obtain reimbursement in a mixed action is, at root, contractual, so that there must be new consideration, in addition to that provided by the contract, to support a reimbursement claim. When Buss was handed down, however, it articulated a rule squarely to the contrary. Buss held that an insurer's right to seek reimbursement is implied in law, as a counterbalance to an insured's right, also implied in law, to have a defense of the whole of an action, [¶] . . . [¶] . . . The whole point of the Buss case is that an insurer's right to reimbursement for defense costs for never-even-potentially-covered-claims is predicated on a legal right `implied in law as quasi-contractual,' not a matter of any agreement between the parties. As the Buss majority stated: `The insurer therefore has a right of reimbursement that is implied in law as quasi-contractual, whether or not it has one that is implied in fact in the policy as contractual.' ( Buss, supra, 16 Cal.4th at p. 51.) [¶] That is the reason Justice Kennard dissented. She believed that unless the right to reimbursement is in the written contract, i.e., exists as a matter of agreement, not law, it shouldn't exist." (Prichard, supra, 84 Cal.App.4th at pp. 904-905, some italics fns. omitted, italics added.)

There is no way that the insured's position here can be squared with Prichard, or Buss for that matter. Given that the insurer's assertion of reimbursement rights in the case before us was necessarily not a matter of contract, the assertion of a reimbursement right by the insurer could not deprive the insured of any "benefits" under the contract. It could only deprive the insured of benefits that the law gave it independent of contract.

So — how does the insured attempt to distinguish Buss? (It makes no attempt to distinguish Prichard.) It quotes two sentences from a passage on page 50 of Buss, which state that an insurance company "may not proceed by means of a `reservation' of its `right' of reimbursement" when claims are potentially covered, because in such a situation it "simply has no such `right' to reserve."⁴⁴

44 Here are the two and one-half sentences, as they appear in the respondent's brief: "[T]he insurer may not proceed by means of a `reservation' of its `right' of reimbursement. It simply has no such `right' to `reserve.' That is true even if the insured agrees to the `reservation.' The creation of a right of reimbursement would amount to a pro tanto supersession of the policy — which would require a separate contract supported by separate consideration."

While the thought is not developed beyond that, one can extract from these quotations this syllogism: Because the MacKinnon case ultimately showed that the district's lawsuit against the insured contained potentially covered claims, the insurer was ipso facto precluded from trying to reserve its right to obtain reimbursement.

We quote the entirety of the passage containing these two sentences from Buss in the margin (which begins on page 49 and ends on page 52 of the official reporter) editing out only the citations, to show that, in context, the Supreme Court was not laying down a rule to the effect that a reservation of a right to obtain reimbursement is itself a wrongful act in cases where subsequent case law may show that (as indeed is the case here) the reservation was necessarily doomed.⁴⁵ The Buss court was, at that point in the opinion, simply saying when a claim for reimbursement is, or is not, appropriate. The key word is "seek," in the sense of actually filing an action for reimbursement, as distinct from merely "reserving" the future possibility of an action for reimbursement. Readers will also note that the whole passage emphasizes that an insurance company's claim for reimbursement is not rooted in the insurance contract itself, but is "implied in law as quasi-contractual," the same as if you pay off a $50 debt with a $100 bill, you have a right implied in law to get $50 in change back. ( Buss, supra, 16 Cal.4th at p. 51.) (That's the very example given in Buss.)

45 Here is the passage, copied in its entirety except for the citations:
"We now turn to the first question — In a `mixed' action, may the insurer seek reimbursement from the insured for defense costs?
"The answer is as follows.
"As to the claims that are at least potentially covered, the insurer may not seek reimbursement for defense costs. Apparently, none of the decisional law considering such claims in and of themselves suggests otherwise.
"The reason is this. Under the policy, the insurer has a duty to defend the insured as to the claims that are at least potentially covered. With regard to defense costs for these claims, the insurer has been paid premiums by the insured. It bargained to bear these costs. To attempt to shift them would upset the arrangement. [Citation.] This would not be the case if the policy itself provided for reimbursement: such a policy would qualify itself. It would also not be the case if there were a separate contract supported by separate consideration: Such a contract would supersede the policy pro tanto. Otherwise, however, the insurer may not seek reimbursement. Surely, it does not have a right of reimbursement implied in fact in the policy, having bargained to bear the costs in question. Neither does it have such a right implied in law. Under the law of restitution, a right of this sort runs against the person who benefits from `unjust enrichment' and in favor of the person who suffers loss thereby. [Citations.] Any `enrichment' of the insured by the insurer through the insurer's bearing of bargained-for defense costs is consistent with the insurer's obligation under the policy and therefore cannot be deemed `unjust.' It follows a fortiori that the insurer may not proceed by means of a `reservation' of its `right' of reimbursement. It simply has no such `right' to `reserve.' That is true even if the insured agrees to the `reservation.' The creation of a right of reimbursement would amount to a pro tanto supersession of the policy — which would require a separate contract supported by separate consideration. [Citation.]
"As to the claims that are not even potentially covered, however, the insurer may indeed seek reimbursement for defense costs. Apparently, all the decisional law considering such claims in and of themselves so assumes. [Citations.] So has it been held: `California law clearly allows insurers to be reimbursed for attorney's fees' and other expenses `paid in defending insureds against claims for which there was no obligation to defend.' [Citation.]
"The reason is this. Under the policy, the insurer does not have a duty to defend the insured as to the claims that are not even potentially covered. With regard to defense costs for these claims, the insurer has not been paid premiums by the insured. It did not bargain to bear these costs. To attempt to shift them would not upset the arrangement. [Citation.] The insurer therefore has a right of reimbursement that is implied in law as quasi-contractual, whether or not it has one that is implied in fact in the policy as contractual. As stated, under the law of restitution such a right runs against the person who benefits from `unjust enrichment' and in favor of the person who suffers loss thereby. The `enrichment' of the insured by the insurer through the insurer's bearing of unbargained-for defense costs is inconsistent with the insurer's freedom under the policy and therefore must be deemed `unjust.' It is like the case of A and B. A has a contractual duty to pay B $50. He has only a $100 bill. He may be held to have a prophylactic duty to tender the note. But he surely has a right, implied in law if not in fact, to get back $50. Even if the policy's language were unclear, the hypothetical insured could not have an objectively reasonable expectation that it was entitled to what would in fact be a windfall. [Citations.] Whatever hopes such an insured may have had based on the decisional law's imposition on the insurer of a duty to defend the entire `mixed' action would have been dispelled by that same law's assumption that the insurer may seek reimbursement for the kind of costs identified above. [Citations.]" ( Buss, supra, 16 Cal.4th at pp. 49-52, fns. italics omitted.)

We should note, in this regard, that the time when a reservation of rights is asserted may make a difference. For example, a reservation to seek recoupment certainly is not bad faith when it is made at a time when the insurance company may reasonably conclude that it is expending costs on behalf of the policyholder for not even potentially covered claims, even if a later, subsequent decision of the Supreme Court rendered the right to recoup nothing but a fantasy.

In the case before us, for example, the reservation was issued in September 2000, and withdrawn (or mostly withdrawn — we will come to the AIG problem in a moment) in October 2002. The October withdrawal was easily well in advance of the August 2003 Supreme Court MacKinnon decision. Thus, up to August 2003, the insurer here could, quite reasonably, assert the legal position that there was no claim in the district's suit against the insured that was even potentially covered.

Consider this: If, for example, our Supreme Court had been persuaded by the plain language rationale articulated by the Supreme Court of Wisconsin in Peace ex rel. Lerner v. Northwestern Nat'l Ins. Co., supra, 228 Wis.2d 106, the insurer here might have had a viable claim for reimbursement extending beyond August 2003. Only by August 2003 could one say with reasonable certainty that the insured had been provided with a defense and indemnity owed under the contract, as distinct from a defense not owed under the contract.

In short, the "may" and "may not" in Buss's statements on pages 49 and 50 of the official reporter about when an insurance company may or may not seek reimbursement must be read in the context of the reasonable viability of the insurance company's legal position at the time of the reservation. Here, even after the reservation was issued in September 2000, the future would still see at least two trial judges, and a majority of justices on two appellate panels, who would agree with the insurer's position on the total pollution exclusion.

Any other reading of the Buss language would mean that an insurance company could not reserve any right at all to recoup what it was never contractually obligated to pay for in the first place unless there was a California Supreme Court case directly on point and favoring its position (and even if the high court had not yet spoken but there were California appellate cases or decisions from supreme courts of other states supporting the insurance company's position). That is an absurdity which finds no support in the Buss decision.

The AIG thread remains: The insured here notes that the pre-MacKinnon withdrawal of the reservation of rights in October 2002 did not expressly withdraw the insurer's right to seek reimbursement from AIG, the excess insurance company. The insured posits a scenario whereby the insurer might have sued AIG for reimbursement of defense costs (ironically, AIG had fronted those costs in the first place!); had the insurer been successful in that suit, AIG might have sued the insured for the insured's $50,000 deductible with the AIG policy. The insured's theory is thus that the insurer here obstinately clung to the hope of reimbursement from the excess insurance company to the detriment of the insured, even after the MacKinnon case ended any reasonable hope of reimbursement.

Again, the theory is at odds with the facts of the case and what the Buss decision actually said about reimbursement. Most basically, Buss said that in order to have any claim for reimbursement, the insurance company must — the Buss court italicized the word must — affirmatively reserve its right to seek it: "We note that the Court of Appeal assumed that, in order to obtain reimbursement for defense costs, the insurer must reserve its right thereto. To the extent that this right is implied in law as quasi-contractual, it must indeed be reserved. [Citation.] Through reservation, the insurer gives the insured notice of how it will, or at least may, proceed and thereby provides it an opportunity to take any steps that it may deem reasonable or necessary in response — including whether to accept defense at the insurer's hands and under the insurer's control [references to earlier footnotes] or, instead, to defend itself as it chooses. To the extent that this right is implied in fact in the policy as contractual, it should be reserved. Through reservation, the insurer avoids waiver." ( Buss, supra, 16 Cal.4th at p. 61, fn. 27, italics omitted, except for the word "must," which was in the original.)

This passage from Buss shows us that it is not enough, if an insurance company is to have any viable claim for reimbursement, simply to craft an incomplete withdrawal of a previous reservation. Such an incomplete withdrawal would not give the policyholder proper "notice" of how the insurance company "will, or at least may, proceed" and thereby give the policyholder the "opportunity to take any steps that it may deem reasonable or necessary in response." ( Buss, supra, 16 Cal.4th at p. 61, fn. 27; see also Scottsdale, supra, 36 Cal.4th at pp. 649, 662 [emphasizing necessity that insurance company "properly" reserve its right to seek reimbursement].)

Here, given that there is no evidence in this case that the insurer ever affirmatively attempted to reserve its rights to obtain reimbursement from AIG, the idea that somehow the insurer was still reserving some claim to reimbursement from AIG that would ultimately redound to the insured's detriment is but a lawyer's trick to impute to the insurer actions it never took.

Or to put it more plainly: After the October 2002 withdrawal of its reservation of rights to seek reimbursement from the insured, the insured had nothing reasonably to worry about, even from AIG.

3. Miscellaneous Brandt Fee Issue

Finally, what about the fees — $8,921.40 — that the insured had incurred in its bad faith case against the insurer up to the time of settlement of the underlying suit and payment of all of its defense costs? Can we at least uphold the judgment to this extent?

No. As explained in part III.A., ante, tort damages against an insurance company are premised on the unreasonable denial of policy benefits. When an insurance company acts reasonably, though incorrectly, only contract damages are available. And Brandt fees are clearly tort damages. (See Cassim v. Allstate Ins. Co., supra, 33 Cal.4th at p. 806 [explaining that when an "insurer fails to act fairly and in good faith . . . such breach may result in tort liability for proximately caused damages" including Brandt fees]; Essex Ins. Co. v. Five Star Dye House, Inc. (2006) 38 Cal.4th 1252, 1257-1258 [ 45 Cal.Rptr.3d 362, 137 P.3d 192] [explaining that Brandt fees are damages as a result of an insurance company's tortious conduct].) Thus, in Cassim, the court observed: "Of course, without a tort judgment, there could be no Brandt fees." ( Cassim, supra, 33 Cal.4th at p. 808.) And in this case, there is no tort judgment.

IV. CONCLUSION AND DISPOSITION

A proper damage award becomes a hard-fought case when the evidence and law supports it. And this certainly has been a hard-fought case. Even so, there is much amiss, most of it tracing from one fatal decision by the trial court, namely to grant a motion in limine holding as a matter of law that the insurer had been unreasonable. As it turned out, the insurer was reasonable.

Accordingly: The judgment is reversed with directions that Griffin Dewatering Company take nothing by way of its complaint. Northern Insurance Company of New York will recover its costs on appeal.

Bedsworth, J., and Aronson, J., concurred.