Determining whether and when the duty to defend ends can be complicated. This is particularly so if only part of the underlying action is resolved or only some defendant insureds are dismissed or the underlying action is not final. The analysis may touch on procedural as well as substantive law governing the underlying action. An insurer withdrawing the defense could find itself facing claims of breach of contract or bad faith—whether or not it acted properly in withdrawing.
This article discusses some common issues that arise in determining whether the duty to defend has ended. It begins with a review of exhaustion and of whether and when the underlying action is deemed final for purposes of the duty to defend. Next, it explores settlement; in particular, settlement of less than the entire underlying action. It then discusses an insurer’s options if the policyholder fails to cooperate. Finally, it addresses withdrawal procedures, including the formal options of declaratory relief and interpleader as well as informal withdrawals by letter. Throughout, the article offers practical tips for insurers that wish to withdraw the defense and for policyholders who learn that their defense will be withdrawn.
Modern commercial general liability (CGL) policies often specify that the duty to defend ends once the policy limits are exhausted by payment of judgments or settlements. The current main CGL form promulgated by the Insurance Services Office (ISO) provides: “Our right and duty to defend ends when we have used up the applicable limit of insurance in the payment of judgments or settlements under Coverages A or B. . . .” Courts have held that this and similar language means what it says—payment of policy limits in judgment or settlement terminates the insurer’s duty to defend.
Courts in several states have concluded that the word “exhausted”—in the auto policy provision that states “[o]ur duty to settle or defend ends when our limit of liability for this coverage has been exhausted”—refers to judgments or settlements, even when the policy does not specify how it must be exhausted. Other courts, while agreeing that exhaustion requires payment by judgment or settlement, have ruled that this or similar language is ambiguous because it does not specify how exhaustion is to be accomplished. The auto policy at issue in Brown v. Lumbermens Mutual Casualty Co. contained the same provision as in Samply v. Integrity Insurance Co., Emcasco Insurance Co. v. Davis, and Anderson v. U.S. Fidelity & Guaranty Co., while the auto policy at issue in Douglas v. Allied American Insurance provided that the insurer “has no obligation to any insured after the applicable limits of the policy have been exhausted by payment.” Each court reviewed the exhaustion provision in conjunction with other parts of the insuring agreement, as well as dictionary definitions. In Brown, the court determined that the insurer’s agreement to “settle or defend” covered claims meant defense coverage continued until settlement or judgment. The Douglas court came to the same conclusion based on the insurer’s agreement to pay those “sums which the insured shall become legally obligated to pay.” Both the Brown and Douglas courts further stated that payment of the policy limits into court in interpleader, without a judgment or settlement, did not relieve the insurer of its duty to defend.
Courts have differed as to whether an insurer whose policy has exhausted by way of partial settlement or judgment can withdraw if covered claims remain pending in the underlying action. However, where the insured has paid the limits to settle, it has been held to have no obligation to defend a subsequent suit based on the same occurrence.
Exhaustion is determined separately for each policy at issue. The West Virginia Supreme Court has held that exhaustion of a primary policy does not terminate an obligation to defend under an umbrella policy. In Noland v. Virginia Insurance Reciprocal, the insurer issued both primary and umbrella medical malpractice policies to the insured hospital. The malpractice plaintiffs settled with the hospital, exhausting the primary limits. The hospital had filed a third-party complaint against a nurse, who was also insured under the policies. The West Virginia Supreme Court concluded that the insurer had an obligation to defend the nurse against the hospital’s third-party complaint under the excess policies. The court rejected the insurer’s argument that exhaustion of the primary policy could terminate a duty to defend under the umbrella policy.
Many cases hold that an insurer is required to defend until the underlying action is final. Determining when the underlying action is final is not always as straightforward as the general rule suggests, however.
It is important to note that finality depends on procedural law and, sometimes, substantive law. Dismissal of the underlying action has been held to constitute finality in several states. In other states, dismissal by the court is not final until there is a final judgment or the deadline for filing an appeal has expired.
In mixed actions, i.e., actions that include both potentially covered and non-covered claims, courts have varied as to whether the underlying action is final when the potentially covered claims are dismissed but the non-covered claims remain pending. Courts in New Mexico and New Jersey have recently stated that the duty to defend continues until all covered claims are eliminated or resolved. A federal court predicted that Indiana would come to the same conclusion. On the other hand, courts in Hawaii and Oregon have held that the duty to defend continues, especially where the potentially covered claims were not dismissed with finality.
Most states that have considered the issue have ruled that a duty to defend encompasses an obligation to fund an appeal. However, a few states have concluded that an obligation to appeal exists only if there are “reasonable grounds” either for the insured’s appeal or to believe that the insured’s interests are best served by an appeal. In Michigan, an insurer “will be expected to proceed with an appeal, if requested by an insured, if it writes a broad ‘duty to defend’ clause into its insurance contracts.” Palmer v. Pacific Indemnity Co. has been criticized on the grounds that it “could place an insurer in the untenable position of being obligated to pursue a totally groundless appeal simply to avoid a charge that it breached its duty to defend.”
A court in Florida held that the insurer was not required to pay for the insured’s appeal, although it appears that the ruling was based on the fact that the potentially covered claims were no longer at issue. Reller, Inc. v. Hartford Insurance Co. of the Southeast involved a mixed underlying action in which the insurer defended the insured through trial and paid the judgment on the potentially covered claim.
As noted above, the current ISO main CGL form provides that the insurer’s duty to defend ends upon payment of the limits to settle. Most states provide that settlement terminates the duty to defend only if the policy language so provides. A minority of courts have stated, as a general rule, that the duty to defend terminates upon settlement, regardless of policy language.
Even though all states have inferred that a duty to settle exists, settling the underlying action simply to terminate the duty to defend may be deemed bad faith.
The insurer must usually obtain a release from the third-party claimant in favor of the insured in order to terminate the duty to defend. In the rare instance where a third-party claimant refuses to sign a release, the insurer is, in most states, insulated from bad-faith claims if it declines to settle.
Underlying actions involving multiple claimants and multiple insureds may pose additional issues, due to the arguably increased possibility for excess exposure.
Nearly all states allow an insurer to exhaust limits by settling with fewer than all third-party claimants. “[A] liability insurer may, in good faith and without notification to others, settle part of multiple claims against its insured even though such settlements deplete or exhaust the policy limits so that remaining claimants have no recourse against [the] insurer.” This rule usually also applies in claims against additional insureds. Under these cases, where an insurer exhausts the policy by settling in good faith with one of multiple claimants, it can withdraw from the defense.
Delaware cases have held that even though some third-party claimants provided releases, the duty to defend did not end until releases were obtained from all third-party claimants. In Florida, so long as the insurer attempts to settle all claims globally and acts reasonably in settling the highest exposure claims, it can in good faith exhaust limits by settling less than all claims.
In claims against multiple insureds under the same policy, many courts permit the insurer to pay limits to settle the claims against one insured, thereby extinguishing the duty to defend the remaining insureds. In California, settling on behalf of only one of several insureds can constitute a breach of the policy as to the remaining insured.
In the context of mixed actions, a minority of states, notably Minnesota and New York, allow an insurer to settle the potentially covered claim or claims and withdraw from the defense, even though this leaves the insured exposed to defense costs for non-covered claims.
Most modern liability policies require the insured to cooperate. ISO’s current main CGL form contains a condition providing that the insured must “[c]ooperate with us in the . . . defense against the ‘suit’.”
Noncooperation situations frequently involve the insured’s refusal to accept a defense by counsel appointed by the insurer. Such a failure to cooperate has been held to relieve the insurer of its duty to defend. An insured’s refusal to cooperate in responding to the complaint or trial preparation has also been held to violate the cooperation condition, allowing the insurer to withdraw from the defense.
In New York, an insurer seeking to withdraw due to noncooperation must show “that (1) it acted diligently in seeking to bring about the insured’s cooperation, (2) its efforts were reasonably calculated to obtain the insured’s cooperation, and (3) the attitude of the insured, after its cooperation was sought, was one of willful and avowed obstruction.” In states that do not require such a stringent showing, an insurer should usually write the insured to warn of the consequences of not cooperating—including that the insurer will have no duty to defend or indemnify—and to give the insured another opportunity to cooperate. Many, if not most, insureds cooperate once they understand the consequences of noncooperation. For those insureds that choose not to cooperate, such a letter can protect the insurer, should the noncooperating insured file a coverage action due to the insurer’s withdrawal of the defense.
Insureds who are represented by coverage counsel, or whose broker is proactive in the claims process, may benefit from hearing from their counsel or broker that cooperation is a policy precondition to coverage, so that they should cooperate, within the bounds recommended by coverage counsel and without revealing attorney-client information, in order to ensure continued availability of coverage—at least so long as the insurer is acting pursuant to its duty to defend.
If the insured is concerned about sharing work product with the insurer, such as litigation plans, budgets, and written communications, a joint defense and common interest agreement may be appropriate. The agreement can state that the sharing of defense materials is not intended to and shall not waive applicable privileges or protections.
A special rule applies in the context of auto policies in most states. “Absolute liability” or “frozen liability” statutes “require payment of the minimum statutorily required insurance benefits, if the law required the policy to be in place, even if the insured has breached the insurance contract,” including by failing to cooperate. Such a rule promotes the public policy of compensating injured third-party claimants. In a minority of states, auto insurers too can decline coverage if the insured fails to comply with policy conditions.
The formal procedures for withdrawing a defense are a declaratory relief action or an interpleader action.
With respect to declaratory relief actions, under the majority rule, they are an optional means of withdrawing from the defense of a pending underlying action. Although Illinois does not require an insurer to seek declaratory relief, it has developed a unique approach, based on its stringent estoppel law. Illinois has a special rule that an insurer that withdraws the defense and does not timely seek declaratory relief will be estopped from asserting any coverage defenses, even if such defenses might have otherwise barred coverage. Thus, courts adhering to the majority rule often comment that seeking declaratory relief is a “safe course of action” for the insurer to determine its coverage obligations. Moreover, filing a non-frivolous declaratory relief action does not constitute bad faith—although insureds frequently respond by filing a counterclaim alleging bad faith.
A small minority of states require an insurer to file a declaratory relief action in order to withdraw from the defense where the underlying action is still pending.
California has two lines of authority as to whether an insurer must obtain a declaration before withdrawing from the defense, depending on the coverage picture for the underlying action. In Ringler Associates, Inc. v. Maryland Casualty Co., the California Court of Appeal stated that an insurer is not required to seek a declaratory relief action in order to withdraw, once it has determined that there is no potential for indemnity—although it is “prudent” for the insurer to seek such relief. In Prichard v. Liberty Mutual Insurance Co., another California Court of Appeal panel stated that in the different context of a mixed underlying action, the insurer by definition has a duty to defend, at least at the outset of the underlying action; thus, the court indicated that a declaration is required for the insurer to withdraw.
Declaratory relief may not be available where the issues to be adjudicated in the underlying action and the declaratory relief action are the same—at least not while the underlying action is pending. Common examples involve whether the insured’s conduct was intentional or accidental and whether someone was acting within the scope of employment. In these situations, one party is likely to seek a stay of the declaratory relief action, and courts will often grant a stay in order to avoid prejudice to the insured’s defense of the underlying action. If the declaratory relief action and underlying action proceed concurrently, the insurer should usually not propound discovery requests to the insured that could prejudice its defense of the underlying action. Strict “corners” states will usually not grant declaratory relief, as a substantive matter, where the allegations in the complaint are potentially covered, even if discovery later reveals facts showing that no coverage exists.
In some states, such as Idaho, the insurer must continue to defend the underlying action while the declaratory relief action is pending, while in other states, such as Kentucky, the insurer may opt not to defend but may later be held liable for the judgment if the underlying action is found to be covered. Some courts have held that even if declaratory relief is granted, the insurer must wait to withdraw the defense until the declaratory relief action is final under the relevant procedural law. On the other hand, if the insured does not seek a stay of the trial court’s ruling while it appeals a grant of declaratory relief, the insurer may withdraw immediately.
With respect to interpleader, an insurer can interplead policy limits either as a plaintiff or, if named as a defendant, by counterclaim or cross-claim. Interpleader can be a useful procedure, particularly where there are multiple third-party claimants or the claims against the insured may exceed policy limits. Although an insurer may interplead, it is not required to do so.
The cases allowing an insurer to interplead limits in order to withdraw the defense have typically required that the policy language allow the insurer to do so. And where the policy language does not permit an insurer to do so, or is ambiguous on that point, courts have not allowed interpleader.
Where an insurer acts promptly and properly under the state’s claims handling standards, courts have rejected insureds’ claims of bad faith for filing an interpleader:
We think the favored approach to managing multiple claims in excess of the policy limits must include some provision for certainty to insureds, insurers, and litigants short of submitting each case to a jury. In that regard, as a matter of Arizona law, we hold that (1) the prompt, good faith filing of an interpleader as to all known claimants with (2) payment of the policy limits into the court and (3) the continued provision of a defense for the insured as to each pending claim, acts as a safe harbor for an insurer against a bad faith claim for failure to properly manage the policy limits (or give equal consideration to settlement offers) when multiple claimants are involved and the expected claims are in excess of the applicable policy limits.
Simply interpleading the limits may not protect the insurer or insured, however, for two main reasons. First, tendering limits is usually insufficient to terminate the duty to defend, even where policy language provides that the duty to defend ends upon exhaustion of limits. This principle may apply to additional insureds as well as the named insured. This principle may also prevent a primary insurer from shifting the duty to defend to an excess insurer. Second, unless the insurer obtains a release from the third-party claimants in favor of the insured, the insured continues to face possible liability. Thus, an insurer may wish to consider obtaining a release before interpleading policy limits.
Filing a declaratory relief action or interpleader is likely to spawn a counterclaim for breach of contract and bad faith. Thus, insurers may wish to document the file carefully, including showing that they have appropriately investigated, communicated with the insured and, if allowed and warranted, with the third-party claimants, the insured’s potential liability and exposure, and, if the insurer is considering interpleader, obtained a release in favor of the insured. Venue and choice of law may be key considerations, and the insurer may wish to be the first to file for those reasons. If applicable, the insurer may wish to communicate with the insured before filing. Such communications may, for example, allow the insured to sign an agreement to be bound by the result of the action, to protect its interests if it has asserted a counterclaim in the underlying action, and to maintain its business relationships.
Insureds on the receiving end of such declaratory relief or interpleader actions should, if they have not already, retain coverage counsel to evaluate the appropriate response. It can become expensive for a small company when, after already retaining defense counsel, it must now hire coverage counsel to defend the declaratory judgment action. But the consequences of not retaining coverage counsel may result in a determination that there is no coverage. Coverage counsel can assess defenses to the declaratory relief action and develop a strategy to safeguard and maximize coverage available to the insured, as well as determine whether to bring a breach of contract or bad faith action (or both). And until advised to do otherwise, the insured should continue to comply with its obligations under the policy (such as the duty to cooperate) in order to avoid an argument by the insurer that it was the insured, not the insurer, that breached the policy and rendered coverage unavailable.
As noted, some states may allow the insurer to informally withdraw the defense. In that situation, the insurer should act promptly, as the longer the underlying action has been pending, the greater the chance the insured will claim prejudice by withdrawal of the defense—and an important consideration in withdrawing is to avoid prejudicing the insured’s defense of the underlying action, especially in timing of the withdrawal. An insurer should usually avoid withdrawing before key depositions and hearings, and as trial nears. Moreover, an insurer should usually give the insured written notice of its intent to withdraw and advise the insured to retain its own defense or (if applicable) that the insured can itself retain existing defense counsel. Thirty days’ notice is generally considered sufficient, although a Florida court commented that “[t]he insurer should make allowances for the time that the insured will need to retain new counsel, and should continue to represent the insured after the settlement, if necessary, until new counsel can be retained.”
The insurer may also wish to invite the insured to provide any documents or information showing that the insurer’s duty to defend has not terminated. If new defense counsel will be retained, the insurer should instruct panel counsel to turn over the file and sign a substitution of attorney form.
If an insured is advised that its insurer intends to terminate the defense while the underlying action is pending, the insured should evaluate whether the duty to defend did actually terminate and should advise the insurer of any reasons that the duty to defend did not in fact terminate. The insured should also inform the insurer of how long it will take to retain new counsel and of any upcoming deadlines or events for which the insured requests the insurer continue to provide a defense. If the insured believes that termination of defense is inappropriate, it may consider bringing a declaratory relief action against the insurer, which may, under appropriate circumstances, include claims of bad faith.
Determining whether and when the duty to defend terminates tends to be much less obvious than the determination of when it attaches. The determination hinges on the policy language, applicable state law, and the particular facts of each case. The easy answer is that exhaustion of limits through a final judgment or an executed settlement of all covered claims, with releases obtained for the insured, is almost universally recognized as terminating the duty to defend. Beyond that, the issues become murkier, and both sides can benefit from careful evaluation of all applicable facts and the procedural setting of the underlying action, as well as communication.
Keywords: litigation, insurance, coverage, underlying action, duty to defend, withdrawal of duty to defend, when does duty to defend end, termination of duty to defend, policy exhaustion, finality of underlying action, settlement of underlying action, duty to cooperate, declaratory relief action, declaratory judgment action, interpleader, interplead limits, interplead policy limits, release
Rina Carmel is a coverage attorney and partner at Zelle McDonough & Cohen LLP, Los Angeles. Karin S. Aldama is a coverage lawyer representing policyholders and a partner at Perkins Coie LLP, Phoenix. Tred Eyerly is a coverage attorney and is of counsel at Damon Key Leong Kupchak Hastert, Honolulu.
