dissenting:
Because I believe the punitive damages award in this case is not “grossly excessive,” I would affirm. In reviewing the size of a punitive damages award, our sole duty is to ensure its imposition does not violate due process. Where an award lies within the bounds of due process, as this one does, we may not substitute a figure we consider more reasonable for one fairly awarded by a jury and properly reviewed by a district court. Therefore, I respectfully dissent.
1. Due Process Review of Punitive Damages
To comport with the Constitution, a punitive damages award must strike the proper balance between the state goals of deterrence and retribution and a defendant’s due process right to be free from arbitrary punishment. See State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408, 416-17, 123 S.Ct. 1513, 155 L.Ed.2d 585 (2003). The Supreme Court has determined the balance is upset at the point an award becomes “grossly excessive,” reasoning that, “[t]o the extent an award is grossly excessive, it furthers no legitimate purpose and constitutes an arbitrary deprivation of property.” Id. at 417, 123 S.Ct. 1513 (citing Pac. Mut. Life Ins. Co. v. Haslip, 499 U.S. 1, 42, 111 S.Ct. 1032, 113 L.Ed.2d 1 (1991)).
But as the majority notes, ante at 1082-83, the Court has shown little inclination to define “grossly excessive” more concretely. See State Farm, 538 U.S. at 424, 123 S.Ct. *10961513; BMW of N. Am., Inc. v. Gore, 517 U.S. 559, 582, 116 S.Ct. 1589, 134 L.Ed.2d 809 (1996). While it has several times hinted at the possibility of establishing a 4 to 1 benchmark ratio of punitive damages to compensatory damages, it has never explicitly done so. See State Farm, 538 U.S. at 425, 123 S.Ct. 1513 (citing BMW, 517 U.S. at 581, 116 S.Ct. 1589; Haslip, 499 U.S. at 23-24, 111 S.Ct. 1032). Instead, the one constitutional limit the Court has identified is that generally found between single-digit and double-digit multipliers. See id. (“[F]ew awards exceeding a single-digit ratio between punitive and compensatory damages, to a significant degree, will satisfy due process.... Single-digit multipliers are more likely to comport with due process, while still achieving the State’s goals of deterrence and retribution, than awards with ratios in range of 500 to 1[or] 145 to 1.” (internal citations omitted)).
The Supreme Court’s reluctance to establish a more concrete limit, or to adopt any other sort of categorical approach, counsels that in cases such as the one at bar, “[t]he judicial function is to police a range, not a point.” Mathias v. Accor Econ. Lodging, Inc., 347 F.3d 672, 678 (7th Cir.2003) (citing BMW, 517 U.S. at 582-83, 116 S.Ct. 1589; TXO Prod. Corp. v. Alliance Res. Corp., 509 U.S. 443, 458, 113 S.Ct. 2711, 125 L.Ed.2d 366 (1993)). We should let this punitive damages award stand unless the BMW factors indicate with some certainty that it was the product of caprice or bias such that its imposition violates Exxon’s right to due process.1 “Assuming that fair procedures were followed, a judgment that is a product of that process is entitled to a strong presumption of validity. Indeed, there are persuasive reasons for suggesting that the presumption should be irrebuttable, or virtually so.” TXO, 509 U.S. at 457, 113 S.Ct. 2711 (plurality opinion) (internal citations omitted).
No procedural concerns are present here that, at the outset, might weaken the “strong presumption of validity” to which this award is entitled. See BMW, 517 U.S. at 586-87, 116 S.Ct. 1589 (Breyer, J., concurring) (citing TXO, 509 U.S. at 457, 113 S.Ct. 2711; Haslip, 499 U.S. at 40-42, 111 S.Ct. 1032); see also id. at 583, 116 S.Ct. 1589 (“In most cases, the ratio will be within a constitutionally acceptable range, and remittitur will not be justified on this basis.”). The jury received thorough, almost prescient, punitive damages instructions.2 And although Exxon is a large *1097corporation, there is no indication that the size of this punitive damages award resulted from an improper “emphasis on the wealth of the wrongdoer” at trial, see TXO, 509 U.S. at 464, 113 S.Ct. 2711, or from an attempt by Plaintiffs or the jury to “make up for the failure of other factors, such as ‘reprehensibility,’ ” see BMW, 517 U.S. at 591, 116 S.Ct. 1589 (Breyer, J., concurring).3
Furthermore, Exxon’s conduct implicates a strong state interest in punishing reckless behavior and deterring its future repetition. Our constitutional review must consider punitive damages in the context of these state interests. See id. at 568, 116 S.Ct. 1589 (“Only when an award can fairly be categorized as ‘grossly excessive’ in relation to these interests does it enter the zone of arbitrariness that violates the Due Process Clause of the Fourteenth Amendment.” (emphasis added)). In both State Farm and BMW, the Court’s guidepost analysis was not an entirely separate endeavor, but instead gave structure to its constitutional concern that the defendants’ due process rights were violated by judgments incorporating punishment for conduct not properly before the awarding court. See State Farm, 538 U.S. at 419-24, 123 S.Ct. 1513 (discussing out-of-state conduct and conduct unrelated to plaintiffs’ injuries); BMW, 517 U.S. at 568-73, 116 S.Ct. 1589 (describing out-of-state conduct).
In stark contrast, there is no concern here that the scope of appropriate state interests has been exceeded. This punitive damages award was imposed pursuant to strong, but properly circumscribed, state interests. As the district court noted, Plaintiffs’ collection of federal and state claims all arise out of harm to “Alaska fisheries, Alaska business,[and] Alaska property” caused by Exxon’s conduct having “a direct nexus with the grounding of the Exxon Valdez on Bligh Reef in Prince William Sound.” See In re Exxon Valdez, 296 F.Supp.2d at 1090-91.
Thus, before engaging in the multi-fac-tored analysis introduced in BMW and reiterated in State Farm, it is important to note that we are not faced here with any of the major constitutional concerns present in those cases.
2. BMW Guidepost Analysis
Although I agree with much of the majority’s analysis under BMW and State Farm, I cannot agree with it all. Despite clear guidance from the Court that reprehensibility is the critical factor, the majority, ante at 1083, 1088-89, gives defining weight to a consideration entirely of its own creation. It then engages, ante at 1093-94, in what appears to be the very “categorical approach” the Supreme Court has consistently rejected. See BMW, 517 U.S. at 582, 116 S.Ct. 1589. An appropriate evaluation of the award in question demonstrates it is constitutionally permissible.
(a) Reprehensibility
In its most recent punitive damages opinion, the Supreme Court gave direct instruction to courts evaluating reprehensibility. State Farm, 538 U.S. at 419, 123 S.Ct. 1513. As the majority correctly notes, ante at 1080-81, we must weigh five factors: (1) whether the harm was solely *1098economic, (2) whether the con duct showed indifference to or reckless disregard for others’ health and safety, (3) whether the conduct’s target was financially vulnerable, (4) whether the conduct involved repeated actions, and (5) whether the harm resulted from intentional malice or mere accident. State Farm, 538 U.S. at 419, 123 S.Ct. 1513. Somewhat inexplicably, though, the majority adds to the State Farm factors one of its own creation — post-tort mitigation. See ante at 1084 (“We must also consider mitigating factors.”); id. at 1088. I do not agree that mitigation should be considered in a reprehensibility analysis. Furthermore, unlike the majority, I believe that all five State Farm factors weigh in favor of finding that Exxon’s reckless conduct was highly reprehensible.
(i) Mitigation
I cannot agree with the majority’s assertion that we must consider Exxon’s post-tort mitigation in evaluating the reprehensibility of its original misconduct. See ante at 1083-84. The majority is correct that when we previously considered Exxon’s conduct, we suggested mitigation should be considered as part of the reprehensibility analysis. See Baker v. Hazelwood (In re the Exxon Valdez), 270 F.3d 1215, 1242 (9th Cir.2001) [hereinafter Punitive Damages Opinion /]. However, subsequent to our decision in Punitive Damages Opinion I, the Supreme Court decided State Farm, which significantly refined the Court’s punitive damages jurisprudence. The analysis of reprehensibility in State Farm differs from our analysis in Punitive Damages Opinion I, and, as intervening controlling authority, gives us reason to reconsider our prior approach. See United States v. Bad Marriage, 439 F.3d 534, 538 (9th Cir.2006) (noting that a court may reexamine an issue it previously decided if “intervening controlling authority makes reconsideration appropriate”).
When we considered mitigation in Punitive Damages Opinion I, Supreme Court precedent provided limited guidance for the reprehensibility analysis. In State Farm, however, the Supreme Court explained that courts should use five specific factors to evaluate reprehensibility. 538 U.S. at 419, 123 S.Ct. 1513. Although there was evidence of mitigation in State Farm, id. at 426, 123 S.Ct. 1513, the Court did not include mitigation as one of the factors in the reprehensibility analysis. Given such explicit guidance, this omission acquires particular significance and suggests we reconsider our prior statement about mitigation.4 As explained below, upon reconsideration I find that including mitigation in the reprehensibility analysis is neither good law nor good policy.
Aside from a single mention of mitigation in Punitive Damages I, the majority’s approach is supported by neither Supreme Court precedent nor our own precedent. The majority cites Swinton v. Potomac Corp., 270 F.3d 794 (9th Cir.2001), as support, even though Swinton, like Punitive Damages Opinion I, was decided prior to State Farm. Therefore, it did not have the benefit of the Supreme Court’s most recent and comprehensive analysis of reprehensibility. Furthermore, Swinton did not consider whether mitigation warrants a reduction in a punitive damages award imposed by a jury. Rather, our analysis was limited to the question of whether the district court erred in excluding evidence of mitigation efforts in an employment dis*1099crimination suit. See id. at 811, 815. We refused in that case to create a generalized rule in the employment context or anywhere else. See id. at 814-15. Instead, we left it to the discretion of the district courts to decide the relevancy of mitigation efforts on a case-by-case basis.
We also expressly rejected the idea that the Supreme Court endorses the categorical relevance of mitigation in punitive damages calculations. See id. at 812 (“We do not interpret the language in BMW and Cooper as relying on evidence of post-occurrence remediation for overturning the punitive damages awards; rather the Court appears simply to have been recounting a full history of the litigation to give a complete picture of the proceedings.”). While post-tort mitigation by a defendant may or may not be relevant to a jury’s determination of whether and in what amount to award punitive damages, Swinton gives no support to the majority’s position that mitigation is properly considered as part of the reprehensibility analysis in a constitutional review.
Additionally, the majority’s approach makes little sense as a matter of policy, for it runs directly counter to the twin goals of punitive damages: deterrence and retribution. See State Farm, 538 U.S. at 416, 123 S.Ct. 1513 (“[Pjunitive damages serve a broader function; they are aimed at deterrence and retribution.”); Theodore Eisen-berg, Damage Awards in Perspective, 36 Wake Forest L.Rev. 1129, 1145 (2001) (“[A] wrongdoing party’s voluntary — to the extent payments are truly voluntary after being ‘caught’ — remediation payment does not reduce the propriety of punishing or deterring.”). While including mitigation in the reprehensibility analysis doubtlessly increases the incentive to remediate, it does so at the expense of undermining deterrence and retribution. The majority’s approach minimizes deterrence by creating a post-tort means of limiting punitive damages. This allows potential tortfeasors to engage in risky behavior, safe in the knowledge they can minimize liability for any resulting harm by prompt payment of foreseeable damages. It also cripples the state’s interest in retribution, as it allows the tortfeasor, rather than the jury, to recharacterize the reprehensibility of its misconduct after a tort has been committed. Cf. Cooper, 532 U.S. at 432, 121 S.Ct. 1678 (recognizing that the “imposition of punitive damages is an expression of [the jury’s] moral condemnation”).
Nonetheless, the majority insists that including mitigation in the reprehensibility analysis is good public policy because it encourages socially beneficial conduct. Ante at 1088. A company in Exxon’s position, however, already has significant incentives to clean up its mess. Had Exxon not taken prompt action to clean up the oil spill and compensate injured parties, see ante at 1072-73, the actual harm caused could well have exceeded the $504.1 million figure we use as the numerator in our ratio analysis. See ante at 1093. Specifically, if eleven billion gallons of oil were left indefinitely in Prince William Sound, and injured parties were without resources to start their lives anew, both economic and social harm would have grown. This would have increased Exxon’s liability not only for compensatory damages, but also for punitive damages. Greater actual harm translates to a larger punitive damages numerator and a higher ceiling for the punitive damages award. Thus, mitigation is already reflected in the calculation of compensatory damages and in our constitutional review of the jury’s punitive damage award.
Moreover, I am not convinced the majority’s approach will ultimately encourage defendants to settle. Cf. Franklin v. Kaypro Corp., 884 F.2d 1222, 1229 (9th Cir.1989) (noting there is an “overriding public *1100interest” in promoting settlement). Instead, I fear it has the unintended consequence of giving tortfeasor defendants a way to reduce the risk of litigation without reaching a settlement with injured parties. Under our past precedent, the threat of a significant punitive damages award created a strong incentive for defendants to pay injured parties in exchange for a release or similar arrangement.5 The majority’s approach, however, allows defendants to limit their exposure to punitive damages by taking unilateral steps, even token ones, to remediate harm. I am concerned this will frequently lead to more protracted litigation, as injured parties will not necessarily be satisfied with defendants’ mitigation efforts, and defendants will have less incentive to reach settlement agreements. Thus, policy implications support the legal conclusion that it is not appropriate to add mitigation to the State Farm factors.
(ii) State Farm Factors
Because I see no basis for the majority’s inclusion of mitigation in our due process reprehensibility analysis, I consider only the five factors outlined by the Supreme Court. I agree with the majority that the first, second, and fourth factors6 suggest Exxon’s conduct was highly reprehensible and capable of supporting a substantial award. However, I cannot agree with the analysis concerning the fifth factor, whether “the harm was the result of intentional malice, trickery, or deceit, or mere accident.” State Farm, 588 U.S. at 419, 123 S.Ct. 1513. As the majority recognizes, Exxon’s decision to put a relapsed alcoholic in charge of a supertanker constituted knowing and reckless misconduct, which was neither intentionally malicious nor a mere accident. Ante at 1088. However, faced with conduct that does not fit squarely in either category mentioned in State Farm, the majority arbitrarily determines this factor weighs against high reprehensibility because Exxon “did not spill the oil on purpose.” Id, at 1088. I cannot agree with this conclusion for two reasons.
First, if we read this State Farm factor to recognize only two categories of conduct, the fact that Exxon’s acts fall in neither category could suggest this is a neutral factor, weighing neither for nor against high reprehensibility. However, if *1101the majority is correct that we must determine whether Exxon’s conduct is more similar to one category or the other,7 I believe it is closer to “intentional malice, trickery, or deceit” than to “mere accident.” State Farm, 538 U.S. at 419, 123 S.Ct. 1513; cf. Black’s Law Dictionary 968 (7th ed.1999) (defining malice as, inter alia, “[rjeekless disregard of the law or of a person’s legal rights”). The jury held Exxon responsible not merely for spilling oil, but rather for knowingly giving command of a supertanker “carrying over 53 million gallons of volatile, toxic, crude oil” to a relapsed alcoholic. See In re Exxon Valdez, 296 F.Supp.2d at 1097. Exxon did so for three years with full knowledge of the tremendous risk of serious harm to the health, safety, and livelihood of many people. See ante at 1086-87. This cannot fairly be described as an accident. Given the extreme recklessness of Exxon’s conduct, I would conclude the fifth factor militates in favor of finding Exxon’s behavior highly reprehensible. Accord Swinton, 270 F.3d at 818 (holding that conduct which was, at most, reckless disregard for others’ health and safety, easily “constitutes highly reprehensible conduct justifying a significant punitive damages award”).
Thus, unlike the majority, I find that all five of State Farm’s reprehensibility factors suggest that Exxon’s reckless conduct in this case- — the malicious endangerment of the property and livelihood of thousands of Alaskans — was highly, if not extremely, reprehensible and capable of “warranting] a substantial penalty.” See BMW, 517 U.S. at 576, 116 S.Ct. 1589.
(b) Ratio
Under the second BMW guidepost, we must analyze “the disparity between the actual or potential harm suffered by the plaintiff and the punitive damages award.” See id. at 418, 123 S.Ct. 1513. While I agree with the majority’s “calculation of harm,” or “numerator,” analysis, ante at 1093,1 cannot agree with its conclusion, id. at 1094, that the Constitution prohibits a ratio in this case above 5 to 1. The majority arrives at this constitutional limit through two steps. First, it uses the “rough framework” of Planned Parenthood of Columbia/Willamette, Inc. v. American Coalition of Life Activists, 422 F.3d 949 (9th Cir.2005), to arrive at the conclusion that the appropriate ratio in this case is above 4 to 1, but no greater than 9 to 1. Ante at 1093-94. However, it then asserts the proper ratio cannot be much greater than 4 to 1 because Exxon’s conduct was not intentional and because Exxon attempted to mitigate the harm it caused. Ante at 1094. I cannot agree with this.
In Planned Parenthood, we established a three-tiered “rough framework” to guide us in determining an appropriate ratio.8 Applying Planned Parenthood to this case, the majority concludes a 4 to 1 benchmark is appropriate based on its determination that the economic damages are “signifi*1102cant.” Ante at 1093-94. As an initial matter, the majority’s assessment of economic damages focuses on a number devoid of its context. An award is significant not because it is numerically large, but rather because it approaches full compensation for the plaintiffs harms. See State Farm, 538 U.S. at 426, 123 S.Ct. 1513 (“The compensatory award in this case was substantial;[the plaintiffs] were awarded $1 million for a year and a half of emotional distress. This was complete compensation.”). I am not convinced that a compensatory damages award that equates to a mere $10,000 per plaintiff is actually “substantial” in the way the Supreme Court uses the term. Cf. id. at 425, 123 S.Ct. 1513 (providing, as an example of “small” economic damages, cases where the injury was hard to detect or not fully economic in nature).
' Even if the majority were correct that the economic damages awarded in this case are “significant,” Planned Parenthood still does not support a 4 to 1 benchmark in this case. In Planned Parenthood, we refused to remit the award to less than a 9 to 1 ratio because not all of the plaintiffs damages were quantifiable, not all of it was compensated, and the plaintiffs were likely to incur further costs. 422 F.3d at 963. All three are true here as well. The oil spill disrupted the social fabric of the plaintiffs’ community. See In re Exxon Valdez, 296 F.Supp.2d at 1094. This type of harm is not easily quantifiable. Moreover, the plaintiffs’ recovery in this case was limited to economic harm. It therefore did not compensate the plaintiffs for harm attributable to increased “social conflict, cultural disruption and psychological stress.” Id. Finally, there is evidence the plaintiffs have incurred substantial further costs. See id. Thus, it cannot be said the compensatory damages in this case are so large or sufficiently comprehensive they warrant a lower punitive damages award.
Nor, in my mind, does the majority find support in Zhang v. American Gem Seafoods, Inc., 339 F.3d 1020 (9th Cir.2003), or Bains LLC v. Arco Products Co., 405 F.3d 764 (9th Cir.2005). That we upheld an award in the 7 to 1 range in Zhang, and remanded for a similar award in Bains— both for intentional racial discrimination in the employment context — says little if nothing about the constitutionality of this award for the reckless endangerment of the property and livelihood of tens of thousands of people. While it is true any given conduct is more reprehensible if intentional than if reckless, it does not necessarily follow that all intentional conduct is more reprehensible than all reckless conduct. Indeed, because we are the first court to review an award for misconduct resulting in harm of the type and scale at issue here, I find it unhelpful to note that our cases to date “have generally reserved high single-digit ratios for the most egregious forms of intentional misconduct, such as threats of violence and intentional racial discrimination.” See ante at 1094. Instead, every indicator in this ease suggests that Exxon’s reckless conduct — leaving for three years a known alcoholic in command of a supertanker in treacherous waters upon which thousands of people depend — is egregious enough to support an award within the 9 to 1 range. Accord Swinton, 270 F.3d at 818-20 (upholding a 28 to 1 ratio despite recognizing that the conduct at issue involved no acts or threats of violence and, therefore, “[did] not amount to the worst kind of tortious conduct a defendant can commit”).
One final consideration convinces me that the 8.93 to 1 ratio in this case does not indicate that Exxon has been subject to a “grossly excessive” punitive damages award. In State Farm, the Supreme Court reiterated that it is appropriate to consider for purposes of ratio calculation not only the actual harm caused, but the potential harm that a defendant’s miscon*1103duct could have foreseeably caused. See 538 U.S. at 418, 123 S.Ct. 1513 (describing the second guidepost as requiring consideration of “the actual or potential harm suffered” (emphasis added) (citing BMW, 517 U.S. at 575, 116 S.Ct. 1589)); accord TXO, 509 U.S. at 460, 113 S.Ct. 2711 (“Taking account of the potential harm that might result from the defendant’s conduct in calculating punitive damages was consistent with the views we expressed in Haslip.” (internal citation omitted)). As the majority recognizes, ante at 1086-87, the potential harm from Exxon’s decision to keep Hazelwood in command of the Exxon Valdez was both massive and foreseeable. But despite the propriety of such consideration, the calculation of harm in this case explicitly incorporates only an estimate of actual, and not of potential, harm. See In re Exxon Valdez, 296 F.Supp.2d at 1103; ante at 1093. Thus, if anything, the jury’s punitive damages award potentially undervalued the harm. Conclusion
In accordance with State Farm and its predecessors, we are required to subject this award to “exacting [de novo] appellate review” in order to ensure it is “based upon an application of law, rather than a decisionmaker’s caprice.” See 538 U.S. at 418, 123 S.Ct. 1513 (internal quotation marks omitted) (quoting BMW, 517 U.S. at 587, 116 S.Ct. 1589 (Breyer, J., concurring)). But that review does not empower us to substitute our own, perhaps more finely-tuned, award for one that was fairly awarded and already lies within the range of constitutional awards. See BMW, 517 U.S. at 583, 116 S.Ct. 1589 (noting that most awards fall within a “constitutionally acceptable range ” (emphasis added)).
After thorough and concerned analysis of this punitive damages award, I conclude that its imposition does not violate Exxon’s constitutional right to due process. The award was levied as a result of fair procedure and in pursuit of the undisputedly strong, and properly circumscribed, state interests in punishing Exxon for its misconduct, and in deterring any similar behavior by Exxon in waters it continues to frequent. While the award is large, it addresses what must be characterized as extremely reprehensible misconduct. There is simply no excuse for allowing a relapsed alcoholic to pilot a supertanker in any waters, much less for three years in the treacherous and treasured waters of Prince William Sound. Exxon’s knowing decision to do so was a malicious one that placed at massive risk, and ultimately seriously injured, the property and livelihood of tens of thousands of Alaskans. There is every indication the award before us reasonably addresses that egregious behavior, and nothing in the record that suggests it resulted from passion, bias, or caprice. I therefore agree with the district court’s assessment that there is no principled means by which this award should be reduced. See In re Exxon Valdez, 296 F.Supp.2d at 1110. Accordingly, and with respect, I dissent.
. The majority correctly recognizes, ante at 1072, that a determination that an award is "grossly excessive” is reviewed de novo. Cooper Indus., Inc. v. Leatherman Tool Group, Inc., 532 U.S. 424, 436, 121 S.Ct. 1678, 149 L.Ed.2d 674 (2001). De novo review, however, is only applied to determine the constitutional upper limit on a punitive damages award in a given case. If the award does not exceed this ceiling, we owe deference to the determination of the district court and jury. See id. at 433-34, 121 S.Ct. 1678 (noting that within substantive limits on an award, the jury has discretion in establishing the precise number). Cooper does not give us free reign to pick the number we would have chosen had we sat as the jury or district court.
. The district court explained the retributive and deterrent purposes of punitive damages and the "appropriate,” i.e., non-environmental, countervailing "Alaska-oriented” interests of the plaintiffs; cautioned the jury that punitive damages must have a rational basis in tire record and bear a reasonable relationship to the harm; admonished the jury not to be arbitrary; and, perhaps most importantly, alerted them that they could take Exxon’s mitigation efforts into account when determining both whether punitive damages were warranted and, if so, the size of the award. See In re Exxon Valdez, 296 F.Supp.2d 1071, 1091 (D.Alaska 2004). Considering that BMW and State Farm were decided after the jury trial, these instructions indeed were, as the majority notes, ante at 1075, "in retrospect, quite forward looking.”
. Indeed to the contrary, there is evidence in the record comparing this award to Exxon’s wealth in a manner that suggests the award was neither capricious nor an instance of over-deterrence. See In re Exxon Valdez, 296 F.Supp.2d at 1105-06 (“[A]fter judgment was entered on the punitive damages award, Exxon’s treasurer advised the court that 'the full payment of the Judgment would not have a material impact on the corporation or its credit quality.' ”).
. The majority suggests State Pann is distinguishable because the dispute concerned an insurance contract rather than a toxic tort. See ante at 1074, 1084. However, the five-part reprehensibility analysis in State Farm is designed to evaluate a broad range of conduct, and nothing in the opinion indicates this framework applies only to insurance cases.
. In this case, the certification of a mandatory punitive damages class meant that individual plaintiffs could not reduce the ultimate punitive damages award by releasing their claims. See In re Exxon Valdez, 229 F.3d 790, 793 (9th Cir.2000) ("Claims for compensatory damages could be easily disposed of by exchanging payment for releases, but a plaintiff's release of its slice of the future lump-sum punitive damages award merely reduced the number of claimants sharing the punitive damages pie, not the size of the pie itself.”). However, several plaintiffs nonetheless used the looming punitive damages award as a bargaining chip by allocating Exxon a portion of any award they might receive. See ante at 1075.
. I am not convinced by the majority’s analysis of the third factor, but I do agree that it plays a relatively small role in this case and therefore does not warrant an extended discussion. The majority classified as neutral the third factor, whether "the target of the conduct had financial vulnerability,” see State Farm, 538 U.S. at 419, 123 S.Ct. 1513. As the majority admits, ante at 1088, by recklessly placing a "relapsed alcoholic in charge of a supertanker,” Exxon knew that it was "imposing a tremendous risk on a tremendous number of people who could not do anything about it.” Not only were many of those people “financially vulnerable” by virtue of being subsistence fishermen, but they were also particularly vulnerable to the specific risk imposed on them by Exxon. See In re Exxon Valdez, 296 F.Supp.2d at 1094-95. Thus, I would find this factor indeed suggests Exxon's reckless conduct was highly reprehensible. See BMW, 517 U.S. at 576, 116 S.Ct. 1589 ("To be sure, infliction of economic injury, especially ... when the target is financially vulnerable, can warrant a substantial penalty.”).
. Contrary to the majority’s assertion, ante at 1084, I do not suggest it views Exxon’s conduct as a largely excusable accident. Rather, I note that in finding this factor “militates against viewing Exxon’s misconduct as highly reprehensible,” ante at 1088, the majority treats Exxon’s reckless misconduct as it would treat an accident. This is not consistent with the majority’s own statement that "the reprehensibility of Exxon’s conduct that produced economic harm to thousands of individuals is high ...” Id.
. Where the economic damages are significant but the behavior not “particularly egregious,” a ratio of less than 4 to 1 is warranted. Planned Parenthood, 422 F.3d at 962. If the economic damages are significant but the behavior “more egregious,” a ratio greater than 4 to 1 might be acceptable. Id. Finally, if the economic damages are insignificant but the behavior is "particularly egregious,” ratios beyond single digits may be appropriate. Id.