In their recent article, The Black Hole Problem in Commercial Boilerplate,Professors Stephen J. Choi, Mitu Gulati, and Robert E. Scott identify a phenomenon found in standardized contracts they describe as “contractual black holes.” The concept of black holes comes from theoretical physics. Under the original hypothesis, the gravitational pull of a black hole is so strong that once light or information is pulled past an event horizon into a black hole, it cannot escape. In recent years, the thesis has been reformulated such that the current thesis is that some information can escape, but it is so degraded that it is virtually useless. In their article, Choi, Gulati, and Scott apply the black hole concept to certain standardized contractual boilerplate provisions.
A contractual black hole is “a boilerplate term that is reused for decades and without reflection merely because it is part of a standard form package of terms, [and is thereby] emptied of any recoverable meaning.”Closely related to contractual black holes are “contractual grey holes.” A contractual grey hole is a meaningless variation of a boilerplate term that has been repeatedly reused over a long period of time such that it “has lost much (but not necessarily all) meaning.” In short, contractual black holes lack any meaning, while contractual grey holes may still contain some meaning, but there is no basis for making a legal distinction between the variations in the language that have appeared over time.
Although their article focuses on pari passu clauses in sovereign debt contracts, Choi, Gulati, and Scott note that “[i]nsurance contracts appear to be another area with the potential for such terms.”They are correct.
Insurance policies are the grandparents of contractual black holes. Insurance traces its origins to 2250 B.C. when Babylonian maritime traders entered “bottomry” contracts, in which a party loaned money to a shipper with the understanding that the money would not be repaid if the ship sank or was pirated.Bottomry contracts eventually evolved into modern insurance and Lloyd’s of London issued the first maritime insurance policies in the 1600s. Almost 100 years ago, insurance policies were the first type of standardized agreement to be called “contracts of adhesion.” Through rote re-usage, many of the terms and conditions contained in insurance policies sold today have been in use for decades. The continual reuse of antiquated policy language is, in part, due to the frequent application of a strict liability version of contra proferentem to the interpretation of insurance policies. Thus, because changing the policy language could be viewed as an admission that the prior language was ambiguous or because the existing language already has been held to be unambiguous, insurers are naturally reluctant to change policy language that has already been interpreted by courts.
Consequently, insurance policies are particularly susceptible to the formation of contractual black holes. Indeed, some courts view insurance policies as massive contractual black holes from which only a few flashes of light (i.e., meaning) escape.To test the hypothesis that insurance policies contain, or even embody, contractual black holes, this Essay considers four provisions found in commercial insurance policies: 1) “Sue and Labor” Clauses, 2) “Ensuing Loss” Clauses, 3) “Non-Cumulation” Clauses, and 4) the “Sudden and Accidental” Pollution Exclusion. These four policy provisions are suitable subjects because they either have generated significant amounts of litigation with inconsistent court rulings or they are facially complex or confusing. An examination of these provisions demonstrates that the hypothesis that insurance policies house many contractual black holes is both confirmed and refuted. Some policy provisions have become contractual black holes, some provisions are only apparent contractual black holes, and some provisions, while on their way to becoming contractual black holes, were saved before their original meaning crossed the event horizon. To understand this conclusion, this Essay proceeds in two parts. Part I explores how the insurance policy drafting process results in the rote reuse of policy language. Part II considers the origins and purposes of the four provisions at issue, and then analyzes whether each provision has become a black hole.
Insurance policies are complex financial instruments drafted by insurers and then sold on a take-it-or-leave-it basis.An insurance organization called the Insurance Services Office, Inc. (ISO) drafts many of the commonly used policy forms and then seeks to have the forms approved by state insurance commissioners. Insurers pay fees for ISO membership, which allows them to use the policy forms drafted by ISO.
Much of the policy language used in ISO’s standard forms was written decades ago, but ISO continues to recycle the same language in subsequent versions of its policies.For example, ISO’s 1973 Comprehensive General Liability (CGL) policy form defines “occurrence” as “an accident, including continuous or repeated exposure to conditions, which results in bodily injury or property damage neither expected nor intended from the standpoint of the insured.” Forty years later, ISO’s 2013 CGL policy form still defines “occurrence” as “an accident, including continuous or repeated exposure to substantially the same general harmful conditions.” Despite being issued 40 years apart, both versions use the phrase “an accident, including continuous or repeated exposure.” The minor variation in the wording of the phrases is an example of what Choi, Gulati, and Scott describe as “encrustation,” where minor changes in language do not really affect its legal meaning. Notably, even though the meaning of the term “accident” has been litigated over and over again for decades, which has produced an array of different court interpretations of the term, ISO has never bothered to define the term in CGL policies.
When it comes to seeking guidance regarding the intent of the drafters, the original drafters typically cannot be called upon to shed light on the meaning of antiquated policy language.They are often unidentifiable or have died during the many years that have elapsed since they drafted the policy language.
Documentation regarding the drafters’ intent is also often unavailable.The lack of documentation may be intentional by insurers because, as a result of the use of a strict liability version of contra proferetem with respect to ambiguities in insurance policies, insurers always take the position that the policy language is unambiguous. If the language is unambiguous, then no extrinsic evidence is needed or used to interpret it. Consequently, ISO and insurers have a good reason not to preserve documentation of intent.
Insurance law essentially dictates this result. Unlike typical contract disputes in which extrinsic evidence regarding the parties’ mutual intent is admissible to resolve disputes regarding ambiguous contract language, there is no mutual intent to discern with respect to insurance policies. Policyholders play no role in the drafting of insurance policies and they do not even get a copy of the policies until after purchasing them.Even the insurers selling the policies typically do not know the intent of policy language because they did not draft it. Thus, because any finding of ambiguity in the policy language almost automatically means the insurer loses, policy language can never be ambiguous from the insurer’s perspective once a coverage dispute arises.
Another disincentive to redrafting and modernizing policy language is that doing so could be viewed as an admission that the older policy language is either unclear or actually covers the types of losses the insurer has been contending it does not.Thus, redrafting and modernizing the language could lead to insurers losing future cases decided under the older policy language. This leads to the rote reuse of the same policy language decade after decade.
Some scholars also have theorized that insurers are reluctant to redraft old policy language because their actuarial data, and thus premiums, are based upon the language already in use.This argument has intuitive appeal, but it may not be empirically correct. Premiums generally are based upon broad factors, such as the nature of the policyholder’s business, the size of the policyholder’s operations, the policyholder’s number of employees, the policyholder’s gross revenues or sales, and the policyholder’s loss history, rather than the granular language of specific policy provisions. With that said, however, insurers routinely add exclusions for certain types of losses they do not want to cover if courts begin to interpret their policies to provide such coverage. So, the idea that insurers have an incentive not to make dramatic changes to policy language may have some merit even if the specific wording of individual policy provisions is not part of the premium calculation.
Finally, the meaning of certain provisions in policies also has been obfuscated by the incredibly complex structure of policies that has developed over time. Consider, for example, ISO’s 2013 CGL policy form. When originally conceived and created, the CGL policy form was the broadest form of liability coverage available under which the insurer agreed to pay “all sums” for which the policyholder became liable for “bodily injuries” or “property damage” caused by an accident.One might expect a policy providing such broad coverage to be a simple, short document. That would be a mistake. The 2013 CGL policy form contains: three sections setting forth the different types of coverage provided, thirty-eight exclusions, nine conditions, and twenty-two definitions (not including the two pages it takes to explain who is an “insured” under the policy).
In sum, the antiquated insurance policy language reused decade after decade, combined with the increasing length and complex organization of the numerous terms, conditions, and exclusions that are cross-referenced throughout, is a recipe for the development of contractual black holes.The next Part tests this hypothesis by considering four policy provisions found in commercial insurance policies.
The first candidate for contractual black hole status is the Non-Cumulation Clause that appears in CGL policies. Non-Cumulation Clauses are implicated when a loss triggers multiple policy periods, and courts and parties must figure out which of the triggered policies are liable and for how much.Non-Cumulation Clauses are a prime example of a policy provision that was first drafted decades ago and, through rote reuse, has lost its meaning when interpreted and applied in current disputes.
The Lloyd’s of London Non-Cumulation Clause, which is the earliest one used in modern occurrence-based insurance policies, was first created in 1960 by Leslie R. Dew and his fellow London underwriters.The clause originally read as follows:
Prior to 1960 when Lloyd’s of London created the first version of modern “occurrence”-based CGL policies, which is the type of CGL policy form most commonly used today, CGL policies were “accident”-based.Under accident-based CGL policies, the coverage-triggering event was an “accident” that gave rise to an injury. The policies did not define the term “accident.” Consequently, courts had to determine what “accident” meant as it was used in the policies. The case law was developing such that some courts had concluded that accidents were not limited in time and space to a single event, but rather, could include situations that took place over longer periods of time and caused ongoing injuries. Consequently, in 1960, the CGL policy form was revised to use a defined term of “occurrence” instead of just the undefined term “accident.”
By changing to occurrence-based insurance, the coverage triggering event became an “occurrence,” which originally was defined as “an accident, including continuous or repeated exposure to conditions, which results in bodily injury or property damage neither expected nor intended from the standpoint of the insured.”Significantly, this change to an occurrence-triggering policy form meant coverage was expressly provided not only for individual injury-causing events, but also for gradual injury-causing situations that could span long periods of time. As one insurer representative in the 1960s explained:
This extension of coverage created a problem for insurers because some injuries simultaneously could be covered under both the older accident-based policies and the new occurrence-based policies.For example, in situations where a court could conclude the coverage triggering “accident” under accident-based policies was the defective manufacture of a product that subsequently caused an injury, under the occurrence-based policies the injuries themselves were the triggering event. This meant the policyholder could recover under both policy forms for the same injury and potentially receive a double recovery. The Non-Cumulation Clause was created to address that problem by preventing the policyholder from receiving a windfall double recovery.
Non-Cumulation Clauses became contractual black holes, however, when they continued to be reused decade after decade with only minor variations despite the dramatically changing legal and scientific landscape. At the time the clause was originally drafted in 1960, the drafters did not conceive of “long-tail” claims such as the asbestos and environmental claims that later arose in the 1970s through 1990s when the delayed manifestation of the injury-causing effects of asbestos exposure became more widely understood and environmental laws were passed that created retroactive strict liability for past waste generators, haulers, and disposers.Nor did, or could, the original drafters intend the clause to apply to such claims. The clause was designed to prevent double recoveries, not to limit recoveries for unforeseen long-tail claims.
When you attempt to apply Non-Cumulation Clauses to modern long-tail claims, where the injury is continuously caused over many years, a contractual black hole appears. The clause reads, in part:
The policy language does not specify how to determine whether a loss is “covered” under a prior-incepting policy or who makes the determination. Nor does it state whether a court judgment is necessary. It also does not address whether the prior insurer needs to admit liability in order to trigger the clause’s application.
Further, it is unclear what constitutes an “amount due” under prior insurance. It could be an amount that a court has adjudicated is due. Or, it could simply be an amount that a subsequent insurer who is attempting to avoid liability merely alleges is due from another insurer that issued a policy in an earlier policy year. Arguably, it should at least be an amount actually paid by a prior insurer for the loss, but the policy does not state that.
In addition, the clause is silent regarding how one should deal with settlements in which a prior incepting insurer denies liability but settles nonetheless. The clause simply does not address whether settlement payments are “amounts due” under a Non-Cumulation Clause. Nor does it state whether later incepting insurance policies get credit for the actual settlement amounts paid or for the full limits of the policies issued by the prior incepting policy. As these examples indicate, Non-Cumulation Clauses become contractual black holes when they are applied to long-tail claims today.
The second candidate for contractual black hole status is the Sue and Labor Clause.Sue and Labor Clauses originated in Lloyd’s of London’s marine insurance policies in the 1600s. A common version of a Sue and Labor Clause provides:
The phrase to “sue, labor and travel for,” or some variation of it, has been used in marine policies for approximately 400 years.The phrase also has been transferred to other types of insurance and is now found in inland property policies.
What does it mean “to sue, labor and travel for, in, and about the defense, safeguard and recovery of the [the property], or any part thereof?” If you own a ship, on what occasion would you have to “sue, labor and travel” to defend or safeguard the ship? Today, the phrase is basically meaningless to most people. The phrase appears to have the qualities of a contractual black hole.
With that said, the Sue and Labor Clause is only an apparent contractual black hole. Although it has been replicated in insurance policies for hundreds of years and appears to be gibberish on its face, the clause has not lost its meaning over time. Courts and insurers know what it means because its meaning and purpose have been preserved through case law. The clause provides coverage to the policyholder for the costs the policyholder incurs in an attempt to avoid or minimize an occurring or impending loss.In short, it is a loss mitigation clause.
The odd language used in the clause is a reflection of the times and circumstances surrounding its original creation. A ship at sea in the 1600s that encountered a storm or pirates was allowed to do anything and everything it could to preserve its cargo and the ship. And, it was allowed to recover from the insurer the costs associated with doing so, including the “labor” costs incurred and the costs associated with “suing” anyone who had salvaged the damaged ship or goods.For most ship owners today, fending off pirates is not as much of problem as it was 400 years ago.
The subsequent transfer of the clause from marine insurance policies to inland property policies without significant revision only heightens the potential contractual black hole quality of the clause because most property owners do not need to worry about fending off pirates or jettisoning cargo during a storm to save property that is on land. Indeed, on its face, the language makes little sense in that context. So, could insurers update the language to make it more understandable and relevant to the circumstances policyholders face today that could result in losses? Absolutely, but that does not mean the language has completely lost its meaning due to rote reuse over the centuries. There is plenty of case law preserving its meaning despite the dated and awkward wording.So, Sue and Labor Clauses should be viewed as only apparent, not real, contractual black holes.
The third candidate for contractual black hole status is the Ensuing Loss Clause that appears in property policies—both homeowners and commercial. Many property policies sold today are “all risk” policies, which means they cover any and all losses unless the peril causing the loss is specifically excluded.Insurers have added exclusions to such policies to avoid covering certain perils such as earthquakes and floods. In addition to specific earthquake and flood exclusions, insurers also often include “anti-concurrent causation” exclusions in their policies. Anti-concurrent causation exclusions purport to exclude coverage for losses caused in any part by an excluded peril: “This policy does not insure loss or damage caused directly or indirectly by any Peril excluded.” Thus, under one literal reading of an anti-concurrent causation exclusion, if an excluded peril plays any role in causing a loss, then the loss arguably is not covered.
The potential for a contractual black hole to develop in this area of insurance arises because policies with anti-concurrent causation (and other) exclusions also often contain an exception to such exclusions known in the insurance world as an Ensuing Loss Clause.Ensuing Loss Clauses have been in existence since the early 1900s when they were created following the 1906 San Francisco fire. Several courts refused to enforce anti-concurrent causation exclusions that purported to exclude coverage for fire damage—a covered peril—that resulted when gas lines were broken by an earthquake—an excluded peril.
One example of an Ensuing Loss Clause provides: “We insure for all risks of physical loss to the property described in Coverage A except for loss caused by: [any of the 6 following excluded perils]. Any ensuing loss from items 1 through 6 not excluded is covered.”Like the term “accident,” insurers have chosen not to define the term “ensuing loss” in their policies. Readers must thus go to other sources to attempt to understand it. A standard dictionary defines “ensue” as: “1. to come afterward; follow immediately” or “2. to happen as a consequence; result.” In other words, the Ensuing Loss Clause reinstates coverage for losses that follow as a result of, at least in part, a covered peril even if an excluded peril is also part of the causation chain, notwithstanding the presence of an anti-concurrent causation exclusion. The confusing, and apparently contradictory, language in Ensuing Loss Clauses combined with inconsistent anti-concurrent causation exclusions creates the right setting for the formation of a contractual black hole.
Indeed, on its face, an Ensuing Loss Clause is a contractual black hole. The wording alone allows almost no light into the meaning for the reader. In a world where every loss is caused by numerous events, how can a loss that is “caused directly or indirectly by any Peril excluded” be excluded from coverage under an anti-concurrent exclusion,while simultaneously be covered under an Ensuing Loss Clause? From this contractual black hole, courts are left with the fruitless task of divining meaning.
Not surprisingly, Ensuing Loss Clauses have flummoxed courts.In determining coverage, some courts simply default to the “efficient proximate cause” doctrine to determine whether a covered or an excluded peril is the first or dominant cause of the loss. Other courts look at whether there was a separate and intervening covered peril that caused the loss. And other courts simply analyze whether a covered peril played any role in causing the loss. If it did, then there is coverage. Interestingly, despite the inconsistency between anti-concurrent causation exclusions and Ensuing Loss Clauses, many courts have held that Ensuing Loss Clauses are unambiguous and, by implication, not contractual black holes. Courts have reached such conclusions even though they have interpreted the clauses inconsistently and made inconsistent coverage determinations when applying them.
In sum, the Ensuing Loss Clause is a better candidate for contractual black hole status than the Sue and Labor Clause. Although the origin of the Ensuing Loss Clause is known, unlike the wording of the Sue and Labor Clause, the wording of an Ensuing Loss Clause is confusing and contradictory when read together with an anti-concurrent causation exclusion. In addition, unlike the Sue and Labor Clause, there is a lack of consensus among courts regarding its meaning and application.
The final candidate for contractual black hole status is the Sudden and Accidental Pollution Exclusion that ISO used in its CGL policy form between 1973 and 1986 and was frequently litigated during the 1990s. The exclusion is worth discussing here, not because the language in the exclusion has lost its meaning due to decades of rote reuse, but rather, as an example of the dynamic that can lead to the creation of insurance contractual black holes.
As discussed in Part II.A, after insurers changed the CGL policy form from accident-based to occurrence-based in the 1960s, CGL policies unquestionably covered injuries that resulted from ongoing injury-causing processes, as opposed to just accidental “events.”Consequently, environmental damage claims were covered so long as they were unexpected and unintended:
Indeed, when insurers changed the CGL policy form from accident-based to occurrence-based in the 1960s, insurers actually marketed the new policies as covering gradual injury-causing situations such as pollution so long as the injury was not expected or intended by the policyholder.
For example, in 1965, Gilbert Bean, a former executive of a major insurer and a member of a committee that was responsible for reviewing and drafting policy language, stated the following with respect to whether the new CGL policy form covered environmental claims: “Manufacturing risks producing insecticides, plant foods, fertilizers, weed killers, paints, chemicals, thermostats or other regulatory devices, to name a few, have created gradual [property damage] exposure. They need this protection and should legitimately expect to be able to buy it, so we have included it.”A year later in 1966, Mr. Bean similarly wrote: “[There is] coverage for gradual [bodily injury] or gradual [property damage] resulting over a period of time from exposure to the insured’s waste disposal. Examples would be gradual adverse effect of smoke, fumes, air or stream pollution, contamination of water supply or vegetation.”
One contemporaneous insurance policy manual that was used to explain the coverage provided under the CGL policy form had the following hypothetical as an example of an “occurrence” that would be covered under the new occurrence-based policy form:
For at least two reasons, the insurers’ appetite for covering environmental claims quickly waned. First, several significant environmental incidents, such as the Torrey Canyon disaster and the Santa Barbara offshore oil spill, created widespread, negative media coverage regarding pollution.Second, in 1970, Congress passed the Federal Water Quality Improvement Act, which imposed strict liability for certain discharges into bodies of water. Thus, widespread negative media attention targeted polluters, and the law began imposing strict liability for certain environmental injuries.
In response to these developments, insurers drafted what is now known as the qualified pollution exclusion or Sudden and Accidental Pollution Exclusion, which first was used as a policy add-on endorsement in 1970 and then became part of the CGL policy form itself in 1973.The Sudden and Accidental Pollution Exclusion is worded as follows:
In general, under this exclusion, claims related to environmental damage are not covered unless the event giving rise to the damage was “sudden and accidental.”
The use of the phrase “sudden and accidental” in the exclusion is an example of rote reuse of boilerplate language in standardized insurance policies. The phrase previously had been used in Boiler and Machinery insurance policies and had a judicially established meaning: “courts uniformly had construed the phrase to mean unexpected and unintended.”Thus, when the phrase was transplanted to CGL policies in the Sudden and Accidental Pollution Exclusion, it already was understood to mean “unexpected and unintended” in the insurance context.
In sum, when the Sudden and Accidental Pollution Exclusion was created, the definition of “occurrence” already limited coverage to injuries or damage that were “neither expected nor intended from the standpoint of the insured”and the phrase “sudden and accidental” was understood to mean unexpected and unintended. Consequently, when seeking approval of the new exclusion, it is unsurprising that insurers told state insurance commissioners across the country that the new exclusion was not a reduction in coverage for pollution claims, but rather, was only a “clarification” of the coverage provided under CGL policies. For example, in a June 10, 1970 letter to the Georgia State Insurance Commissioner, the insurance industry stated:
As a mere “clarification” regarding the existing scope of coverage provided under CGL policies for pollution claims, insurers did not provide a reduction in premiums in exchange for the addition of the new exclusion.
The insurers’ position regarding the meaning of “sudden and accidental” changed, however, when they were confronted with countless lawsuits with hundreds of billions of dollars at stake in widespread environmental insurance coverage litigation in the 1980s and 1990s.The onslaught of environmental insurance coverage litigation occurred because the landscape regarding liability for environmental claims dramatically changed within a few years of the addition of the Sudden and Accidental Pollution Exclusion to CGL policies.
In 1976, Congress passed the Resource Conservation and Recovery Act (RCRA).In 1980, the Comprehensive Environmental Response Compensation and Liability Act (CERCLA) was passed and, in 1986, the Superfund Amendments and Reauthorization Act (SARA) was passed (collectively, these environmental statutes are known as the “Superfund” laws). The Superfund laws imposed retroactive, strict, and joint and several liability for the cleanup of environmental injuries on a variety of entities: (1) the current owners and operators of disposal facilities, (2) the owners or operators of disposal facilities during the time of the disposal, (3) the entities that arranged for the disposal or treatment of hazardous materials, and (4) the transporters of hazardous materials. These new environmental laws created hundreds of billions of dollars of liabilities for policyholders almost overnight. Policyholders, in turn, demanded that their CGL insurers pay such liabilities.
When faced with a bill for hundreds of billions of dollars, insurers took the position that the Sudden and Accidental Pollution Exclusion unambiguously precluded coverage for any and all environmental liabilities unless such liabilities resulted from “abrupt” releases of contaminants.Because the strict liability version of contra proferentem that often applies in insurance disputes means insurers lose if the policy language at issue is ambiguous, the law essentially forced insurers to take the position that the exclusion was unambiguous. As a corollary to that maxim, insurers also refused to produce any documents or allow discovery regarding the original drafters’ intent regarding the meaning of the exclusion because extrinsic evidence should not be relevant or discoverable if the policy language was unambiguous.
Policyholders, on the other hand, disputed that “sudden” unambiguously means “abrupt” by pointing out that “sudden” also can mean “unexpected.”Policyholders then requested documents from state insurance commissioners and successfully moved to compel the production of the drafting history from ISO and insurers regarding the exclusion to see whether the insurers’ litigation position was consistent with: 1) the original intent and purpose of the exclusion, and 2) insurers’ statements to state insurance commissioners regarding the exclusion. Of course, once obtained, the actual historic record regarding the origins of the exclusion belied the insurers’ litigation position that the term “sudden” in the Sudden and Accidental Pollution Exclusion exclusively and unequivocally means “abrupt.”
The story of the Sudden and Accidental Pollution Exclusion is an exemplar regarding the formula for the creation of contractual black holes. First, information regarding the original intent and purpose of a phrase is lost through concealment or the passage of time. Then, the drafters of the contractual language—ISO in this instance—elect not to revise the language despite a patent or latent ambiguity in the language. Indeed, ISO continued to decline to revise the Sudden and Accidental Pollution Exclusion for over a decade while numerous courts construed the exclusion in completely inconsistent ways.ISO finally changed the language in the exclusion after the insurers’ litigation position regarding the meaning of the language had been rejected by numerous courts and insurers had been held liable for billions of dollars associated with environmental cleanups.
Ultimately, the massive litigation regarding the Sudden and Accidental Pollution Exclusion prevented the exclusion from becoming a contractual black hole because the insurers’ position regarding the origin and meaning of the language in the exclusion was proven to be inconsistent with reality. The result, however, could have been completely different. The exclusion could have become another contractual black hole through rote reuse of language and the passage of time. It did not, however, because the litigation regarding the meaning of the Sudden and Accidental Pollution Exclusion occurred fairly soon after the exclusion was drafted. In addition, the policyholders had the resources and tenacity to force the insurance industry to produce the documentation regarding the original intent and meaning of the exclusion.
The potential for contractual black holes to appear in standardized commercial contracts is real. Insurance policies are fertile ground for the creation of contractual black holes. Many policies are drafted by a centralized organization—ISO—and the policy language is reused by rote decade after decade. As non-drafters of the policy language, the insurers that use the ISO policy forms often do not even know what the policy language means themselves.
The rote reuse of policy language then becomes a self-perpetuating cycle because a strict liability version of contra proferentem often applies in insurance disputes. This dictates that insurers always take the position that policy language is unambiguous. Consequently, there is a disincentive for insurers to revise policy language because any changes to it could be viewed as an admission that the old language was ambiguous. Over time, as the policy language becomes antiquated and begins to lose meaning, it is reused nonetheless.
This dynamic, combined with the increasingly complex structure and organization of policies, has resulted in policies, as a whole or in part, appearing to be contractual black holes. An examination of the Non-Cumulation Clause, Sue and Labor Clause, Ensuing Loss Clause, and the Sudden and Accidental Pollution Exclusion demonstrates that some policy provisions have become contractual black holes, some provisions are only apparent contractual black holes, and some provisions were saved before they became black holes.
† Christopher C. French is a Professor of Practice at Penn State Law School; J.D., Harvard Law School; B.A., Columbia University.
 Stephen J. Choi, Mitu Gulati & Robert E. Scott, The Black Hole Problem in Commercial Boilerplate, 67 Duke L.J. 1 (2017).
 Id. at 5.
 Id. at 3 n.2.
 Id. at 3.
 Id. at 4.
 Id. at 4 n.3. This Essay refers to both contractual black and grey holes as contractual black holes despite the minor difference between the two.
 Id. at 7 n.16 (citing Christopher C. French, Understanding Insurance Policies as Noncontracts: An Alternative Approach to Drafting and Construing These Unique Financial Instruments, 89 Temp. L. Rev. 535, 547–48 (2017)).
 See, e.g., Robert H. Jerry, II & Douglas R. Richmond, Understanding Insurance Law 16 (5th ed. 2012) (noting the earliest traces of risk transference resembling insurance can be found within ancient Babylonian society).
 Id. at 16–17.
 See Cheryl B. Preston & Eli McCann, Llewellyn Slept Here: A Short History of Sticky Contracts and Feudalism, 91 Or. L. Rev. 129, 131 (2012) (citing Edwin W. Patterson, The Delivery of a Life-Insurance Policy, 33 Harv. L. Rev. 198, 222 (1919) (describing standardized life insurance policies as contracts of adhesion)).
 See, e.g., John F. Dobbyn & Christopher C. French, Insurance Law in a Nutshell 64 (5th ed. 2015) (“[M]any of the terms and conditions contained in standard form policies were drafted many years ago and are reused each time a new version of the policy form is issued.”); Donald S. Malecki & David D. Thamann, Commercial General Liability Coverage Guide 363-662 (11th ed. 2015) (reproducing the various iterations of the Insurance Services Office, Inc.’s standard Commercial General Liability policy form that have been used for the past 40 years, which reveals the various iterations of the policies contain many provisions that are identical or substantially similar); Michelle E. Boardman, Contra Proferentem: The Allure of Ambiguous Boilerplate, 104 Mich. L. Rev. 1105, 1113 (2006) (arguing that the predictability in the interpretation of policy language by courts incentivizes not changing policy language); see also Christopher C. French, The “Non-Cumulation Clause”: An “Other Insurance” Clause by Another Name, 60 U. Kan. L. Rev. 375, 388 (2011) [hereinafter French, “Non-Cumulation Clause”] (discussing the transfer of the language contained in the non-cumulation clause drafted in 1960 to the 1971 version of the policy that is still found in some policies today).
 See Kenneth S. Abraham, A Theory of Insurance Policy Interpretation, 95 Mich. L. Rev. 531, 538 (1996) (“The strict liability approach to ambiguity is the principal feature of the hornbook statement of contra proferentem . . . . If a policy provision is ‘ambiguous’—reasonably susceptible to more than one interpretation by the ordinary reader of the policy—then the . . . interpretation more favorable to the insured governs . . . .”); Boardman, supra note 13, at 1113; French, supra note 9, at 557–58 (“Unlike in typical contract disputes where contra proferentem a tiebreaker when ambiguous policy language cannot be conclusively clarified by extrinsic evidence, most courts simply construe any ambiguities in the policy language against the insurer and in favor of coverage. [Thus,] contra proferentem . . . in insurance cases has been described as strict liability for the insurer.” (footnotes omitted)).
 Although they do not use the term “contractual black hole,” some courts have described insurance policies as “incomprehensible” and “a mere flood of darkness and confusion.” See, e.g., Storms v. U.S. Fid. & Guar. Co., 388 A.2d 578, 580 (N.H. 1978) (“[I]nsurance policies [contain such] complex verbiage that ‘they would not be understood by men in general, even if [the policies were] subjected to a careful and laborious study. . . . [The policy] would, unless he were an extraordinary man, be an inexplicable riddle, a mere flood of darkness and confusion.’” (second alteration in original) (third DeLancy v. Rockingham Farmers’ Mut. Fire Ins. Co., 52 N.H. 581, 587–88 (N.H. 1873))); S.C. Ins. Co. v. Fid. & Guar. Ins. Underwriters, 489 S.E.2d 200, 206 (S.C. 1997) (“Ambiguity and incomprehensibility seem to be the favorite tools of the insurance trade in drafting policies. Most are a virtually impenetrable thicket of incomprehensible verbosity.” (quoting Universal Underwriters Ins. Co. v. Travelers Ins. Co., 451 S.W.2d 616, 622 (Ky. Ct. App. 1970))).
 “Apparent contractual black holes,” as the phrase is used in this Essay, are provisions that do not appear to have meaning, but a consensus regarding their meaning can be found in case law.
 See, e.g., 1 Jeffrey W. Stempel & Erick S. Knutsen, Stempel and Knutsen on Insurance Coverage § 4.06[b], at 4–65 (Wolters Kluwer 4th ed. Supp. 2017) (“In a sense, the typical insurance contract is one of ‘super-adhesion’ in that the contract is completely standardized and not even reviewed prior to contract formation.”); Michelle Boardman, Insuring Understanding: The Tested Language Defense, 95 Iowa L. Rev. 1075, 1091 (2010) (describing the “hyperstandardization” of insurance policies); Susan Randall, Freedom of Contract in Insurance, 14 Conn. Ins. L.J. 107, 125 (2007) (“[I]n some lines of insurance, all insurance companies provide identical coverage on the same take-it-or-leave-it basis.”); Kent D. Syverud, The Duty To Settle, 76 Va. L. Rev. 1113, 1153 (1990) (“[P]roperty owner’s liability insurance contracts are standardized across insurers in a form few insureds have the power or experience to bargain around.”).
 See, e.g., Hartford Fire Ins. Co. v. California, 509 U.S. 764, 772 (1993) (“[A]n association of approximately 1,400 domestic property and casualty insurers[, ISO] is the almost exclusive source of support services in this country for CGL insurance. ISO develops standard policy forms . . . ; most CGL insurance written in the United States is written on these forms.” (citation omitted)); U.S. Fire Ins. Co. v. J.S.U.B., Inc., 979 So. 2d 871, 879 n.6 (Fla. 2007) (“The Insurance Services Office, Inc., also known as ISO, is an industry organization that promulgates various standard insurance policies that are utilized by insurers throughout the country . . . .”).
 See, e.g., Kenneth S. Abraham & Daniel Schwarcz, Insurance Law and Regulation 36–37 (6th ed. 2015).
 See supra notes 13–14 and accompanying text.
 See Ins. Servs. Office, Inc., Form No. GL 00 02 01 73, Commercial General Liability Coverage Form, Definitions (1973), reprinted in Malecki & Thamann, supra note 13, at 363.
 See Ins. Servs. Office, Inc., Form No. CG 00 01 04 13, Commercial General Liability Coverage Form, Definition No. 13 (2013), reprinted in Malecki & Thamann, supra note 13, at 617. The “expected or intended” language contained in the definition of “occurrence” in the 1973 policy was moved to the exclusions section of the policy form in 2013: “This insurance does not apply to . . . ‘bodily injury’ or ‘property damage’ expected or intended from the standpoint of the insured.” Id. at 604 (Exclusion 2.a).
 See Choi et al., supra note 1, at 5.
 See, e.g., State Farm Fire & Cas. Co. v. CTC Dev. Corp., 720 So. 2d 1072, 1075 (Fla. 1998) (“The difficulty in precisely defining the scope of coverage in liability policies providing coverage for ‘accidents’ is not a problem of recent vintage. As Judge Van Nortwick observed . . . few insurance policy terms have ‘provoked more controversy in litigation than the word “accident.”’” (quoting CTC Dev. Corp. v. State Farm Fire & Cas. Co., 704 So. 2d 579, 581 (Fla. Ct. App. 1997) (Van Nortwick, J., concurring))).
 See e.g., Dobbyn & French, supra note 13, at 64; see also French, “Non-Cumulation Clause,” supra note 13, at 386–89. In an attempt to understand the original drafter’s intent, the author relied upon the deposition testimony of an insurance policy drafter, who incorporated earlier policy language into a new policy form, regarding the original drafter’s intent because the original drafter was dead. See id.
 See id.
 See, e.g., Dobbyn & French, supra note 13, at 64 (“Documentation regarding the intent of the drafters also rarely exists. Consequently, it often is impossible to discern the original intent of the drafters of standard form policy language.”); French, “Non-Cumulation Clause,” supra note 13, at 386–89 (relying upon secondhand deposition testimony to understand the drafter’s intent regarding the non-cumulation clause because of a lack of documentary evidence).
 See supra note 14 and accompanying text.
 See Boardman, supra note 13, at 1120; French, supra note 9, at 537, 548 (citing Eugene R. Anderson & James J. Fournier, Why Courts Enforce Insurance Policyholders’ Objectively Reasonable Expectations of Insurance Coverage, 5 Conn. Ins. L.J. 335, 363 (1998)).
 See supra notes 18–20, 25–26 and accompanying text; see also Restatement (Second) of Contracts § 211 cmt. b. (Am. Law Inst. 1981) (“Employees regularly using a form often have only a limited understanding of its terms and limited authority to vary them.”); Anderson & Fournier, supra note 29, at 364.
 See, e.g., Michelle Boardman, Blank, Black, and Grey Holes in Insurance Contracts 17 (Mar. 2017) (unpublished manuscript), http://www.law.nyu.edu/sites/default/files/Boardman%20Black%20Holes%20in%20Insurance%20Contracts.pdf [https://perma.cc/2N9Z-8CEN] (“[I]nsurers are generally unwilling to add specific exclusions in future policies if their position is that the previous policies did not cover the loss either.”).
 See, e.g., id. at 16 (“Changing contract language may require letting go of, or weakening the predictive power of, valuable actuarial data.”); see also Boardman, supra note 13, at 1116 (“[T]he cost of each clause becomes increasingly clear as actuarial data is collected and pooled.”).
 See, e.g., What Goes into the Price of a General Liability Insurance Policy for Small Businesses?, Insureon Blog (Mar. 12, 2014), http://www.insureon.com/blog/post/2014/03/12/general-liability-price-factors.aspx [https://perma.cc/9LH9-UJWK] (listing the 8 factors that determine premium rates as: 1) “Size and Condition of Your Business Premises, 2) “Type of Business Operations/Industry,” 3) “Experience in Your Profession, Field, or Business,” 4) Number of Employees,” 5) “Location of Your Business,” 6) “Limits and Deductibles,” 7) “Policy Features,” and 8) “Claims History”); see also Cont’l Cas. Co. v. Amerisure Ins. Co., 226 F. Supp. 3d 537, 542 (W.D.N.C. 2017) (“[The insurer] calculated the final premium (i.e. price) charged to [the policyholder] for CGL insurance provided under the [insurer’s] policies based on [the policyholder’s] payroll on all of [the policyholder’s] operations. . . .”); Monkey Ridge, LLC v. Unigard Ins. Co., No. 2:16-CV-0213, 2016 WL 5864428, at *2 (E.D. Wash. Oct. 6, 2016) (“[The insurer] calculated the premiums due on the CGL Policy part of the Primary Policy and the Umbrella Policy based on the number of properties identified and the acreage of those properties.”); Essex Ins. Co. v. Ragland Mills, Inc., No. 06-0737-CV-W, 2008 WL 351014, at *8 (W.D. Mo. Feb. 6, 2008) (“This premium was calculated based upon the hazards and gross receipts of elevator inspection.”).
 See, e.g., Christopher C. French, The Role of the Profit Imperative in Risk Management, 17 U. Pa. J. Bus. L. 1081, 1096–1114 (2015) (discussing insurers’ additions of exclusions for pollution claims, asbestos claims, terrorism claims, Y2K claims, and mold claims after losses associated with such claims began to materialize).
 See, e.g., Jeffrey W. Stempel, Assessing the Coverage Carnage: Asbestos Liability and Insurance After Three Decades of Dispute, 12 Conn. Ins. L.J. 349, 358 (2006). Notably, when the CGL policy form was first introduced in the 1940s it was called “Comprehensive General Liability” insurance. ISO reduced the coverage provided under CGL policies by adding more and more exclusions to the policy form over the decades and renamed the policy “Commercial General Liability” insurance in 1986, but retained the CGL acronym. Id. at 355.
 See Malecki & Thamann, supra note 13, at 603–18.
 See Boardman, supra note 17, at 1119 (“It is not just the language of insurance policies that makes for difficult reading. The order of the language, the parachronistic structure of the policy, and the intimate connection between clauses found in separate ‘sections’ pages apart, [make the policy difficult to understand. Thus,] consumer[s] . . . often miss . . . controlling clauses.”).
 See French, “Non-Cumulation Clause,” supra note 13, at 375–77.
 See id. at 386.
 Id. (quoting Randolph M. Fields, The Underwriting of Unlimited Risk: The London Market Umbrella Liability Policy 1950 to 1970, Morgan Owen Medal Essay Submission to the Chartered Insurance Institute (1994) (Exhibit 16) (LRD 60 Form Policy, Condition C)).
 Id. at 387; Stempel, supra note 35, at 363.
 French, “Non-Cumulation Clause,” supra note 13, at 387.
 Stempel, supra note 35, at 363–64.
 See, e.g., Shipman v. Emp’rs. Mut. Liab. Ins. Co., 125 S.E.2d 72, 75–76 (Ga. Ct. App. 1962) (explaining that an “accident” may take place over time).
 See, e.g., French, “Non-Cumulation Clause,” supra note 13, at 387; Stempel, supra note 35, at 364; John J. Tarpey, The New Comprehensive Policy: Some of the Changes, 33 Ins. Couns. J. 223, 223 (1966) (“The principal reason given for revision of the [the policy form] was adverse court decisions.”).
 See Ins. Servs. Office, Inc., Form No. GL 00 02 01 73, Commercial General Liability Coverage Form, Definitions (1973), reprinted in Malecki & Thamann, supra note 13, at 363.
 Stempel, supra note 35, at 368 (emphasis omitted) (quoting Norman Nachman, The New Policy Provisions for General Liability Insurance, 10 CPCU Annals 196, 199–200 (1965)). Nachman was the manager of casualty insurance and multiple lines insurance at the National Bureau of Casualty Underwriters, a predecessor to ISO.
 See French, “Non-Cumulation Clause,” supra note 13, at 387.
 See id.
 Id. at 387–88.
 French, “Non-Cumulation Clause,” supra note 13, at 386.
 See generally Boardman, supra note 31, at 7–8 (describing the origins and history of Sue and Labor Clauses and considering whether they are contractual black holes).
 Id. at 7.
 Ocean Towing Co. da Venezeula v. Cont’l Ins. Co., No. 89-2819, 1992 WL 40788, at *6 (E.D. La. Feb. 27, 1992).
 See Boardman, supra note 31, at 8.
 See id.
 See id.; see also Barry R. Ostrager & Thomas R. Newman, Handbook on Insurance Coverage Disputes § 21.02[e], at 1739 (18th ed. 2017).
 See, e.g., Albany Ins. Co. v. Anh Thi Kieu, 927 F.2d 882, 894 (5th Cir. 1991) (“Sue and labor expenses are sums spent by the assured in an effort to mitigate damages and loss. ‘The purpose of the sue and labor clause [in an insurance contract] is to reimburse the insured for those expenditures which are made primarily for the benefit of the insurer . . . .’” (alteration in original) (quoting Blasser Bros., v. N. Pan–Am. Line, 628 F.2d 376, 386 (5th Cir.1980))); Armada Supply Inc. v. Wright, 858 F.2d 842, 853 (2d Cir. 1988) (“Sue and labor expenses are those reasonable costs borne by the assured to mitigate the loss and thus reduce the amount to be paid by the underwriter.”).
 See Boardman, supra note 31, at 8.
 See supra note 62.
 See, e.g., Jeff Katofsky, Subsiding Away: Can California Homeowners Recover from Their Insurer for Subsidence Damages to Their Homes?, 20 Pac. L.J. 783, 785 (1989) (“In an ‘all-risk’ policy, all losses except those specifically excluded are covered. This is the broadest form of coverage and has been so interpreted by the courts.” (footnote omitted)).
 See, e.g., Ins. Servs. Office, Inc., Form No. HO 00 03 05 11, Exclusion A.2 (2010), reprinted in Abraham & Schwarcz, supra note 19, at 197 (excluding coverage for “loss caused directly or indirectly by . . . [e]arthquake . . . ” and “flood”); Ins. Servs. Office, Inc., Form No. CP 10 20 06 07, Commercial Property Broad Form, Exclusion (b) (2007), reprinted in Bruce J. Hillman, Commercial Property, 404–05 (4th ed. 2009) (noting that commercial all risk property policies exclude coverage “for loss or damage caused directly or indirectly by . . . [an] [e]arthquake, including any earth sinking, rising or shifting related to such event”); Christopher C. French, Insuring Floods: The Most Common and Devastating Natural Catastrophes in America, 60 Vill. L. Rev. 53, 61 (2015) (“Almost uniformly, [insurers] have refused to insure flood losses for non-commercial entities despite selling ‘all risk’ homeowners property policies.”).
 See Christopher C. French, The “Ensuing Loss” Clause in Insurance Policies: The Forgotten and Misunderstood Antidote to Anti-Concurrent Causation Exclusions, 13 Nev. L.J. 215, 216 (2012).
 Blaine Constr. Corp. v. Ins. Co. of N. Am., 171 F.3d 343, 346 (6th Cir. 1999) (quoting insurance policy at issue).
 See French, supra note 67, at 217.
 See id. at 216–17.
 Roberts v. State Farm Fire & Cas. Co., 705 P.2d 1335, 1336 (Ariz. 1985) (en banc).
 Ensue, Webster’s New World College Dictionary (5th ed. 2014).
 Blaine Constr. Corp., 171 F.3d at 346 (quoting insurance policy at issue).
 See French, supra note 67, at 228–34 (collecting cases).
 As one court stated:
The efficient proximate cause rule operates as an interpretive tool to establish coverage when a covered peril “sets other causes into motion which, in an unbroken sequence, produce the result for which recovery is sought.” The opposite proposition, however, is not a rule of law. When an excluded peril sets in motion a causal chain that includes covered perils, the efficient proximate cause rule does not mandate exclusion of the loss. “[T]he efficient proximate cause rule operates in favor of coverage. A converse rule would, of course, operate in favor of no coverage. . . . Because policies should normally be construed in favor of coverage, because there is no settled law favoring this argument, contrary to the insurer’s claim, and because the insurer does not offer any further justification or authority supporting such a rule, we decline to adopt the rule urged by the insurer.”
Vision One, LLC v. Phila. Indem. Ins. Co., 276 P.3d 300, 309 (Wash. 2012) (en banc) (first quoting McDonald v. State Farm Fire & Cas. Co., 837 P.2d 1000, 1004 (Wash. 1992) (en banc); then quoting Key Tronic Corp. v. Aetna (Cigna) Fire Underwriters Ins. Co., 881 P.2d 201, 206 (Wash. 1994) (en banc)).
 See, e.g., Swire Pac. Holdings, Inc. v. Zurich Ins. Co., 845 So. 2d 161, 167–68 (Fla. 2003) (holding that the repair of structural deficiencies due to design defects was not an ensuing loss because there was no property damage separate from the defects themselves).
 See, e.g., Roberts v. State Farm Fire & Cas. Co., 705 P.2d 1335, 1337 (Ariz. 1985) (en banc) (holding that damages caused by bees, an excluded peril, were covered due to an ensuing loss provision because the damage caused by honey leaking from the bees’ hive “ensued”—resulted—after the bees had been exterminated).
 Compare Roberts, 705 P.2d at 1337, with Swire Pac. Holdings, 845 So. 2d at 166.
 See supra Part II.A.
 New Castle Cty. v. Hartford Accident & Indem. Co., 933 F.2d 1162, 1197 (3d Cir. 1991), abrogated by N. Ins. Co. of N.Y. v. Aardvark Assocs., Inc., 942 F.2d 189 (3d Cir. 1991).
 See, e.g., Just v. Land Reclamation, Ltd., 456 N.W. 2d 570, 574 (Wis. 1990) (“At least with respect to environmental claims, contemporaneous industry commentary on the 1966 CGL policy indicates that there was no intent to avoid coverage for unexpected or unintended pollution.”); Morton Int’l, Inc. v. Gen. Accident Ins. Co. of Am., 629 A.2d 831, 849–71 (N.J. 1993) (discussing the evidence, commentators’ views, and case law regarding coverage for environmental injuries under the 1966 CGL policy form); see also Thomas Reiter, David Strasser & William Pohlman, The Pollution Exclusion Under Ohio Law: Staying The Course, 59 U. Cin. L. Rev. 1165, 1191–93 (1991) (discussing coverage for environmental claims under occurrence-based CGL policies and the history of insurers’ positions regarding such coverage).
 Robert Saylor & David Zolensky, Pollution Coverage and the Intent of the CGL Drafters: The Effect of Living Backwards, 1 Mealey’s Litig. Rep. (Ins.) 4425, 4432 (1987) (quoting Gilbert Bean, New Comprehensive General and Automobile Program, The Effect on Manufacturing Risks, presented at Mutual Insurance Technical Conference (Nov. 15–18, 1965)).
 Id. at 4438 n.34 (1987) (quoting Gilbert Bean, Summary of Broadened Coverage Under New CGL Policies With Necessary Limitation To Make This Broadening Possible (1966)).
 Stempel, Assessing The Coverage Carnage, supra note 35, at 372 (quoting [Anonymous Insurer], The Comprehensive General Liability Policy Workbook 11–12 (1973)).
 See, e.g., Warren Brockmeier, Pollution—The Risk and Insurance Problem, 12 For Def. 77, 77–78 (1971) (discussing changes to CGL coverage after environmental disasters in the 1960s); James Hourihan, Insurance Coverage for Environmental Damage Claims, 15 Forum 551, 553 (1980) (“Pollution claims burst on the insurance scene following the Torrey Canyon disaster and the Santa Barbara off-shore drilling oil spills in 1969.”).
 Pub. L. No. 91-224, § 102, 84 Stat. 91 (1970) (formerly codified at 33 U.S.C. § 1161) (superseded by Federal Water Pollution Control Act Amendments of 1972, § 2, Pub. L. No. 92-500, 88 Stat. 816 (1972) (codified at 33 U.S.C. § 1251 et seq. (2012))).
 See, e.g., Reiter et al., supra note 81, at 1196–1200; Ins. Servs. Office, Inc., Form No. GL 00 02 01 73, Commercial General Liability Coverage Form, Exclusion (f) (1973), reprinted in Malecki & Thamann, supra note 13, at 366.
 Ins. Servs. Office, Inc., Form No. GL 00 02 01 73, Commercial General Liability Coverage Form, Exclusion (f) (1973), reprinted in Malecki & Thamann, supra note 13, at 366.
 New Castle Cty. v. Hartford Accident & Indem. Co., 933 F.2d 1162, 1197 (3d Cir. 1991), abrogated by N. Ins. Co. of N.Y. v. Aardvark Assocs., Inc., 942 F.2d 189 (3d Cir. 1991); see Steven Plitt et al., 10A Couch on Insurance § 150:30 (3d ed. 2017) (“When coverage is limited to a sudden ‘breaking’ of machinery, the word ‘sudden’ should be given its primary meaning as a happening without previous notice or as something coming or occurring unexpectedly as unforeseen or unprepared for. That is, ‘sudden’ is not to be construed as synonymous with instantaneous.”)).
 See supra note 21 and accompanying text.
 See, e.g., Reiter et al., supra note 81, at 1200–05 (discussing the insurance industry’s representations regarding the scope of the Sudden and Accidental Pollution Exclusion); Morton Int’l, Inc. v. Gen. Accident Ins. Co. of Am., 629 A.2d 831, 848–53 (N.J. 1993) (same); Joy Techs., Inc. v. Liberty Mut. Ins. Co., 421 S.E.2d 493, 498–99 (W. Va. 1992) (same).
 Morton, 629 A.2d at 853 (second and third alterations in original) (quoting Claussen v. Aetna Cas. & Sur. Co., 676 F. Supp. 1571, 1573 (S.D. Ga. 1987)).
 See, e.g., id. at 848, 853; Reiter et al., supra note 81, at 1202.
 See, e.g., Reiter et al., supra note 81, at 1171 (noting that the estimated industry liability for the environmental cleanup was $150 billion to $700 billion).
 Solid Waste Disposal Act Amendments of 1980, Pub. L. No. 96-482, 94 Stat. 2334; Hazardous and Solid Waste Amendments of 1984, Pub. L. No. 98-616, 98 Stat. 3221 (current versions of both amendments can be found at 42 U.S.C. §§ 6901–6992 (2012)).
 42 U.S.C. §§ 9601–9675 (2012).
 See, e.g., O’Neil v. Picillo, 883 F.2d 176, 178 (1st Cir. 1989) (discussing CERCLA’s retroactive and joint and several liability); United States v. R.W. Meyer, Inc., 889 F.2d 1497, 1506 (6th Cir. 1989) (same); United States v. Stringfellow, 661 F. Supp. 1053, 1060, 1062 (C.D. Cal. 1987) (discussing CERCLA’s imposition of strict, and joint and several liability).
 See, e.g., Reiter et al., supra note 81, at 1171.
 See, e.g., E.I. du Pont de Nemours & Co. v. Admiral Ins. Co., 711 A.2d 45, 52 (Del. Super. Ct. 1995) (“[Insurers] argue the term ‘sudden’ in the exception to the pollution exclusion has a temporal meaning synonymous with ‘abrupt’ . . . .”); Morton Int’l, Inc. v. Gen. Accident Ins. Co. of Am., 629 A.2d 831, 852 (N.J. 1993) (noting that insurers’ position was that CGL policies only covered pollution if the releases causing the pollution were abrupt); Sinclair Oil Corp. v. Republic Ins. Co., 929 P.2d 535, 538 (Wyo. 1996) (stating that insurers contend “sudden incorporates a temporal element”); Reiter et al., supra note 81, at 1174 (noting that insurers generally argue that “sudden” means “abrupt” or “happening quickly”).
 See supra note 14 and accompanying text.
 See, e.g., E.I. du Pont de Nemours & Co.,711 A.2d at 54 (noting that the policyholder only obtained drafting history documents regarding the Sudden and Accidental Pollution Exclusion after successfully moving to compel their production).
 See, e.g., id. at 52 (“DuPont argues the term ‘sudden’ is ambiguous, and the [c]ourt should interpret ‘sudden’ to mean ‘unexpected.’”); Sinclair Oil Corp., 929 P.2d at 538 (stating that the policyholder “contends the term is ambiguous because although sudden can mean ‘abrupt’ or ‘happening quickly’ it can also mean ‘unexpected’”).
 See, e.g., Hoechst Celanese Corp. v. Nat’l Union Fire Ins. Co. of Pittsburgh, Pa., 623 A.2d 1128, 1130 (Del. Super. Ct. 1992) (“ISO produced approximately 250,000 to 275,000 pages of responsive documents, which had been previously collected for production by ISO and contained material related to the development of CGL language prior to March of 1983 and pollution coverage and exclusion language prior to December 1985.”); Morton, 629 A.2d at 848–53 (discussing the documentation regarding insurers’ statements to state insurance commissioners about the regulatory approval of the Sudden and Accidental Pollution Exclusion); Joy Techs., Inc. v. Liberty Mut. Ins. Co., 421 S.E.2d 493, 498–99 (W. Va. 1992) (same).
 See, e.g., Morton, 629 A.2d at 875 (applying regulatory estoppel to prevent the insurers from taking a position regarding the meaning of “sudden and accidental” that was inconsistent with their representations to state insurance commissioners); Joy, 421 S.E.2d at 500 (same).
 Compare Hecla Mining Co. v. N.H. Ins. Co., 811 P.2d 1083, 1092 (Colo. 1991) (en banc) (“Although ‘sudden’ can reasonably be defined to mean abrupt or immediate, it can also reasonably be defined to mean unexpected and unintended. Since the term ‘sudden’ is susceptible to more than one reasonable definition, the term is ambiguous, and we therefore construe the phrase . . . against the insurer . . . .”), and Claussen v. Aetna Cas. & Sur. Co., 380 S.E.2d 686, 690 (Ga. 1989) (“In sum, we conclude that the pollution exclusion clause is capable of more than one reasonable interpretation. The clause must therefore be construed in favor of the insured to mean ‘unexpected and unintended.’”), with Am. Motorists Ins. Co. v. ARTRA Grp., Inc., 659 A.2d 1295, 1308 (Md. 1995) (“We agree with the interpretation of the pollution exclusion clause adopted in numerous other cases. . . . Under those interpretations, the language of such an exclusion provides coverage only for pollution which is both sudden and accidental. It does not apply to gradual pollution. . . .”), and Upjohn Co. v. N.H. Ins. Co., 476 N.W.2d 392, 397 (Mich. 1991) (“We find persuasive the recent opinions . . . which find the terms of the pollution exclusion to be unambiguous. We conclude that when considered in its plain and easily understood sense, ‘sudden’ is defined with a ‘temporal element that joins together conceptually the immediate and the unexpected.’” (footnote omitted) (citations omitted)).
 See Ins. Servs. Office, Inc., Form No. CG 00 01 11 85, Commercial General Liability Coverage Form, Exclusion (f) (1986), reprinted in Malecki & Thamann, supra note 13, at 374 (reflecting the change from the Sudden and Accidental Pollution Exclusion to the Absolute Pollution Exclusion in ISO’s 1986 CGL policy form); see also Ostrager & Newman, supra note 61, at 1896–1900 (citing decisions in 14 states where courts rejected the insurers’ litigation position regarding the Sudden and Accidental Pollution Exclusion); Jeffrey W. Stempel, Reason and Pollution: Correctly Construing the “Absolute” Exclusion in Context and in Accord with Its Purpose and Party Expectations, 34 Tort & Ins. L.J. 1, 1–2 (1998) (“Responding to the flurry of environmental litigation over the application of the “sudden and accidental” pollution exclusion, the insurance industry during the mid-1980s largely adopted new standard pollution exclusion language for commercial general liability (CGL) policies. Since the mid-1980s, the standard form CGL has included the so-called absolute pollution exclusion. . . .” (footnote omitted)).