Stonewall Insurance v. City of Palos Verdes Estates

46 Cal. App. 4th 1810 (1996)

STONEWALL INSURANCE COMPANY, Plaintiff, Cross-defendant and Appellant,
v.
CITY OF PALOS VERDES ESTATES, Defendant, Cross-defendant and Appellant; ADMIRAL INSURANCE COMPANY et al., Defendants, Cross-complainants and Appellants; CANADIAN INDEMNITY COMPANY, Defendant, Cross-complainant and Respondent; FIREMAN'S FUND INSURANCE COMPANY et al., Defendants, Cross-defendants and Respondents. CITY OF PALOS VERDES ESTATES, Plaintiff, Cross-defendant and Appellant;
v.
STONEWALL INSURANCE COMPANY, Defendant, Cross-defendant and Appellant; PURITAN INSURANCE COMPANY, Defendant, Cross-complainant and Appellant; FIREMAN'S FUND INSURANCE COMPANY et al., Defendants, Cross-defendants and Respondents.

Docket Nos. B023805, B045183.

Court of Appeals of California, Second District, Division Seven.

June 19, 1996.

COUNSEL

Wilson, Kenna & Borys and Robert L. Wilson for Plaintiff, Cross-defendant and Appellant in No. C439984 and for Defendant, Cross-defendant and Appellant in No. SWC66204.

Wise, Wiezorek, Timmons & Wise, George E. Wise, Michael J. Pearce, Kane, Ballmer & Berkman, Stephanie Scher, Burke, Williams & Sorensen, Douglas C. Holland, Brian C. Pierik for Defendant, Cross-defendant and Appellant in No. C439984 and for Plaintiff, Cross-defendant and Appellant in No. SWC66204.

Horvitz & Levy, Mitchell C. Tilner, Peter Abrahams, Hillsinger & Costanzo, John H. Walsh, Provizer, Lichtenstein, Pearlman & Phillips, Noel F. Beck, Bell & Weissman and Garland O. Bell, Jr., for Defendants, Cross-Complainants and Appellants in No. C439984.

Cotkin, Collins & Franscell, Joan M. Dolinsky and Lori S. Blitstien for Defendant, Cross-complainant and Respondent in No. C439984.

Nelsen, Tang, Thompson, Pegue, Thornton & Spurling, Knapp, Petersen & Clarke and K. Stephen Tang for Defendant, Cross-complainant and Appellant in No. SWC66204.

Black, Compean, Hall & Lenneman, Robert H. Black, Michael D. Compean, Clausen & Campbell, Robert H. Black, Michael D. Compean, Sonnenschein Nath & Rosenthal, Paul E.B. Glad, Thomas Holden, Kaufman & Logan, Jeffrey Kaufman, Haight, Brown & Bonesteel, Harold Hansen Brown, Gary C. Ottoson, Roy G. Weatherup, Michael J. Leahy, J.R. Seashore, Denis J. Moriarty, Jules Solomon Zeman, Caitlin Doyle Berfield, Roper & Folino, John B. Larson, Joseph L. Stark, Keesal, Young & Logan, Robert D. Feighner, E. Scott Douglas, Sedgwick, Detert, Moran & Arnold, Scott M. Kolod, O'Melveny & Myers, John G. Niles and Martin S. Checov for Defendants, Cross-defendants and Respondents in Nos. C439984 and SWC66204.

OPINION

GOLD, J.^[*] —

I. APPELLATE BACKGROUND

This division filed its decision in the within cause on May 15, 1992, and modified it on June 11, 1992, on denial of rehearing. Said decision suggested that in light of the then recent granting of review by the California Supreme Court of Montrose Chemical Corp. v. Admiral Ins. Com.^[†] (Cal. App.), grant of review of this case by the Supreme Court would contribute to certainty of California law.

On August 27, 1992, the Supreme Court did grant review of this case.

On July 3, 1995, the Supreme Court filed its decision in the Montrose case: Montrose Chemical Corp. v. Admiral Ins. Com. (1995) 10 Cal. 4th 645 [42 Cal. Rptr. 2d 324, 897 P.2d 1]. The Supreme Court's decision in the Montrose case is hereinafter referred to as "Montrose."

This case is again before this court by virtue of an order of the Supreme Court filed October 19, 1995. By that order, the Supreme Court transferred the cause to this court with directions to vacate our prior decision herein and to reconsider the cause in light of Montrose.

II. THE PROCEEDINGS BELOW

A. The Underlying Action (the Papworth case)

In an underlying action, Michael T. Papworth (Papworth) sued the City of Palos Verdes Estates (the City), claiming that as a consequence of a continuous and repeated course of conduct of the City from 1971 to 1981 (namely, improper design and maintenance of a City storm drain adjoining property purchased by Papworth in August of 1971), Papworth's property was damaged and ultimately became worthless. Papworth's complaint sounded in negligence, nuisance and inverse condemnation.

The jury awarded Papworth $1,188,791.57 as damages for negligence and nuisance and $1,881,946.70 as damages for inverse condemnation. Judgment was entered for $1,881,946.70. Pending appeal, the underlying action was settled by payment of $1.6 million together with a stipulation (confirmed in an order of court) vacating the judgment "for all purposes." Of the $1.6 million settlement, $350,000 was paid by the City, $300,000 by The Jefferson Insurance Company of New York (Jefferson) and $950,000 by Stonewall Insurance Company (Stonewall). Other insurers of the City refused to contribute toward the settlement.

B. The Two Cases on Appeal

The appeals that are the subject of the within decision arise in the two cases that followed settlement of the underlying action.

The City suit: In the first of these two cases (Los Angeles County Superior Court case No. SWC66204, hereinafter referred to as the City suit), the City sued all insurers who had issued primary or excess liability policies to it from 1971 through 1983. In the City suit the City sought (i) on a breach of contract theory, to recover the $350,000 it paid to settle the Papworth claim, less a $1,000 deductible; and (ii) damages against certain of its insurers for bad faith failure to settle the Papworth claim.

The Stonewall suit: In the second case (Los Angeles County Superior Court case No. C439984, hereinafter referred to as the Stonewall suit), Stonewall sued the City and the other insurers who issued liability insurance to the City during the period of exposure to damage resulting from the City's conduct, seeking a return of the $950,000 Stonewall paid toward the settlement of the underlying action, claiming that it was not liable for any of that sum. Its complaint and the cross-complaints filed by the other insurers also seek a declaration of the relative liability of each of the insurers and apportionment of that liability among the insurers.

The trial court initially consolidated the two cases. Ultimately it bifurcated the proceedings and deferred action on what it denominated as "Phase II" (the City's bad faith claims) until after determination of "Phase I" (all of the other issues). We deal herein with appeals from a summary judgment order in favor of one insurer (Canadian Indemnity Company, hereinafter Canadian) and purported appeals from certain of the trial court's subsequent judgments in Phase I. Phase II has not yet been tried.

After the consolidated cases had been bifurcated, the trial court heard Canadian's motion for summary judgment and on November 19, 1986, signed an order granting that motion.

The trial of Phase I thereafter ensued. Following opening statements, the trial court granted nonsuit motions of Central National Insurance Company of Omaha (Central National) and Employers Reinsurance Corporation (Employers). The trial court also granted a motion of Covenant Mutual Insurance Company (Covenant) for judgment pursuant to Code of Civil Procedure section 631.8. Separate judgments in favor of Covenant and Employers were filed on September 28, 1988. A separate judgment in favor of Central National was filed on June 12, 1989. At the end of the City's case-in-chief, the trial court granted the motions of Fireman's Fund Insurance Companies (Fireman's) and Central Mutual Insurance Company (Central Mutual) for judgment pursuant to Code of Civil Procedure section 631.8 on the ground that their policies covered only periods prior to September 2, 1976, and that claims for damage occurring prior to September 2, 1976, were barred by Government Code section 911.2.

Upon the completion of the trial of Phase I, the trial court held that the City's liability was covered by insurance except for $53,000 in deductibles and self-insured retention; and it imposed that liability jointly and severally upon Jefferson and Admiral Insurance Company (Admiral). Jefferson and Admiral had issued policies the trial court deemed primary for the period from November 1, 1975, to July 1, 1980. Stonewall was exonerated from liability and adjudged entitled to recover from Jefferson and Admiral the $950,000 it had paid. The trial court reasoned that as an excess carrier, Stonewall had no obligation to indemnify the City because primary carriers Jefferson and Admiral in the aggregate had coverage in amounts sufficient to fully indemnify the City. All other excess carriers were similarly exonerated. Primary carriers who had issued policies covering periods after February 1, 1980, were exonerated because of the trial court's conclusion that a stipulation in the Papworth case fixing February 1, 1980, as the date of taking for inverse condemnation purposes also had the effect of fixing the date when the Papworth property was a total loss. On August 16, 1989, judgment (hereinafter sometimes referred to as the Phase I judgment) was entered accordingly.

III. SUMMARY OF OUR CONCLUSIONS

The following is a summary of our conclusions concerning the Phase I judgment and the summary judgment in favor of Canadian:

1. This case presents the issue of what event or events activate or "trigger" the obligations of the respective insurers to indemnify the City. We conclude that the "continuing injury" trigger of coverage applied in Montrose should also be applied in this case — that is, that there was a "continuing injury" (a continuous "occurrence," using the language of the policies) throughout the period from the beginning of the damage to the Papworth property for which the City was at fault until that damage became complete — so that all insurers whose policies were in force during any portion of that period covered the loss to the city arising out of that damage.

2. The fact that the Papworth judgment was vacated does not render the inverse condemnation exclusions in the insurance policies issued to the City either all-important or of no significance. An apportionment of the $1.6 million settlement between inverse condemnation liability and negligence/nuisance liability must be made. All carriers other than Jefferson and possibly Fireman's exclude coverage for inverse condemnation and therefore do not cover the inverse condemnation proportion of the $1.6 million settlement. Jefferson cannot invoke its inverse condemnation exclusion because it waited too long to assert that exclusion. The efficacy of Fireman's exclusion will have to be addressed again by the trial court because the trial court improperly granted Fireman's Code of Civil Procedure section 631.8 motion.

3. The trial court erred in granting Canadian's summary judgment motion. It improperly determined that no damage occurred to the Papworth property after February 1, 1980 (a date prior to the inception of the Canadian policy), based upon its conclusion that a stipulation as to the date of taking for inverse condemnation purposes (a date prior to the inception of the Canadian policy period) was binding on all parties to these consolidated cases even though the stipulation was entered into only between the City and Papworth. There was a factual showing of substantial damage to the Papworth property in 1981, after the inception of the Canadian policy, thereby creating a triable issue of material fact and requiring that Canadian's motion be denied.

4. The trial court erred in exonerating carriers who issued policies providing only coverage for periods prior to September 2, 1976. Although Government Code section 911.2 required that Papworth file a claim against the City not later than one year of the accrual of his cause of action, and although Papworth did not file a claim against the City until September 2, 1977, section 911.2 did not bar Papworth's claim against any carrier because the aforementioned one-year period did not begin to run until the situation as to the Papworth property had stabilized — a date clearly after Papworth filed his claim.

5. Because of the error just described in applying Government Code section 911.2, the trial court did not determine the date on which damage to the Papworth property for which the City was culpable first occurred. It will be necessary for the trial court to make such a determination — or at least to determine during which primary carrier's policy period such damage first occurred. Also, because (as mentioned above) the trial court erroneously determined that no such damage occurred after February 1, 1980, it will be necessary for the trial court to determine the latest date on which such damage occurred — or at least to determine during which primary carrier's policy period such damage last occurred. Once such determinations have been made, if (as seems extremely likely) the aggregate of the policy amounts of the primary policies in effect from the beginning of such damage to the end of such damage is adequate to cover the City's obligation to Papworth, then no excess carrier is liable for any of that obligation.

6. Admiral was a primary carrier.

7. Coverage is not eliminated by the "known loss" rule, sometimes called the "loss-in-progress" rule (Ins. Code, §§ 22 & 250).

8. Coverage is not eliminated by the policies' "expected or intended" limitation.

9. The Jefferson policy covered three separate years, each with a separate $300,000 limit and a separate $1,000 deductible — so that Jefferson had, subject to an aggregate of $3,000 in deductibles, an aggregate of $900,000 in coverage for the continuing injury in the case at bar.

10. The City is entitled to recover at least a portion of the $350,000 it paid to Papworth (plus interest thereon) from Jefferson and at least another portion from Admiral. The specific amounts which Jefferson and Admiral (and any other insurer from whom the City is entitled to recover) are to pay the City are to be determined by giving effect to the policy limits and the deductible, retention and "other insurance" clauses of the policies of those carriers from whom the City is entitled to recover. The obligations of those carriers to reimburse the City are not necessarily joint and several.

11. The trial court shall allocate among the carriers themselves the burden of the loss ($1.6 million, possibly less deductibles and/or retentions), plus interest, according to a method determined by the trial court to be fair and reasonable. The trial court shall use what we hereinafter call the "qualified time on the risk" method unless another method would be more equitable.

12. In Phase II the trial court is to consider the claim of the City that some of the insurers are now barred from asserting any exclusions from coverage because they did not take the opportunity to settle the Papworth claim for an amount less than the actual award.

IV. THRESHOLD ISSUES

A. Appealability

1. The Order for Summary Judgment in Favor of Canadian

(1) On December 2, 1986, Puritan Insurance Company filed a notice of appeal from the November 19, 1986, order granting Canadian's motion for summary judgment. While the record before us discloses no judgment in favor of Canadian ever entered on its summary judgment motion, the November 19, 1986, order is appealable because it disposes of all claims by and against Canadian raised in either of the consolidated cases. Decisional law clearly establishes that such an order is appealable. (Justus v. Atkinson (1977) 19 Cal. 3d 564, 568 [139 Cal. Rptr. 97, 565 P.2d 122]; Etienne v. DKM Enterprises, Inc. (1982) 136 Cal. App. 3d 487, 489 [186 Cal. Rptr. 321].)

On October 20, 1986, before the Puritan Insurance Company notice of appeal was filed and even before the November 19, 1986, "attorney order" granting Canadian's motion was filed, Admiral filed a notice of appeal. That notice of appeal purported to appeal "from the judgment rendered on September 25, 1986 ... in favor of defendant, cross-defendant and cross-complainant CANADIAN INDEMNITY COMPANY and against plaintiff STONEWALL INSURANCE COMPANY and all remaining defendants, cross-defendants and cross-complainants...." As noted above, there is no indication in the record that any such judgment was ever rendered, and certainly no indication that any such judgment was rendered on September 25, 1986. What did happen on September 25, 1986, was that the trial court, upon hearing Canadian's summary judgment motion, issued a minute order granting that motion but directing that a formal "attorney order" thereafter be prepared and submitted by Canadian. (In fact, after Canadian submitted such an "attorney order" — the one the trial court eventually signed on November 19, 1986, objections and counterproposals and responses thereto were filed by various parties before the trial court finally elected to sign the proposed "attorney order" Canadian initially submitted.) The September 25, 1986, minute order, not being a final order, is of course nonappealable; and the October 20, 1986, notice of appeal was premature as of the date of its filing. However, in view of the fact that Admiral's notice of appeal obviously was intended to apply to the November 19, 1986, "attorney order" ultimately filed and in view of the fact that the parties have briefed Admiral's appeal from the granting of Canadian's summary judgment motion on the merits without raising any objection to Admiral's notice of appeal during the ensuing nine and one-half years, the requisite good cause appears; and we shall treat Admiral's notice of appeal as filed immediately after the November 19, 1986, order (see Cal. Rules of Court, rules 2(c), 2(d)) and as directed toward that order (see, e.g., Smith v. Smith (1954) 126 Cal. App. 2d 194, 195 [272 P.2d 118]; Seven Up Bottling Co. v. Grocery Drivers Union (1950) 97 Cal. App. 2d 623, 624 [218 P.2d 41]).

2. The Judgments Following Nonsuits

(a) The Judgment in Favor of Covenant

No party has challenged the judgment in favor of Covenant before this court, so we shall leave that judgment undisturbed.

(b) The Judgment in Favor of Central National

As noted above, the judgment in favor of Central National was entered on June 12, 1989. That date is less than 180 days before the City's notice of appeal herein was filed and less than 180 days before Stonewall's notice of cross-appeal herein was filed. However, no notice of appeal filed herein refers to the June 12, 1989, judgment. Stonewall's notice of cross-appeal expressly appealed from "those portions of the judgment rendered on August 16, 1989 ... declaring judgment in favor of ... CENTRAL NATIONAL INSURANCE COMPANY OF OMAHA ... against STONEWALL INSURANCE COMPANY and all prior rulings incorporated therein and superceded thereby." (Italics added.) However, the August 16, 1989, judgment did not by its terms "incorporate" or "supercede" the June 12, 1989, judgment. This is not a situation in which it is appropriate to construe a notice of appeal as relating to a judgment or order other than the one it purports to appeal from. Accordingly, to the extent that any party to the proceedings before us claims that an appeal has been perfected from the June 12, 1989, judgment, we reject such claim and dismiss any such appeal.^[1],^[2]

(c) The Judgment in Favor of Employers

The judgment in favor of Employers was entered on September 28, 1988. It became final long before the filing of any notice of appeal or notice of cross-appeal that could conceivably be construed as relating to it. Accordingly, to the extent that any party to the proceedings before us claims that an appeal has been perfected from the September 28, 1988, judgment, we also reject such claim and dismiss any such appeal.^[3],^[4]

3. The Phase I Judgment

Prior to issuing our 1992 opinion herein, we directed a letter to all parties requesting additional briefing on the question of the appealability of the Phase I judgment in view of the fact that Phase II of these consolidated cases has not as yet been tried. It appeared to us that it was at least arguable that the single judgment rule might bar an appeal herein until Phase II also had gone to judgment. All parties who responded argued that the appeal should go forward.

In our 1992 opinion we held as a threshold matter that the trial court did not abuse its discretion in severing for a later trial the bad faith claims. Next, based upon authority then in existence (Guntert v. Stockton (1974) 43 Cal. App. 3d 203 [117 Cal. Rptr. 601] and Highland Development Co. v. City of Los Angeles (1985) 170 Cal. App. 3d 169 [215 Cal. Rptr. 881]), we concluded in our 1992 opinion that the one final judgment rule did not bar appeal from a judgment in a properly severed portion where hearing the appeal on the merits contributes to judicial efficiency or avoids hardship to the parties. We held that the Phase I judgment was appealable as to those of the insurers not involved in Phase II, and we further concluded that resolution of the appeal as to all parties (even those involved in Phase II) rather than piecemeal as to some only would promote judicial economy. Accordingly, we proceeded to hear the appeal from the Phase I judgment as to all parties to that judgment.

However, almost two years to the day after our 1992 opinion herein was filed, the California Supreme Court decided Morehart v. County of Santa Barbara (1994) 7 Cal. 4th 725 [29 Cal. Rptr. 2d 804, 872 P.2d 143] (hereafter sometimes Morehart), specifically disapproving both Highland Development Co., supra, 170 Cal. App. 3d 169, and two cases that cited Guntert, supra, 43 Cal. App. 3d 203, (Day v. Papadakis (1991) 231 Cal. App. 3d 503 [282 Cal. Rptr. 548]; and DeGrandchamp v. Texaco, Inc. (1979) 100 Cal. App. 3d 424 [160 Cal. Rptr. 899]). (2a) Morehart held that "... an appeal cannot be taken from a judgment that fails to complete the disposition of all of the causes of action between the parties even if the causes of action disposed of by the judgment have been ordered to be tried separately, or may be characterized as `separate and independent' from those remaining." (Morehart, supra, 7 Cal.4th at p. 743.)

Morehart did not purport to overturn the judicially created exception to the one final judgment rule that permits appeal from a judgment in a multiparty case determining all issues between certain parties even though issues remain to be resolved between other parties. (See, e.g., Johnson v. Hayes Cal Builders, Inc. (1963) 60 Cal. 2d 572, 578 [35 Cal. Rptr. 618, 387 P.2d 394]; Aetna Cas. etc. Co. v. Pacific Gas & Elec. Co. (1953) 41 Cal. 2d 785, 788-789 [264 P.2d 5, 41 A.L.R. 2d 1037]; Nicholson v. Henderson (1944) 25 Cal. 2d 375, 379-381 [153 P.2d 945].) Unfortunately, even though our prior opinion alluded to finality and therefore appealability as to insurers who are not involved in Phase II, upon closer scrutiny there appear to be no such insurers. The cross-complaint of Puritan Insurance Company in the City suit names as a cross-defendant every insurer who is an appellant or respondent in the within appeal from the Phase I judgment; and that cross-complaint seeks, among other things, contribution toward and indemnity against the claims asserted by the City in the City suit — claims which include those to be tried in Phase II. (3a) Accordingly, the judgment involved in the within appeal does not determine all issues between any parties to the within consolidated cases and therefore does not come within the aforementioned judicially created exception to the one final judgment rule surviving Morehart.

Therefore, in light of Morehart, we conclude that the Phase I judgment is not appealable.

B. Treating Appeal From Phase I Judgment as Petition for Writ

Notwithstanding the nonappealability of the Phase I judgment, we shall nevertheless decide the merits of that judgment. The means by which we are permitted to do so are reflected in Morehart. (2b) In concluding that the judgment in Morehart was not appealable, the Supreme Court observed that "[a] petition for a writ, not an appeal, is the authorized means for obtaining review of judgments and orders that lack the finality required by Code of Civil Procedure section 904.1, subdivision (a) [the one final judgment rule]." (Morehart v. County of Santa Barbara, supra, 7 Cal.4th at pp. 743-744.)

(3b) This case clearly meets the criteria for treating the notices of appeal as petitions for a peremptory writ of mandate. As in Morehart, "[t]he briefs and record before us contain all [of] the elements prescribed by rule 56 of the California Rules of Court for an original mandate proceeding. The functional equivalents of any necessary verifications ... are supplied ... by the certifications of the clerk's transcript by the clerk of the trial court and of the reporter's transcript by the clerk and the reporter[s]. There is no indication that the trial court as [a] respondent would wish to appear separately or become more than a nominal party to a writ proceeding." (Morehart v. County of Santa Barbara, 7 Cal.4th at pp. 745-746; see also Olson v. Cory (1983) 35 Cal. 3d 390, 401 [197 Cal. Rptr. 843, 673 P.2d 720].) Furthermore, as in Morehart, "... the case in its present posture presents unusual circumstances making it appropriate to ascertain from the record whether there are substantive errors that [we] should, by writ, order the trial court to correct" (Morehart, 7 Cal.4th at p. 746):

1. All parties who responded to our request for additional briefing on the subject of appealability prior to issuance of our 1992 opinion herein argued that the appeal of the Phase I judgment should go forward. Moreover, we ourselves, based on the state of the law as we perceived it when we issued our 1992 opinion, concluded that the Phase I judgment was appealable.

2. The merits of the issues decided by the trial court have been thoroughly briefed before appellate courts in these cases not once but three times: first, back in 1991-92 when the cases were initially before us; next, in 1992-1994 before the Supreme Court; and again, in 1995-96 before us. Further proceedings in the trial court would be unlikely to improve upon the record or briefing now before us.

3. In no other respect would judicial economy be served in this case by deferring resolution of the issues decided by the trial court on Phase I until final judgment on Phase II. To proceed through the trial of Phase II and at some point of time thereafter ascertain through the appellate process that the trial court committed reversible error in Phase I so that both Phase I and potentially Phase II need to be retried would be too kafkaesque a result. The judgment in the Papworth case is now approximately 13 1/4 years old. The Phase I judgment is now almost seven years old. The parties are entitled to know now whether the Phase I judgment is correct.

V. THE PHASE I JUDGMENT

We proceed first to a consideration of the Phase I judgment. The resolution of the appeals from the Canadian summary judgment order will depend upon principles articulated in our analysis of the Phase I judgment and therefore will be discussed later.

A. Insurance Coverage

1. Relevant Facts

The Papworth property, situated on a plateau overlooking the sea and bounded by a cliff leading to the beach, is located in the Bluff Cove area of the City. It was purchased by Papworth in August of 1971. The property adjoins a steep channel on "parkland" owned by the City, on which the City during the relevant period operated a storm drain. The evidence establishes that activity of the City beginning at least as early as 1966 in the design and maintenance of the storm drain periodically and consistently contributed both to relatively minor erosion damage to the Papworth property in 1978 and to activation of a deep-seated landslide which effectively destroyed the property in 1981. Dating from the year prior to the time of Papworth's acquisition of the property into 1979 the City periodically but negligently engaged in futile efforts to mitigate the deficiencies in its drainage system.

During the relevant period the City was insured against liability for property damage by general comprehensive liability or municipal liability policies issued by the following carriers, all parties to the consolidated actions:^[5]

(a) Fireman's, which issued a primary insurance policy for the period July 1, 1971, to July 1, 1974, with limits of $500,000 per year and an excess policy for the same period with a limit of $5 million.

(b) Central Mutual, which issued a primary policy for the period July 1, 1974, to November 1, 1975, with limits of $300,000 per occurrence and in the aggregate.

(c) Employers, which issued an excess policy for the period July 1, 1974, to July 1, 1975, with a limit of $5 million.

(d) Jefferson, which issued a primary policy for the period November 1, 1975, to November 1, 1978, with limits of $300,000 and a deductible of $1,000 per claim (which, as we explain later in this opinion, are yearly limits and deductibles).

(e) Admiral, which issued a policy for the period November 1, 1978, to July 1, 1980, with limits of $1 million per occurrence per policy year over a $50,000 per occurrence retention.

(f) Central National, which issued an excess policy for the period July 1, 1975, to November 1, 1976, with a limit of $10 million.

(g) Stonewall, which issued an excess policy for the period November 1, 1976, to November 1, 1977, with a limit of $4.7 million.

(h) Maine Bonding & Casualty Company (Maine Bonding), which issued an excess policy for the period November 1, 1977, to November 1, 1978 with a limit of $5 million.

(i) Puritan Insurance Company (Puritan), which issued an excess policy for the period November 1, 1978, to November 1, 1979, with a limit of $4 million.

(j) Canadian, which issued a policy with limits of $1 million per occurrence over a retention of $50,000 for the period July 1, 1980, to July 1, 1982.

With insignificant variations in language, all of the insurance policies included or incorporated as "following form" language drawn from forms developed by the National Bureau of Casualty Writers and the Mutual Insurance Rating Bureau in 1966 and revised in 1973 by their successor, the Insurance Services Office (ISO). (See 3 Cal. Insurance Law & Practice §§ 49:01[2], 49.04[2], pp. 49-6.1-49-7, 49-13.) The pertinent language of the Fireman's primary policy provides:

"The Company will pay on behalf of the insured all sums which the insured shall become legally obligated to pay because of ... property damage to which this insurance applies caused by an occurrence." "Property damage" is defined in the policy as "injury to or destruction of tangible property." "Occurrence" is defined as:

"[A]n accident, including injurious exposure to conditions which results, during the policy period, in ... property damage neither expected nor intended from the standpoint of the insured."

The Fireman's excess policy covering the same period as its primary policy, as well as the other insurance policies here involved, contain with insignificant variations a definition of "occurrence" different from the above quoted definition contained in Fireman's primary policy. (Fireman's makes no argument that the definition of "occurrence" contained in its primary policy should be construed differently than the definition contained in its excess policy.) As exemplified by the Jefferson policy, the term "occurrence" is defined as: "[A]n event or a continuous or repeated exposure to conditions which causes ... damage to property during the policy period that is neither knowingly nor intentionally caused by or at the direction of the insured.... All ... property damage arising out of a continuous or repeated exposure to substantially the same general conditions shall be considered one occurrence."

2. Coverage Afforded Against Third Party Property Damage Claims by Successive Insurance Policies in the Situation of Repeated Exposure to a Continuing Condition

(4) Understandably, various insurer parties argue for different triggers of coverage,^[6] each asserting the position which benefits it. At the point in time when we issued our 1992 opinion herein, the issue of which of various possible triggers applied in a context involving (a) a third party property damage claim where there were (b) successive insurance policies and the liability of the insured involved (c) continuous and repeated harms was a relatively open question. That issue has now been definitively resolved by Montrose. After discussing various possible triggers, evaluating prior case law in California and other jurisdictions, and analyzing the typical expectations of the insured and the insurers in such a context, the Supreme Court in Montrose held that a "continuing injury" trigger of coverage applies: all carriers on the risk from inception of harm to the time the loss is no longer contingent cover the risk.

It is true that Montrose involved a dispute arising at an earlier stage of the third party litigation, namely, a dispute as to whether a particular carrier had a duty to defend its insured against the third party's claims. A duty to defend of course can arise where there is merely a potential for coverage. In the cases at bar the issue is whether there is a duty to indemnify; and the test of whether there is a duty to indemnify is whether there is actual coverage, not merely a potential for coverage.

However, the Montrose court in effect decided the actual coverage question in this case in the course of determining whether there was a potential for coverage in that case. Even though the Montrose case only involved the issue of the duty to defend, the Montrose opinion stated what criteria were to be used in determining whether there was actual coverage because the determination of whether there was a duty to defend necessarily involves whether there was a potential for actual coverage. Montrose noted (in fn. 9 at 10 Cal.4th at p. 659) that although an insurer may have a duty to defend, it may ultimately be found not to have a duty to indemnify either because no damages are awarded in the underlying action or because the actual judgment is for damages not covered under the policy. Those contingencies are inapplicable in a case such as the one at bar, where the issue of actual coverage comes before the trial court after resolution of the underlying case. Nothing in this case distinguishes it from Montrose as to what the appropriate trigger of actual coverage should be, and we hold that all parties before us who issued policies with policy periods including any date subsequent to the commencement of harm to the Papworth property from the City's negligent efforts to remedy the deficiencies in its drainage system and prior to the date the Papworth property was effectively destroyed covered the City's liability to Papworth.

However, our conclusions just stated do not resolve a subsidiary question involved in the cases at bar: In what proportions do the aforementioned insurers bear the City's liability to Papworth? Montrose did not involve that question, because only one of the series of successive insurers was before the Montrose court and, as noted above, the issue was only whether there was potential coverage under the policy issued by that one insurer. Here we have concluded that there is actual coverage under the policies issued by a number of different insurers. We shall discuss the proportionate liability of the subject insurers after considering a number of other issues.

B. The Impact of Inverse Condemnation Exclusions

1. Relevant Facts

The Fireman's policies exclude from coverage liability "arising under Article I Section 14 of the Constitution of the State of California."^[7] The other relevant insurance policies are more specific with respect to exclusion of inverse condemnation liability. Minutes of a meeting of the Palos Verdes Estates City Council in 1974 disclose that the City declined to purchase inverse condemnation coverage because it involved an additional premium of $1,073. The City was aware that after 1974 no carrier issued coverage for inverse condemnation.

The City tendered defense of the Papworth claim to Jefferson shortly after Papworth filed his initial complaint on March 4, 1980. Jefferson undertook the defense of the claim without a reservation of rights and retained the attorney who represented the City in the Papworth case. Jefferson did not inform the City of a reservation of rights because of its inverse condemnation exclusion until September 2, 1982, three weeks before the trial of the Papworth case was scheduled to begin.

In its statement of decision, the trial court found and concluded:

"All carriers take the position that there is no coverage for damages for inverse condemnation. The Court agrees that inverse condemnation was validly excluded from coverage, that the City was aware that it had no such coverage, and that the City could not reasonably have expected such coverage....

"... [T]he court finds that [the] attempted disclaimer [of inverse condemnation coverage by Jefferson] was not effective. A carrier cannot accept the defense without a reservation of rights or disclaimer and shortly before trial, when it would be impractical for the insured to obtain counsel to effectively represent it, inform the insured that it will not pay damages attributable to a specific claim of which it had knowledge for some two years."

2. The Efficacy of the Inverse Condemnation Exclusions

(5) Except with respect to the Fireman's policies, the trial court's finding that the policy provisions excluding inverse condemnation coverage were effective as to all insurers other than Jefferson is fully supported by evidence such as that referred to in the above quoted passage from the trial court's statement of decision.^[8]

With respect to the Fireman's policies, however, it is questionable whether the trial court's finding is supported by substantial evidence. An identical exclusion of liability arising under article I section 14 of the California Constitution has been held ineffective to exclude liability for inverse condemnation. (General Ins. Co. of America v. City of Belvedere (N.D.Cal. 1984) 582 F. Supp. 88.) There the court reasoned: "[The l]anguage in the policy is to be construed as a layman might read it, not as an attorney or insurance agent might read it.... [¶] ... [¶] ... [I]t is questionable whether the reasonable person of ordinary education and intelligence, upon being referred ... to `Article I, Section 19' [as renumbered from 14] would emerge with any conviction that what was meant was inverse condemnation. Although such actions do indeed `arise under' that provision, neither the old Section 14 nor the present Section 19 makes specific reference to inverse condemnation. In fact, the words appear neither in the Constitution nor ... in the index to the Constitution. In order to inform himself that the Section covers inverse condemnation actions, the reasonable layman would either have to consult an attorney or familiarize himself with California appellate law." (582 F. Supp. at pp. 89-90.)

We are cited to nothing in the record that indicates that inverse condemnation coverage was unavailable at the time of the inception (1971) of the Fireman's policies or that it was available only at an increased premium. The aforementioned city council meeting took place some three years after the Fireman's policies went into effect. If Fireman's were bound by the record before us, we might well conclude that Fireman's policies did not exclude liability for inverse condemnation.

However, Fireman's never had the opportunity to introduce evidence on its own behalf concerning the impact of the language excluding coverage for liability under section 14 of article I of the California Constitution. Instead, the trial court granted Fireman's motion for judgment under Code of Civil Procedure section 631.8 on the ground (hereinafter discussed) that Government Code section 911.2, requiring the filing of claims against governmental entities, cut off liability of carriers providing coverage only before August 19, 1976. (Fireman's was such a carrier.) The granting of the section 631.8 motion effectively cut off Fireman's opportunity to offer evidence that might demonstrate that its "Section 14 of Article I" exclusion might apply in this case — evidence, for example, of some expression of the mutual intention of the parties that the clause in question exclude coverage for inverse condemnation liability. (See, e.g., City of Mill Valley v. Transamerica Ins. Co. (1979) 98 Cal. App. 3d 595, 602-603 [159 Cal. Rptr. 635] [contemplating the possibility that evidence of the mutual intention of the parties might demonstrate that a policy did not cover inverse condemnation liability].) Because we hold in part V.C. of this opinion that the trial court erred in granting Fireman's section 631.8 motion, in the proceedings below following issuance of our writ of mandate the trial court must give Fireman's an opportunity to offer any evidence it may have tending to demonstrate that its "Section 14 of Article I" exclusion applies in this case.

3. The Impact of Jefferson's Delay in Notifying the City of Its Reservation of Rights

(6a) The trial court's holding that Jefferson cannot assert the inverse condemnation exclusion contained in its policy is supported by substantial evidence and will be upheld.

(7) An insurer's assumption of the defense of its insured without giving notice of a reservation of rights may preclude the insurer from denying coverage because of that which could have been reserved. "An insurer can waive policy provisions which would otherwise defeat coverage. [Citation.] Where an insurer reserves its right to claim noncoverage under the policy, notice of the reservation must be given to the insured or the reservation is deemed waived. `... [I]f a liability insurer with knowledge of a ground of ... noncoverage under the policy, assumes and conducts the defense of an action brought against the insured, without disclaiming liability and giving notice of its reservation of rights, it is thereafter precluded in an action upon the policy from setting up such ground of ... noncoverage. In other words, the insurer's unconditional defense of an action brought against its insured constitutes a waiver of the terms of the policy and an estoppel of the insurer to assert such grounds.' (14 Crouch on Insurance (2d ed. 1965) § 51:77, pp. 579-582, fns. omitted.)" (Miller v. Elite Ins. Co. (1980) 100 Cal. App. 3d 739, 754 [161 Cal. Rptr. 322]; see also Ins. Co. of the West v. Haralambos Beverage Co. (1987) 195 Cal. App. 3d 1308, 1320 [241 Cal. Rptr. 427]; Val's Painting & Drywall, Inc. v. Allstate Ins. Co. (1975) 53 Cal. App. 3d 576 [126 Cal. Rptr. 267].)

(6b) Here the record supports both an inference of waiver and an inference that the City detrimentally relied on Jefferson's nonassertion of a reservation of rights. By not reserving its rights with respect to inverse condemnation, Jefferson placed the counsel it had selected in a conflict of interest position in the preparation and pretrial conduct of the City's defense of the Papworth case. Upon assertion of a reservation of rights the City would have been entitled to have separate counsel represent it at Jefferson's expense. (E.g., San Diego Federal Credit Union v. Cumis Ins. Society, Inc. (1984) 162 Cal. App. 3d 358 [208 Cal. Rptr. 494, 50 A.L.R. 4th 913].) The City, being unaware of any reservation of rights by Jefferson, did not request representation by separate counsel. By the time the notice of reservation of rights was sent, trial was at hand and it would have been too late for an attorney representing solely the interests of the insured to have replaced the attorney who was representing the insurer's interests as well.

Jefferson argues that the City was not prejudiced because it knew that the Jefferson policy did not cover its inverse condemnation liability. That argument misses the point. What the City did not know was that Jefferson was going to assert the inverse condemnation exclusion. So far as the City knew, Jefferson was not; and the City was entitled to rely upon that justified assumption.

4. The Significance of the Fact That the Papworth Judgment Was Vacated Pursuant to the Settlement of the Papworth Case

(8a) As to the insurers other than Jefferson and perhaps Fireman's — those with a valid and enforceable inverse condemnation exclusion — the question remains as to the impact of that exclusion upon the obligation of those insurers to indemnify the City for its liability to Papworth. The trial court found as follows with respect to this issue: "[T]he judgment [in the Papworth case] was vacated and the funds paid in settlement of the Papworth claim were to dispose of the case and were not designated as an inverse condemnation judgment. The parties had their own reasons to want the judgment vacated, and those reasons may have been compelling, but the effect of the stipulation and order vacating the judgment is that the funds paid in settlement cannot now be said to have been paid to satisfy a judgment for inverse condemnation."

In support of these findings, the City argues that the court order based on the stipulation settling the Papworth case pending its appeal vacated the Papworth judgment for all purposes and that the judgment and the jury verdict on which it was based are nullities. Persuasive authorities are to the contrary:

(9) Under the doctrine of collateral estoppel, issue preclusion requires only that the prior adjudication upon which it is based be "`sufficiently firm to be accorded [preclusive] effect.'" (7 Witkin, Cal. Procedure (3d ed. 1985) Judgment, § 216, p. 653, quoting Rest.2d Judgments, § 13.) Where an appeal is settled favorably to the plaintiff and then dismissed "the Restatement analysis and reason itself dictate that the trial court judgment reemerges with sufficient finality to permit application of collateral estoppel." (Sandoval v. Superior Court (1983) 140 Cal. App. 3d 932, 937 [190 Cal. Rptr. 29]; see also Producers Dairy Delivery Co. v. Sentry Ins. Co. (1986) 41 Cal. 3d 903, 911 [226 Cal. Rptr. 558, 718 P.2d 920]; McClain v. Rush (1989) 216 Cal. App. 3d 18 [264 Cal. Rptr. 563].)

(8b) Where an underlying action against a governmental entity proceeds upon the theory of inverse condemnation, a judgment is entered against the entity, and pending appeal the case is settled with a stipulation that the judgment be vacated and the appeal is thereupon abandoned, collateral estoppel precludes the entity's denial that its liability was in inverse condemnation. (City of Laguna Beach v. Mead Reinsurance Corp. (1990) 226 Cal. App. 3d 822 [276 Cal. Rptr. 438].) The City of Laguna Beach case, decided after the trial court's judgment herein, demonstrates that the trial court's conclusion that the settlement of the Papworth action pending appeal eliminated its preclusive effect was erroneous. The fact that the $1.6 million settlement amount exceeded the amount the jury verdict attributed to negligence/nuisance (apparently even when that amount is enhanced by costs and postjudgment interest recoverable thereon) makes this conclusion even more inescapable.

However, the question still remains: now that we have determined that the inverse condemnation exclusions (other than those in the Jefferson policy and possibly the Fireman's policies) are enforceable, what is their effect?

5. The Effect of the Enforceable Inverse Condemnation Exclusions

The judgment in the Papworth case recites jury verdicts awarding Papworth damages of $1,188,791 on his negligence and nuisance claims against the City and damages of $1,881,946 on his inverse condemnation claim. The difference is largely attributable to prejudgment interest and attorney fees and other costs allowable on the inverse condemnation claim. Judgment was entered for the larger amount, and for this reason the record is silent with respect to costs allowable on the negligence and nuisance claims. The nature of the jury verdict raises the question of the scope of the inverse condemnation exclusion in the relevant insurance policies: does the exclusion apply to the entire Papworth $1.6 million settlement, only to a proportion of that settlement, or to none of that settlement.

(10) The City argues that it is impossible to attribute any specific part of the settlement to either inverse condemnation or negligence/nuisance liability, so that the exclusion is applicable to none of that settlement. This argument ignores the fact that the record includes a jury verdict from which allocation between insured and uninsured liability may be inferred once allowable costs attributable to the negligence and nuisance claims are determined.

Although arguing that their inverse condemnation exclusions apply to the entire Papworth case settlement, the insurer parties whose policies have effective exclusions argue in the alternative that the exclusions apply to a proportion of that settlement. We conclude that their alternative argument is the correct one.

In arguing for application of the inverse condemnation exclusions to the entire settlement, the insurers rely upon City of Laguna Beach, supra, There the Court of Appeal held that the municipality was not entitled to produce evidence of its negligence and that the inverse condemnation exclusion applied to the entire judgment in the underlying case. However, in City of Laguna Beach all causes of action other than the one for inverse condemnation were dismissed by the plaintiff in the underlying case prior to submission of that case to the jury, and the jury returned a verdict only on the theory of inverse condemnation. (226 Cal. App.3d at p. 827.) While of no preclusive effect against the insurers (who were neither parties to nor represented by a party on the issue in the underlying action), the jury verdict in the underlying action is part of that which results in issue preclusion favorable to the insurers. Benefited by issue preclusion incident to the jury verdict supporting the vacated judgment, the insurers must take the bad (a judgment generated in part by negligence and nuisance findings and therefore in part not subject to the inverse condemnation exclusions) with the good (a judgment generated in part by an inverse condemnation finding and therefore in part subject to those exclusions).

Inverse condemnation is sometimes the sole basis of a cause of action, as where liability is imposed without fault because it is caused by the use of public improvements deliberately planned and built. (See, e.g., Yee v. City of Sausalito (1983) 141 Cal. App. 3d 917, 919-920 [190 Cal. Rptr. 595].) In other cases inverse condemnation is an alternative to other theories of recovery. (See, e.g., Bozaich v. State of California (1973) 32 Cal. App. 3d 688, 693, fn. 2 [108 Cal. Rptr. 392] [negligence].) In the latter instance the inverse condemnation theory is significant only as it relates to the remedy — to the measure of damages. To the extent of the measure of damages for negligence and nuisance the City incurred a legal obligation to compensate Papworth irrespective of inverse condemnation, and this form of legal obligation is covered by the relevant insurance policies and is not subject to the inverse condemnation exclusions.

The net result is that Jefferson covers the entire $1.6 million of the Papworth settlement, Fireman's may or may not (depending upon the trial court's decision hereafter as to the effectiveness of Fireman's "Section 14 of Article I" exclusion), but the other insurers whose policies apply cover only that portion of the $1.6 million which was paid to settle the jury's award of damages for negligence/nuisance as adjusted upward for costs. It therefore is appropriate that we order the trial court to determine what portion of the $1.6 million was paid to settle the negligence/nuisance verdict and costs on that verdict. We suspect (and, as noted above, the City argues) that it is impossible to attribute any part of the settlement to either inverse condemnation or negligence/nuisance liability. If such be the case, upon issuance of our writ of mandate the trial court should conclude that the insurers (other than Jefferson and possibly Fireman's) whose policies apply cover only that proportion of the $1.6 million which the jury's award of damages for negligence/nuisance as adjusted upward for costs bears to the entire judgment in the Papworth case.

C. The Effect of the Limitations Period in Government Code Section 911.2

Government Code section 911.2 requires that a claim of the sort that Papworth had against the City be presented not later than one year after the accrual of the cause of action.

The trial court held: "(1) ... [T]he [Papworth] claim dated August 19, 1977 established the parameters of the liability of the City; and (2) ... since no damage which occurred prior to August 19, 1976 could be recovered (the claim was filed September 2, 1977 and that date would control, but it is inconsequential here) carriers providing coverage prior to that date are not liable."

(11) This holding is erroneous. It ignores authority establishing that in a context such as presented here (one involving continuous and repeated damage incident to a public improvement), the limitations period does not begin to run until the situation has stabilized. (Pierpont Inn, Inc. v. State of California (1969) 70 Cal. 2d 282, 291-294 [74 Cal. Rptr. 521, 449 P.2d 737], relying upon United States v. Dickinson (1947) 331 U.S. 745 [91 L. Ed. 1789, 67 S. Ct. 1382] [periodic increases in water levels created by a dam; statute of limitations did not begin to run as damage occurred, but only when situation became stabilized].) The evidence in the case at bar supports only one conclusion on the issue of whether the situation concerning the Papworth property had become stabilized prior to September 2, 1976 (one year prior to the filing of Papworth's claim): it had not. As noted in part V.A. 1., ante, as a result of the City's ongoing periodic design, maintenance and mitigation activities, relatively minor erosion damage to the Papworth property was occurring in 1978 and the deep-seated landslide was activated which effectively destroyed the property in 1981. Accordingly, the City's liability for pre-September 2, 1976, damages was not cut off by Government Code section 911.2, and carriers providing coverage before that date are not exonerated from liability by section 911.2.

D. The Limitation That Only a Contingent Risk Is Insurable

(12) Insurance Code section 22 defines insurance as "a contract whereby one undertakes to indemnify another against loss, damage, or liability arising from a contingent or unknown event." Insurance Code section 250 limits insurability to events which are "contingent or unknown ..., whether past or [present]." Insurance Code sections 22 and 250 coalesce in what has come to be known as the "loss-in-progress" or "known loss" rule. (See Montrose, 10 Cal.4th at pp. 689-690.) This rule holds that insurance is not available to indemnify against losses which are not contingent. (Ibid.) The insurers contend that this rule bars the City from recovering any of the Papworth loss from them.

1. The Facts Relevant to Contingency of Risk

There was substantial evidence before the trial court to the following effect:

(a) The City was aware of deficiencies in its storm drain system for at least eight years prior to Papworth's 1971 acquisition of his property. In 1963 a storm drain deficiency study conducted by the City found in general that the system lacked adequate capacity and in particular that the portion of the storm drain system located in the City parkland adjacent to the Papworth property (the parkland drain) was deficient. Two years later a geological study commissioned by the City found that the geology of the steep walled gully in which the parkland drain was located posed a minimum hazard although deterioration of berms and drainage devices could lead to destructive erosion of the slope below the lot subsequently purchased by Papworth. The recommendations of the 1963 study were not implemented because of the City's determination that funds were not available.

(b) In 1970 Papworth's predecessor in interest complained to the City of erosion in the parkland and requested that the corrugated storm drain pipe there located be extended. In response to this complaint, shortly before Papworth purchased his property the drain was extended from its original outlet to a point on the beach.

(c) Papworth purchased the relevant property in August of 1971. In the winter of 1971-1972 there was a break in the storm drain pipe. Papworth became concerned that the drain was causing erosion in the parkland area and communicated his concern to the City. Due to the City's prioritization of funds to other purposes which precluded more effective repair, and as a temporary measure, the outlet of the drain was extended with gunited mesh. In 1972 cracks appeared in the Papworth dwelling but were attributed by him to settling of fill, a condition for which the City was not responsible. No problems were encountered with the gunite-extended drain in 1972 and 1973, and in August of 1973 Papworth informed the City that the temporary repair to the drain seemed sufficient.

(d) The same year Papworth, having retained a soils engineer, advised the City that if the storm drain situation were not permanently corrected by relocating the drain the Papworth home would eventually be destroyed by erosion. The then city engineer was of a contrary opinion. Later the gunite began to deteriorate, and Papworth repeatedly over the next four years complained to the City that the drain was deficient. Over this period there was further cracking in the walls and foundation of the Papworth house and cracking in the adjacent street as well. This was also attributed to settlement of fill.

(e) On September 2, 1977, Papworth filed a claim against the City pursuant to Government Code section 905 et seq. Founded in negligence and nuisance, the claim asserted that the storm drain was undermining the Papworth property and eroding the cliff adjacent to it. The claim asserted further that the erosion would continue absent appropriate action by the City. It sought compensation of $150,000 for the estimated loss in value of the Papworth property because of erosion in the parkland and the lack of a clear prospect that the erosion would be stabilized and the consequent risk that the Papworth house would be undermined and destroyed. The filing of the claim was motivated by Papworth's "five years of frustration" in dealing with the City on the problem of the storm drain. Because of the claim the City expected to be sued.

(f) The winter of 1977-1978 was one of exceptionally heavy rainfall. The parkland drain pipe burst, the guniting was destroyed, and erosion extended from the parkland onto the Papworth property, destroying six feet of the backyard and undermining the back fence. Papworth believed this was a "one time event." The City repaired the burst pipe and regunited the storm drain channel.

(g) On February 1, 1978, Papworth filed suit on his claim against the City, asserting the theories of negligence and nuisance, among others, but not including a cause of action for inverse condemnation. The complaint also sought an injunction to compel an adequate storm drain. This complaint was not served.

(h) A 1978 repair guniting the storm drain failed in 1979 and was repeated after the 1979 failure. The 1979-1980 season was also one of unusually heavy rain. Water from the storm drain ran over the Papworth property and damaged fences, uprooted trees and undermined its lateral support. In January of 1980 there was further damage to the Papworth property from a burst underground concrete pipe, and further cracking of the residence. A geological report prepared in 1980 described its findings of the cause of the damage to the residence as inconclusive. On March 4, 1980, Papworth filed and subsequently served an amended complaint which added a cause of action for inverse condemnation. By 1981 the Papworth home was effectively destroyed. The bedroom wing built beneath the remainder of the house had moved by more than one foot, and the house had deformed dramatically. In the trial of Papworth's action against the City the parties stipulated that for the purposes of inverse condemnation the date of taking was February 1, 1980. Apparently because of the stipulation, the trial court in these consolidated cases made no findings with respect to poststipulation damage.

(i) Late in 1981 further geological examination developed for the first time the fact that the damage to the Papworth house and the repeated failures of the storm drain system were caused by a large, deep-seated landslide that had become active as early as 1966. The landslide had been activated because of deficiencies in the storm drain system as constructed in the 1920's. A failure to extend the drain to the beach resulted in both leakage of water into the underground structure and removal of lateral support, both of which accelerated the slide and consequent eventual destruction of the Papworth property. The landslide "picked up speed" in years of heavy rainfall and had gained in intensity during the heavy rains from 1977 into 1980.

On this evidence the trial court concluded and found: "The landslide which ultimately caused the destruction of the Papworth home was a contingent and unknown event and could be insured. Even before Papworth purchased the property in 1971, the City was on notice that the drain in the gully next to the property was causing erosion. However, until 1981 ... no one was aware of the existence of the landslide which caused the damages ... which Papworth recovered against the City.... This known erosion could have continued for many years without destroying the Papworth residence had it not triggered the undiscovered (until 1981) landslide."

No party questions this finding as lacking evidentiary support.

2. The Law Relevant to the Question of Contingency

The Supreme Court in Montrose considered in depth the application of the loss-in-progress rule to a case such as the one at bar, namely, a third party case involving continuous or progressively deteriorating property damage. The majority opinion reviewed conflicting (or at least seemingly conflicting) decisions of other courts, evaluated the policy considerations involved, and concluded: "We therefore hold that, in the context of continuous or progressively deteriorating property damage or bodily injury insurable under a third party CGL policy, as long as there remains uncertainty about damage or injury that may occur during the policy period and the imposition of liability upon the insured, and no legal obligation to pay third party claims has been established, there is a potentially insurable risk within the meaning of sections 22 and 250 for which coverage may be sought. Stated differently, the loss-in-progress rule will not defeat coverage for a claimed loss where it had yet to be established, at the time the insurer entered into the contract of insurance with the policyholder, that the insured had a legal obligation to pay damages to a third party in connection with a loss." (10 Cal.4th at p. 693.)

(13) Applying the above quoted principles to the case at bar, it is clear that the loss-in-progress rule does not bar recovery against any insurer who is a party to the within appellate proceedings. The earliest date anyone could conceivably argue that it was established that the City had a legal obligation to pay damages to Papworth was November 24, 1982, when the jury verdict in the Papworth case was rendered. The judgment in the Papworth case was not entered until January 11, 1983. No insurer who is a party to the within appellate proceedings issued a policy after, or with a policy period commencing after, November 24, 1982. (The insurer with the latest policy period commencement date was Midland Insurance Company, which issued an excess policy for the one-year period commencing July 1, 1981; and Midland went into receivership and ceased to be a party to this litigation.)

3. The Consequences to Contingency of the Papworth Government Claim and the Subsequent Filing of the Papworth Lawsuit Against the City

(14) The insurer parties with policies incepting after September 2, 1977, argue that of the City's loss incident to the damage caused to Papworth prior to that date, whatever might have been contingent became noncontingent when Papworth's government claim against the City was filed on that date. This argument is founded primarily upon the definition of contingency contained in Fireman's Fund Ins. Co. v. Aetna Casualty & Surety Co. (1990) 223 Cal. App. 3d 1621, 1627 [273 Cal. Rptr. 431], to the effect that it is the damage which must be contingent or unknown and not the liability of the insured or the cause of the damage. The argument finds support in the fact that the City expected that it would be sued on the claim.

However, as indicated in part V.D. 2., ante, we have concluded that the appropriate test of insurability under Montrose is lack of definitive establishment of the insured's liability. Montrose's holding in essence seems to have disapproved the Fireman's Fund Ins. Co. v. Aetna Casualty & Surety Co. definition of contingency.

Furthermore, here Papworth's government claim asserted a liability of the City of a wholly different character than that established in Papworth's action. The claim was based upon erosion within the parkland channel which diminished the value of the Papworth property by the threat of erosion to it and the removal of lateral support. The trial court found that there was a loss when the claim was filed due to stress cracks and removal of lateral support. But when the claim was filed it was unknown to everyone that the real threat was not merely of these harms, but rather that the entire property including the improvements would be devastated by the landslide. The difference in degree is so great as to be a difference in kind. Even if the applicable test were whether the insured knew of the risk of liability at the time the policy was issued, the risk that in fact materialized to cause the great bulk of the Papworth damage was contingent when all of the subject policies were issued.

(15) Admiral and other insurers whose policies incepted after the Papworth government claim argue that they should be relieved of the obligation to indemnify the City because the claim was not disclosed in the City's applications for insurance. It is certainly possible that these insurers might have been entitled to escape liability by rescinding their policies upon an appropriate showing. (Ins. Code, §§ 331-338.) But the insurers asserting this proposition do not accompany it with the requisite showing that their pleadings in the trial court raised the issue of rescission. As to Admiral the trial court concluded: "No such defense was raised in the pleadings, nor was any prejudice shown." This finding is not attacked in the proceedings before us.

E. The "Expected or Intended" Limitation

(16) Some of the relevant insurance policies limit the definition of "occurrence" to damage which is neither intended nor expected from the standpoint of the insured. California law is clear that in the instant case the Papworth damage was not intended by the City. Even a conscious disregard of the safety of others does not render the resulting harm intentional so as to preclude insurability. (Peterson v. Superior Court (1982) 31 Cal. 3d 147, 158 [181 Cal. Rptr. 784, 642 P.2d 1305].) Here the City's ineffectual efforts to correct the deficiencies in its storm drain were negligent. However, these efforts nevertheless support an inference of lack of intent to cause harm.

Nor can it be said on this record that the Papworth loss for which the City was held liable in the underlying action was expected within any relevant policy period prior to sometime in 1981. The relevant policy language treats the expectation as that of the insured. There is testimony of the then city manager and the then city engineer that they did not expect the loss. This and the proposition that the true cause of the destruction of the Papworth property was unknown until 1981 support the trial court's conclusion that the loss was unexpected.

We recognize the presence of a counterargument. There is evidence in the record that the futile efforts of the City to stem the effect of erosion were due to its fiscal decisions from as far back as 1963. A public policy encouraging governmental entities to face fiscal responsibility conceivably could counsel a different interpretation of the expected damage limitation in the situation of municipal liability insurance. No legislative or Supreme Court enunciation of such a policy has been called to our attention, however; and it is not our place as an intermediate court to take this step.

F. The Period(s) and Deductible(s) of the Jefferson Policy

The Jefferson policy is for the three-year period from November 1, 1975, to November 1, 1978. It covers liability for property damage with limits "per Endorsement CSL-1 (3/75) attached." There are three separate endorsements for the years 1975 through 1978, each including a limit of $300,000 per occurrence and in the aggregate and a deductible of $1,000 per claim. There are three separate "Declarations," each for a separate policy period.

The trial court concluded that the Jefferson policy covered three separate periods with a limit of $300,000 for each period, an aggregate of $900,000 in coverage. (17) Jefferson argues that its policy included one $300,000 limit applicable to the entire three-year period.

We conclude that the trial court's determination was correct. At the least, the policy is ambiguous and the ambiguity must be construed against Jefferson. Moreover, the ambiguity is resolved against Jefferson in a stipulation between it and Maine Bonding providing: "The subject policies of insurance issued by ... Jefferson ... provided coverage of $300,000 per occurrence per year as respects property damage...."

Jefferson makes no claim that it sought in the trial court to be relieved of the stipulation.

Jefferson claims also that its policy language defining occurrence limits its exposure to $300,000. This language (like that in the other relevant policies) states that all damage arising from continuous and repeated exposure is deemed a single occurrence. This argument ignores two points: (1) the policy covers liability for occurrences within a policy period; and (2) the Jefferson policy covers three separate periods. Reading (as we must) the Jefferson policy in a fashion resolving ambiguities against it, the language on which Jefferson relies must be construed as referring to a single occurrence in a policy period. The Jefferson policy covering three policy periods, the policy language amounts to a $300,000 per period limitation — or, in the context of this case (involving a continuous trigger), a $900,000 limitation.

An additional consequence of our conclusion that the Jefferson policy covers three policy periods is that the Jefferson policy has a separate $1,000 deductible for each policy period — or, in the context of this case, $3,000 in deductibles.

G. The Liability of Excess Insurers to the City

1. The Principles Governing Allocation of Liability as Between Primary and Excess Insurers

The liability of excess insurers to the City is a question of contract. (Montrose, supra, 10 Cal.4th at p. 681, fn. 19.) Fireman's excess policies state: "The Company shall be liable for payment under this policy only after the insured and the insurers affording coverage under the primary policies ... have paid or become obligated to pay the applicable amount or amounts of such insurance following final judgment ... or by written agreement of the insured, the claimant, the Company and the aforesaid insurers."

Some of the other excess policies contain similar language and some expressly state that the primary policies are those listed on the pertinent declarations. There is, however, a latent ambiguity in these provisions because none covers the situation here — a situation in which several primary policies cover the eventual liability of the insured.

The fact that the total amount of primary insurance covering the loss exceeds the amount contemplated in the excess policy does not subject the excess carrier to liability. (18) "[L]iability under a secondary [excess] policy will not attach until all primary insurance is exhausted, even if the total amount of primary insurance exceeds the amount contemplated in the secondary policy." (Olympic Ins. Co. v. Employees Surplus Lines Ins. Co. (1981) 126 Cal. App. 3d 593, 600 [178 Cal. Rptr. 908]; see also North River Ins. Co. v. American Home Assurance Co. (1989) 210 Cal. App. 3d 108, 112 [257 Cal. Rptr. 129]; Hartford Accident & Indemnity Co. v. Sequoia Ins. Co. (1989) 211 Cal. App. 3d 1285, 1302 [260 Cal. Rptr. 190].)

Here it is almost certain that there is adequate primary coverage to fund the City's liability to Papworth:

(a) There is a maximum of $1.6 million in liability of insurers to the City.^[9]

(b) The trial court held that the burden of the City's loss^[10] should be borne by Jefferson and Admiral and included in its statement of decision a finding that some of the damage for which the City was liable to Papworth "occurred before 1976" and that the loss "was continuous until February 1, 1980, the date on which it was stipulated that the property was taken by the City." As far as they go, those findings are supported by substantial evidence before the trial court and are binding on all parties. Because the trial court concluded that Government Code section 911.2 cut off liability of carriers whose policies covered only periods before August 19, 1976, it did not need to determine the issue of whether any damage occurred before August 19, 1976, for which the City was liable; and its statement of decision is not sufficiently clear on that subject to warrant our holding that the trial court determined that any damage for which the City was liable occurred before "before 1976" — literally, "before December 31, 1975." (The trial court's finding that damage occurred "before 1976" is no more than a finding that damage occurred on December 31, 1995, or earlier.) In part V.C. of this opinion we hold that the trial court erred in exculpating carriers whose policies covered only periods before August 19, 1976; and upon issuance of our writ of mandate, the trial court will need to decide the issue of whether any damage occurred before December 31, 1975.

(c) Similarly, because the trial court concluded that the stipulation as to the February 1, 1980, date of taking of the Papworth property by the City fixed the date on which the loss was complete and because the trial court had previously ordered summary judgment in favor of Canadian (the only insurer — other than Midland Insurance Company, whose policy is now irrelevant [see pt. V.D. 2., ante] — that issued a policy incepting after February 1, 1980), the trial court did not need to determine at the Phase I trial the issue of whether any damage for which the City was liable occurred after February 1, 1980; and its statement of decision at the conclusion of the Phase I trial is not sufficiently clear on that subject to warrant our holding that the trial court determined that issue. And even if it were, it could not be binding on Canadian, who did not participate in the trial because of the granting of an order for summary judgment in its favor. In part VI of this opinion we hold that the trial court erred in ordering summary judgment in favor of Canadian; and upon remand the trial court will need to address the issue of whether any damage for which the City was liable occurred after February 1, 1980.

(d) However, from the foregoing analysis it is clear that at least Jefferson and Admiral, who issued primary policies with policy periods including dates on or after August 19, 1976, and on or before February 1, 1980, covered the City's liability to Papworth.

(e) Accordingly, under the continuous trigger analysis prescribed by Montrose, coverage was continuously triggered during at least the period from December 31, 1975, to February 1, 1980 — a period of more than four years — at least one thousand four-hundred ninety-three days. This is the shortest conceivable period of coverage. Utilizing the preferred, "qualified time on the risk" method of apportionment discussed at part V.H., post, prorating a maximum potential liability of $1.6 million over 1,493 days would yield a maximum potential liability of approximately $1,071.67 per day of coverage or approximately $391,159.55 per year. The $1 million per occurrence per year coverage^[11] under the Admiral primary policy (having a policy period of November 1, 1978, to July 1, 1980) is more than adequate to cover its pro rata share of this per diem liability. The $300,000 per occurrence per year (approximately $891.92 per day) primary policy or policies issued by Jefferson for the aggregate of three years commencing November 1, 1975, could theoretically fall short of the approximate $391,159.55 yearly liability hypothesized above by $91,159.55 per year for each of the last two of the three policy years and by $27,931.02 for the last three hundred and six days of the first policy year ($1,071.67 per day of coverage for 306 days = $327,931.02, less the $300,000 limit of coverage for the first policy year) — a theoretical shortfall of approximately $210,250.12. The issue would then become whether Stonewall and/or Maine Bonding as excess carriers during Jefferson's policy period^[12] on the one hand or Admiral as the primary carrier after Jefferson's policy period on the other hand are/is liable for this approximately $210,250.12 shortfall. We decline to express an opinion on this issue today because there seems to be very little likelihood that all factors necessary to create this shortfall in primary coverage during 2.836 of the years covered by the Jefferson policy will converge.^[13]

(19) We do, however, today hold that if the limits of liability on the policies of primary insurance that are determined to cover the City's liability to Papworth are adequate to cover their pro rata share of that liability (that pro rata liability to be determined consistent with the views expressed herein), no excess carrier is liable to the City for any of the City's liability to Papworth. In substance we adopt the "horizontal allocation of the risk" approach to liability as between primary and excess carriers, rather than the "vertical" approach. To begin with, it seems clear from the City's assertion that all of its primary insurers covered its liability that the City's reasonable expectations treated the excess policies as a secondary source. Moreover, the "horizontal" approach seems far more consistent with Montrose's continuous trigger approach. That is, if "occurrences" are continuously occurring throughout a period of time, all of the primary policies in force during that period of time cover these occurrences, and all of them are primary to each of the excess policies; and if the limits of liability of each of these primary policies is adequate in the aggregate to cover the liability of the insured, there is no "excess" loss for the excess policies to cover.

2. The Status of Admiral as a Primary or Excess Carrier

Admiral contends that having issued a policy to the City with a $50,000 self-insured retention, it is an excess carrier. While there is support for Admiral's position (see, e.g., Nabisco, Inc. v. Transport Indemnity Co. (1983) 143 Cal. App. 3d 831 [192 Cal. Rptr. 207]), the trial court found that Admiral was referred to as a primary carrier in an agreement (to which Admiral was a party) entered into at the time of the Papworth settlement. This finding supports the trial court's treatment of Admiral as a primary carrier and was not challenged by Admiral until its petition for rehearing following the filing of our earlier opinion herein. We decline to consider the challenge of that finding because of the challenge's untimeliness and the absence of appearance of good cause for its untimeliness. (E.g., CAMSI IV. v. Hunter Technology Corp. (1991) 230 Cal. App. 3d 1525, 1542 [282 Cal. Rptr. 80]; People ex rel. Dept. of Public Works v. Mascotti (1962) 206 Cal. App. 2d 772, 779 [23 Cal. Rptr. 846].) Our holding that Admiral was a primary carrier of course does not render the retention provided for in its policy wholly irrelevant.

H. Allocation of Liability Among the Primary Insurers

The issue of how the liability of the City to Papworth should be apportioned among the primary carriers in a third party property damage case involving repeated exposure to a continuing condition and successive insurance carriers has not been addressed definitively in any reported California case.^[14]

In reality, two separate questions are involved in determining how the liability of the City to Papworth should be apportioned among the primary carriers:

Question 1: How much of a judgment is the City entitled to recover from each primary insurer?

Question 2: In what proportions are the primary carriers to share in whatever payments any of them made or makes toward the $1.6 million loss?

So long as all the primary carriers are solvent and amenable to jurisdiction to enforce the judgment herein, the answer to question 1 is of no practical significance. The primary carriers will simply pay the City their respective portions of the aggregate amount the City is entitled to recover from them. However, if carrier X is obligated (as among the primary carriers) to bear a share of the aggregate liability to the City and carrier X is insolvent (or if a judgment against carrier X is otherwise unenforceable), there is practical significance to question 1: it provides the answer as to whether there are one or more carriers who must pay the City the share of carrier X. We do not know whether there is a "Carrier X" among the primary insurers who must share (as among themselves) the liability to the City.^[15] However, the City does claim on this appeal that the trial court erred by failing to hold all carriers on the risk liable to the City (to the extent of their respective policy limits and subject to deductibles and retentions) for the full $350,000 paid by the City toward the Papworth settlement; and for whatever reason this may be important to the City, it is entitled to an answer to question 1. The carriers of course are entitled to an answer to question 2.

1. Question 1 — Who Must Pay the City the $350,000?

Neither question 1 nor question 2 was directly answered by Montrose. In Montrose the matter before the Supreme Court was the insured's appeal from a summary judgment against the insured and in favor of one of seven insurers, holding that that one insurer owed neither a duty to defend nor a duty to indemnify the insured.^[16]Montrose therefore did not directly involve the issue of allocation as among the primary insurers of their liability to the insured. However, we find the answer to question 1 in Montrose's analysis: With one important qualification, all primary carriers on the risk are liable to the City (up to the limits of their respective policies, less any applicable deductibles or retentions) for the full $350,000. Inherent in Montrose's conclusion that in cases such as the one at bar a "continuing injury" trigger of coverage applies is the principle that damage was occurring throughout the period in question and that all carriers issuing primary policies for dates within that period are fully liable to the insured for the entire loss.^[17] Once an injury triggers coverage, according to the language of the policies in the case at bar (and the standard CGL (comprehensive general libility) policy), the insurer must indemnify the insured for "all sums" which the insured becomes obligated to pay, whether during the period of the policy issued by that insurer or after; the policy language does not limit the insurer's liability to "all sums which the insured becomes liable to pay during the policy period." (See Montrose, supra, 10 Cal.4th at p. 665.) Under a continuing injury trigger, coverage for a manifested loss is not terminated by the expiration of the policy; coverage continues until the damage is complete. (Id. at pp. 678, 680-681, 686; accord, Armstrong World Industries, Inc. v. Aetna Casualty & Surety Co., supra, 45 Cal. App.4th at p. 50 (Armstrong); Keene Corp. v. Ins. Co. of North America (D.C. Cir.1981) 667 F.2d 1034, 1048 [215 App.D.C. 156].)

However, as suggested above, there is an important potential qualification to this principle, arising out of the "other insurance" exclusions included in at least some of the primary policies. Some background concerning the law seems appropriate:

According to Montrose, the first reported California case to discuss the triggering of potential coverage under third party liability insurance policies where continuous or progressively deteriorating property damage was involved was California Union Ins. Co. v. Landmark Ins. Co. (1983) 145 Cal. App. 3d 462 [193 Cal. Rptr. 461]. (See Montrose, supra, 10 Cal.4th at p. 678.) California Union adopted the "continuing injury" trigger approach and held each of the two successive carriers on the risk in that case to be jointly and severally liable to the insured for the full amount of damage. While generally embracing California Union, the Montrose court disagreed with one aspect of the California Union decision: "We do not endorse that aspect of the California Union court's holding that both insurers in that case were jointly and severally liable for the full amount of damage occurring during the successive policy period. (California Union, supra, 145 Cal. App.3d at p. 478.) Allocation of the cost of indemnification once several insurers have been found liable to indemnify the insured for all or some portion of a continuing injury or progressively deteriorating property damage requires application of principles of contract law to the express terms and limitations of the various policies of insurance on the risk. (See Keene Corp. v. Ins. Co. of North America (D.C. Cir.1981) 667 F.2d 1034, 1051 [215 App.D.C. 156] (Keene); [Ins. Co. North America v. Forty-Eight Insulations (6th Cir.1980) 633 F.2d 1212, 1225, clarified 657 F.2d 814, cert. den. (1981) 454 U.S. 1109 (70 L. Ed. 2d 650, 102 S. Ct. 686)].)" (10 Cal.4th at p. 681, fn. 19, italics omitted.)

The Supreme Court thus criticized the "joint and several" holding of California Union for failing to take "the express terms and limitations" of the policies into account. We construe that criticism as relating not only to any deductible and retention provisions of the policies but also to any "other insurance" clauses in the policies.

The modern comprehensive general liability policy typically contains such a clause. It may take the form of outright denial of coverage under the policy if there is other insurance covering the risk (an "escape" clause), a provision that the policy provides only excess coverage if there is other insurance covering the risk (an "excess" clause), a provision that if there is other insurance covering the risk the loss will be shared on some proportional basis with that other insurance (a "pro rata" clause), or some other variation upon the same theme. (See Peerless Casualty Co. v. Continental Casualty Co. (1956) 144 Cal. App. 2d 617, 621-623 [301 P.2d 602].) When such clauses conflict in such a way as to leave the insured unprotected, they are not enforced according to their literal terms. (Continental Casualty Co. v. Pacific Indemnity Co. 134 Cal. App. 3d 389 [184 Cal. Rptr. 583]; CSE Ins. Group v. Northbrook Property & Casualty Co. (1994) 23 Cal. App. 4th 1839 [29 Cal. Rptr. 2d 120].) However, that fact does not mean that they are totally ignored. Typically, neither insurer is held liable to the insured for 100 percent of the loss; instead, as both of the cases last cited illustrate, the court fashions some sort of pro rata formula of liability to the insured.

Jefferson and Admiral and some or all of the other primary insurers who issued policies covering the period during which liability was triggered have "other insurance" clauses in their policies. The Phase I judgment awarded the City a joint and several recovery of $297,000 against Jefferson and Admiral. As the last-quoted passage from Montrose indicates, the $297,000 joint and several award to the City was error: For example, assuming (without deciding) that only Jefferson and Admiral are obligated to pay the City anything, in view of the existence of an "other insurance" clause in each of their policies the obligation of paying the City the $297,000 portion of its payout to Papworth remaining after deducting the $3,000 in Jefferson deductibles and the $50,000 Admiral retention should be imposed partly upon Jefferson severally and partly upon Admiral severally.

Neither the statement of decision nor the judgment in Phase I in any way suggests that the trial court took any "other insurance" clauses into account in arriving at its $297,000 joint and several award to the City. In the proceedings below following issuance of our writ of mandate, the trial court must do so. The language and implications of those clauses have not been briefed adequately for us to attempt to decide their proper interpretation and application in the first instance as a matter of law.

There is an additional apportionment issue inherent in question 1: is any insurer obligated to pay the City any part of the $3,000 in deductibles on the Jefferson policy or policies or any part of the $50,000 retention on the Admiral policy? For ease of reference, we shall characterize this as "the $53,000 question."^[18]

To begin with, since their respective policy provisions control, it is obvious that Jefferson is not obligated to pay the City any of the $3,000 in deductibles under its policy or policies and that Admiral is not obligated to pay the City any of the $50,000 retention under its policy.

However, is Admiral or any other carrier liable to the City for the $3,000 in deductibles under the Jefferson policy or policies; and is Jefferson or any other carrier liable to the City for the $50,000 retention under the Admiral policy?^[19]

At this point we know that there are at least two insurers on the risk, Jefferson and Admiral. If these are the only insurers on the risk, a further examination of the $53,000 question is unnecessary: at oral argument, counsel for the City indicated that in connection with the Phase I judgment, if only Jefferson and Admiral are liable to the City, (1) the City is seeking $297,000 plus interest thereon, and (2) the City is not contending that it is entitled to the remaining $53,000 that it paid to Papworth. Accordingly, if the trial court concludes on remand that only Jefferson and Admiral are liable to the City, it shall not award the City anything by reason of the $3,000 in deductibles on the Jefferson policy or policies or by reason of the $50,000 retention on the Admiral policy.

However, if on remand the trial court concludes that there are insurers in addition to Jefferson and Admiral that are liable to the City, the resolution of the $53,000 question will become quite difficult. There are two very recent decisions of different divisions of the First District Court of Appeal that reach conclusions that are difficult to reconcile on questions similar to the $53,000 question; and each collects cases from other jurisdictions reaching the same and opposite conclusions.

In the context of a dispute concerning costs of defending against liability for toxic waste, Aerojet-General Corp. v. Transport Indemnity Co. (1996) 50 Cal. App. 4th 354, 379-383 [53 Cal. Rptr. 2d 398] review granted September 25, 1996 (S054501), held that the insured was required to bear a proportion of its own defense costs by reason of the fact that some of the primary policies that covered the insured's liability provided that the insured would bear its own defense costs. (The policies containing such provisions had policy periods encompassing eight of the approximately thirty years during which "occurrences" [a]llegedly occurred.) Aerojet-General (a) quoted Montrose to the effect that "allocation of the cost of indemnification ... requires application of principles of contract law to the express terms and limitations of the various policies of insurance on the risk" (Montrose, 10 Cal.4th at p. 681, fn. 19), (b) found that the nature of the policies issued with such provisions unequivocally demonstrated that the insured had no reasonable expectation of coverage for defense costs for "occurrences" which happened during the policy periods of those policies, and (c) held that no insurer was liable for the proportion of the defense costs chargeable to the aforementioned eight-year period.

In the context of a dispute concerning coverage for asbestos-related claims, Armstrong, supra, 45 Cal. App. 4th 1, 55-57, held that insurers on the risk must pay the insured 100 percent of the insured's liability to third parties even though the insured was uninsured or self-insured for a portion of the time during which "occurrences" were occurring. Where the issue is the liability of the insurer(s) to the insured, Armstrong appears to construe Montrose's mandate to consider policy language as not requiring consideration of the "other insurance" clause portion of the policy language.

We believe that the Supreme Court intended by Montrose that in determining disputes between an insured and insurer(s), the policy language to be considered includes the "other insurance" clauses of the policies on the risk. As we have already noted, the language and implications of the "other insurance" clauses in the various primary policies issued to the City have not been briefed adequately for us to attempt to determine their proper interpretation and application in the first instance as a matter of law. Accordingly, if on remand the trial court concludes that insurers in addition to Jefferson and Admiral are liable to the City, the trial court shall determine whether any carrier other than Jefferson is liable to the City for the $3,000 in deductibles and whether any carrier other than Admiral is liable to the City for the $50,000 retention, and in doing so shall take into consideration the language of the respective policies, including but not limited to the "other insurance" clauses of the policies of the insurers potentially liable for the deductibles or the retention.

2. Question 2 — As Among the Primary Insurers, Who Should Bear What Proportion of the $1.6 Million (Plus Interest on the $350,000) Total Payout?

Question 2 involves allocation among the primary carriers on the risk of not only the $350,000 paid to Papworth by the City (plus interest thereon), but also the $950,000 paid Papworth by Stonewall and the $300,000 paid Papworth by Jefferson. As may already be obvious, some of the same considerations involved in answering question 1 are applicable to question 2, but some different considerations also apply.

Because we must consider both principles of equity and principles of public policy, formulating a principle of allocation is no easy task. Equity indicates that all insurers whose policies covered the loss should participate in the cost of indemnifying the insured. Public policy may point in another direction.

Privately obtained liability insurance is the mechanism through which our tort system is primarily financed. The transaction costs of this system are such that almost one dollar of such costs is incurred in order to pay one dollar of compensation. (Kakalic & Pace, Costs and Compensation in Tort Litigation, Rand Corp. Annual Rep. 1985-1986.) A principle of allocation which uses a bright line designation of a single carrier liable for the payment has the advantage of reducing transaction costs by avoiding intrainsurer litigation. If data support the proposition that over the long haul various insurers are likely in different cases to be so positioned that in the end the financial consequences to them of a bright line rule will have essentially the same overall financial impact as equitable apportionment in individual cases, then a bright line rule may be in order. This data is not before us, and neither the Legislature nor our Supreme Court has declared the public policy consideration to which we refer. In this state of affairs, recognition of the role of our court counsels that we not reach the conclusion that public policy controls.

In Home Ins. Co. v. Landmark Ins. Co. (1988) 205 Cal. App. 3d 1388 [253 Cal. Rptr. 277], a first party property damage insurance case, the Court of Appeal opted for a bright line rule. It held that "as between two first-party insurers, one of which is on the risk on the date of the first manifestation of property damage, and the other on the risk after the date of first manifestation of damage, the first insurer must pay the entire claim." (205 Cal. App.3d at p. 1393.) However the Home Insurance court reached this result not in reference to the sort of data to which we have referred but rather by reasoning: "Our approach promotes certainty in the insurance industry and allows insurers to gauge premiums with greater accuracy. Presumably this should reduce costs for consumers because insurers will be able to set aside proper reserves for well-defined coverages and avoid increasing such reserves to cover potential financial losses caused by uncertainty in the definition of coverage...." (Id. at pp. 1395-1396.)

Home Insurance differs from the instant case in that it involves first party insurance, a distinction deemed significant by our Supreme Court in both Montrose and Prudential-LMI Com. Insurance v. Superior Court (1990) 51 Cal. 3d 674, 698 [274 Cal. Rptr. 387, 798 P.2d 1230], although in a different context. More significantly the Home Insurance opinion does not include support for its statement of the impact of its conclusion upon insurer reserve practices and consumer costs. In the situation of third party liability insurance involved in the case at bar, this impact is at best unknown absent supporting data; and we have not been furnished with any such data. Under the Home Insurance approach current insurers would be required to establish reserves to account for the uncertainty that they would be solely liable for long ago unknowable occurrences in other insurers' policy periods. Whether this will result in lesser reserves and lower premiums than would be the case if liability is apportioned among all insurers on the risk is at best problematic. We conclude therefore that with no empirical support for the policy statement in Home Insurance, its rule should not be considered as precedent in the third party insurance context.

Having explored the possibility of a result dictated by public policy and finding none, we return to the application of ordinary principles of equity:

(a) We have previously concluded that by reason of the contractual provisions of the excess policies, liability should rest, to the fullest extent consistent with those provisions, upon the primary carriers.

(b) As indicated in our discussion of "Question 1," in view of the continuous trigger doctrine of Montrose and the language contained therein, each of the primary carriers covering the liability of the City is potentially liable for all of the City's liability, up to the limits of the particular primary policy issued by that carrier and subject to the deductible, retention and "other insurance" clauses of that policy. Accordingly, equity dictates that each primary carrier should bear some proportion of the ultimate burden of that liability.

(c) There are any number of possible ways to apportion that ultimate burden, including but not necessarily limited to the following:

(1) Apportionment based upon the relative policy limits of each primary policy (the "policy limits" method). (See Argonaut Ins. Co. v. Transport Indem. Co. (1972) 6 Cal. 3d 496, 507 [99 Cal. Rptr. 617, 492 P.2d 673]; Employers Reinsurance Corp. v. Phoenix Ins. Co. (1986) 186 Cal. App. 3d 545, 557 [230 Cal. Rptr. 792]; CNA Casualty of California v. Seaboard Surety Co. (1986) 176 Cal. App. 3d 598, 620 [222 Cal. Rptr. 276].)

(2) Apportionment based upon the relative duration of each primary policy as compared with the overall period during which the "occurrences" "occurred" (the "time on the risk" method). (See, e.g., Ins. Co. North America v. Forty-Eight Insulations (6th Cir.1980) 633 F.2d 1212, clarified 657 F.2d 814, cert. den. (1981) 454 U.S. 1109 [70 L. Ed. 2d 650, 102 S. Ct. 686].)

(3) Apportionment based upon those relative durations, but with the respective durations multiplied by the amount of the respective limits of the primary policies — so that the insurers issuing primary policies with higher limits would bear a greater share of the liability per year than those issuing primary policies with lower limits. (See Armstrong, supra, 45 Cal. App.4th at pp. 52-53; Owens-Illinois, Inc. v. United Ins. Co. (1994) 138 N.J. 437 [650 A.2d 974, 993-995].)

(4) Apportionment based on the premiums paid. (See Insurance Co. of Tex. v. Employers Liability Assur. Corp. (S.D.Cal. 1958) 163 F. Supp. 143, 147, 151.)

(5) Apportionment among each carrier in equal shares. (See Reliance Ins. Co. v. St. Paul Surplus Lines Ins. (4th Cir.1985) 753 F.2d 1288, 1292.)

(6) Apportionment among each carrier in equal shares up to the policy limits of the policy with the lowest limits; then among each carrier other than the one issuing the policy with the lowest limits, in equal shares up to the policy limits of the policy with the next-to-lowest limits; and so on in the same fashion, until the loss has been apportioned in full (the "maximum loss" method). (See, e.g., Mission Ins. Co. v. Allendale Mut. Ins. Co. (1981) 95 Wash. 2d 464, 465-468 [626 P.2d 505].)

(d) It seems to us that the "time on the risk" method described in subparagraph (2) above is the approach likely to lead to the fairest result in most cases:

(1) Both the "policy limits" method and the permutation thereof described in subparagraph (3) above have the disadvantage of subjecting the insurer with high policy limits to a greater proportion of the loss in low-loss cases even though all carriers have agreed to cover the loss up to the limits of the policy with the lowest limits.^[20] The "maximum loss" method of apportionment avoids that problem, but both it and the "policy limits" method would have carriers that issue policies having identical limits but drastically different policy periods (and whose policies therefore are likely to have generated drastically different premiums and involved drastically different levels of risk exposure) bear equal shares of the liability.

(2) Apportionment based on the premiums paid has some superficial attractiveness (sort of an "as ye have reaped, so shall ye sew" approach). However, that approach ignores the multitude of different factors that can create differences in liability insurance premiums — factors such as differences in coverages in respects irrelevant to the liability actually incurred, inflation, differences in the level of exposure to the loss ultimately incurred at the time the policies are written, etc. These factors can justify a higher premium for policy X than for policy Y without in any way justifying imposition upon the insurer issuing policy X more of the liability to be apportioned than is imposed upon the insurer issuing policy Y.

(3) Apportionment among each carrier in equal shares is so arbitrary that its potential unfairness is patent. Under this approach a company that issued a primary policy with a duration of the first year of an 11-year continuous occurrence and collected a premium accordingly would have to bear the same share of the liability incurred for that 11-year occurrence as a company that issued a policy with identical coverage except with a duration of the last 10 years of the 11-year period and that collected a premium accordingly.

(4) Of course, the "time on the risk" method cannot be applied in such a way as to impose upon one carrier a greater share of the liability than its policy limits. If the liability to be apportioned is so great that its apportionment pursuant to the "time on the risk" rule exhausts the liability limits of a policy, once that level of apportionment of the liability has been reached, the remaining liability can only be apportioned (again on the basis of "time on the risk") among the carriers whose policy limits have not yet been exhausted. In such a situation the approach being utilized is really a "time on the risk" method qualified by a variation on the "maximum loss" method.

We believe that this "qualified time on the risk" method is the one which ordinarily should be used. However, as is suggested by Montrose, 10 Cal.4th at page 687, Signal Companies, Inc. v. Harbor Ins. Co., supra, 27 Cal. 3d 359, 369, and CNA Casualty of California v. Seaboard Surety Co., supra, 176 Cal. App. 3d 598, 620, there can be situations in which there are more equitable approaches.^[21] Accordingly, we do not direct the trial court here to use this rule but only to use it unless it determines that use of another approach would be more equitable.

(20) Accordingly, liability as among the insurers here involved should be apportioned as follows. First, the excess insurers are wholly free from liability (either to the City or to any of the other carriers) if in the aggregate there was adequate primary insurance in force from the beginning to the end of the period during which the Papworth property sustained damage caused by fault on the part of the City. (See pt. V.F., ante) Second, only Jefferson and possibly Fireman's is/are responsible for that portion of the City's obligation to Papworth attributable to the proportionate excess of inverse condemnation liability over negligence/nuisance liability. (See pt. V.B., ante.). Third, the remaining indemnification should be paid by and apportioned among all the primary insurance carriers whose policies incepted during the aforementioned period of damage to the Papworth property for which the City was at fault. Fourth, subject to the impact of applicable deductibles and retentions, the City is entitled to recover the $350,000 it paid to Papworth (plus interest thereon) from (i) Jefferson and possibly Fireman's to the extent of the proportion thereof representing inverse condemnation liability and (ii) all of the just-mentioned primary insurers with respect to the remainder, based upon the policy limits and the deductible, retention and "other insurance" clauses of those insurers' respective policies. Fifth, the trial court possesses discretion to apportion among the primary carriers just mentioned the burden of the $1.6 million total payout (plus interest on the $350,000 paid by the City),^[22] utilizing a method which is on the whole fair and reasonable. The "qualified time on the risk" method of apportionment described above is to be used unless another method would be more equitable.

I. The Failure of Insurers to Settle the Papworth Claim

The City contends that because some of its insurance carriers did not take the opportunity before trial to settle the Papworth claim for an amount less than the actual award, these carriers are now barred from asserting any exclusions from coverage. This issue remains open, in Phase II of the litigation. It must be resolved there. (We again hold, as we did in our 1992 opinion, that the trial court did not abuse its discretion in severing the Phase II claims for a later trial.)

J. Reallocation of Defense Costs as Between Jefferson and Other Carriers

Jefferson contends that the trial court erred in not allocating some portion of its defense costs to Admiral. Jefferson did not raise the issue in the trial court. It cannot now raise the issue on appeal. (Doers v. Golden Gate Bridge etc. Dist. (1979) 23 Cal. 3d 180, 184-185 [151 Cal. Rptr. 837, 588 P.2d 1261].)

VI. CANADIAN'S MOTION FOR SUMMARY JUDGMENT

(21) The trial court granted Canadian's motion for summary judgment on the theory that it was undisputed that the Papworth property was a total loss in February of 1980, a date before the inception of the Canadian policy. This conclusion was based upon the trial court's determination that the stipulated date of taking for the purposes of Papworth's inverse condemnation claim established the time of total loss. We conclude that as against anyone other than the parties to it (the City and Papworth), that stipulation is not conclusive;^[23] and opposition to the motion included a factual showing of substantial damage to the Papworth property in 1981, after the inception of Canadian's policy, thereby creating a triable issue as to a material fact. The trial court thus erred in granting the motion for summary judgment. (Code Civ. Proc. § 437c, subds. (c), (g).)

There is, however, a substantial remaining problem not raised in light of the trial court ruling. Canadian may not be liable, but for a different reason than that given by the trial court. It may be that undisputed facts disclose that the stipulated date of taking accomplished a merger of the City's obligations to Papworth. If the constructive taking encompassed all of Papworth's interest, it would seem that any damage thereafter accruing could not have been damage sustained by a third party (Papworth) for which the City could be held liable.

The trial court must vacate its order granting Canadian's summary judgment motion and enter an order denying that motion. Whether the trial court wishes to entertain a second summary judgment motion brought on the theory last mentioned is a matter within the trial court's discretion. (Gailing v. Rose, Klein & Marias (1996) 43 Cal. App. 4th 1570, 1579 [51 Cal. Rptr. 2d 381].)

Canadian did not participate in the trial of the action. However, its briefs disclose a position on coverage issues similar to that asserted by other carriers and discussed in this opinion. We leave to the trial court the question of the nature of further proceedings with respect to Canadian if no second summary judgment motion by Canadian is made, entertained or granted. It may be that absent an offer of proof by Canadian of new and different evidence with respect to the nature and cause of the Papworth loss, a full trial will not be necessary.

VII. DISPOSITION

Pursuant to the Supreme Court's order filed October 19, 1995, herein, our prior decision filed herein on May 15, 1992, as modified on June 11, 1992, upon denial of rehearing, is hereby vacated.

The order for summary judgment in favor of Canadian is hereby reversed with directions to the trial court to enter an order denying Canadian's summary judgment motion.

Any purported appeal from the judgment in favor of Central National entered June 12, 1989, is hereby dismissed.

Any purported appeal from the judgment in favor of Employers entered September 28, 1988, is hereby dismissed.

The notices of appeal from the Phase I judgment or aspects thereof are attempts to appeal from a nonappealable judgment. We treat those notices of appeal as petitions for a peremptory writ of mandate. Those petitions are granted in the following respects only: A peremptory writ of mandate shall issue ordering the trial court to do all of the following:

1. Set aside the Phase I judgment except to the extent that it is in favor of Central National and Employers.

2. Determine, in accordance with the principles stated in part V.B. of this opinion, whether Fireman's policies exclude inverse condemnation coverage.

3. Determine, in accordance with the principles stated in parts V.B., V.F., V.G., V.H. and VI. of this opinion, which carriers still in this litigation^[24] must reimburse the City for the $350,000 it paid to Papworth, less applicable deductibles and retentions, plus interest thereon.

4. Determine, in accordance with the principles stated in parts V.B., V.F., V.G., V.H. and VI. of this opinion, how the burden of the sums paid to Papworth by Stonewall and Jefferson^[25] and of the $350,000 to be paid to the City (less applicable deductibles and retentions), plus interest thereon, is to be allocated among the carriers still in this litigation.

5. Except with respect to Covenant, Central National and Employers, redetermine who shall be entitled to recover from whom costs incurred before the trial court.

6. Enter a new judgment on Phase I that implements the determinations the trial court makes pursuant to the foregoing provisions of the writ of mandate hereby ordered. The trial court may, of course, take additional evidence to the extent necessary to make the determinations described above.

In all other respects, the petitions for writ of mandate are denied.

The City, Stonewall, Employers, Central National, Maine Bonding, Puritan, and Covenant if it incurred these costs, shall recover costs on appeal and in these mandate proceedings payable jointly and severally by the other parties to the purported appeals from the Phase I judgment.

Lillie, P.J., and Woods, J., concurred.

A petition for a rehearing was denied July 12, 1996, and the petitions of all insurance companies for review by the Supreme Court were denied October 23, 1996.

NOTES

[*] Judge of the Los Angeles Superior Court, assigned by the Chief Justice pursuant to article VI, section 6 of the California Constitution.

[†] Reporter's Note: Review granted, May 21, 1992 (S026013).

[1] The fact that we decline to review the trial court's exoneration of Central National is quite unlikely to make any practical difference in the result in these consolidated cases, in view of the conclusions expressed in part V.G. 1., post, wherein we note that the possibility that any excess carrier will be obligated to contribute toward any of the City's liability to Papworth seems extremely remote.

[2] The August 16, 1989, judgment contains certain new provisions (relating to the recovery of costs incurred in these consolidated cases) concerning Central National. Our dismissal of any purported appeal from the June 12, 1989, judgment of course does not prevent us from considering provisions concerning Central National contained in the August 16, 1989, judgment but not contained in the June 12, 1989, judgment.

[3] Again, the fact that we decline to review the trial court's exoneration of Employers is quite unlikely to make any practical difference in the result in these consolidated cases, in view of the conclusions expressed in part V.G. 1., post, wherein we note that the possibility that any excess carrier will be obligated to contribute toward any of the City's liability to Papworth seems extremely remote.

[4] Again, the August 16, 1989, judgment contains certain new provisions (relating to the recovery of costs incurred in these consolidated cases) concerning Employers; and our dismissal of any purported appeal from the September 28, 1988, judgment of course does not prevent us from considering provisions concerning Employers not contained in the September 28, 1988, judgment but contained in the August 16, 1989, judgment.

[5] As noted above, a judgment in favor of Covenant, which issued a special public entity difference in condition comprehensive liability insurance policy covering the period November 1, 1976, to November 1, 1977, was entered by the trial court following the granting of Covenant's motion for judgment pursuant to Code of Civil Procedure section 631.8. As indicated above, no party has challenged that ruling before this court, so we shall not disturb that judgment.

Furthermore, other carriers providing excess coverage over Canadian's policy either were granted summary judgment without opposition and no appeal therefrom was taken or went into liquidation and ceased to be a party to the litigation.

[6] As indicated in Montrose, "[i]n the third party liability insurance context, `trigger of coverage' has been used by insureds and insurers alike to denote the circumstances that activate the insurer's defense and indemnity obligations under the policy." (10 Cal.4th at p. 655, fn. 2.)

[7] Section 14 of article I has since been renumbered section 19.

[8] Admiral asserts for the first time in its reply brief that the Canadian policy covered inverse condemnation liability by reason of an endorsement allegedly qualifying Canadian's inverse condemnation exclusion. We decline to consider the effect of that exclusion in view of the untimely assertion of the contention concerning it and the absence of any appearance of good cause for the untimeliness. (E.g., Neighbours v. Buzz Oates Enterprises (1990) 217 Cal. App. 3d 325, 335 [265 Cal. Rptr. 788]; Balboa Ins. Co. v. Aguirre (1983) 149 Cal. App. 3d 1002, 1010 [197 Cal. Rptr. 250].)

[9] This number of course does not include interest on the $1.6 million to those that paid it, but that fact is unimportant for purposes of the current discussion because each carrier who is liable will of course owe interest on its liability and the policy limits of the various policies of course do not limit the issuer's liability for interest on those policy limits.

[10] Other than deductibles and self-insured retentions.

[11] We shall address later in this opinion how the City's $50,000 per occurrence retention under the Admiral policy is to be taken into account, if at all.

[12] As indicated above, Central National also issued an excess policy with a policy period that included a portion of the period from December 31, 1975, to February 1, 1980. However, as also indicated above (in part IV. A. 2. (b)), we will not disturb the trial court's exoneration of Central National.

[13] For example, if (as contended by the City) the proportion of the $1.6 million payment to Papworth generated by the negligence/nuisance claims augmented by the taxable costs related thereto is a fraction in which the numerator is $1,391,412.57 and the denominator is $1,881,946.70, the pro rata liability of Jefferson and Admiral would be less than $300,000 per year — although Jefferson could of course potentially be liable for the full $300,000 because of the hereinabove-discussed waiver of its policy's inverse condemnation exclusion.

[14] That issue has recently been addressed exhaustively by Division One of the First District in an asbestos-related personal injury context. (Armstrong World Industries, Inc. v. Aetna Casualty & Surety Co. (1996) 45 Cal. App. 4th 1, 48-57 [52 Cal. Rptr. 2d 690].) While dictum in Armstrong indicates a different preference than we express as to the formula or method to be used in apportioning liability among insurers (we prefer what we call the "qualified time on the risk" method whereas the Armstrong Court apparently prefers a method that takes policy limits into account — see 45 Cal. App.4th at pp. 52-53), we both agree that the trial court has discretion to select a method that will produce the most equitable results. (Id. at p. 53; see Signal Companies, Inc. v. Harbor Ins. Co. (1980) 27 Cal. 3d 359, 369 [165 Cal. Rptr. 799, 612 P.2d 889, 19 A.L.R. 4th 75].) Also, it may be that in an asbestos or other bodily injury context a different method of apportionment is generally more equitable than in a property damage context, although we do not now so hold because no bodily injury claim is now before us.

[15] A recent attempted withdrawal by counsel for one primary carrier does contain information suggesting that that carrier may be insolvent.

[16] The summary judgment was reversed because the Supreme Court concluded that a triable issue existed as to the material fact of whether the insured in whose favor the summary judgment had been granted owed a duty to defend and therefore that it was error to enter summary judgment in favor of that insurer and against the insured. As we have hereinabove indicated, the issue of whether there was actual coverage therefore was never reached by the Supreme Court.

[17] Again, subject to the limits of the policies and to deductible and retention provisions.

[18] It may in fact be two questions, in that it is at least conceivable that the $3,000 in deductibles should be treated differently than the $50,000 retention.

[19] So far as we are aware, other than the Jefferson and Admiral policies, the only policy that is potentially a primary policy on the risk and had a deductible or retention provision in it is the Canadian policy. That policy had a $50,000 retention. As is set forth in part VI, post, we shall reverse the order for summary judgment in favor of Canadian. Upon remand, if the trial court finds that any "occurrence" occurred during the period of the Canadian policy, the $53,000 question will become a $103,000 question because the trial court will have to determine whether any carrier other than Canadian must pay anything to the City by reason of the retention in the Canadian policy.

[20] The Supreme Judicial Court of Maine eloquently pointed out this disadvantage in Carriers Ins. Co. v. Am. Policyholders' Ins. Co. (Me. 1979) 404 A.2d 216:

"[The policy limits method, the premiums paid method and the maximum loss method each are] grounded on the premises, often unarticulated, that on equitable principles the loss should be shared among the insurers either on the basis of the risk that they have undertaken or the benefit they have received. In its clearest expression, the [policy limits] rule has been justified on the theory that `the burden imposed on each insurer is generally proportional to the benefit which he received, since the size of the premium is most always directly related to the size of the policy.' [Citation.] On precisely these grounds, the [policy limits] rule has been criticized since `[i]t is commonly known that the cost of liability insurance does not increase proportionately with the policy limits.' Cosmopolitan Mutual Insurance Company v. Continental Casualty Co. [(1959) 28 N.J. 554,] note 4 at 564, 147 A.2d [529,] 534. Once minimum coverage has been obtained, significant supplemental coverage can be [obtained] at only a modest increase in cost.

"On the other hand, if the [policy limits] rule is less equitable than that ... approach which apportions on the basis of premiums received, it has the advantage of facile application. Unless the multiple policies cover the identical risks, there would be too many variables affecting the premiums to permit them to serve as a benchmark for an equitable adjustment. [Citation.]

"............................

"... [The policy limits] rule unfairly discriminates against the larger policy by apportioning the loss in proportion to the respective policy limits, utterly forgetting that both insurers, by their contracts, have in fact agreed to cover a loss up to the limits of the lesser policy. Until that point is reached, the [policy limits] rule amounts to no more than an unacceptable subsidy from the high-coverage to the low-coverage carrier....

"`The [policy limits] method of prorating operates inequitably in its differentiating treatment of the high-loss and low-loss insurer. In return for a greater premium the insurer providing higher coverage has undertaken to protect the insured against accidents involving high losses. Yet because of this undertaking to protect against high loss the larger insurer is in an unfavorable position vis-a-vis the other insurer even in cases of low loss, since under the [policy limits] method of prorating the insurer affording the greater maximum coverage pays the greater segment of any loss incurred, regardless of the amount of the loss. This seems inequitable since both insurers have equally undertaken to insure against the low-loss accident.'

"The [policy limits] rule would hardly encourage an insurer from increasing its coverage where it is aware that there is a lesser policy. It would increase the insurer's potential liability not only in the high-risk situation [in] which the additional premiums are presumably meant to recompense, but it would have the untoward effect of increasing liability in the more likely to occur low-risk situation. Carried to [the] extreme, it would further increase the cost of additional insurance thereby reducing the likelihood that an insured would choose such coverage.... The Court would be reluctant to adopt a rule which would seemingly have little social utility." (404 A.2d at pp. 221-222.)

[21] Indeed, one such situation is described in the excerpt from the trial court's decision quoted in the Armstrong opinion:

"`Typically a pro rata "other insurance" clause provides for proration according to the "applicable limit of liability."'" (Armstrong, 45 Cal. App.4th at p. 52.)

To the extent that policies involved in a case have pro rata "other insurance" clauses (as distinguished from other types of "other insurance" clauses), it may well be appropriate to take policy limits into account in apportioning among the insurers the ultimate burden of the insured's liability.

[22] Again, adjusted for applicable deductibles and retentions.

[23] As noted in part IV. A. 1., ante, Puritan and Admiral have appealed from the summary judgment in favor of Canadian.

[24] So that there will be no misunderstanding, Covenant, Central National and Employers will no longer be in this litigation once this decision becomes final, nor, of course, will any of the other carriers referred to in footnote 5 of this opinion.

[25] To the extent that Stonewall and Jefferson have overpaid their respective shares of the burden, they of course will be entitled to recover not only the excess amounts they paid but also interest thereon.