Opinion

Frontier Oil Corporation (Frontier) and its wholly owned subsidiary, Wainoco Oil & Gas Company (Wainoco), appeal a summary judgment in favor of RLI Insurance Company (RLI). Frontier and Wainoco contend RLI has a duty to defend them in several personal injury actions arising from the operation of an oil and gas production facility adjacent to Beverly Hills High School in Beverly Hills, California. Frontier’s predecessor and RLI’s predecessor had entered into a liability insurance policy in Texas. The parties dispute whether RLI agreed under the terms of the policy to defend pollution claims and whether a duty to defend the underlying actions has arisen. Each of these questions presents a choice-of-law issue concerning the law governing our determination. Each choice-of-law issue in turn presents the threshold question of the applicable choice-of-law rule.

We conclude that notwithstanding the application of the governmental interest analysis to other choice-of-law issues, Civil Code section 1646 is the choice-of-law rule that determines the law governing the interpretation of a contract. Section 1646 states that a contract is to be interpreted according to the law and usage of the place it is to be performed if the contract “indicate^] a place of performance” and according to the law and usage of the place it was made if the contract “does not indicate a place of performance.” A contract “indicate^] a place of performance” within the meaning of section 1646 if the contract expressly specifies a place of performance or if the intended place of performance can be gleaned from the nature of the contract and its surrounding circumstances. California, as the location of the risk insured under the policy, was the state where RLI would be obligated to perform its defense obligations under the policy, and the contracting parties knew this at the time the policy was issued. Indeed, the policy included several endorsements reflecting the existence of a covered risk located in California. The law of California therefore governs the interpretation of the policy. Interpreting the policy under California law, we will hold that it includes a contractual duty to defend and that the facts alleged in the underlying complaints were sufficient to create a potential for coverage giving rise to a duty to defend. We will therefore reverse the judgment.

FACTUAL AND PROCEDURAL BACKGROUND

1. The Insurance Policy

Underwriters Indemnity Company, RLI’s predecessor in interest, issued a commercial general liability insurance policy to Wainoco Oil Corporation, Frontier’s predecessor in interest, in January 1988. Wainoco Oil Corporation, acting through an insurance brokerage, and Underwriters Indemnity Company entered into the insurance contract in Texas. The policy was effective from October 1, 1987, to October 1, 1988.

The policy’s liability insuring clause states; “We will pay those sums that the insured becomes legally obligated to pay as damages because of ‘bodily injury’ or ‘property damage’ to which this insurance applies. . . . The ‘bodily injury’ or ‘property damage’ must be caused by an ‘occurrence’ ... .We will have the right and duty to defend any ‘suit’ seeking those damages.” (Italics added.) “Exclusion of,” in the same coverage form, provided for an “absolute” pollution exclusion: “This insurance does not apply to: . ‘Bodily injury’ or ‘property damage’ arising out of the actual, alleged or threatened discharge, dispersal, release or escape of pollutants . . . .”

The policy includes several contemporaneously issued endorsements, including one entitled “Oil and Gas Lease Operators’ Pollution Liability Coverage” (the pollution liability endorsement), which states at the top of its first page, “This endorsement forms a part of the policy to which attached, effective on the inception date of the policy unless otherwise stated herein.” The pollution liability endorsement deletes “exclusion of’ from the policy and states: “We will pay those sums that the ‘insured’ becomes legally obligated to pay as compensatory damages because of ‘bodily injury’ or ‘property damage’ to which this insurance applies, caused by a ‘pollution incident’.” The endorsement defines “pollution incident” as “the sudden and accidental emission, discharge, release or escape of pollutants into or upon land or the atmosphere, provided that such emission, discharge, release or escape emanates from operations conducted on land and results in ‘environmental damage’.” “Environmental damage” is defined as “the injurious presence in or upon land or the atmosphere of solid, liquid, gaseous or thermal contaminants, irritants or pollutants.” The endorsement, however, does not mention a duty to defend.

The policy also includes three additional endorsements relating specifically to the oil and gas operations in Beverly Hills, California. One adds the City of Beverly Hills as an additional insured (“with respect to claims arising out of the following project: Oil and Gas Operations at 9865 Olympic Blvd., City of Beverly Hill[s], CA”), apparently as a public entity that had issued a permit to conduct oil and gas operations at the site; the second adds the Department of Transportation of the City of Los Angeles as an additional insured, also apparently as a public entity that had issued a permit; and the third is a “Waiver of Transfer Rights of Recovery Against Others” made out in favor of the City of Beverly Hills with respect to claims arising from the same project.

Finally, an endorsement for certain “Texas Changes” apparently conforms this policy issued in Texas with Texas law with respect to three specific areas, none of which is relevant to the issues raised in this appeal: (1) the “notice prejudice” rule, (2) policy cancellation, and (3) policy renewal. That endorsement also states that the insured may complain to the Texas State Board of Insurance if any dispute concerning the premium or a claim is not resolved.

2. Underlying Actions and Tender of Defense

Lori Lynn Moss and numerous other plaintiffs filed a complaint against Frontier, Wainoco, and other oil and gas industry defendants in June 2003 (Moss v. Venoco, Inc. (Super. Ct. L.A. County, No. BC297083)). The plaintiffs alleged that the defendants’ oil and gas operations at “Drill-Site #1” and other locations at the Beverly Hills site caused releases of toxic chemicals into the environment resulting in personal injuries and deaths. Other plaintiffs filed similar complaints against Frontier, Wainoco, and others alleging the same operative facts in six additional actions filed from July 2003 to May 2005. Frontier and Wainoco tendered defense of all of these actions to several primary liability insurers, including RLI. RLI responded that it would investigate the matter.

In January 2004, Frontier and Wainoco made a written demand on numerous insurers, including RLI, to provide a defense in the underlying actions. On February 12, 2004, RLI filed a complaint in the United States District Court for the Southern District of Texas seeking a declaratory judgment that it had no duty to defend or indemnify Frontier or Wainoco in the underlying actions (RLI Insurance Co. v. Wainoco Oil & Gas Co. (S.D.Tex., No. H-04-0553)). RLI sent a letter to Frontier and Wainoco the next day denying coverage and a defense.

3. Trial Court Proceedings

Frontier and Wainoco filed their complaint in this action against RLI and other insurers in February 2004, alleging counts for (1) declaratory relief, (2) breach of contract, and (3) breach of the implied covenant of good faith and fair dealing, all relating to the insurers’ refusal to defend the underlying actions. The Texas federal district court stayed RLI’s declaratory relief action in April 2004 under its broad discretionary authority to abstain from ruling on a declaratory relief action where there is a pending state court action in which the matters in controversy might be fully resolved. Frontier and Wainoco later resolved this action with respect to all of the insurer defendants except RLI.

RLI moved for summary judgment against Frontier in November 2005, arguing that it had no duty to defend Frontier in the underlying actions. RLI argued that Texas law governed the dispute pursuant to Civil Code section 1646 and Code of Civil Procedure section 1857 and that under Texas law, RLI had no duty to indemnify or defend Frontier. Specifically, RLI argued that under Texas law (1) the pollution liability endorsement does not promise a defense of pollution claims, so the policy provides for indemnity of pollution claims but not a defense; and, in any event (2) the allegations in the underlying complaints do not create a potential for coverage. RLI moved for summary judgment against Wainoco on the same grounds, and also argued that Wainoco was entitled to no relief because it was not a named insured under the policy.

The trial court concluded that Civil Code section 1646 determined whether the law of California or Texas governed the dispute. Noting the “Texas Changes” endorsement, the court stated, “The insurance policy as a whole shows the intent of the parties at the time the insurance contract was made was that Texas law would apply to any disputes arising out of the contract.” The court concluded, “the insurance contract was made and accepted between a Texas insurer and a Texas-based insured in Texas and was to be performed under Texas law.” The court therefore held that Texas law governed this dispute. The court acknowledged that “[t]he major risk identified by the parties in their contract was the insured’s petroleum drilling operation in California,” but stated that such circumstance alone did not compel the conclusion that California law governed.

The trial court concluded that, under Texas law, the policy did not include a duty to defend pollution claims because the pollution liability endorsement did not expressly promise a defense. The court therefore held that RLI had no duty to defend Frontier or Wainoco regardless of the allegations in the underlying actions. The court granted the summary judgment motions against both Frontier and Wainoco and entered a judgment in favor of RLI. Frontier and Wainoco filed a timely appeal.

CONTENTIONS

Frontier and Wainoco contend (1) the governmental interest analysis is the choice-of-law rule that determines whether California law or Texas law governs the parties’ rights under the policy, and under that analysis California law applies; (2) even under Civil Code section 1646 and Code of Civil Procedure section 1857, the contemplated place of performance was California, so California law applies; (3) interpreted in accordance with California law, the policy includes a contractual duty to defend that extends to claims arising from pollution incidents; (4) the allegations of the third party complaints establish a potential for coverage under the policy and therefore create a duty to defend; and (5) RLI has a duty to defend under Texas law, even if it applies.

RLI disputes these contentions and argues that (1) Civil Code section 1646 and Code of Civil Procedure section 1857 establish the choice-of-law rule that determines whether California law or Texas law governs the interpretation of the policy and, under that analysis, Texas law applies because the policy does not expressly specify a place of performance and the contract was made in Texas; (2) even under the governmental interest analysis, Texas law applies; (3) interpreted in accordance with Texas law, the policy does not include a duty to defend pollution claims; (4) the allegations of the third party complaints do not establish a potential for coverage under the policy and therefore cannot support a duty to defend under Texas law; (5) if California law applies, the matter should be remanded to the superior court to consider extrinsic evidence that was not considered in ruling on the prior summary judgment motion; and (6) Wainoco is not an insured under the policy, so the judgment should be affirmed as to Wainoco.

DISCUSSION

1. Standard of Review

A party is entitled to summary judgment only if there is no triable issue of material fact and the party is entitled to judgment as a matter of law. (Code Civ. Proc., § 437c, subd. (c).) A defendant moving for summary judgment must show that one or more elements of the plaintiff’s cause of action cannot be established or that there is a complete defense. (Id., subd. (p)(2).) A defendant can satisfy its burden by presenting evidence that negates an element of the cause of action or evidence that the plaintiff cannot reasonably obtain needed evidence. (Kahn v. East Side Union High School Dist. (2003) 31 Cal.4th 990, 1003 [ 4 Cal.Rptr.3d 103 , 75 P.3d 30 ].) If the defendant meets this burden, the burden shifts to the plaintiff to set forth “specific facts” showing that a triable issue of material fact exists. (Code Civ. Proc., § 437c, subd. (p)(2).) We review the trial court’s ruling de novo, liberally construe the evidence in favor of the opposing party, and resolve all doubts concerning the evidence in favor of the opposing party. (Miller v. Department of Corrections (2005) 36 Cal.4th 446, 460 [ 30 Cal.Rptr.3d 797 , 115 P.3d 77 ].)

2. Choice-of-law Doctrine

Choice of law refers to the determination of which state’s or other jurisdiction’s law applies to a particular issue. Choice of law is one branch of the larger doctrine of conflicts of laws, which also includes the questions in which jurisdiction a suit can be brought and the effect of a foreign judgment. (See Weintraub, Commentary on the Conflict of Laws (5th ed. 2006) § 1.1, p. 1; Rest.2d Conflict of Laws, § 2, com. a, pp. 2-3.) Choice-of-law rules can be statutory or based on common law. (See Rest.2d Conflict of Laws, § 5, coms. b, c, p. 9; Leflar, Choice-of-Law Statutes (1977) 44 Tenn. L.Rev. 951.) Courts and commentators have long struggled with differing approaches to challenging choice-of-law problems.

The determination whether RLI has a duty to defend under the policy presents two principal issues. The first is whether the policy includes a duty to defend pollution claims. This is a question of contract interpretation. The second is whether the underlying personal injury actions trigger the duty to defend. Each of these questions presents a choice-of-law issue as to whether the law of California or another state governs. Each choice-of-law issue in turn presents the threshold question of which choice-of-law rule properly applies.

We first will decide that Civil Code section 1646 is the choice-of-law rule that determines the law governing the interpretation of the policy. We then will apply section 1646, conclude that California law governs, and interpret the policy under California law. Such interpretation will require us to conclude that the policy includes a duty to defend pollution claims. Next, we will consider the choice of law with respect to whether the underlying actions trigger the duty to defend. We will hold that California law governs under either section 1646 or the governmental interest analysis. Applying California law, we will conclude that RLI has a duty to defend Frontier and Wainoco in the underlying actions.

3. Civil Code Section 1646 Is California’s Choice-of-law Rule That Determines the Law Governing the Interpretation of a Contract

Civil Code section 1646 (see fn. 1, ante) was first enacted in 1872 based on an identical provision in the Field Code (Field’s Draft N.Y. Civ. Code (1865) § 811) and remains unchanged today. Published opinions by California courts have applied the statute as a choice-of-law rule determining the law governing the interpretation of a contract very infrequently, and those opinions (discussed post) provide little guidance as to the meaning of the critical phrase “indicate a place of performance” (Civ. Code, § 1646).

Our fundamental task in construing a statute is to ascertain the legislative intent so as to effectuate the purpose of the law. (Hassan v. Mercy American River Hospital (2003) 31 Cal.4th 709, 715 [ 3 Cal.Rptr.3d 623 , 74 P.3d 726 ].) Because the statutory language ordinarily is the most reliable indicator of legislative intent, we begin by examining the words of the statute. (Ibid.) We give the words of the statute their ordinary and usual meaning and construe them in the context of the statute as a whole and the entire scheme of law of which it is a part. (State Farm Mutual Automobile Ins. Co. v. Garamendi (2004) 32 Cal.4th 1029, 1043 [ 12 Cal.Rptr.3d 343 , 88 P.3d 71 ].) If the language is clear and a literal construction would not result in absurd consequences that the Legislature did not intend, we presume that the Legislature meant what it said and the plain meaning governs. (Coalition of Concerned Communities, Inc. v. City of Los Angeles (2004) 34 Cal.4th 733, 737 [ 21 Cal.Rptr.3d 676 , 101 P.3d 563 ].) If the language is ambiguous, we may consider a variety of extrinsic aids, including the purpose of the statute, legislative history, and public policy. (Ibid.)

Civil Code section 1646, by its express terms, prescribes a choice-of-law rule concerning the interpretation of contracts. It states that a contract is to be interpreted according to the law and usage of the place where it is to be performed, but only if the contract “indicate^] a place of performance,” and is to be interpreted according to the law and usage of the place where it was made if the contract “does not indicate a place of performance.” Frontier and Wainoco construe the quoted language to mean that a contract must be interpreted according to the law of the place where it is to be performed if the contract contemplates performance in a particular place, while RLI construes the same language to mean that the law of the place where the contract is to be performed governs only if the contract expressly specifies a place of performance.

The apparent purpose of Civil Code section 1646 is to determine the choice of law with respect to the interpretation of a contract in accordance with the parties’ presumed intention at the time they entered into the contract. This was the explanation given by courts and commentators for the former common law rule on which section 1646 apparently was based. Judge Joseph Story, in his highly regarded treatise on conflict of laws, described an exception to the general rule that the law of the place of making governed the validity and interpretation of a contract: “[W]here the contract is, either expressly or tacitly, to be performed in any other place, there the general rule is in conformity to the presumed intention of the parties that the contract as to its validity, nature, obligation, and interpretation is to be governed by the law of the place of performance.” (Story, Conflict of Laws (7th ed. 1872) § 280, p. 325.) Similarly, Kent’s Commentaries explained that the law of the place a contract is to be performed should govern because, “The rights of the parties are to be judged of by that law by which they intended, or rather by which they may justly be presumed to have bound themselves. [Citations.]” (2 Kent, Commentaries on American Law (12th ed. 1873) p. 622 (459) fn. 1.) Numerous contemporary judicial opinions followed this rule based on the parties’ presumed intention. (See, e.g., Vanzant Jones & Co. v. Arnold, Hamilton & Johnson et al. (1860) 31 Ga. 210 ; McDaniel v. The Chicago & Northwestern Railway Co. (1868) 24 Iowa 412 ; Dyke v. Erie Railway (1871) 45 N.Y. 113, 116-117 ; Waverly Bk., Appellant, v. Hall (1892) 150 Pa. 466, 472-473 [ 24 A. 665 ].)

In our view, Civil Code section 1646 was intended to give effect to the parties’ presumed intention that the law of the place a contract is to be performed should govern its interpretation. The parties’ intention as to the place of performance sometimes can be gleaned from the nature of the contract and the surrounding circumstances, even if the contract does not expressly specify a place of performance. This was Story’s view. (Story, Conflict of Laws, supra, § 280, p. 325 [“where the contract is, either expressly or tacitly, to be performed in any other place, . . .” (italics added)]; see also Pomeroy v. Ainsworth (N.Y. Sup. 1856) 22 Barb. 118, 128 [“If no place of performance is expressly stated, or implied from the terms of the contract, the law of the place where it was made will govern” (italics added)].) Moreover, the use of the word “indicate” in section 1646 in lieu of a more restrictive word such as “specify” or “state” suggests that the Legislature intended a less restrictive meaning. Accordingly, we conclude that a contract “indicate^]” a place of performance within the meaning of section 1646 if the intended place of performance can be gleaned from the nature of the contract and its surrounding circumstances. The intended place of performance is a question of contract interpretation for the court to decide, except to the extent the answer may depend on the credibility of extrinsic evidence. (Parsons v. Bristol Development Co. (1965) 62 Cal.2d 861, 865 [ 44 Cal.Rptr. 767 , 402 P.2d 839 ]; American Alternative Ins. Corp. v. Superior Court (2006) 135 Cal.App.4th 1239, 1245 [ 37 Cal.Rptr.3d 918 ].)

4. Prior California Opinions Offer Little Guidance on the Application of Civil Code Section 1646 As a Choice-of-law Rule

The California Courts of Appeal have cited Civil Code section 1646 in several opinions involving choice-of-law issues, but those opinions offer little or no meaningful guidance as to the meaning of “indicate a place of performance.”

Blochman etc. Bank v. Ketcham (1918) 36 Cal.App. 284 [ 171 P. 1084 ] (Blochman) was the first opinion by a California court to expressly invoke Civil Code section 1646 as a choice-of-law rule. Blochman involved a promissory note executed and delivered in Mexico, with the place of payment left blank. The plaintiff later inserted in the blank a location in California. (Blochman, supra, at p. 285 .) Blochman concluded that the parties intended to allow the holder to designate a place for payment. The court stated that this was the law of California when a blank was left in a note for the holder to fill in, stated that there was nothing in the record to indicate that Mexican law was to the contrary, and therefore presumed that Mexican law was the same. (Id. at pp. 286-287.) Although this discussion concerned the choice of law with respect to the interpretation of the contract, Blochman did not cite section 1646 in this context. Instead, Blochman cited section 1646 in support of its conclusion that California law, rather than the law of Mexico, governed the validity and enforceability of the note because the note was payable in California. ( 36 Cal.App. at p. 287 .) Blochman did not discuss whether there was any indication that the contracting parties anticipated that the note would be made payable in California.

Blair v. New York Life Ins. Co. (1940) 40 Cal.App.2d 494 [ 104 P.2d 1075 ] (Blair) was an action by an insured for payment of disability insurance benefits. The defendant executed and issued the policy in New York, and delivered it to the insured in the State of Washington. The insured moved to California several years later. The insured received disability benefits in Washington and California before the defendant discontinued the payment of benefits based on the insured’s failure to disclose material facts concerning her physical condition. (Id. at p. 496.) The choice-of-law issue concerned the interpretation of an incontestability clause. (Id. at p. 500.) Blair quoted Civil Code section 1646 and stated that the statute “means that the place where the contract is made is of importance when the contract does not indicate a specific place of performance.” (Blair, supra, at p. 498 .) After an intervening discussion, Blair stated, “The law of the place of performance of an insurance contract controls as to its legal construction and effect, but that of the place where made governs on all questions of execution and validity (Flittner v. Equitable Life Assur. Soc. [(1916)] 30 Cal.App. 209 [ 157 Pac. 630 ]), unless the terms of the contract provide otherwise or the circumstances indicate a different intention.” (Id. at p. 499.) Blair did not cite section 1646 with regard to this last quoted statement.

The terms of the insurance policy in Blair provided for payment of premiums at the defendant’s home office in New York or to an authorized agent elsewhere. (Blair, supra, 40 Cal.App.2d at p. 498 .) Blair concluded, “[u]nder this provision, the contract was to be performed partly in New York and partly in any other place where the premiums were paid or other necessary business could be transacted.” (Id. at p. 499; but see Flittner v. Equitable Life Assur. Soc., supra, 30 Cal.App. at p. 213 , reaching a different conclusion on similar facts.) Blair stated that the contract was performed partly in New York and partly in Washington, that the defendant waived the right to apply Washington law by paying benefits to the plaintiff in California, and that the law of California governed because the breach occurred in California. (Blair, at p. 499; see also Braun v. New York Life Ins. Co. (1941) 46 Cal.App.2d 335, 344-346 [ 115 P.2d 880 ] [followed Blair and declined to discuss the argument that Blair failed to give effect to Civ. Code, § 1646].) The holding in Blair apparently was based on the stated choice-of-law rule that in the circumstances of that case, liability for breach of contract was governed by the laws of the state where the breach occurred; the holding apparently was not based on section 1646.

Shippers Dev. Co. v. General Ins. Co. (1969) 274 Cal.App.2d 661 [ 79 Cal.Rptr. 388 ] (Shippers) involved the interpretation of an insurance policy exclusion. (Id. at pp. 673-674.) Shippers stated that the insurer relied on Civil Code section 1646, quoted the statute, and then stated: “The insurer’s argument fails to consider the following rule: ‘The law of the place of performance of an insurance contract controls as to its legal construction and effect, but that of the place where made governs on all questions of execution and validity [citation], unless the terms of the contract provide otherwise or the circumstances indicate a different intention. In this case, the circumstances do not indicate that in the matter of the interpretation of the contract, the parties intended to be restricted to the laws of the State of Washington, where the contract was consummated by the delivery of the policy.’ (Blair v. New York Life Ins. Co.[, supra,] 40 Cal.App.2d [at p.] 499 [ 104 P.2d 1075 ].) It was contemplated that the policy would be effective and might have to be performed outside the State of Washington. The law of this state in interpreting the policy will govern the extent of the coverage with respect to losses occurring within this state. [Citation.]” (Shippers, supra, at p. 674 , fn. omitted.) Thus, it appears that rather than construe and apply section 1646, Shippers followed dicta from Blair . Although it appears that the dicta may be somewhat consistent with section 1646 with respect to the law governing contract interpretation, we must construe and apply the statute rather than a purported rule of law of some other origin.

Some California opinions citing Civil Code section 1646 have concluded that there was no conflict between the laws of the states involved, and therefore have not determined whether the contract indicated a place of performance. (See, e.g., Gitano Group, Inc. v. Kemper Group (1994) 26 Cal.App.4th 49 , 56-57 & fn. 4 [ 31 Cal.Rptr.2d 271 ]; International Service Ins. Co. v. Gonzales (1987) 194 Cal.App.3d 110, 116 [ 239 Cal.Rptr. 341 ].) Other opinions have cited section 1646 to support conclusions on choice-of-law issues other than the law governing the interpretation of a contract (e.g., Hambrecht & Quist Venture Partners v. American Medical Internat., Inc. (1995) 38 Cal.App.4th 1532, 1540-1541 [ 46 Cal.Rptr.2d 33 ] [statute of limitations]; Klein v. Superior Court (1988) 198 Cal.App.3d 894, 902 [ 244 Cal.Rptr. 226 ] [formation of an oral contract]; Henderson v. Superior Court (1978) 77 Cal.App.3d 583, 592-593 [ 142 Cal.Rptr. 478 ] [validity and enforceability of a contract, statute of limitations, statute of frauds, and parol evidence rule]; Wheeling v. Financial Indem. Co. (1962) 201 Cal.App.2d 36, 40 [ 19 Cal.Rptr. 879 ] [validity of an insurance policy exclusion]; Bracker v. American Nat. Food, Inc. (1955) 133 Cal.App.2d 338, 344 [ 284 P.2d 163 ] [measure of contract damages]; Hardy v. Musicraft Records, Inc. (1949) 93 Cal.App.2d 698, 702 [ 209 P.2d 839 ] [validity of a stock sale under securities laws]; Blochman, supra, 36 Cal.App. at p. 287 [validity of a promissory note]), without explaining why a statute that by its express terms establishes a choice-of-law rule only as to the interpretation of a contract should determine other choice-of-law issues. These and other California opinions offer little or no analysis of the meaning of “indicate a place of performance” in section 1646.

5. The Governmental Interest Analysis Does Not Supplant Civil Code Section 1646 As to the Law Governing the Interpretation of Contracts

The most prevalent modem choice-of-law mle in California is the governmental interest analysis. It replaced former common law choice-of-law rules, and is sometimes labeled California’s modem choice-of-law mle. Some courts and commentators (discussed post) have inferred or suggested that the judicially developed governmental interest analysis supplants the choice-of-law rale stated in Civil Code section 1646. Our careful review of the governmental interest analysis as developed by the California Supreme Court, however, causes us to conclude that it does not supplant the legislative command of section 1646.

Under the governmental interest analysis, the court first determines whether the applicable rales of law of the potentially concerned jurisdictions are the same or different. If the applicable rales of law are identical, the court may apply California law. If the applicable rales of law differ materially, the court proceeds to the second step, which involves an examination of the interests of each jurisdiction in having its own law applied to the particular dispute. If each jurisdiction has an interest in applying its own law to the issue, there is a “true conflict” and the court must proceed to the third step. In the third step, known as the comparative impairment analysis, the court determines which jurisdiction has a greater interest in the application of its own law to the issue or, conversely, which jurisdiction’s interest would be more significantly impaired if its law were not applied. The court must apply the law of the jurisdiction whose interest would be more significantly impaired if its law were not applied. (Kearney v. Salomon Smith Barney, Inc. (2006) 39 Cal.4th 95, 107-108 [ 45 Cal.Rptr.3d 730 , 137 P.3d 914 ] (Kearney); Washington Mutual Bank v. Superior Court (2001) 24 Cal.4th 906, 919-920 [ 103 Cal.Rptr.2d 320 , 15 P.3d 1071 ] (Washington Mutual).)

The California Supreme Court first definitively adopted the governmental interest analysis for use in tort actions in Reich v. Purcell (1967) 67 Cal.2d 551 [ 63 Cal.Rptr. 31 , 432 P.2d 727 ] (Reich), a wrongful death action arising from the collision of two automobiles in Missouri driven by residents of California and Ohio. (Id. at p. 552.) The choice-of-law issue involved the amount of tort damages. A Missouri statute limited the amount of damages in a wrongful death action, while there were no such limitations under the laws of California and Ohio. (Id. at pp. 552-553.) Reich rejected the former common law rule that the law of the place of the wrong governed in tort actions. (Id. at pp. 553, 555.) Reich stated, “The forum must search to find the proper law to apply based upon the interests of the litigants and the involved states” (id. at p. 553), and stated that the former rigid rule “must be subordinated to the objective of proper choice of law in conflict cases, i.e., to determine the law that most appropriately applies to the issue involved [citation]” (id. at p. 555). Reich concluded that California had no interest in applying its law to the benefit of the defendant because California did not limit the amount of damages awardable, and that Missouri had no interest in applying its law limiting damages to the benefit of a defendant who was not a resident of that state as of the date of injury. The court therefore concluded that the law of Ohio governed the amount of damages and that the Missouri limitation on damages did not apply. (Id. at pp. 555-556.)

The California Supreme Court subsequently applied the governmental interest analysis in several other cases involving questions of tort and contract law. Travelers Ins. Co. v. Workmen’s Comp. App. Bd. (1967) 68 Cal.2d 7 [ 64 Cal.Rptr. 440 , 434 P.2d 992 ] was an action for workmen’s compensation benefits. The plaintiff originally entered into an oral employment contract in California, and later was injured in Utah. (Id. at pp. 9-11.) The plaintiff was entitled to benefits under California’s workmen’s compensation statute only if he were injured while working under an employment contract that was entered into in California. (Id. at p. 11.) The choice-of-law issues concerned the place of formation of the employment contract, whether an employment agency in Colorado acted as the employer’s agent for purposes of transmitting an employment offer to the plaintiff, and whether the parties had rescinded the oral contract and formed a new employment contract in Wyoming. (Ibid.) Travelers decided each of these questions by considering the governmental interests of the states involved, stating, “California has rejected the traditional mechanical solutions to choice-of-law problems and adopted foreign law only when it is appropriate in light of the significant interests in the particular case.” (Id. at p. 11, fn. omitted.) Travelers concluded that California had a strong interest in applying its own laws to determine whether an injured Californian was entitled to benefits under the workmen’s compensation act and that the other states involved had no significant interest in applying their laws to the dispute. (Id. at pp. 11-14.) The court therefore concluded that California law governed the place-of-contracting, agency, and rescission issues. (Id. at p. 14.)

Hurtado v. Superior Court (1974) 11 Cal.3d 574 [ 114 Cal.Rptr. 106 , 522 P.2d 666 ] was a wrongful death action arising from a collision in California involving three automobiles driven by residents of California and Mexico. (Id. at p. 578.) The choice-of-law issue involved the amount of tort damages. Hurtado concluded that Mexico had no interest in the application of its law limiting the amount of damages in a wrongful death action to the benefit of defendants who were not residents of Mexico, so there was no true conflict and the measure of damages of California, as the forum state, should apply. (Id. at pp. 580-582.)

Bernhard v. Harrah’s Club (1976) 16 Cal.3d 313 [ 128 Cal.Rptr. 215 , 546 P.2d 719 ] (Bernhard) involved a collision in California between an automobile driven by a California resident and a motorcycle driven by the plaintiff, who was also a California resident. The automobile driver allegedly was intoxicated after drinking liquor served by the defendant at a gambling casino in Nevada. (Id. at pp. 315-316.) The choice-of-law issue concerned the existence of tort liability. The allegations of the complaint supported negligence liability under California law, but did not support liability under Nevada law. (Id. at p. 317.) Bernhard concluded that each state was interested in the application of its own law, so there was a true conflict, and that the true conflict should be resolved by a comparative impairment analysis. (Id. at pp. 318-321.) Bernhard concluded that California’s interest in preventing the sale of alcoholic beverages to intoxicated persons whose intoxicated state is likely to contribute to injuries in California would be more significantly impaired than would Nevada’s interest in protecting its tavern keepers from unlimited liability. (Id. at pp. 322-323.) The court therefore concluded that California law applied and that the complaint alleged facts sufficient to state a cause of action under California law. (Id. at pp. 323, 325.)

Offshore Rental Co. v. Continental Oil Co. (1978) 22 Cal.3d 157 [ 148 Cal.Rptr. 867 , 583 P.2d 721 ] (Offshore Rental) was a negligence action arising from a personal injury suffered by an officer of the plaintiff corporation while on the defendant’s premises in Louisiana. The plaintiff sought to recover damages for injuries to its business interests resulting from the loss of the officer’s services. (Id. at pp. 160-161.) The choice-of-law issue concerned the existence of tort liability. Offshore Rental assumed that California law (Civ. Code, § 49) allowed a cause of action for negligent injury to a business employee, and concluded that Louisiana law did not allow such a cause of action. (Offshore Rental, supra, at pp. 162-163.) Offshore Rental noted that no California court had squarely held that the California statute supported such a cause of action, while a Louisiana appellate court nine years earlier had held that Louisiana law did not allow a cause of action by a corporate plaintiff for the loss of services of its officer. (Id. at pp. 162, 168.) Offshore Rental characterized the California statute as “unusual and outmoded” and concluded that California’s interest in establishing liability under the statute was less strong than Louisiana’s interest in the application of its “ ‘prevalent and progressive’ ” law. (Id. at p. 168.) The court therefore concluded that Louisiana law applied and that the trial court’s dismissal of the cause of action was proper. (Id. at pp. 169-170.)

Washington Mutual, supra, 24 Cal.4th 906 , was an action against a home mortgage lender for breach of contract, breach of the implied covenant of good faith and fair dealing, unfair business practices under the unfair competition law (UCL) (Bus. & Prof. Code, § 17200 et seq.), unjust enrichment, and conversion. The deed of trust included a choice-of-law provision stating that the security instrument was governed by federal law and the law of the jurisdiction where the secured property was located. (Washington Mutual, supra, at p. 912.) The trial court certified the action as a nationwide class action without determining which law would apply to the class members. (Id. at p. 913.) The Washington Mutual court stated that a consideration of the law applicable to the class members’ claims was essential to an informed decision on class certification. (Id. at p. 915.) It also stated that the trial court must apply the analysis set forth in Nedlloyd Lines B.V. v. Superior Court, supra, 3 Cal.4th 459 , to determine whether the class causes of action were subject to enforceable choice-of-law agreements. (Washington Mutual, supra, at pp. 915-916.) The court also stated that if the choice-of-law agreements were unenforceable or did not apply to the class causes of action and the party opposing class certification continued to assert that the law of another state applied to nonresident class members, the trial court must apply the governmental interest analysis to determine which state’s law to apply. (Id. at pp. 918-919.) Thus, the choice-of-law issues concerned the law governing the plaintiffs’ causes of action for breach of contract, breach of the implied covenant, unfair business practices, unjust enrichment, and conversion.

Washington Mutual, supra, 24 Cal.4th at page 919 , stated, “California follows a three-step ‘governmental interest analysis’ to address conflict of laws claims and ascertain the most appropriate law applicable to the issues where there is no effective choice-of-law agreement. [Citations.]” After describing the governmental interest analysis, the court concluded, “[t]hese rules apply whether the dispute arises out of contract or tort [citations], and a separate conflict of laws inquiry must be made with respect to each issue in the case [citations].” (Id. at p. 920.) Washington Mutual reversed the judgment by the Court of Appeal with directions to issue a peremptory writ of mandate directing the trial court to vacate its certification order and conduct the choice-of-law analyses described in the opinion. (Id. at p. 929.)

In Kearney, supra, 39 Cal.4th 95 , clients of the defendant brokerage firm alleged that the defendant had surreptitiously recorded their telephone conversations without their consent in violation of a California statute (Pen. Code, § 632). The plaintiffs, in a class action complaint, alleged violations of the statute and unfair business practices under the UCL. ( Kearney, supra, at pp. 102-103 & fn. 1.) The choice-of-law issue concerned the law governing those statutory causes of action. The court stated that the allegations of the complaint supported the alleged causes of action under California law, but did not support liability under Georgia law. (Id. at pp. 120-122.) Kearney concluded that California’s interest in protecting individuals in California from the secret recording of their confidential telephone conversations would be more significantly impaired than would Georgia’s interest in protecting the right of a business to record telephone conversations for legitimate business reasons. (Id. at pp. 124-128.) The court therefore held that California law applied and that the complaint alleged facts sufficient to state a cause of action under California law. (Id. at pp. 323, 325.)

Some federal courts have concluded or suggested that the governmental interest analysis has judicially supplanted the choice-of-law rule stated in Civil Code section 1646 or that the state of California law on this point is uncertain. The Ninth Circuit in Strassberg v. New England Mut. Life Ins. Co. (9th Cir. 1978) 575 F.2d 1262 applied the governmental interest analysis to determine the law governing whether an insurance policy had lapsed due to nonpayment of insurance premiums. Strassberg stated, “California conflicts law has developed significantly since the original enactment of California Civil Code § 1646,” and “California law moved away from a mechanical choice of law process to employ the ‘governmental interest analysis.’ ” (Id. at p. 1263.) Later, in Arno v. Club Med Inc. (9th Cir. 1994) 22 F.3d 1464 , the Ninth Circuit applied the governmental interest analysis to determine the law governing the existence of an implied in fact contract. Amo stated that section 1646 would yield the same result and, “[t]here appears to be some difference of opinion as to whether California’s choice of law rule for contracts is the governmental interest test... or the test of Cal.Civ. Code § 1646.” (Arno, at pp. 1468-1469 & fn. 6.) Although section 1646 states a choice-of-law rule only as to contract interpretation, Strassberg and Amo, in making these statements, did not distinguish contract interpretation from other choice-of-law issues.

The Sixth Circuit in Northland Ins. Co. v. Guardsman Products, Inc. (6th Cir. 1998) 141 F.3d 612 applied the governmental interest analysis to determine the law governing the interpretation of an insurance policy, stating, “the California Supreme Court ‘has moved away from the mechanical application of § 1646’ ” and “ ‘now employs what is commonly termed the “governmental interest analysis” approach to choice-of-law questions.’ [Citation.]” (I d. at p. 617.) Similarly, Vestrock Partners v. California Energy Co. (S.D.N.Y., Aug. 26, 1993, No. 92 Civ. 5690) 1993 WL 328912 stated with regard to the governmental interest analysis, “California courts normally employ the choice of law test in contracts cases notwithstanding § 1646.” (Id. at p. *5; see also Reppy, Choice of Law Problems Arising when Unmarried Cohabitants Change Domicile (2002) 55 SMU L.Rev. 273, 292, fn. 90 [stating, “the California Supreme Court’s adoption of interest analysis had worked a judicial repeal of the . . . choice of law rule in section 1646”].) Again, in making these statements, these authorities did not distinguish contract interpretation from other choice-of-law issues.

After reviewing all of these authorities, and in light of the fact that our Supreme Court has not addressed this issue, we are unable to conclude that California has “judicially abrogated” the express legislative mandate of Civil Code section 1646 relating to the interpretation of contracts. Instead, we hold that the choice-of-law rule in Civil Code section 1646 determines the law governing the interpretation of a contract, notwithstanding the application of the governmental interest analysis to other choice-of-law issues. The California Supreme Court has never applied the governmental interest analysis to determine the law governing the interpretation of a contract and has never stated or suggested that section 1646 does not determine the law governing the interpretation of a contract. Broad statements declaring the governmental interest analysis California’s choice-of-law rule and rejecting traditional choice-of-law rules should be viewed in light of the particular choice-of-law issues that were before the court in those cases.

The opinions by the California Courts of Appeal cited by Frontier and Wainoco are not inconsistent with our conclusion that Civil Code section 1646 determines the law governing contract interpretation notwithstanding the application of the governmental interest analysis to other choice-of-law issues. Robert McMullan & Son, Inc. v. United States Fid. & Guar. Co. (1980) 103 Cal.App.3d 198 [ 162 Cal.Rptr. 720 ] was a declaratory relief action against a liability insurer. The trial court determined that the insurer had a duty to defend an action against the insured, and the question on appeal was whether the insured was entitled to recover its attorney fees incurred in the declaratory relief action. Applying the governmental interest analysis, the McMullan court concluded that the laws of California and the other states involved did not differ on that question and that even if the laws differed, only California had an interest in applying its own laws to the dispute. (Id. at pp. 203-206.) The choice-of-law issue concerned the right to recover attorney fees in an action to enforce the insured’s rights under the policy and did not involve an issue of contract interpretation.

Stonewall Surplus Lines Ins. Co. v. Johnson Controls, Inc. (1993) 14 Cal.App.4th 637 [ 17 Cal.Rptr.2d 713 ] (Stonewall) was a declaratory relief action by excess liability insurers against their insured. The trial court concluded that the insurers were not required to indemnify the insured for a punitive damages award because California law, for reasons of public policy, prohibits indemnity for punitive damages. (Id. at pp. 640-641.) The question on appeal was whether the insurers could be required to indemnify the insured for punitive damages. Applying the governmental interest analysis, the Court of Appeal concluded that the laws of California and Wisconsin differed on that question and that California had a greater interest in applying its own laws to the dispute. (Id. at pp. 645, 649.) In deciding that California law governed, Stonewall placed particular importance on the location of the insured risk. (Id. at pp. 646-647, citing Rest.2d Conflict of Laws, § 193 & com. f, pp. 613-614.) The choice-of-law issue concerned the existence of a right of indemnity for punitive damages and did not involve an issue of contract interpretation.

6. The Intended Place of Performance of a Liability Insurance Policy Is the Place of the Insured Risk

The liability insurance policy at issue in this case includes both indemnity and defense obligations. An indemnity obligation “entails the payment of money in order to resolve liability. [Citations.]” (Buss v. Superior Court (1997) 16 Cal.4th 35, 46 [ 65 Cal.Rptr.2d 366 , 939 P.2d 766 ] (Buss).) A defense obligation, in contrast, “entails the rendering of a service, viz., the mounting and funding of a defense [citations].” (Ibid.) An insurer performs its defense obligation by providing defense services through an attorney. If the insured risk involves operations at one or more fixed locations, as here, a third party complaint against the insured arising from those operations typically is prosecuted in the jurisdiction where the operations are located. The insurer provides defense services in the jurisdiction where the suit is prosecuted. There is little doubt that this was the understanding of the parties at the time they entered into the insurance contract in this case.

The policy provides general liability coverage and specifically refers to claims arising from oil and gas operations at “Drill-Site #1” in Beverly Hills, California. Two policy endorsements name the City of Beverly Hills and the Department of Transportation of the City of Los Angeles as additional insureds “with respect to claims arising out of... [f] Oil or Gas Operations” in the City of Beverly Hills, California. In another endorsement, RLI waives its right of recovery against the City of Beverly Hills for certain payments made under the policy with respect to the same specified claims. These three endorsements clearly demonstrate that the parties intended the policy to provide coverage for the insureds’ oil and gas operations in Beverly Hills. Accordingly, we conclude that the parties anticipated that a suit arising from those operations in Beverly Hills could be prosecuted in California and that RLI would be obligated to provide a defense in California if the claims were potentially covered under the policy.

We therefore conclude that California was the intended place of performance of the contract with respect to those claims, that the policy thus “indicate[s] a place of performance” within the meaning of Civil Code section 1646 with respect to such claims, and that California law governs the interpretation of the policy. Accordingly, we have no need to apply a governmental interest analysis or give consideration to Texas law with respect to the interpretation of the policy.

7. Interpreted Under California Law, the RLI Policy Includes a Duty to Defend Pollution Claims Arising from “Sudden and Accidental” Releases

We interpret an insurance policy under California law using the same rules of interpretation applicable to other contracts. (Palmer v. Truck Ins. Exchange (1999) 21 Cal.4th 1109, 1115 [ 90 Cal.Rptr.2d 647 , 988 P.2d 568 ].) “The mutual intention of the contracting parties at the time the contract was formed governs. (Civ. Code, § 1636; Palmer, supra, at p. 1115 .) We ascertain that intention solely from the written contract if possible, but also consider the circumstances under which the contract was made and the matter to which it relates. (Civ. Code, §§ 1639, 1647.) We consider the contract as a whole and interpret the language in context, rather than interpret a provision in isolation. (Id., § 1641.) We interpret words in accordance with their ordinary and popular sense, unless the words are used in a technical sense or a special meaning is given to them by usage. (Id., § 1644.) If contractual language is clear and explicit and does not involve an absurdity, the plain meaning governs. (Id., § 1638.)” (American Alternative Ins. Corp. v. Superior Court, supra, 135 Cal.App.4th at p. 1245 .)

We interpret an insuring clause broadly, consistent with the reasonable expectations of the insured. (MacKinnon v. Truck Ins. Exchange (2003) 31 Cal.4th 635, 648 [ 3 Cal.Rptr.3d 228 , 73 P.3d 1205 ]; AIU Ins. Co. v. Superior Court (1990) 51 Cal.3d 807, 822 [ 274 Cal.Rptr. 820 , 799 P.2d 1253 ].) An exclusion from coverage otherwise within the scope of an insuring clause must be clear and unmistakable to be given effect. (MacKinnon, supra, at p. 648 .) “ ‘The exclusionary clause “must be conspicuous, plain and clear.” ’ [Citation.]” (Ibid.) An exception to an exclusion is treated in the same manner as a coverage provision and therefore is interpreted broadly, consistent with the insured’s reasonable expectations. (TRB Investments, Inc. v. Fireman’s Fund Ins. Co. (2006) 40 Cal.4th 19, 27-28 [ 50 Cal.Rptr.3d 597 , 145 P.3d 472 ].)

An endorsement modifies the basic insuring forms of the policy and is an integral part of the policy. (Adams v. Explorer Ins. Co. (2003) 107 Cal.App.4th 438, 450-451 [ 132 Cal.Rptr.2d 24 ].) “Standing alone, an endorsement means nothing. ‘Endorsements on an insurance policy form a part of the insurance contract [citation], and the policy of insurance with the endorsements and riders thereon must be construed together as a whole [citation].’ ” (Id. at p. 451.) An endorsement can expand or restrict the coverage otherwise provided by the policy. If there is any conflict between an endorsement and the body of a policy, the endorsement controls, provided that any reduction in coverage reasonably expected under the body of a policy must be “ ‘conspicuous, plain and clear.’ ” (Haynes v. Farmers Ins. Exchange (2004) 32 Cal.4th 1198, 1208 [ 13 Cal.Rptr.3d 68 , 89 P.3d 381 ].)

The policy here expressly promises both indemnity and a defense. The insuring clause in the commercial general liability coverage form provides coverage for sums that the insured becomes legally obligated to pay as damages because of bodily injury or property damage caused by “an ‘occurrence,’ ” and states that RLI has the right and duty to defend any suit seeking those damages. “Exclusion of’ in the same coverage form states that the insurance does not apply to bodily injury or property damage caused by the actual or threatened release of pollutants at premises owned, occupied, or used by the insured. The pollution liability endorsement deletes exclusion of and provides coverage for sums that the insured becomes legally obligated to pay as damages because of bodily injury or property damage caused by “a ‘pollution incident.’ ” The endorsement does not mention the duty to defend, does not clearly and unmistakably exclude pollution claims from the duty to defend stated in the body of the policy, and therefore does not exclude pollution claims from the contractual duty to defend.

The insuring clause in the pollution endorsement provides coverage for damages arising from “a ‘pollution incident,’ ” which is defined in relevant part as a “sudden and accidental” release resulting in environmental damage. California courts interpret the phrase “sudden and accidental” when used in this context to mean abrupt in time, unexpected, and unintended. (Shell Oil Co. v. Winterthur Swiss Ins. Co. (1993) 12 Cal.App.4th 715, 755 [ 15 Cal.Rptr.2d 815 ].)

Interpreting the policy under California law, we conclude that RLI promised to defend claims for damages arising from pollution incidents with respect to the oil and gas operations in Beverly Hills notwithstanding that the pollution liability endorsement does not mention a duty to defend.

8. The Allegations of the Third Party Complaints Create a Duty to Defend

a. The Applicable Choice-of-law Rule

The fundamental principles under California law regarding whether a duty to defend arises were established in Gray v. Zurich Insurance Co. (1966) 65 Cal.2d 263 [ 54 Cal.Rptr. 104 , 419 P.2d 168 ] (Gray). Gray rejected the argument that an insurer had a duty to defend under the policy only if the insured was entitled to indemnity, and held that the duty to defend was independent of and broader than the duty to indemnify. (Id. at pp. 274—275.) Gray stated that uncertainties concerning policy interpretation must be resolved in favor of the reasonable expectations of the insured, and that in light of the policy language, the insured reasonably would expect the insurer to defend a suit involving a loss of the nature and kind covered by the policy. (Ibid.) Gray therefore concluded that an insurer must defend a suit that potentially seeks damages covered by the policy. (Id. at p. 275.)

Gray seemed to state that these conclusions reflected the court’s interpretation of policy provisions, as distinguished from a rule of law imposed on the parties for public policy reasons regardless of their intentions upon entering into the insurance contract. (Gray, supra, 65 Cal.2d at p. 274 .) This suggests that the fundamental principles discussed above regarding whether a duty to defend arises are rules of law regarding policy interpretation and that Civil Code section 1646 determines the appropriate choice of law governing these issues. On the other hand, it was not necessary for the court in Gray to decide whether these rules of law were rules of policy interpretation for purposes of section 1646; these rules could be characterized as public policy choices that do not necessarily reflect the actual or presumed intention of the parties. Moreover, Gray stated further that the duty to defend arises whenever the insurer ascertains facts that create a potential for liability under the policy, whether the insurer learns those facts from the complaint, the insured, or other sources, and did not explain that conclusion as a question of policy interpretation or refer to the reasonable expectations of the insured. (Gray, supra, at pp. 276-277 .) We need not decide whether these or other rules regarding the duty to defend are rules of policy interpretation for purposes of section 1646 because our conclusion under either section 1646 or the governmental interest analysis is that California law governs, as we will explain.

b. The Applicable Rules of Law of California and Texas Do Not Materially Differ

The first step in the governmental interest analysis is to determine whether the applicable rules of law of the potentially concerned jurisdictions materially differ. (Washington Mutual, supra, 24 Cal.4th at p. 919 .) If there is no material difference, there is no choice-of-law problem and the court may proceed to apply California law. (See id. at p. 920.) The party arguing that foreign law governs has the burden to identify the applicable foreign law, show that it materially differs from California law, and show that the foreign law furthers an interest of the foreign state. (Id. at p. 919; Bernhard, supra, 16 Cal.3d at pp. 317-318.)

A liability insurer has a duty to defend its insured under California law if facts alleged in the complaint, or other facts known to the insurer, potentially could give rise to coverage under the policy. (Scottsdale Ins. Co. v. MV Transportation (2005) 36 Cal.4th 643, 654-655 [ 31 Cal.Rptr.3d 147 , 115 P.3d 460 ] (Scottsdale); Gray, supra, 65 Cal.2d at pp. 275-277.) The facts need only “raise the possibility” that the insured will be held liable for covered damages. (Montrose, supra, 6 Cal.4th at p. 304.) An insurer has a duty to defend even if the claims against the insured are “ ‘groundless, false, or fraudulent.’ ” (Waller v. Truck Ins. Exchange, Inc. (1995) 11 Cal.4th 1, 19 [ 44 Cal.Rptr.2d 370 , 900 P.2d 619 ].) “Any doubt as to whether the facts establish the existence of the defense duty must be resolved in the insured’s favor.” (Montrose, supra, at pp. 299-300.)

A duty to defend arises under California law upon the tender to the insurer of a potentially covered claim and continues until the lawsuit is concluded or until the insurer shows that facts extrinsic to the third party complaint conclusively negate the potential for coverage. (Scottsdale, supra, 36 Cal.4th at p. 655; Montrose, supra, 6 Cal.4th at pp. 298-300.) If a duty to defend arises, the insurer must defend the action in its entirety, including claims that are not potentially covered. ( Buss, supra, 16 Cal.4th at pp. 48-49.) If a duty to defend has arisen by virtue of the existence of a potential for coverage but is later extinguished, it is extinguished prospectively only, and not retroactively. (Id. at p. 46.)

A liability insurer has a duty to defend its insured under Texas law if facts alleged in the complaint potentially could give rise to coverage under the policy. (GuideOne Elite v. Fielder Rd Baptist Church (Tex. 2006) 197 S.W.3d 305, 310 (GuideOne); Heyden Newport Chem. Corp. v. Southern Gen. Ins. Co. (Tex. 1965) 387 S.W.2d 22, 26 (Heyden).) “A plaintiff’s factual allegations that potentially support a covered claim is all that is needed to invoke the insurer’s duty to defend . . . [citation].” (GuideOne, supra, at p. 310.) “ ‘Where the complaint does not state facts sufficient to clearly bring the case within or without the coverage, the general rule is that the insurer is obligated to defend if there is, potentially, a case under the complaint within the coverage of the policy. Stated differently, in case of doubt as to whether or not the allegations of a complaint against the insured state a cause of action within the coverage of a liability policy sufficient to compel the insurer to defend the action, such doubt will be resolved in insured’s favor.’ ” (Heyden, supra, at p. 26 ; accord, Nat. Union Fire v. Merch. Fast Motor Lines (Tex. 1997) 939 S.W.2d 139, 141 (Nat. Union).) Under Texas law, the duty to defend is determined based on the policy terms and the allegations of the complaint, “without regard to the truth or falsity of those allegations” and without regard to extrinsic evidence. (GuideOne, supra, at p. 308; see also Heyden, supra, at p. 24 .) Any doubt as to whether the complaint alleges facts that give rise to a duty to defend must be resolved in favor of the insured. (Allstate Ins. Co. v. Hallman (Tex. 2005) 159 S.W.3d 640, 643 ; Heyden, supra, at p. 26 .) If the duty to defend arises, the insurer must defend the action in its entirety, including claims that are not potentially covered. (Warrantech Corp. v. Steadfast Ins. Co. (Tex.App. 2006) 210 S.W.3d 760, 766 ; Nokia, Inc. v. Zurich American Ins. Co. (Tex.App. 2006) 202 S.W.3d 384, 388 .)

Texas law differs from California law in that under Texas law, facts extrinsic to the complaint cannot give rise to a duty to defend. (GuideOne, supra, 197 S.W.3d at p. 308 ; Heyden, supra, 387 S.W.2d at p. 24 .) Under California law, in contrast, extrinsic facts known to the insurer give rise to a duty to defend if the facts create a potential for coverage under the policy. (Scottsdale, supra, 36 Cal.4th at p. 654; Gray, supra, 65 Cal.2d 263, 276-277 .) Because the complaints in the underlying actions allege facts that create a potential for coverage without regard to extrinsic evidence, as explained post, this difference between California law and the law of Texas is not relevant here. Apart from this distinction, the foregoing Texas rules of law do not materially differ from California law, and no party has shown that any other Texas rule of law applicable to these facts materially differs from California law. We therefore conclude that the applicable rules of law of California and Texas are substantially identical with respect to a liability insurer’s duty to defend; thus, even if the governmental interest analysis was the choice-of-law rule that we were to use to determine if a duty to defend ever arose, California law would be applied. (Washington Mutual, supra, 24 Cal.4th at p. 920 .)

c. The Third Party Complaints Allege Claims That Are Potentially Covered Under the Policy

The complaints allege that defendants, including Frontier and Wainoco, conducted oil and gas exploration, production, processing, and storage activities at Drill-Site #1. They allege that as a result of those operations, hazardous substances were “spilled, emitted, released, [and] discharged” into the environment. They allege further that the operations resulted in “releases, discharges, fugitive emissions, leaks and spills.” The complaints allege damage of a nature and kind covered by the policy and do not foreclose the possibility that the damage was caused by a sudden and accidental release, and therefore create a potential for coverage under the policy. (Waller v. Truck Ins. Exchange, Inc., supra, 11 Cal.4th at p. 19 ; Gray, supra, 65 Cal.2d at pp. 274—275.) This is sufficient to establish RTFs duty to defend. (Montrose, supra, 6 Cal.4th at pp. 299-300; Horace Mann Ins. Co. v. Barbara B. (1993) 4 Cal.4th 1076, 1082-1083 [ 17 Cal.Rptr.2d 210 , 846 P.2d 792 ].) RLI potentially could extinguish its defense duty prospectively, but cannot extinguish it retroactively. (Scottsdale, supra, 36 Cal.4th at p. 655; Buss, supra, 16 Cal.4th at p. 46 .) RLI failed to show that facts extrinsic to the third party complaints conclusively negated the potential for coverage before the termination of the underlying actions and therefore failed to extinguish the duty to defend before the termination of those actions. Because RLI failed to establish the absence of a duty to defend, RLI was not entitled to summary judgment against Frontier and Wainoco on that basis.

9. RLI Failed to Establish That Wainoco Is Not an Insured

The policy declarations identify the named insured as “Wainoco Oil Corporation, et al.” (Capitalization omitted.) An endorsement referenced on the declarations page and incorporated in the policy lists several additional named insureds, including “Wainoco Oil & Gas Company, and the Northwestern Mutual Life Insurance Co., A Joint Venture” and several “Wainoco Appalachian” partnerships. The commercial general liability coverage form states that if the named insured is a partnership or joint venture, the insured’s members and partners “are also insureds, but only with respect to the conduct of [the partnership’s or joint venture’s] business.”

RLI argued in its summary judgment motion against Wainoco and argues again on appeal that Wainoco is not an insured with respect to the underlying actions because the alleged conduct does not pertain to Wainoco’s joint venture activities. As the party moving for summary judgment, RLI had the burden to present either evidence that negates an element of the cause of action or evidence that Frontier and Wainoco cannot reasonably obtain needed evidence. (Kahn v. East Side Union High School Dist., supra, 31 Cal.4th at p. 1003.) RLI cited no evidence in support of its argument that Wainoco is not an insured apart from the policy and the third party complaints. A policy endorsement specifically identifies the Beverly Hills site and supports a reasonable inference that the parties intended the policy to provide coverage for Wainoco’s operations at the site. The third party complaints allege that Wainoco conducted oil and gas operations at the Beverly Hills site. Nothing in either the policy or the third party complaints forecloses the possibility that Wainoco’s operations at the Beverly Hills site pertained to its joint venture with Northwestern Mutual Life Insurance Company. Absent evidence to compel the conclusion that Wainoco’s operations at the site did not pertain to its joint venture business, RLI failed to establish that Wainoco is not an insured.

DISPOSITION

The judgment is reversed. The matter is remanded for further proceedings not inconsistent with the views expressed herein. Frontier and Wainoco are entitled to recover their costs on appeal.

Klein, R J., and Kitching, J., concurred.

On September 5, 2007, the opinion was modified to read as printed above. Respondent’s petition for review by the Supreme Court was denied November 14, 2007, S156343. George, C. J., did not participate therein.

Civil Code section 1646 provides: “A contract is to be interpreted according to the law and usage of the place where it is to be performed; or, if it does not indicate a place of performance, according to the law and usage of the place where it is made.” The effect of this endorsement was to remove the absolute pollution exclusion from the original policy and provide in its place a limited promise of coverage for claims arising from any “sudden and accidental emission, discharge, release or escape of pollutants.” The trial court also apparently relied on Code of Civil Procedure section 1857 (see fn. 4, post). Code of Civil Procedure section 1857 states, in terms somewhat similar to Civil Code section 1646: “The language of a writing is to be interpreted according to the meaning it bears in the place of its execution, unless the parties have reference to a different place.” In light of our conclusion that Civil Code section 1646 determines the law governing the interpretation of a contract and the apparent similarities between the two statutes, we need not decide whether Code of Civil Procedure section 1857 also states a choice-of-law rule or only an interpretive rule concerning usage. Civil Code section 1646 prescribes both a choice-of-law rule concerning the interpretation of contracts and a rule of interpretation regarding word usage. The California Supreme Court has cited section 1646 in several opinions to support the admissibility of evidence of usage for the purpose of interpreting a contract. (Callahan v. Stanley (1881) 57 Cal. 476, 479 ; Burns v. Sennett & Miller (1893) 99 Cal. 363, 371-372 [ 33 P. 916 ]; Law v. Northern Assurance Co. (1913) 165 Cal. 394, 407 [ 132 P. 590 ]; Alta Planing Mill Co. v. Garland (1914) 167 Cal. 179, 183 [ 138 P. 738 ]; Buckner v. Leon & Co. (1928) 204 Cal. 225, 227 [ 267 P. 693 ]; Alamitos Land Co. v. Shell Oil Co. (1935) 3 Cal.2d 396, 404 [ 44 P.2d 573 ]; Ermolieff v. R.K.O. Radio Pictures (1942) 19 Cal.2d 543, 550 [ 122 P.2d 3 ]; Beneficial etc. Ins. Co. v. Kurt Hitke & Co. (1956) 46 Cal.2d 517, 525-526 [ 297 P.2d 428 ].) Those opinions typically cited section 1646 together with other statutes stating that the words of a contract or writing are to be interpreted in accordance with any special meaning intended by the parties (e.g., Civ. Code, §§ 1644, 1645; Code Civ. Proc., § 1861), and did not discuss the meaning of “indicate a place of performance” in section 1646. The California Supreme Court has never cited section 1646 as a choice-of-law rule. The former common law rule described by these commentators was not limited to the choice-of-law issue of contract interpretation, as is Civil Code section 1646. If the parties state their intention in an express choice-of-law clause, California courts ordinarily will enforce the parties’ stated intention unless (1) the chosen state has no substantial relationship to the parties or their transaction, and there is no other reasonable basis for the parties’ choice, or (2) the chosen state’s law is contrary to a fundamental policy of the state whose law otherwise would apply, and the latter state has a materially greater interest in the matter than does the chosen state. (Nedlloyd Lines B.V. v. Superior Court (1992) 3 Cal.4th 459 , 464-466 & fns. 5 & 6 [ 11 Cal.Rptr.2d 330 , 834 P.2d 1148 ].) Both of these authorities were among the references cited in the notes to section 811 of the Field’s Draft New York Civil Code, which was the source for the identical language in Civil Code section 1646. Civil Code section 1646 states that a contract is to be interpreted according to the law and usage of the place it is to be performed if the contract indicates a place of performance, but does not state that the validity and enforceability of a contract are to be determined according to the law of the intended place of performance. Flittner v. Equitable Life Assur. Soc., supra, 30 Cal.App. 209 held that a minor could rescind a life insurance policy and recover the premiums paid, applying California law. (Id. at pp. 215-216.) The choice-of-law issue was whether the law of California or New York governed the minor’s right to rescind the contract. The contract was made in California, and Flittner concluded that the contract contemplated performance in New York. (Id. at p. 213.) Flittner stated: “All matters connected with the performance of a contract are regulated by the law of the place where the contract by its terms is to be performed. All matters bearing upon the execution, the interpretation, and the validity of contracts including the capacity of the parties to contract, are determined by the law where the contract is made (Union Nat. Bank v. Chapman, 169 N. Y. 538 , [ 88 Am.St.Rep. 614 , 57 L. R. A. 513, 62 N. E. 672 ]; Phoenix Mut. Life Ins. Co. v. Simons, 52 Mo.App. 357 ; 2 Wharton on Conflict of Laws, sec. 401; note in Mayer v. Roche, 77 N.J.L. 681 [ 75 A. 235 , 26 L. R. A. (N. S.) 767]; Hauck Clo. Co. v. Sharpe, 83 Mo.App. 385 ).” (Id. at p. 215, italics added.) Flittner did not cite Civil Code section 1646 with respect to the law governing the interpretation of a contract and cited no California authority in support of the quoted statement. Blair, supra, 40 Cal.App.2d at page 499 stated: “In recognizing California as the place for the payment of benefits by delivery thereof in this state, defendant waived any right to insist that the Washington law governed. The waiver is not one of jurisdiction, but of the right to apply the Washington law to an action tried in California. The intention of the parties, as gathered from attending circumstances, is controlling in determining the place of trial and the governing laws. Liability for the breach of a contract partly performed in one state, and made and partly performed in another, is fixed by the laws of the state wherein the breach occurred. (17 C. J. S., p. 343.) The breach in this case was defendant’s refusal to pay further disability income benefits to plaintiff in California.” Wheeling v. Financial Indem. Co., supra, 201 Cal.App.2d 36 and Estate of Grace (1948) 88 Cal.App.2d 956 [ 200 P.2d 189 ] reflect differing views on the meaning of “indicate a place of performance” (Civ. Code, § 1646). Wheeling stated that an automobile liability insurance policy “was made and delivered in California, and did not specify a place of performance,” and therefore applied California law to a claim arising from an accident in Virginia. (Wheeling, supra, at p. 40 .) In contrast, Estate of Grace stated that an adoption contract made in both Texas and California “necessarily would be performed wherever the Graces might take the child,” and held that the contract was enforceable in California although void in Texas. (Estate of Grace, supra, at pp. 962-963, 966.) Bernkrant v. Fowler (1961) 55 Cal.2d 588 [ 12 Cal.Rptr. 266 , 360 P.2d 906 ], a precursor to Reich, supra, 67 Cal.2d 551 , was an action to cancel a promissory note pursuant to an oral contract. (Bernkrant, supra, at pp. 590-591 .) Bernkrant concluded that California had no interest in applying its own statute of frauds to a transaction made and performed in Nevada involving the refinancing of a loan secured by land in Nevada, and therefore held that the Nevada statute of frauds governed. (Id. at pp. 595-596.) See also Shannon-Vail Five Inc. v. Bunch (9th Cir. 2001) 270 F.3d 1207, 1210-1213 (Nev. law governed the viability of a usury cause of action under either Civ. Code, § 1646 or the governmental interest analysis), and Costco Wholesale Corp. v. Liberty Mut. Ins .Co. (S.D.Cal. 2007) 472 F.Supp.2d 1183, 1197-1198, 1207 (Conn. law governed the scope of an insurance policy exclusion under either § 1646 or the governmental interest analysis). In contrast, Royal Indemnity Group v. Travelers Indemnity Co. (N.D.Cal., Sept. 6, 2005, No. C-04-00886) 2005 WL 2176896 , pages *4—*5 concluded that the California courts had not abrogated the choice-of-law rule in section 1646. Whether this differentiated approach is either wise or desirable is a question best addressed to the Legislature, which has the sole authority to repeal a statute. Washington Mutual, supra, 24 Cal.4th at page 919 (“California follows a three-step ‘governmental interest analysis’ to address conflict of laws claims and ascertain the most appropriate law applicable to the issues where there is no effective choice-of-law agreement”); Offshore Rental, supra, 22 Cal.3d at page 161 (“Questions of choice of law are determined in California, as plaintiff correctly contends, by the ‘governmental interest analysis’ rather than by the trial court’s ‘most significant contacts theory’ ”). Travelers Ins. Co. v. Workmen’s Comp. App. Bd., supra, 68 Cal.2d at p. 11 (“California has rejected the traditional mechanical solutions to choice-of-law problems and adopted foreign law only when it is appropriate in light of the significant interests in the particular case”). We cited Stonewall, supra, 14 Cal.App.4th 637 , on this point in Downey Venture v. LMI Ins. Co. (1998) 66 Cal.App.4th 478, 514 [ 78 Cal.Rptr.2d 142 ] (Downey Venture), and stated, “[a] liability insurance policy issued on a nationwide basis may be construed in accordance with the law of the jurisdiction in which a particular claim arises. [Citation.] Thus, the same policy language may receive different construction and application in different jurisdictions.” We held in Downey Venture that Insurance Code section 533 precluded indemnification for malicious prosecution under California law but did not relieve an insurer of the duty to defend a malicious prosecution claim. (Downey Venture, supra, at pp. 506, 509.) We concluded that despite the bar of indemnification under California law, an express promise to provide coverage for malicious prosecution was not illusory, not only because the insurer was required to defend such claims, but also because a California court could enforce the laws of other states that might not preclude indemnification. (Id. at pp. 514-516.) Our statement with regard to such a choice-of-law decision by a California court concerned the enforceability of an express contractual promise, rather than contract interpretation. Moreover, the pollution liability endorsement here expressly states that it “forms a part of the policy to which attached.” Gray, supra, 65 Cal.2d at page 274 , stated: “Since we must resolve uncertainties in favor of the insured and interpret the policy provisions according to the layman’s reasonable expectations, [fn. omitted] and since the effect of the exclusionary clause is neither conspicuous, plain, nor clear, we hold that in the present case the policy provides for an obligation to defend and that such obligation is independent of the indemnification coverage.” (Italics added.) In Montrose Chemical Corp. v. Superior Court (1993) 6 Cal.4th 287 [ 24 Cal.Rptr.2d 467 , 861 P.2d 1153 ] (Montrose), the Supreme Court made clear that its analysis in Gray, supra, 65 Cal.2d 263 , was not limited to policy interpretation: “Because the policy at issue in Gray was ambiguous, and could be read either to exclude or to provide for coverage, we held that ordinary principles of insurance contract interpretation required it be construed in the insured’s favor, according to his reasonable expectations. ( Gray, supra, 65 Cal.2d at pp. 271-272; [citation].) As a distinct, separate, and alternative basis for our decision, we recognized that the insured is entitled to a defense if the underlying complaint alleges the insured’s liability for damages potentially covered under the policy, or if the complaint might be amended to give rise to a liability that would be covered under the policy. ( Gray, supra, 65 Cal.2d at pp. 275-276.) H] The alternative holding in Gray, supra, 65 Cal.2d 263 , establishes the rule that the insurer must defend in some lawsuits where liability under the policy ultimately fails to materialize; this is one reason why it is often said that the duty to defend is broader than the duty to indemnify. [Citation.]” (Montrose, supra, at p. 299 .) We explain our conclusion under Civil Code section 1646 that California law governs the interpretation of the policy in parts 3 through 6, ante. RLI quotes language from Nat. Union, supra, 939 S.W.2d at page 142 , stating: “We will not read facts into the pleadings. [Citation.] Nor will we look outside the pleadings, or imagine factual scenarios which might trigger coverage.” The complaint against the insured in that case alleged that the insured’s driver negligently discharged a firearm while driving a truck, killing a passenger in another vehicle. (Ibid.) The Texas Supreme Court concluded that the facts alleged in the complaint did not indicate that the death resulted from use of the insured vehicle, as required for coverage under the policy. (Id. at p. 142.) The court stated, “the facts alleged in the pleadings do not suggest even a remote causal relationship between the truck’s operation and Gonzalez’s injury.” (Ibid.) It was in this context that the court, in the language quoted ante, refused to augment the allegations of the complaint in the guise of a liberal construction. Wainoco Oil Corporation was predecessor in interest to Frontier, as stated ante, and is to be distinguished from Frontier’s wholly owned subsidiary, Wainoco.