Emory Law Journal

Exxon Shipping Co. v. Baker: The Perils of Judicial Punitive Damages Reform
Jeff Kerr J.D. Candidate, Emory University School of Law (2010). I am especially grateful to Professors William Buzbee, Joanna Shepherd, and Alexander “Sasha” Volokh for their guidance and insightful feedback throughout the writing process. I am also grateful to the editing team at the Emory Law Journal. And finally, I would like to thank my wife Lindy.

Abstract

The Supreme Court’s recent decision in Exxon Shipping Co. v. Baker established a conservative one-to-one cap on the ratio of punitive to compensatory damages in maritime law. This decision raises the question whether the Court will apply a similar constitutional limit in future punitive damages cases. In the meantime, lower courts have already begun to rely on Exxon Shipping as persuasive authority for limiting punitive damages further than the Supreme Court’s previous cases require. This Comment argues that Exxon Shipping’s one-to-one cap in maritime cases is inconsistent with key principles of punitive damages law, advises against the application of Exxon Shipping’s one-to-one cap in non-maritime cases, and explains why the Supreme Court should not enact a similar cap on punitive damages in future constitutional cases.

Punitive damages are too important to be capped at a one-to-one ratio with compensatory damages. This Comment explains that such a cap has the potential to create significant economic inefficiencies. Moreover, a one-to-one cap undermines the retributive role of punitive damages since reprehensible conduct often may not result in substantial compensatory damages. The rule of Exxon Shipping will likely remain the law in admiralty, but, as this Comment argues, courts should not to expand the rule of Exxon Shipping beyond maritime cases.

Introduction

In Exxon Shipping Co. v. Baker, the U.S. Supreme Court held that punitive damages for reckless conduct should be limited to a one-to-one ratio with compensatory damages as a matter of maritime common law. 1128 S. Ct. 2605, 2633 (2008).Exxon Shipping represents a departure from the Court’s prior punitive damages cases such as State Farm Mutual Automobile Insurance Co. v. Campbell 2538 U.S. 408 (2003). and BMW of North America, Inc. v. Gore, 3517 U.S. 559 (1996). where the Court evaluated punitive damages awards under the Due Process Clause. 4 Id. at 562 (“The Due Process Clause of the Fourteenth Amendment prohibits a State from imposing a ‘grossly excessive’ punishment on a tortfeasor.” (quoting TXO Prod. Corp. v. Alliance Res. Corp., 509 U.S. 443, 454 (1993))). In some respects, Exxon Shipping is an extension of these earlier cases because while the Court ostensibly based its holding on maritime law, it did not rely on maritime law precedent. 5 See Exxon Shipping, 128 S. Ct. at 2619–34. Rather, the Court based its decision upon fairness considerations, policy analysis, and statistical studies of punitive damages. This Comment argues that the Court’s reasoning fails to justify its strict limitation of maritime punitive damages. Building upon and revising some of the conclusions of earlier scholarship, this Comment demonstrates that punitive damages often need to exceed a one-to-one ratio with compensatory damages to deter future harms and provide retributive justice. 6 See Joanna M. Shepherd, Tort Reforms’ Winners and Losers: The Competing Effects of Care and Activity Levels, 55 UCLA L. Rev. 905, 910 (2008) (“Deterrence is the function of tort law by which the law creates incentives that induce people to avoid inappropriately dangerous activities.”). Finally, just as the Court criticizes “outlier” punitive damages awards, 7 Exxon Shipping, 128 S. Ct. at 2633 (“[T]he unpredictable outlier cases . . . call the fairness of the system into question.”). this Comment argues that Exxon Shipping itself should be viewed as an “outlier” case and should not be treated as persuasive authority for placing further limits on non-maritime punitive damages.

This Comment focuses on the potentially far-reaching implications of Exxon Shipping’s one-to-one cap on punitive damages for reckless conduct. Although punitive damages are a controversial feature of American law, 8Alex Sienkiewicz, Towards a Legal Land Ethic: Punitive Damages, Natural Value, and the Ecological Commons, 15 Penn St. Envtl. L. Rev. 91, 99 (2006). this Comment shows that punitive damages are too important to be rendered ineffectual by overly stringent caps. By increasing liability to an amount in excess of what is required to compensate the plaintiff, a punitive award goes beyond the traditional goal of making the plaintiff whole. 9 See Racich v. Celotex Corp., 887 F.2d 393, 396 (2d Cir. 1989). Important deterrence and retributive rationales justify this extra-compensatory penalty. 10 Exxon Shipping, 128 S. Ct. at 2621 (“Regardless of the alternative rationales over the years, the consensus today is that punitives are aimed . . . at retribution and deterring harmful conduct.”); Marc Galanter & David Luban, Poetic Justice: Punitive Damages and Legal Pluralism, 42 Am. U. L. Rev. 1393, 1428 (1993). Although these rationales occasionally conflict, 11Mark A. Drumbl, Collective Violence and Individual Punishment: The Criminality of Mass Atrocity, 99 Nw. U. L. Rev. 539, 577 (2005) (“[T]he goals of deterrence . . . may well conflict with those of retribution . . . .”). this Comment argues that they frequently overlap—especially when the defendant’s conduct is reckless. This analysis reveals that the Court’s adoption of a one-to-one cap in cases of reckless conduct lacks support. To be useful, punitive damages often must be awarded at a higher ratio. 12Galanter & Luban, supra note 10, at 1396 (“[I]f punitive damages are pared back too drastically, civil law may be underenforced.”).

Part I of this Comment examines facts of the Exxon Valdez oil spill that were glossed over in the mainstream media and demonstrates why, from a retributivist perspective, the Exxon corporation may have deserved punishment. Part II demonstrates that punitive damages are a vital part of the common law, many states have already limited punitive damages, and the Supreme Court’s constitutional jurisprudence on punitive damages adequately protects defendants from egregious punitive awards. The heart of this Comment, Part III, presents an economic analysis of various situations in which punitive damages should be awarded at a greater than one-to-one ratio, explains the retributivist approach to punitive damages, and illustrates how the two can be reconciled. Part IV critiques the Court’s reasoning in Exxon Shipping, emphasizing its failure to take account of the role of punitive damages in providing deterrence and retribution. Finally, Part V argues that Exxon Shipping should not be viewed as persuasive authority for limiting punitive damages in non-maritime cases.

I. Factual Background

Although the Valdez oil spill was one of the most publicized anthropogenic environmental disasters in history, 13 James W. Dearing & Everett M. Rogers, Agenda-Setting 39 (1996) (noting the “massive media attention given to the Exxon Valdez incident”); Hannah Lendon & Brian Martin, Environmental Disasters, in Brian Martin et al., Justice Ignited 99, 103 (2006) (“Of all spills, the Exxon Valdez is most well-known.”). many of the facts surrounding the incident were deliberately obscured by Exxon’s public relations experts and are not well-known. 14 See Riki Ott, Not One Drop: Betrayal and Courage in the Wake of the Exxon Valdez Oil Spill 19, 66–67 (2008) (“Exxon launched an aggressive public relations campaign to quiet ‘rumors’ of extensive damage from its spill.”). These facts reveal that the spill was the result of Exxon’s foolish decision to allow a captain who was a “relapsed alcoholic” to pilot the Valdez. 15 Exxon Shipping, 128 S. Ct. at 2640 (Breyer, J., concurring and dissenting) (“The jury could reasonably have believed that Exxon knowingly allowed a relapsed alcoholic repeatedly to pilot a vessel filled with millions of gallons of oil through waters that provided the livelihood for the many plaintiffs in this case.”); see also id. at 2612 (majority opinion) (detailing Captain Hazelwood’s consumption of alcohol). In addition to the circumstances of the spill itself, Exxon’s promises before the spill and its behavior after the spill underscore the justification for retributivist damages. 16 Charles Perrow, Normal Accidents 374–75 (1999) (describing the contingency plans that oil companies were required to submit before being allowed to operate in Prince William Sound as “fantasy documents”).

In January 1968, America’s largest oil field 17 Ott, supra note 14, at 273. was discovered 250 miles north of the Arctic Circle in Prudhoe Bay, Alaska. 18 U.S. Gen. Accounting Office, Alaskan North Slope Oil: Limiting Effects of Lifting Export Ban on Oil and Shipping Industries and Consumers 12 (1999), available at http://www.gao.gov/archive/1999/rc99191.pdf. Because the surrounding ocean is frozen much of the year at this latitude, several oil companies proposed to build the 800-mile Trans-Alaska Pipeline System (TAPS) to transport oil from Prudhoe Bay to Valdez, Alaska, where it would be pumped into tankers for marine transport. 19 Id. It was clear at an early stage that the environmental risks of the project included the possibility of massive oil spills that could jeopardize the ecology and economy of Prince William Sound and disrupt the subsistence lifestyles of Alaskans and Native Americans living in the area. 20Brief of the Alaska Legislative Council et al. as Amici Curiae Supporting Respondents at 10, Exxon Shipping Co. v. Baker, 128 S. Ct. 2605 (2008) (No. 07-219), available at http://www.abanet.org/publiced/preview/briefs/pdfs/07-08/07-219_RespondentAmCuAlaskaLegCouncil.pdf.

To alleviate fears of an environmental catastrophe, the oil companies involved, including Exxon, promised both the public and Congress (whose approval was required) 21 Ott, supra note 14, at 24. that they would adhere to high standards of care to curtail or even eliminate the risk of major oil spills. 22Brief of the Alaska Legislative Council et al., supra note 20, at 10; Ott, supra note 14, at 273 (“[In 1972] oilmen and/or [the] Nixon administration repeatedly promised that state-of-the-art construction, tankers, navigational procedures, and oil spill response equipment will make ‘operations in Port Valdez and Prince William Sound the safest in the world.’”). These promises convinced Congress, and in the summer of 1977, the first tanker carrying oil from Prudhoe Bay cast off from the Port of Valdez into the waters of Prince William Sound. 23 U.S. Gen. Accounting Office, supra note 18, at 12.

Within two years, the oil companies were beginning to disregard their earlier promises to Congress, including the pledge to use double-bottomed tankers, which would have limited the impact of a spill by as much as fifty percent. 24 Ott, supra note 14, at 2; see also U.S. Coast Guard, Oil Spill Response Research & Development Program: A Decade of Achievement 37 (2003). The companies also fell behind in contingency planning and preparedness. After the Valdez spill, an investigative team found that Exxon’s “emergency plan” contained no contingency planning specifically tailored to conditions at Prince William Sound. 25 Samuel K. Skinner & William K. Reilly, Nat’l Response Team, The Exxon Valdez Oil Spill: A Report to the President 8 (1989) [hereinafter Nat’l Response Team Report], available at www.uscg.mil/History/.../ExxonValdezNRT1989Report.pdf. Apparently, Exxon’s only on-shore response equipment consisted of a van and some sampling gear. 26 Id. This lack of preparedness ensured that effective cleanup would be nearly impossible in the event of a spill. 27Lendon & Martin, supra note 13, at 103. However, even if Alyeska Pipeline Service Company, Exxon, and the Coast Guard had been more prepared, an effective cleanup would have been unlikely. See Eliot Marshall, Valdez: The Predicted Oil Spill, Science, April 1989, at 20.

The Valdez spill occurred several minutes after midnight on March 24, 1989, when the Valdez struck a reef in Prince William Sound, tearing open eleven of the ship’s cargo tanks with gashes that extended along its full length. 28 Nat’l Response Team Report, supra note 25, at 3. Shortly afterwards, Coast Guard investigators discovered that the ship’s captain, a known alcoholic, had taken command that night after consuming five double vodkas. 29 In re Exxon Valdez, 270 F.3d 1215, 1236 (9th Cir. 2001). Exxon officials were aware of the captain’s alcohol problems. Robert M. Schoch, Case Studies in Environmental Science 44 (1996). Within several weeks of the spill, oil had spread to cover one thousand square miles of pristine ocean. 30 Nat’l Response Team Report, supra note 25, at 13.

Prince William Sound is a highly sensitive marine environment, 31 Id. at 26–27, 31. and it is especially vulnerable to the long-term effects of an oil spill because cold water temperatures in the Sound result in slower-than-usual weathering and biodegradation of oil. 32 Id. at 25. The spilled oil killed marine birds, mammals, 33 Id. at 27. and fish. 34 Id. at 28. In addition, the spill affected the livelihoods of roughly one-third of Alaska’s twelve thousand commercial fishermen, 35 Id. at 31. and fishery closings caused by the spill affected an estimated three to four thousand workers in the area’s fish processing industry. 36 Id. The spill resulted in serious financial losses and psychological stresses for the inhabitants of the area, as well as severe and long-term damage to the environment. 37Brief of Sociologists, Psychologists & Law & Economics Scholars as Amici Curiae Supporting Respondents passim, Exxon Shipping Co. v. Baker, 128 S. Ct. 2605 (2008) (No. 07-219), available at http://www.abanet.org/publiced/preview/briefs/pdfs/07-08/07-219_RespondentAmCuSocPsychLawEcoScholars.pdf.

II. Punitive Damages Law: Tradition and Reform

Turning from the facts of the spill, this Part describes the adequacy and reasonableness of punitive damages law and its reforms prior to the Court’s decision in Exxon Shipping. Section A reveals that punitive damages, far from being a modern invention, have long been part of the common law 381 Linda L. Schlueter, Punitive Damages 5 (5th ed. 2005). and have served important functions. Section B shows that state statutes frequently limit punitive damages—proving that legislatures are capable of reforming punitive damages without intervention by courts. Although some of these statutes have harmful effects, 39 See, e.g., Lucinda M. Finley, The Hidden Victims of Tort Reform: Children, Women, and the Elderly, 53 Emory L.J. 1263, 1307 (2004) (“[C]aps on punitive damages that tie them to the amount of economic loss only can have a disparate impact on injured women.”). they nonetheless provide valuable data on the costs and benefits of tort reform. 40Mark C. Weber, Mass Jury Trials in Mass Tort Cases: Some Preliminary Issues, 48 DePaul L. Rev. 463, 473 (1998). Finally, Section C argues that the punitive damages cases where the Supreme Court relied on the Due Process Clause place reasonable constitutional limits on punitive damages. Taken together, common law principles, statutory limits, and constitutional interpretations show that Exxon Shipping’s one-to-one rule is overly restrictive in any context—but should certainly not be construed as applying beyond maritime law.

A. Common Law Punitive Damages

The first English case to provide an explicit articulation of punitive damages was Wilkes v. Wood, decided in 1763. 41 Schlueter, supra note 38, at 5. In Wilkes, the plaintiff argued that “trifling damages would put no stop at all” to the defendant’s conduct, and the court agreed. 42 Wilkes v. Wood, (1763) 98 Eng. Rep. 489, 490 (KB).Wilkes thus recognized an important economic justification for punitive damages: They are sometimes necessary to deter the defendant from engaging in similar conduct in the future. 43 Accord Mathias v. Accor Econ. Lodging, Inc., 347 F.3d 672, 677 (7th Cir. 2003) (“[Compensatory damages] would be too slight to give the victim an incentive to sue . . . . [T]o limit the plaintiff to compensatory damages would enable the defendant to commit the offensive act with impunity provided he was willing to pay . . . .”). A later English case, Merest v. Harvey, exemplifies a second economic justification for punitive damages. 44 Schlueter, supra note 38, at 9. In Merest, the court remarked that it was appropriate to award punitive damages in a case involving a defendant who had provocatively knocked off the plaintiff’s hat, 45 Id. at 10. reasoning that awards in such cases served to “prevent the practice of dueling.” 46 Merest v. Harvey, (1814) 128 Eng. Rep. 761, 761 (Heath, J.).

Linda Schlueter, author of the treatise Punitive Damages, shows that by the mid-eighteenth century, established legal doctrine held that punitive damages could be used to punish the defendant in cases of “malice, oppression, or gross fraud.” 47 Schlueter, supra note 38, at 6. Unlike today, 48David Fink, Note, Best v. Taylor Machine Works, The Remittitur Doctrine, and the Implications for Tort Reform, 94 Nw. U. L. Rev. 227, 228 (1999) (“The remittitur doctrine exists in virtually every jurisdiction nationwide . . . .”). juries were given “unfettered discretion” to decide the amount of damages. 49 Schlueter, supra note 38, at 6. However, in certain cases a grand jury would review the jury’s findings, and if the grand jury overturned these findings, members of the jury would be subject to severe penalties. Id. at 6 n.31. Schlueter notes that punitive damages were also used to compensate for injuries that were not otherwise compensable in English law at the time, 50 Id. at 8. such as “hurt feelings, wounded dignity, or insult.” 51 Id. Across the Atlantic, early American courts also recognized benefits of punitive damages. As Justice Story remarked in Boston Manufacturing Co. v. Fiske, courts commonly used punitive damages to penalize “offending parties,” even in maritime cases. 52 2 Mason 119, 121 (1820). Concurrent with a trend toward allowing compensatory damages for emotional suffering, American courts in the nineteenth century began to limit the purposes of punitive damages to punishment and deterrence. 53 Id. at 16. In 1851, the Supreme Court explained the extra-compensatory role played by punitive damages in Day v. Woodworth: “It is a well established principle of the common law that . . . a jury may inflict . . . exemplary, punitive, or vindictive damages upon a defendant, having in view the enormity of his offence rather than the measure of compensation to the plaintiff.” 54Day v. Woodworth, 54 U.S. (13 How.) 363, 371 (1851). Thus, the Supreme Court recognized the common law view of punitive damages more than 150 years ago. This basic view remained settled until the era of tort reform in the last decades of the twentieth century. 55John T. Nockleby & Shannon Curreri, 100 Years of Conflict: The Past and Future of Tort Retrenchment, 38 Loy. L.A. L. Rev. 1021, 1021 (2005) (“In the 1970s insurance companies, tobacco interests, and large industry launched a political campaign . . . . Unlike previous reform efforts that sought to change rules of law through case-by-case adjudication in the courts, the self-styled tort ‘reform’ movement pursued a much grander vision: transforming the cultural understanding of civil litigation . . . by attacking the system itself . . . . [A]dvocates seek to persuade the public through advertising and lobbying that the civil justice system is corrupted . . . .” (footnote omitted)). In sum, the history reveals that punitive damages were approved by the highest courts and served vital functions such as discouraging illegal retaliation for provocative behavior and deterring defendants from repeating tortious conduct.

B. Recent Trends in State Regulation of Punitive Damages Law

In response to a perceived explosion in punitive damages liability in the late twentieth century, many states enacted statutes limiting punitive damages in various ways. 56John A. Albers, Note, State of Confusion: Substantive and Procedural Due Process with Regard to Punitive Damages After TXO Production Corp. v. Alliance Res. Corp., 26 U. Tol. L. Rev. 159, 175 (1994) (“[S]ince 1986 many states enacted statutes that either placed caps upon the recovery of punitive damages or stiffened the burden of proof applicable to such awards.”). Such approaches include (1) limiting punitive damages to the amount of compensatory damages (or to some multiple thereof), (2) capping punitive damages at specific dollar amounts, (3) prohibiting more than one award of punitive damages based on the same conduct, (4) requiring that a fixed percentage of all punitive damages awards go to a state victims’ fund, and (5) requiring that punitive damages be determined in a proceeding separate from one used to determine compensatory damages. 57 See BMW of N. Am., Inc. v. Gore, 517 U.S. 559, 615–19 (1996) (Ginsburg, J. & Rehnquist, C.J., dissenting) (providing comprehensive survey of state legislation affecting punitive damages). As Justice Ginsburg noted in her dissenting opinion in BMW of North America, Inc., these statutes demonstrate that legislatures are capable of limiting punitive damages without help from courts. 58 Id. at 613–14 (“[T]he reexamination prominent in state courts and in legislative arenas . . . serves to underscore why the Court’s enterprise is undue.” (footnote omitted)). In addition, legislative reforms of punitive damages are more flexible than common law reforms of the type provided in Exxon Shipping because legislation can be repealed or modified without running afoul of stare decisis. 59 Percival E. Jackson, Dissent in the Supreme Court 521 (1969) (“Ordinary legislation, not disguised as constitutional interpretation, is flexible and subject to ready change in response to public opinion without the tug-of-war which rules of stare decisis generate. Nor need it overcome the obduracy of men with life tenure who, like most men, are not given to confess error.”).

C. Constitutional Limits on Punitive Damages Awards

In 1996, the Supreme Court struck down a state court award of punitive damages in the landmark case of BMW of North America v. Gore. 60517 U.S. 559. In 2003, the Court struck down another state court award in State Farm Mutual Automobile Insurance Co. v. Campbell, this time clarifying which awards of punitive damages violate due process rights. 61538 U.S. 408 (2003). In this dramatic episode in the history of punitive damages, the Court recognized a constitutional right under the Due Process Clause of the Fourteenth Amendment to be free from excessive and arbitrary punitive awards. 62 BMW, 517 U.S. at 568; State Farm, 538 U.S. at 416. This limit applies irrespective of the process provided (as with other rights recognized as substantive due process 63Coniston Corp. v. Vill. of Hoffman Estates, 844 F.2d 461, 465 (7th Cir. 1988) (“The other objection to the due process route in a case such as the present one is that it depends on the idea of ‘substantive’ due process. This is the idea that depriving a person of life, liberty, or property can violate the due process clause of the Fifth and Fourteenth Amendments even if there are no procedural irregularities . . . .”). rights). 64 See BMW, 517 U.S. at 582 (limiting punitive damages despite right to a jury trial and appeals); State Farm, 538 U.S. at 416 (limiting punitive damages). These pivotal holdings are summarized in the following paragraphs.

Although the Supreme Court’s due process cases recognize the value of punitive damages awards as both a deterrent and a form of retributive justice, 65 State Farm, 538 U.S. at 416. the Court imposed “procedural and substantive constitutional limitations” on such awards. 66 Id. According to the Court, the Due Process Clause of the Fourteenth Amendment “prohibits the imposition of grossly excessive or arbitrary punishments on a tortfeasor.” 67 Id. Additionally, the Court held that compensatory damages should be presumed to make the plaintiff whole, so punitive damages should only be awarded in cases where the defendant’s conduct is sufficiently reprehensible to warrant further penalties. 68 Id. at 419. Against this presumption, the Court held that, of all the factors to be considered, the reprehensibility of the defendant’s conduct was the most relevant in evaluating the reasonableness of an award of punitive damages. 69 BMW, 517 U.S. at 575.

The Court provided numerical guidelines to indicate when the amount of punitive damages in a particular case may be constitutionally excessive. Few awards exceeding a single-digit ratio of punitive to compensatory damages will satisfy the requirements of constitutional due process. 70 State Farm, 538 U.S. at 425. The single-digit ratio may be exceeded in cases where the harm is difficult to detect (a view the economic literature amply supports) 71 E.g., A. Mitchell Polinsky & Steven Shavell, Punitive Damages: An Economic Analysis, 111 Harv. L. Rev. 869, 874 (1998). or in which an “egregious act has resulted in only a small amount of economic damages.” 72 State Farm, 538 U.S. at 425. Although nine-to-one is the presumptive ceiling of the ratio of punitive to compensatory damages, the Court signaled that awards should not cluster at this upper bound. 73 Id. According to the Court, in many cases a four-to-one ratio “might be close to the line of constitutional impropriety,” 74 Id. (citing Pac. Mut. Life Ins. Co. v. Haslip, 499 U.S. 1, 23–24 (1991)). but when compensatory damages are “substantial,” 75 Id. The Court has never defined the amount of compensatory damages that should be considered “substantial,” however. a one-to-one ratio may be the maximum acceptable ratio. 76 Id. On the other hand, the Court emphasized that these guidelines are not inflexible or absolute, 77 Id. (“Our jurisprudence and the principles it has now established demonstrate, however, that, in practice, few awards exceeding a single-digit ratio between punitive and compensatory damages, to a significant degree, will satisfy due process.”). and several appellate courts have deployed creative arguments to support large punitive awards. 78 See, e.g., Action Marine, Inc. v. Cont’l Carbon Inc., 481 F.3d 1302 (11th Cir. 2007) (upholding a punitive damages award); Mathias v. Accor Econ. Lodging, Inc., 347 F.3d 672 (7th Cir. 2003) (upholding a punitive damages award in excess of nine-to-one ratio).

Traditional punitive damages and their reform before the decision in Exxon Shipping represent a useful and still-evolving body of law under which punitive damages have been substantially limited. The patchwork of state statutory limits and the Supreme Court’s overarching constitutional framework provide multiple levels of control over potentially erratic awards. If extended to non-maritime cases, Exxon Shipping’s one-to-one cap on punitive damages for reckless conduct would override many state statutes and represent a major shift away from the flexibility of the Supreme Court’s constitutional jurisprudence governing punitive damages.

III. Against One-to-One Caps

This Part analyzes punitive damages according to both deterrence and retributivist rationales. Section A analyzes the economic and deterrence-based reasons for awarding punitive damages. This argument is limited to the context of unintentional torts because focusing on negligence, recklessness, and strict liability shows why the Court’s decision in Exxon Shipping is flawed from an economic perspective. Section B presents non-economic and retributivist arguments for punitive damages. Finally, Section C shows how economic and retributivist rationales frequently overlap and suggests solutions for those cases in which the rationales seem to point in different directions.

A. Economic & Deterrence-Based Arguments

Although foreseeable and preventable, 79 See Marshall, supra note 27, at 20. the Exxon Valdez disaster was an accident and not an intentional tort. Although scholars have written extensively about the justifications for punitive damages in the context of intentional torts, 80 See, e.g., David Friedman, Law’s Order: What Economics Has to Do with Law and Why It Matters 206–11 (2001) (discussing economic rationales for applying punitive damages in various intentional tort scenarios); Richard A. Posner, Economic Analysis of Law 206–08 (7th ed. 2007) (discussing various rationales for punitive damages). this analysis will focus exclusively on the reasons for applying punitive damages in unintentional tort cases. The first subsection introduces four basic economic concepts: (1) the burden of taking precautions, (2) the probability of harm, (3) the gravity of harm, and (4) the concept of efficient precautions. The second subsection applies these concepts by investigating the incentives faced by a hypothetical firm, “ChemShip,” that transports chemicals across the country with a fleet of eighteen-wheeled trucks. This example, explored in detail, makes it possible to examine what happens when certain variables change, such as the amount of care taken by the company and the probability of paying damages, while other variables remain constant. The ChemShip hypothetical demonstrates the need for punitive damages in various recurring situations. For instance, under a strict liability regime, punitive damages are warranted when ChemShip will sometimes be able to escape paying damages even though its conduct is tortious. 81 See Polinsky & Shavell, supra note 71, at 874. The third subsection uses the ChemShip scenario to demonstrate that in a negligence regime the possibility of escaping detection does not necessarily warrant an award of punitive damages. Instead, courts should make the decision to award punitive damages on a case-by-case basis. 82This is a departure from Judge Posner’s theory, which holds that punitive damages would be inefficient in both negligence and strict liability cases. Posner, supra note 80, at 206. Additionally my argument that the analysis of punitive damages differs when the legal standard is negligence sets it apart from Polinsky and Shavell’s study. See Polinsky & Shavell, supra note 71, at 886 (“Because damages should equal harm under the strict liability rule, and because we assume that damages should equal harm under the negligence rule for the reasons given, we generally will not distinguish between the rules in our subsequent discussion.”). In both negligence and strict liability regimes, a tortfeasor’s attempt to conceal harms or otherwise reduce the probability of paying damages justifies an award of punitive damages. The fourth subsection demonstrates that when a tortfeasor’s conduct is reckless, an award of punitive damages may be warranted regardless of the probability of paying damages. 83This conclusion is in partial agreement with existing literature but is based on different reasoning. See Posner, supra note 80, at 207. Finally, the fifth subsection critiques the argument that punitive damages should not be awarded when the defendant is a corporation.

1. The Hand Formula & Efficient Precautions

As the common law has long recognized, punitive damages help to punish wrongdoers, deter future harms, and avoid other social ills. 84Browning-Ferris Indus. of Vt., Inc. v. Kelco Disposal, Inc., 492 U.S. 257, 275 (1989) (“[P]unitive damages advance the interests of punishment and deterrence . . . .”); see Mathias v. Accor Econ. Lodging, Inc., 347 F.3d 672, 676–77 (7th Cir. 2003) (“[O]ne function of punitive-damages awards is to relieve the pressures on an overloaded system of criminal justice by providing a civil alternative to criminal prosecution of minor crimes. . . . [An] award of punitive damages . . . serves the additional purpose of limiting the defendant’s ability to profit from its fraud by escaping detection . . . .”). Economic theory reveals that punitive damages are particularly important if a tortfeasor has a chance of escaping judgment or behaves recklessly. 85 See Posner, supra note 80, at 206–07. The law and economics movement shows that harmful accidents can be avoided through changes in the degree of caution with which activities are performed and through changes in the overall volume of risky activities. 86Shepherd, supra note 6, at 911. Liability regimes, such as tort law, force potential tortfeasors to take account of the accident costs they impose on others. 87Gideon Parchomovsky & Alex Stein, Torts and Innovation, 107 Mich. L. Rev. 285, 314 (2008) (“Tort law is commonly thought of as a mechanism of assigning liability to wrongdoers and thereby forcing them to internalize the costs they impose on others. This, indeed, is its primary effect.”). Ideally, those engaged in risky activities will take efficient precautions by increasing care levels (and sometimes by decreasing activity levels) as long as the anticipated benefit from these additional precautions is not outweighed by their cost. 88Russell B. Korobkin & Thomas S. Ulen, Law and Behavioral Science: Removing the Rationality Assumption from Law and Economics, 88 Cal. L. Rev. 1051, 1096 (2000) (“Traditional law and economics suggests that precaution is efficient when its benefits outweigh its costs . . . .”).

Judge Learned Hand’s negligence formula sums up this economic wisdom with three simple variables: B for the burden of taking precautions, P for the probability of harm, and L for the gravity of harm. 89 See United States v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947) (“[I]n algebraic terms: if the probability be called P; the injury, L; and the burden, B; liability depends upon whether B is less than L multiplied by P: i.e., whether B [is less than] PL.”). The formula has proven extremely useful in the economic analysis of law, and courts occasionally apply it. 90 See Posner, supra note 80, at 168 n.1. It is important to note how the variables of the Hand Formula interact with each other. Increasing B results in a decrease in P because an accident is less likely to occur when precautions are taken. 91Michelle J. White, The Economics of Accidents, 86 Mich. L. Rev. 1217, 1219 (1988) (reviewing Steven Shavell, Economic Analysis of Accident Law (1987)) (“Economists assume that accidents occur less often and involve less damage when potential injurers and victims use higher levels of care.”). In addition, the Hand Formula shows how the prospect of liability should encourage rational actors to increase care levels to the optimal amount. 92 Cf. Jennifer H. Arlen, Compensation Systems and Efficient Deterrence, 52 Md. L. Rev. 1093, 1096–97 (1993) (“Standard economic analysis has shown that in the unilateral risk context, strict liability rules can be used to induce both efficient caretaking and efficient activity levels, because strict liability can be employed to force injurers to bear the full social cost of any risks they create.”). When potential tortfeasors take optimal precautions, they minimize costs to society—calculated as the sum of the risk (P x L)—and the amount of resources devoted to precautions (B), or (P x L) + B. 93 Id. at 1096 (describing the “level at which the total social cost of accidents is minimized—that is, the level that minimizes the cost of reducing (or eliminating) the risk in question, plus the expected cost to the members of society of the resulting injuries”).

2. Punitive Damages & Strict Liability

This Comment applies the concepts of burden, probability, and gravity of harm in a simple example that will demonstrate why a tortfeasor’s liability must sometimes exceed the amount of harm that the tortfeasor causes—in other words, when a punitive award must be added to compensatory damages. For this example, assume that a company, “ChemShip,” owns and operates a fleet of eighteen-wheeled trucks that transport chemicals across the United States. Assume that if ChemShip spends no money on precautionary measures such as thicker tanks, safety valves, and driver rest requirements, then the probability that a hazardous chemical spill will occur over the course of a year is 0.05. If a spill occurs, the resulting harm will be $1 million. Also assume that safety measures are available and not prohibitively expensive. According to a model (which utilizes an equation to relate the money that ChemShip might spend on precautions to the probability of a spill), 94The model used here is not intended to reflect the real-world probability of accidents in the chemical transportation industry, nor does it realistically depict the amount of money spent on precautionary measures. The model does, however, accurately illustrate the relationship between the amount spent on safety precautions and the probability of a harmful accident occurring. an investment of $10,000 will decrease the yearly probability of a spill to 0.0153, an investment of $20,000 will reduce the probability to 0.0093, and an investment of $30,000 will reduce the probability to 0.0069. Note that the incremental reduction in the probability of harm decreases with each additional $10,000 spent. 95The first $10,000 spent will decrease the probability by 0.0357. The second $10,000 spent will decrease the probability by 0.0060. And the third $10,000 spent will decrease the probability by 0.0025. This occurs because reducing risk with additional safety measures becomes progressively more expensive. 96 Posner, supra note 80, at 168 (“[I]nputs of care are scarce and therefore their price rises as more and more are bought.”). Consequently, the amount of care per dollar decreases as more “care” is purchased. Additionally, the equation presupposes that as long as ChemShip is engaged in shipping activities, there will always be at least a 0.001 probability of a spill in any given year, no matter how much is spent on precautions. 97This relationship is captured in a function that takes as its argument the dollar amount spent on precautions and whose output is the probability of harm. The equation used for this example is f(x) = 200 / (x + 4000) + 0.001. It was designed to have a y-intercept of ~ 0.05, a slope that is initially negative but that approaches 0, and a value for y that approaches 0.001 as x approaches infinity. Although the numbers presented in this example are hypothetical, they are representative of the situations that the tort doctrines relating to accidents seek to control—situations involving risky activities where precautions can be taken to reduce risk, but at an increasing marginal cost. 98 Posner, supra note 80, at 168. The following graph illustrates this scenario:

Graph & Table 1: Total Costs (Private and Social) When Paying Damages Is Certain 99The single line in Graph 1 represents both private and social costs, which are equal in this case.

kerr-fig1

Cost of Precautions (B)

Prob. of Harm

(P)

Potential Liability/Harm (L)

Expected Liability/Harm

(P x L)

Total Cost

B + (P x L)

$0.00

0.0510

$1,000,000.00

$51,000.00

$51,000.00

*$10,000.00

0.0153

$1,000,000.00

$15,285.71

*$25,285.71

$20,000.00

0.0093

$1,000,000.00

$9,333.33

$29,333.33

$30,000.00

0.0069

$1,000,000.00

$6,882.35

$36,882.35

*Represents the socially optimal investment in precautions

Assuming that L represents the actual harm caused by an accident and that L will always be paid in the event of a spill, a liability scheme based on charging the shipper $1 million each time an accident occurs will efficiently manage risks. This scheme aligns private and social costs. 100Paul G. Mahoney, Precaution Costs and the Law of Fraud in Impersonal Markets, 78 Va. L. Rev. 623, 628 (1992) (“The optimal deterrence framework holds that liability rules minimize social cost by forcing defendants to choose levels of activity and care that reflect social costs and benefits rather than the defendants’ own private costs and benefits.”). Imposing liability gives the shipper an incentive to invest roughly $10,000 in precautions because doing so minimizes the shipper’s private costs. This level is also socially optimal because it minimizes the sum of the precautionary costs expended by the shipper (which are themselves a cost to society) and the social costs generated by the risk.

However, the well-functioning liability scheme described above goes awry if the shipper is sometimes able to escape paying damages when accidents occur. 101Polinsky & Shavell, supra note 71, at 887–89. Perhaps spills are difficult to detect, or perhaps the harms from spills only manifest long after an accident occurs. 102 See, e.g., Kirk Johnson, Ex-Grace Officials on Trial in Asbestos Poisoning, N.Y. Times, Feb. 18, 2009, at A15 (“Charlie Welch, 55, who worked as a guard at the mine for a time and says he too suffers from asbestosis . . . remembers the trains carrying vermiculite in open rail cars, billowing dust plumes through town as they rumbled out to the wider world.”). The example of shipping asbestos-laden material in open rail cars proves that tortious conduct may be both difficult to detect and slow to manifest. Id. Assume that the probability that the shipper will have to pay $1 million for a given accident is 0.25. The following table shows how this causes private and social costs to diverge:

Graph & Table 2: Private and Social Costs When the Probability of Paying Damages Is 0.25.

kerr-fig2

kerr-tbl2

† Indicates the level of precautionary spending that minimizes the shipper’s costs

* Represents the socially optimal investment in precautions

Here, ChemShip will choose to invest about $3,000 in precautions because this minimizes the shipper’s private costs, but social costs would be minimized, as before, if ChemShip were to spend roughly $10,000. Thus, the reduced probability of detection changes the shipper’s cost schedule, causing a substantial decrease in social welfare. When, as in the previous example, the probability of paying damages is 1.0, social costs are $25,202.64, but with the decreased probability of detection they are $32,242.07. Moreover, the distributional effect is especially problematic since the shipper is much better off in this scenario, with private costs of $10,310.52 rather than the $25,202.64 he would have to pay under the previous scenario. Perhaps the most troubling implication of this comparison is that ChemShip will have a high incentive to conceal accidents or engage in other conduct to escape liability. When the probability of detection drops from 1.0 to 0.25, ChemShip’s private costs drops by nearly 60%. Accordingly, tortfeasors may attempt to capture the benefits of reduced liability by attempting to conceal accidents or by pressuring victims not to sue. This analysis suggests that because of the negative consequences associated with reductions in the probability of paying damages, tortfeasors who attempt to conceal accidents should be charged additional penalties. 103 See Paul H. Rubin, John E. Calfee & Mark F. Grady, BMW v Gore: Mitigating The Punitive Economics of Punitive Damages, 5 Sup. Ct. Econ. Rev. 179, 189 (1997).

3. Punitive Damages & Negligent Torts

The Hand Formula shows that if an increase in B would result in a decrease in P such that the benefit (in terms of a reduction in expected accident costs) would exceed the cost of the precaution, then the injurer is negligent. 104 Posner, supra note 80, at 168 (“Hand wrote that a potential injurer is negligent if but only if B < PL . . .”). The “marginal Hand Formula” described by Judge Posner is similar to the formula used in these examples. Id. at 168 n.2 (explaining how the “marginal Hand Formula” is derived). Setting aside other factors (such as proximate cause and pure economic loss), the negligent actor will pay L in compensation to the accident victim when B is less than P x L. 105 See United States. v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947). The formulation “If B < PL then the actor is negligent” is in fact an oversimplification. See infra note for a discussion of a more accurate method for assessing negligence based on the “marginal Hand Formula.” In contrast, if the injurer is not negligent, then the injurer is not liable—no matter what harm results.

Since the negligence doctrine completely removes liability when an injurer takes “due care,” 106 W. Page Keeton et al., Prosser and Keeton on Torts § 30, at 164, § 31, at 169 (W. Page Keeton ed., 5th ed. 1984). the liability-diminishing effects of a reduction in the probability of paying damages may not be large enough to induce a potential tortfeasor to spend less than the optimal amount on precautions. In the present example, where liability depends on a showing of negligence, even if the probability of detection is as low as 0.25, the shipper will continue to spend the socially optimal amount of $10,000, and punitive damages will be unnecessary. 107The model presented in this Part reveals that this may often be the case. The following graph reveals that the shipper still attains minimal private costs (very narrowly) at the socially efficient level of precautions:

Graph & Table 3: Private and Social Costs in Negligence When the Probability of Paying Damages Is 0.25

kerr-fig3

kerr-tbl3

* Represents the socially optimal investment in precautions

The “carrot” of zero liability in a negligence regime provides a sufficient incentive for ChemShip to exercise due care. In this case, even the cost-distorting effects of reductions in the probability of detection do not necessarily lead to inefficient outcomes. True, when the probability of detection is 0.25, only one in four of those injured by negligent tortfeasors will be compensated, but punitive damages will not solve that problem. Instead, policies should be directed toward increasing the probability of detection so that all parties injured by negligence may be compensated. This may be accomplished by providing additional punishment for tortfeasors who take measures to decrease the probability of detection of their actions, in effect reducing potential tortfeasors’ incentive to conceal harms. 108Fed. Deposit Ins. Corp. v. W.R. Grace & Co., 877 F.2d 614, 623 (7th Cir. 1989) (“The most straightforward rationale for punitive damages, as for fines and other criminal punishments that exceed the actual injury done by (or profit obtained by) the tortfeasor or criminal, is that they are necessary to deter torts or crimes that are concealable.”). Such tortfeasors would obviously include those who attempt to conceal harms, but should also include tortfeasors who conduct their operations in such a way that tracing harms is difficult.

However, reducing the probability of paying damages from 0.25 to 0.1 dramatically alters the shipper’s incentives and demonstrates that a negligence regime will not always cause the shipper to take due care. When the probability of detection drops to 0.1, the model predicts that the shipper will take few or no precautions.

Graph & Table 4: Private and Social Costs in a Negligence Regime When the Probability of Paying Damages Is 0.1

kerr-fig4

kerr-tbl4

† Indicates the level of precautionary spending that minimizes the shipper’s costs

* Represents the socially optimal investment in precautions

The result is a serious misalignment between the level of care that is optimal for the shipper and the level optimal for society. Here, punitive damages must be assessed to restore the alignment between social and private costs. The conclusion, therefore, is not that punitive damages are always or never appropriate in negligence regimes, but that they must be assessed on a case-by-case basis to determine whether circumstances or misconduct have made it more profitable for the tortfeasor to spend less than the socially optimal amount on precautions. 109This determination could require some rather complex calculations, but this Comment maintains that only a case-by-case method for assessing punitive damages in negligence cases will achieve efficient results. Fortunately, when a divergence of optimums 110For example, a divergence of optimums occurs when social and private costs vary in such a way that the potential injurer is better off taking a socially inefficient level of precaution. occurs—whether in strict liability or in negligence—the alignment of costs can be restored by multiplying the damages payment by the reciprocal of the probability of detection, thus charging the tortfeasor “punitive” damages each time the tortfeasor is caught. 111Polinsky & Shavell, supra note 71, at 874. Interestingly, in some cases a punitive award along these lines will be justified even if the tortfeasor’s conduct implies no culpability. In such situations, there is a genuine disconnect between the rationales of retribution and deterrence. See infra Part III.C. Multiplying the damages in Table 2 by four would undo the effects of the 0.25 probability of not paying damages, essentially recreating the scheme in Table 1. If punitive damages were calculated in this manner, the shipper would once again spend the optimal amount on precautions. Even after this correction, a distributional problem remains because the shipper will pay four times as much to a quarter as many plaintiffs, while three-quarters of the victims receive nothing. But interestingly, the distributional inequalities among the injured may lead to an increase in the probability of detection: publicity from lawsuits with high damages, combined with communication among plaintiffs and others who have been injured, will perhaps give the remaining three quarters of potential plaintiffs sufficient information and incentive to sue. 112 Cf. Barbara Pressley Noble, At Work; The Legacy of Jack McGann, N.Y. Times, Nov. 15, 1992, § 3, at 27 (reporting that a single case filed against an employer for discriminatory termination of insurance benefits helped turn the issue into a cause célèbre). When it becomes apparent that the probability of detection has increased, punitive damages should be lowered to avoid over-deterrence and unnecessarily large penalties.

4. Punitive Damages for Reckless Conduct

Economically speaking, conduct is “reckless” when a tortfeasor refuses to take basic precautions despite very high expected accident costs. 113 Friedman, supra note 80, at 208 (“Part of what makes us describe a tort as reckless is the failure of the tortfeasor to take even the simplest and most obvious precautions.”); Posner, supra note 80, at 207. Non-economic formulations of recklessness are less precise. See Keeton et al., supra note 106, § 34, at 214 (“[T]here is often no clear distinction at all between such [willful, wanton, or reckless] conduct and ‘gross’ negligence, and the two have tended to merge and take on the same meaning, of an aggravated form of negligence, differing in quality rather than in degree from ordinary lack of care.”). In terms of the familiar variables, B is relatively low while P x L is high. 114 Posner, supra note 80, at 207–09 (“B is positive but extremely low, while P and L are both extremely high.”). Judge Posner’s assertion that both P and L are extremely high does not make much sense. Even in his example, it seems unlikely that P would be “extremely” high. It seems more reasonable to judge recklessness based on the percentage difference between B and P x L. Suppose that in the ChemShip scenario described above, the probability of an accident occurring could be reduced tenfold with an expenditure of only $5,000, but ChemShip does not make this investment. Here, the presupposition of rationality collapses because it seems clear that the shipper is not minimizing its private costs. Perhaps company managers consider safety precautions to be contrary to a “macho” business ethic, or perhaps management is risk-loving. 115Craig S. Lerner & Moin A. Yahya, “Left Behind” After Sarbanes-Oxley, 44 Am. Crim. L. Rev. 1383, 1386 (2007) (contrasting “bean counters” with “swashbucklers—that is, people who are risk-neutral with those who are possibly risk-loving with respect to business matters and legal compliance”). Similarly, profits may be so high that company decision makers do not consider $1 million in damages to be worth their attention. 116 Cf. Avishalom Tor, The Fable of Entry: Bounded Rationality, Market Discipline, and Legal Policy, 101 Mich. L. Rev. 482, 514 (2002) (“Profitability and growth are frequently the focus of entrants’ attention, the very factors that lead many of them to embark on new ventures. Although entrants will not analyze these factors perfectly, given their bounded rationality, entrants will nevertheless focus much attention on them. Entrants are likely to ignore other background variables . . . especially if these variables do not affect entrants’ ability to embark on a new venture but ‘only’ the venture’s prospects. The analysis of such variables, if done at all, would therefore be more likely to fall prey to the processes of overconfidence, leading entrants to exhibit a relative insensitivity to their presence.”). If, for example, the company is earning $1 billion in yearly profits, then as long as managers can be sure they will have to pay only $1 million in the case of an accident, they may not spend time worrying about such contingencies. Even with no precautions, expected damages would only be $50,000, and assuming punitive damages are not awarded, this is indeed a “trifling” sum 117 See supra note 42 and accompanying text. compared to $1 billion in profits. Management’s attention is not an unlimited resource, 118 James G. March, A Primer on Decision Making: How Decisions Happen 10 (1994) (“Time and capabilities for attention are limited. Not everything can be attended to at once. Too many signals are received. Too many things are relevant to a decision. Because of these limitations, theories of decision making are often better described as theories of attention or search than as theories of choice. They are concerned with the way in which scarce attention is allocated.”). and if investments of attention in matters other than controlling damages seem likely to pay a higher reward, safety may not be prioritized. Thus, a scarcity of managerial attention could explain a company’s failure to take obviously efficient precautions.

Since reckless conduct is by definition inefficient, concerns with over-deterrence do not come into play. 119 Friedman, supra note 80, at 208 (“Part of what makes us describe a tort as reckless is the failure of the tortfeasor to take even the simplest and most obvious precautions. That suggests that his behavior was clearly inefficient, so we need not worry about over deterring it.”). Reckless conduct is also easy to avoid, since B is low. 120 Posner, supra note 80, at 207 (“Take the case of recklessness. I decide to rest my eyes while driving, and plow at high speed into a flock of pedestrians. B is positive but extremely low, while P and L are both extremely high.”). The cognitive errors described above, including the irrational attitude towards safety, but especially an extremely profitable firm’s failure to consider inefficiencies that are small in comparison to profits, can be overcome by assessing punitive damages. 121 See Moin A. Yahya, Deterring Roper’s Juveniles: Using a Law and Economics Approach to Show that the Logic of Roper Implies that Juveniles Require the Death Penalty More than Adults, 111 Penn St. L. Rev. 53, 69 (2006) (“Now, suppose that the individual is risk-loving. . . . [T]here exists a penalty that will deter a risk-loving individual, but the penalty required to deter [is] about 70% higher than what [is] needed to deter the risk-averse individual.”). The amount of punitive damages should be based on contextual factors such as whether the tortfeasor demonstrated risk-loving behavior or failed to take precautions because profits were so high that harms seemed minimal in comparison. Punitive damages should be calibrated to send a signal to the tortfeasor and to other similarly situated actors that will overcome their cognitive biases by showing that reckless disregard for safety measures can result in serious financial consequences.

5. Punitive Damages Against Corporations

Some scholars have suggested that punitive damages should not be assessed against corporations 122 E.g., W. Kip Viscusi, The Social Costs of Punitive Damages Against Corporations in Environmental and Safety Torts, 87 Geo. L.J. 285, 335 (1998) (arguing that punitive damages fail to incentivize risk-reducing behavior in corporations). because (1) punishment of corporate entities fails to punish the responsible parties, 123Polinsky & Shavell, supra note 71, at 949. (2) the desire to punish a “legal fiction” such as a corporation is incoherent, 124 Id. (“We find this conception of the punishment goal unappealing . . . because it necessitates believing that people would, after reflecting on the matter, want to impose a penalty on what ultimately is an artificial legal construct. The notion that individuals would want to punish firms per se strikes us as not entirely different from the idea that individuals would want to punish inanimate objects for causing harm . . . .”). and (3) punitive damages do not deter corporations. 125Viscusi, supra note 123, at 288. This subsection will explain why each of these claims are unpersuasive.

The first argument states that punitive damages assessed against corporations fail to serve a retributive function since the wrong individuals—stockholders and customers—bear the brunt of the punishment rather than the truly responsible parties—managers and directors. 126Polinsky & Shavell, supra note 71, at 949. There are several objections to this argument. First, customers may not be hurt at all because firms (especially large corporations able to charge monopoly or quasi-monopoly prices) may be both willing and able to reduce prices after litigation and attendant bad publicity in order to lure back customers. 127 See Bryan Mercurio, Resolving the Public Health Crisis in the Developing World: Problems and Barriers of Access to Essential Medicines, 5 Nw. U. J. Int’l Hum. Rts. 1, 27 (2006) (“[P]harmaceutical companies routinely sell pharmaceuticals at heavily reduced prices to developing countries in order to promote goodwill and, in all probability, to counter the negative publicity the industry has received in recent years.”). Second, stockholders are not completely blameless because they are able to choose whose stock to buy, and they have power over the management of the corporation. 128In fact, some have noted that shareholder voting power is on the rise. See, e.g., Lisa M. Fairfax, Shareholder Democracy on Trial: International Perspective on the Effectiveness of Increased Shareholder Power, 3 Va. L. & Bus. Rev. 1, 5 (2008) (“This examination reveals not only that shareholders have become more active within recent years, but also that their activism has had an impact on corporate affairs.”). Accordingly, stockholders share a portion of the responsibility, and a drop in share prices is the deserved consequence of investing in a company with a poor safety record. 129Moreover, allowing punitive damages to be imposed against corporations that act reprehensibly could encourage ethical investing. Cf. Einer Elhauge, Sacrificing Corporate Profits in the Public Interest, 80 N.Y.U. L. Rev. 733, 784 (2005) (“An increasing number of investors now put their money in funds committed to avoid investments in corporations that create environmental harms . . . .”). Finally, large punitive damages assessments against corporations will in fact result in the punishment of responsible parties whether the responsible parties are managers or low-level employees. If a manager was responsible for the harm that resulted in punitive damages liability, a rational board of directors would fire the manager, 130 Cf. Stephanie E. Tanger, Enforcing Corporate Responsibility for Violations of Workplace Immigration Laws: The Case of Meatpacking, 9 Harv. Latino L. Rev. 59, 79 (2006) (“In response to the lawsuit, Tyson fired the managers who pled guilty and claimed to redouble its efforts . . . .”). and, similarly, a low-level employee would likely be fired 131Often, low-level employees will be judgment-proof. This in turn provides part of the economic justification for the rule of respondeat superior. See Posner, supra note 80, at 188 (“[M]ost employees lack the resources to pay a judgment . . . .”). and possibly imprisoned if the acts were criminal. 132For instance, Exxon fired Captain Hazelwood soon after the Valdez oil spill, and he also faced criminal prosecution. Exxon Valdez Trial to Be Held in Anchorage, N.Y. Times, Oct. 29, 1989, § 1, at 27.

The second argument against awarding punitive damages for corporate misconduct states that the desire to punish a corporation is incoherent since corporations are legal constructs and cannot be “responsible” in the same manner as natural persons. 133Polinsky & Shavell, supra note 71, at 949. Both legal history and common experience belie this argument. As Justice Oliver Wendell Holmes, Jr., argued in The Common Law, long before modern concepts of liability emerged, early legal systems sometimes “punished” inanimate objects. 134 Oliver Wendell Holmes, Jr., The Common Law 8–38 (Dover 1991) (1881). As Justice Holmes explained:We have now followed the development of the chief forms of liability in modern law for anything other than the immediate and manifest consequences of a man’s own acts. . . . We have seen a single germ multiplying and branching into products as different from each other as the flower from the root. It hardly remains to ask what that germ was. We have seen that it was the desire of retaliation against the offending thing itself. . . . A consideration of the earliest instances will show, as might have been expected, that vengeance, not compensation, and vengeance on the offending thing, was the original object. The ox in Exodus was to be stoned. The axe in the Athenian law was to be banished. The tree, in Mr. Tylor’s instance, was to be chopped to pieces.Id. at 34. However, regardless whether ancient law saw fit to punish inanimate objects, a corporation is, in fact, much closer to a human being than to an axe or a stone. 135Business law also gives corporations many of the legal rights enjoyed by natural persons. See Santa Clara County v. S. Pac. R.R. Co., 118 U.S. 394, 396 (1886) (“The court does not wish to hear argument on the question whether the provision in the Fourteenth Amendment to the Constitution, which forbids a State to deny to any person within its jurisdiction the equal protection of the laws, applies to these corporations. We are all of opinion that it does.” (quoting Chief Justice Waite)) (syllabus). This adds strength to the argument that corporate “persons”—and not only natural persons—should face the possibility of economic punishment through punitive damages. Thus, a fortiori, punishing a corporate entity—with its ability to buy and sell, to produce goods, to own property, to “remember” through institutional memory, and to “think” through its managers—is far from incoherent. Punishment alters corporate behavior, just as it does the conduct of a natural person. 136 See Bert Swart, International Trends Towards Establishing Some Form of Punishment for Corporations, 6 J. Int’l Crim. Just. 947, 951–52 (2008) (“[C]orporations have their own institutional memories. They can remember things that happened in the past and learn from their experiences. They can correct their policies if they have made mistakes, or if circumstances make it desirable for them to change them. So far as their relationship with the law is concerned, they can make free choices about whether or not to comply with the law. Their freedom of choice in this respect makes it both possible and justifiable to hold them accountable for their choices. Experience also shows that sanctions imposed on corporations for having violated the law often induce them to adjust their goals or policies in order to prevent repetition.”).

The third argument against assessing punitive damages for corporate misconduct states that punitive damages do not induce corporations to reduce risks. 137Viscusi, supra note 123, at 288. W. Kip Viscusi has used this argument to buttress the claim that punitive damages awards should be eliminated against corporations in cases involving risk and environmental decisions. 138 Id. at 335. Using cost–benefit analysis to assess the value of punitive damages, 139 Id. at 286–87. Viscusi contends that (1) “punitive damages have no significant deterrent effect,” 140 Id. at 288. (2) “eliminating risk becomes inordinately costly,” 141 Id. at 307. (3) “compensatory damages are generally adequate for deterrence,” 142 Id. at 310. (4) “market forces promote safety,” 143 Id. at 315. and (5) “punitive damages cause economic harm.” 144 Id. at 322. On closer examination, Viscusi’s arguments are self-contradictory. The claim that “punitive damages have no significant deterrent effect” is incompatible with the claim that “[t]he high stakes and high variability of punitive damage awards are of substantial concern to companies.” 145 See id. at 285. Viscusi, however, argues that punitive damages are assessed more or less at random, making it impossible for a corporation to respond to the threat of punitive damages liability by spending more on precautions. Id. at 309. When profit-maximizing companies are substantially concerned about something that could affect their bottom line, they generally take steps to prevent it from happening. 146 See, e.g., David A. Super, Privatization, Policy Paralysis, and the Poor, 96 Cal. L. Rev. 393, 408 (2008) (describing the “laser focus on the bottom line” that characterizes private business). This is how deterrence works to minimize risks in the business context. 147Eric A. Posner & Cass R. Sunstein, Dollars and Death, 72 U. Chi. L. Rev. 537, 554 (2005) (“To understand how deterrence works, one must take an ex ante perspective. . . . As is well known, people can be given optimal incentives to take care if they are required to pay damages for any financial losses that they cause . . . .” (footnotes omitted)). Besides, since reckless or intentional conduct is generally a prerequisite for an award of punitive damages, 148 See Keeton et al., supra note 106, § 2, at 9–10 (“Something more than the mere commission of a tort is always required for punitive damages. There must be circumstances of aggravation or outrage . . . or such a conscious and deliberate disregard of the interests of others that the conduct may be called willful or wanton.” (footnotes omitted)); id. § 35, at 212–13 (explaining that conduct labeled as “willful,” “wanton,” or “reckless” remains, “at essence, negligent, rather than actually intended to do harm,” yet “is held to justify an award of punitive damages” (footnote omitted)). it is hard to see how, if punitive damages impose a “catastrophic threat,” 149Viscusi, supra note 123, at 285. punitive damages would not deter reckless—and therefore easily avoidable—conduct. Furthermore, the claim that punitive damages will not prompt corporations to be more cautious is especially doubtful in light of evidence suggesting that legal liability is perhaps the greatest factor inducing corporations to design safer products. 150 See Sandra F. Gavin, Stealth Tort Reform, 42 Val. U. L. Rev. 431, 438 (2008) (“In 1983, the Rand Institute studied the ‘serious public policy problem, namely the manufacture of products that may have been unreasonably dangerous to their users[,]’ to determine what external pressures had the greatest influence on promoting products safety and concluded, ‘[o]f all the various external social pressures, product liability has the greatest influence on product design decisions.’” (alteration in original) (quoting George Eads & Peter Reuter, Designing Safer Products, Corporate Responses to Product Liability Law and Regulation, at iii–viii (1983))).

A final consideration regarding punitive damages and corporate conduct relates to corporate attitudes toward risk. Some scholars maintain that corporations are “risk-neutral,” 151Polinsky & Shavell, supra note 71, at 887 n.44. while most individuals are “risk-averse.” Accordingly, damages need not be as high for individuals to ensure the appropriate level of deterrence. 152 Id. at 886–87. Thus, it may be economically justifiable to set damages at a higher level when the defendant is a corporation.

B. Retributivist Arguments

The literature on retributivist justifications for punitive damages is extensive and provides a refreshing (and perhaps necessary) contrast to the economic literature; 153 See, e.g., Galanter & Luban, supra note 10; Dan Markel, Retributive Damages: A Theory of Punitive Damages as Intermediate Sanction, 94 Cornell L. Rev. 239 (2009); Catherine M. Sharkey, Punitive Damages as Societal Damages, 113 Yale L.J. 347 (2003); Benjamin C. Zipursky, A Theory of Punitive Damages, 84 Tex. L. Rev. 105 (2005). indeed, it seems unlikely that economic theories will ever adequately explain punitive damages if such damages are to retain a genuinely punitive character. 154George P. Fletcher, Paradoxes in Legal Thought, 85 Colum. L. Rev. 1263, 1264 (1985) (“The criteria of crime, criminal responsibility, and punishment have yet to receive an adequate account in the literature of law and economics.”). Even Judge Posner admits that wealth-maximizing theories of law and economics are limited as a tool for the explanation and generation of legal rules. 155 Posner, supra note 80, at 216 (“But the fact that any sort of rape license is even thinkable within the framework of the wealth-maximization theory that guides so much of the analysis in this book is a limitation on the usefulness of that theory. What generates the possibility of a rape license is that fact that the rapist’s utility is weighted the same as his victim’s utility. If it were given a zero weight in the calculus of costs and benefits, a rape license could not be efficient. The only persuasive basis for such a weighting, however, would be a moral principle different from efficiency.” (emphasis added)). For other examples of moral atrocities that a theory based purely on economic efficiency would recommend, see id. at 11, 12, 27. Thus, retributivist modes of understanding punitive damages are essential to a complete understanding of the role of punitive damages in tort law.

1. Standing Up to Giants

In Poetic Justice: Punitive Damages and Legal Pluralism, Marc Galanter and David Luban provide an eloquent defense of punitive damages without relying on economic concepts. 156 Galanter & Luban, supra note 10. In their view, punitive damages are “perhaps the most important instrument in the legal repertoire for pronouncing moral disapproval of economically formidable offenders.” 157 Id. at 1428. All other legal sanctions, they argue, fail to carry moral force adequate to address the implications of the underlying conduct. 158 Id. Although Galanter and Luban agree with economically minded reformers who maintain that punitive damages should not be “utterly discretionary and without limits,” their reasoning differs. 159 Id. at 1461. In their view, punitive damages should be limited “not because completely discretionary punitive damages are economically harmful, as tort reformers typically claim, but rather because retribution demands penalties that bear a significant relation to the nature of the wrongdoing.” 160 Id. In a rebuff to the economic logic of punitive damages, Galanter and Luban claim that “[c]ivil punishments . . . reinforce the notion of law as a realm of moral achievement rather than technical adjustment.” 161 Id. at 1428 (emphasis added).

Galanter and Luban are directly on point when they argue that retributive concerns, which address the “heinousness of the offense,” are fundamentally undermined when punitive awards are keyed to the amount of compensatory damages 162 Id. at 1432 (“A retributivist scales punishment to the heinousness of the offense, and that is not measured by the magnitude of harm.”). because heinous acts may not cause significant compensatory damages. 163 Id. For instance, “cold-bloodedly throwing a child out of a skyscraper window may result in very little harm because the child’s suspenders miraculously catch on a flagpole.” 164 Id. Because of these concerns, Galanter and Luban maintain that retribution and a one-to-one cap are incompatible, and they reject proposals to cap punitive damages at a multiple of compensatory damages. 165 Id. at 1461 (“In our view, the limits of punitive damages have to do entirely with the heinousness of the wrongful act; they have nothing to do with the size of compensatory awards. Thus, we oppose proposals to cap punitive damages at some small multiple of compensatory damages.”). The Court’s endorsement of the retributivist rationale in Exxon Shipping thus rings hollow when compared with Galanter and Luban’s thoughtful analysis of the moral implications of punitive damages.

2. The Benefits of “Wild” Awards

According to Jeffrey White, punitive damages serve a useful social function in the law of torts because they allow the community to express its moral condemnation of the defendant’s misconduct. 166Jeffrey R. White, State Farm and Punitive Damages: Call the Jury Back, 5 J. High Tech. L. 79, 79 (2005) (“Nowhere does the civil jury speak louder than when it awards punitive damages against a defendant who has violated our common understanding of acceptable behavior. The jury verdict speaks as the conscience of the community.”). What is more, judges help to limit the potential economic harm of excessive jury awards. According to David Partlett, “Judges, as repeat players, may exercise the corrective function by outlying verdicts in accord with the equity notion that like cases should be treated alike.” 167David F. Partlett, The Republican Model and Punitive Damages, 41 San Diego L. Rev. 1409, 1427 (2004). Moreover, “Judges . . . in motions for remitter and on appeal, may act as an appropriate corrective.” 168 Id.

Similar to White and Galanter and Luban, Partlett also remarks that a jury’s assertion of “community outrage about the flouting of a right is not a symptom of the illness of the system[,]” because “[e]ven the outrageous punitive award establishes a healthy dialogue where the wronged citizen is accorded respect and the wrongdoer suffers punishment for his or her misdeeds.” 169 Id. In conclusion, scholars and judges who ignore the retributivist rationale of punitive damages ignore a well-established line of legal tradition that is closely tied to common notions of fairness. 170 See Cass R. Sunstein, On the Psychology of Punishment, 11 Sup. Ct. Econ. Rev. 171, 171 (2004) (“Jurors are intuitive retributivists . . . .”).

C. Synthesis

The foregoing analysis indicates that punitive damages must occasionally exceed compensatory damages by a considerable margin. Exxon Shipping’s one-to-one cap unduly fetters punitive damages and will interfere with deterrence and retributivist goals. Accordingly, a one-to-one cap is inappropriate if deterrence and retribution are taken as goals of tort law. Although tensions exist between deterrence and retribution, 171Drumbl, supra note 11, at 577 (“[T]he goals of deterrence . . . may well conflict with those of retribution . . . .”). in most cases the two rationales overlap—especially in cases of reckless conduct.

1. Overlapping Rationales

Since punitive damages are generally restricted to situations in which the tortfeasor has behaved recklessly or intentionally, 172 See Keeton et al., supra note 106, § 2, at 9 (“Something more than the mere commission of a tort is always required for punitive damages.”). the divide between retributivist and economic rationales for punitive damages can easily be bridged. As argued above, one of the least problematic arguments for assessing punitive damages applies when the defendant’s conduct is reckless. 173 See supra Part III.A.3. Similarly, reckless conduct is blameworthy from a retributivist perspective because it suggests that the defendant places a low value on the safety of others. 174 See, e.g., Reeves v. Carlson, 969 P.2d 252, 256 (Kan. 1998) (“To be reckless, conduct must be such as to show disregard of or indifference to consequences, under circumstances involving danger to life or safety of others.”). Thus, proponents of both the retributivist rationale and the economic rationale agree that punitive damages should be awarded for reckless conduct. Economic arguments would additionally support an assessment of punitive damages for less culpable defendants who are frequently able to escape liability. 175 See supra Part III. In such cases, where less culpable conduct nevertheless supports an award of punitive damages, judges could instruct juries that a punitive damages award will cause the defendant, and potential defendants, to exercise due care and that the award eventually may help other victims to receive compensation.

2. Quantitative Concerns

In Punitive Damages: An Economic Analysis, Polinsky and Shavell argue that when deterrence and retributivist concerns support differing amounts of punitive damages, the best solution is to find a compromise. 176Polinsky & Shavell, supra note 71, at 955 (“It is evident that the best level of punitive damages should be a compromise between the levels that are optimal when each objective is considered independently.”). However, genuine difficulties exist in calculating the amount of punitive damages when retribution is the purpose. As Galanter and Luban suggest, “The magnitude of punishment must reflect the magnitude and, if possible, the nature of the asserted inequality between wrongdoer and victim.” 177Galanter & Luban, supra note 10, at 1432. Clarifying this point, they add that “[a] more heinous act expresses more contempt for the victim’s value relative to the wrongdoer’s, and so the retributivist believes that a more decisive defeat must be visited on the wrongdoer to reassert the public’s judgment of the victim’s worth.” 178 Id. at 1433. Unfortunately, Galanter and Luban provide little guidance for arriving at the specific dollar amount necessary to effect a “decisive defeat.” However, given that outrageous yet unintentional conduct is likely to be at least reckless, even an excessive award by the jury does not raise substantial economic concerns—the concept of over-deterrence does not apply to reckless conduct because the optimal level of reckless conduct is zero. Thus, the jury’s reasoned choice of an amount of damages, though unguided by numerical rules, will not cause economic harm in cases of recklessness.

IV. Exxon Shipping Co. v. Baker: A Suspect Analysis

This Part returns to the Court’s decision in Exxon Shipping. The first section summarizes the relevant portions of the majority opinion. The second section critiques the majority opinion and suggests how the Court went wrong in crafting its overly restrictive one-to-one rule.

A. The Majority Opinion

At trial, the Alaska federal district court certified a mandatory class for the more than 32,000 plaintiffs seeking punitive damages against Exxon. 179Exxon Shipping Co. v. Baker, 128 S. Ct. 2605, 2613 (2008). The jury awarded $287 million in compensatory damages to the class of commercial fishermen 180 Id. at 2614. and $5 billion in punitive damages. 181 Id. The jury awarded $5,000 in punitive damages against the Valdez’s captain. Id. On appeal, the Ninth Circuit remanded the case twice because the punitive damages award did not meet the due process standards set by the Supreme Court in its constitutional punitive damages cases. 182 Id. Ultimately, the circuit court remitted the punitive award to $2.5 billion. 183 Id.

The Supreme Court granted certiorari to consider several issues, including whether the $2.5 billion award was excessive as a matter of maritime common law, 184 Id. at 2611. a branch of law “which falls within a federal court’s jurisdiction to decide in the manner of a common law court, subject to the authority of Congress to legislate otherwise if it disagrees with the judicial result.” 185 Id. at 2619. Although the Court noted that the question of punitive damages in the context of maritime law was an issue of first impression, 186 Id. the majority opinion frequently cited constitutional punitive damages cases, not maritime cases or principles. 187 Id., passim.

In a lengthy survey of punitive damages law, 188 Id. at 2620–25. the Court emphasized that the “real problem” with punitive damage awards is their “stark unpredictability.” 189 Id. at 2625. In its review of the Ninth Circuit’s $2.5 billion award, the Court specified that its analysis was based on maritime law, and it was therefore unnecessary to apply the constitutional doctrines outlined in State Farm. 190 Id. at 2626. The Court did not apply the Gore guideposts—reprehensibility, disparity, comparison to civil penalties—as required by the due process punitive damages cases. See BMW of N. Am., Inc. v. Gore, 517 U.S. 559, 575 (1996) (setting forth the guideposts). The Court then reasoned that “the unpredictability of high punitive awards . . . is in tension with the function of the awards as punitive . . . .” 191 Exxon Shipping, 128 S. Ct. at 2627. According to the majority, an “eccentrically high punitive verdict” carries “an implication of unfairness . . . in a system whose commonly held notion of law rests on a sense of fairness in dealing with one another.” 192 Id.

The Court cited Justice Holmes’s essay, The Path of the Law, 193O.W. Holmes, The Path of the Law, 10 Harv. L. Rev. 457, 459 (1897). Justice Holmes was then serving on the Supreme Judicial Court of Massachusetts. Id. at 457. in support of the proposition that “a penalty should be reasonably predictable in its severity, so that even Justice Holmes’s ‘bad man’ can look ahead with some ability to know what the stakes are in choosing one course of action or another.” 194 Exxon Shipping, 128 S. Ct. at 2627. This argument is not universally accepted. According to Jeffrey R. White:Business practices and corporate decisions require cost planning and consideration of potential liability exposure. Unpredictable punitive damage awards are an unwelcome wild card.This argument received a cool reception in state courts in the 1980s. Many took the view expressed by the Maine Supreme Court [in Tuttle v. Raymond, 494 A.2d 1353, 1359 (Me. 1985),] that “the lack of any precise formula by which punitive damages can be calculated is one of the important assets of the doctrine.” As the New Jersey Supreme Court explained [in Fischer v. Johns-Manville Corp., 512 A.2d 466, 477 (N.J. 1986)]:Anticipation of these damages will allow potential defendants, aware of dangers of a product, to factor those anticipated damages into a cost-benefit analysis and to decide whether to market a particular product. The risk and amount of such damages can, and in some cases will, be reflected in the cost of a product, in which event the product will be marketed in its dangerous condition.Thus, [according to Palmer v. A.H. Robins Co., 684 P.2d 187, 218 (Colo. 1984)], ‘[i]f punitive damages are predictably certain, they become just another item in the cost of doing business.’ That result defeats the purpose of punitive damages in deterring misconduct and turns them into a user fee that permits defendants to continue their misconduct for a price.White, supra note 167, at 88 (footnotes omitted). In searching for a rule of law that would reduce or remove entirely the unpredictability that made punitive damages awards problematic, the Court rejected the verbal formulae adopted by various state courts 195 Exxon Shipping, 128 S. Ct. at 2628–29. and remarked that “eliminating unpredictable outlying punitive awards by more rigorous standards than the constitutional limit” would likely be best achieved in a way analogous to the “criminal-law pattern of quantified limits.” 196 Id. at 2629. The Court eventually ruled that “a [one-to-one] ratio, which is above the median award, is a fair upper limit in such maritime cases.” 197 Id. at 2629–33.

B. Critiques

This section argues that the Court’s reasoning in Exxon Shipping suffers from serious flaws. First, the majority reasoned from a single anecdote (and a lack of confirming studies) that punitive damages are unpredictable and thus in need of reform. Second, the majority misread and misused statistics to justify capping punitive damages at a one-to-one ratio. Finally, the majority largely ignored deterrence and retributivist rationales for punitive damages.

1. Drawing an Inference from an Anecdote

One of the more puzzling aspects of the Court’s opinion is its crafting of a new rule to address a situation that even the Court seemed to identify as unproblematic. The Court noted that although punitive damages have “been the target of audible criticism in recent decades . . . . [a] survey of the literature reveals that discretion to award . . . punitive damages has not mass-produced runaway awards . . . .” 198 Id. at 2624. Moreover, “The figures thus show an overall restraint . . . .” 199 Id. at 2624–25. One study cited by the Court 200 Id. at 2625. observed that “[p]unitive damages are infrequent, typically for small sums, and concentrated primarily in contract-related cases.” 201Brian J. Ostrom, et al., A Step Above Anecdote: A Profile of the Civil Jury in the 1990s, 79 Judicature 233, 238 (1996). According to the Court, however, these studies obscured the more subtle problem of the unpredictability of punitive damages. 202 Exxon Shipping, 128 S. Ct. at 2625.

The Court stated that variability in the size of punitive awards, and in the ratio between punitive and compensatory damages, could only be acceptable if it resulted from judges and juries adopting a measured approach that tended to produce both consistent and optimal levels of damages in cases with similar facts. 203 Id. However, the facts in two cases will never be exactly the same. See George Priest, Introduction to Cass R. Sunstein et al., Punitive Damages: How Juries Decide 1, 2 (2002) (“The magnitude of punitive damages verdicts appears to vary substantially across juries. But this judgment, too, is problematic. In some sense, no two cases are alike. Thus, there is an inherent difficulty in evaluating one verdict against another . . . .”). However, the Court noted that “anecdotal evidence suggests that nothing of that sort is going on.” 204 Exxon Shipping, 128 S. Ct. at 2625–26. This assertion was based on the examination of a single anecdote from prior case law: An Alabama jury awarded a plaintiff $4 million in punitive damages, while in a different Alabama case with “strikingly similar facts,” no punitive damages were awarded. 205 Id. at 2626. The case involving the $4 million award is BMW of North America Inc. v. Gore, 517 U.S. 559 (1996), the first case in which the Supreme Court struck down a state court award of punitive damages. The Supreme Court, 1995 Term—Leading Cases, 110 Harv. L. Rev. 135, 145, 150 (1996). Not only is it a weak form of argumentation to rely on anecdotal evidence—much less a single anecdote—but the Court avoided mentioning that the Alabama Supreme Court remitted the damages to $2 million, and then, after the U.S. Supreme Court struck down the award, the Alabama Supreme Court again remitted the award, this time to just $50,000. Thus, the Court exaggerated the true disparity between awards in the two cases. In addition, selecting a case decided before the Court clarified its punitive damages jurisprudence in State Farm Mutual Automobile Insurance Co v. Campbell, 538 U.S. 408 (2003), is misleading. In Exxon Shipping, the Court assessed the supposed inadequacy of the law in 2008, and BMW likely would have been decided differently if it had been heard after State Farm. See W. Kip Viscusi, The Blockbuster Punitive Damages Awards, 53 Emory L.J. 1405, 1420–26 (2004) (discussing the effect of State Farm on large punitive damages awards). Remarkably, on the basis of this single anecdote and the Court’s further observation that it was not aware of any “scholarly work pointing to consistency across punitive awards in cases involving similar claims and circumstances,” 206 Exxon Shipping, 128 S. Ct. at 2626. the Court concluded that the unpredictability of punitive damages was an established fact. 207 Id. at 2625–34.

However, the Court’s conclusion that punitive damages are unpredictable simply does not follow from these facts. Even if the ratio of punitive to compensatory damages varied wildly, this would not prove that punitive damages are unpredictable. Some other factor (or set of factors) could be responsible for the variation, and the level of punitive damages could perhaps be predicted quite accurately if these other factors were properly understood. The Court’s reasoning could be compared to an argument stating that because the height of law students is not correlated with academic performance, then student performance in law school is unpredictable.

Moreover, the Court’s emphasis on predictability is misplaced, since punitive damages could be predictable but based on the wrong factors. 208 See Jonathan M. Karpoff & John R. Lott, Jr., On the Determinants and Importance of Punitive Damage Awards, 42 J.L. & Econ. 527, 529 (1999) (“An additional concern is not just whether punitive damages are predictable, but whether they are predictable for the right reasons.”). For example, if a study were to find that defendants’ wealth is the key determinant of punitive damages, then—assuming considerations of a defendant’s wealth are unprincipled or against public policy—there would be a serious reason for reforming the system even though damages were highly predictable. Nevertheless, the Court failed to show that punitive damages actually are unpredictable. A single anecdote is not enough, and since the Court’s one-to-one rule is supposedly justified by the need to eliminate the unpredictability of punitive damages awards, the rule itself is called into question.

2. The Court’s Misuse of Statistical Evidence

Commentators have already criticized the Court’s use of statistics in Exxon Shipping. 209The Supreme Court has frequently relied on statistical studies. See Adam Liptak, From One Footnote, a Debate over the Tangles of Law, Science and Money, N.Y. Times, Nov. 25, 2008, at A16 (“The Supreme Court has often considered academic studies in its decisions, starting with Louis D. Brandeis’s famous 1908 brief collecting medical and other evidence to support laws limiting work hours. Lawyers still call such submissions ‘Brandeis briefs.’ The court’s signal triumph, Brown v. Board of Education in 1954, cited studies from psychologists and others, and citations to empirical work are commonplace these days.”). However, the Court’s use of the studies it cited in Exxon Shipping has already drawn heavy criticism:“The opinion reads like a bad joke,” said Jeffrey J. Rachlinski, a law professor at Cornell. “They say they know of no study showing punitive damages are orderly in any way, and yet they cite” a study by Theodore Eisenberg, a prominent empirical legal studies scholar at Cornell, “showing punitive damages are pretty orderly.”Professor Eisenberg struggled to stay respectful about the [C]ourt’s approach to his work . . . . He finally settled on this phrase: “I believe the [C]ourt went seriously astray” in concluding that his work supported a reduced award.Id. The following statement shows how the Court misused statistics: “[B]y most accounts the median ratio of punitive to compensatory awards has remained less than [one-to-one] . . . . The figures . . . suggest that in many instances a high ratio of punitive to compensatory damages is substantially greater than necessary to punish or deter.” 210 Exxon Shipping, 128 S. Ct. at 2624–25 (emphasis added). This statement exemplifies the use of a measure of central tendency (the median ratio of compensatory to punitive damages) to indicate what is desirable from an instrumentalist perspective. To see the fallacy of this reasoning, consider that punitive damages serve the functions of punishment and deterrence and that for many of the reasons described in this Comment (such as a low probability of paying damages, recklessness, or morally outrageous disregard for others’ safety), a higher ratio is often required.

The Court embraces this fallacy in holding that punitive damages in cases of recklessness should be capped at one-to-one. The Court’s argument is simple: Since most of the time judges and juries award punitive damages such that the ratio of punitive to compensatory damages is less than one-to-one, then a one-to-one ratio is a “fair upper limit.” 211 Id. at 2633. As stated before, this is an illogical argument. The Court’s justification for capping the ratio at one-to-one is formally equivalent to an argument that since the average criminal goes to jail for fifty months, 212 Sentencing-Guideline Study Finds Continuing Disparities, N.Y. Times, Nov. 27, 2004, at A11 (describing study by U.S. Sentencing Commission finding that the average federal prison inmate’s sentence is fifty months). then no criminal shall serve a longer sentence. In conclusion, the Court’s assertion that punitive damages are unpredictable is unwarranted based on the data the Court examined. Moreover, the Court compounds the error by creating a rule in which a ratio based on the median is applied in a sweeping manner to all cases of reckless conduct.

3. The Court’s Failure to Consider Either Deterrence or Retribution

In Exxon Shipping, the Court failed to consider the consequences of its one-to-one cap. As the economic analysis of punitive damages in Part III of this Comment reveals, total damages often must be increased by adding punitive damages when the probability is less than 1.0 that the defendant will pay compensatory damages. Regrettably, the Court’s one-to-one rule will only provide adequate deterrence when the chance of paying damages is fifty percent or greater. 213A one-to-one cap will only allow total damages to be twice the amount of compensatory damages. As explained in Part III, when the probability of the defendant paying damages is only 0.5, then (assuming the reward for negligence does not already provide the defendant an adequate incentive to take efficient precautions) compensatory damages must be multiplied by two. If the odds of paying damages were any less than 0.5, then a one-to-one cap would prevent the tort system from charging the defendant enough to ensure that private and social costs do not continue to diverge. Worse, a one-to-one cap will diminish the ability of punitive damages to minimize reckless conduct. 214The one-to-one cap will be especially harmful when compensatory damages are low or when the defendant’s taste for risk warrants significantly higher damages. See supra Part III. In sum, the Court in Exxon Shipping almost completely ignored the economic and deterrence-based rationales for punitive damages.

Additionally, the Court refused to acknowledge the strong retributive concerns present in the case. The opinion ignores the well-documented devastation of the Alaskan environment and the oil spill’s damaging effects on the lives and livelihoods of those Alaskans and Native Americans living nearby. Given that the Court has endorsed retribution as a justification for punitive damages, 215 Exxon Shipping, 128 S. Ct. at 2621 (“[T]he consensus today is that punitives are aimed not at compensation but principally at retribution and deterring harmful conduct.”). it is puzzling that it would fail to mention facts that show the blameworthiness of the defendant. 216Some felt that the Court hardly acknowledged Exxon’s blameworthiness. See Adam Liptak, Damages Cut Against Exxon in Valdez Case, N.Y Times, June 26, 2008, at A1 (“Jeffrey L. Fisher, a lawyer for the plaintiffs, said there was ‘a great deal of sadness’ among his clients. ‘What is painful,’ Mr. Fisher said, ‘is that there seems to have been some disagreement between the dissenters and the majority on how reprehensible Exxon’s conduct was.’”). Some scholars have argued that the Court was already ignoring the retributivist rationale for punitive damages before Exxon Shipping was decided. See Paul J. Zwier, The Utility of a Nonconsequentialist Rationale for Civil Jury-Awarded Punitive Damages, 54 U. Kan. L. Rev. 403, 404 (2006) (describing “the harm the Court inflicts by minimizing the retributive justification and usurping the discretion of the jury” and arguing that “the Court backhandedly endorsed the law and economics, or deterrence, model of punishment and ignored the broader effects on social norms and values that result from taking the retribution analysis out of the hands of a common law jury”). Moreover, since the district court created a mandatory punitive damages class, it is difficult to see why Exxon’s punishment-worthy conduct should have been so narrowly circumscribed. 217 Exxon Shipping, 128 S. Ct. at 2613. Even if there was a procedural bar to considering certain elements of Exxon’s conduct for purposes of increasing or maintaining the punitive damages award, the company’s reckless acts still warranted more detailed description in the opinion. An examination of facts the Court failed to mention reveals a much more shocking picture of the human and environmental costs of the spill, 218See supra Part I. as well as Exxon’s obstructionist legal strategy, 219 See Denise E. Antolini, Punitive Damages in Rhetoric and Reality: An Integrated Empirical Analysis of Punitive Damages Judgments in Hawaii, 1985–2001, 20 J.L. & Pol. 143, 148 (2004) (“[C]onsumer advocate Ralph Nader decried Exxon’s failure to pay the award more than a decade after the verdict, calling the company’s response to the award ‘a legal war of attrition, while thousands of Alaskans and others suffered.’” (quoting Ralph Nader, Justice Delayed Is Justice Denied, CommonDreams.org, Oct. 9, 2000, http://www.commondreams.org/views/101000-109.htm)). and the deceitfulness of the company toward those who were harmed. 220 See, e.g., Ott, supra note 14, at 23 (reporting that in the effort to get the Trans-Alaska Pipeline System permit, “Exxon’s Ken Fountain told lobbyists, ‘I don’t care if every goddamn fish dies, get that [authorization] permit . . . .’”). Some of Exxon’s methods for cleaning up the spill oil were harmful to the environment and to the health of cleanup workers, and an Exxon official admitted that the actual cleanup was secondary to “its public image aspects.” Id. at 52 (describing how the high-pressure “hot-water wash” used to clean the shoreline “cooked clams, mussels, and other intertidal animals that had survived the spill and destroyed sea plants and animals that supported the entire coastal ecosystem”). On the other hand, the Court notes that Exxon spent billions on voluntary cleanup efforts 221Exxon Shipping, 128 S. Ct. at 2613 (“In the aftermath of the disaster, Exxon spent around $2.1 billion in cleanup efforts.”). (though some commentators have described these efforts as counterproductive, unsafe, and predominately motivated by public relations concerns). 222Some consider Exxon’s cleanup efforts to have been ineffective and mishandled, and claim the cleanup imposed serious health risks on the cleanup workers. Ott, supra note 14, at 52. Some critics think that Exxon’s expenditure of $2 billion on cleanup may have been more harmful to Prince William Sound than doing nothing. Schoch, supra note 29, at 44–45. If so, it is a valid question whether the cleanup should have been considered as a mitigating factor at all. If evidence of the resources spent on cleanup was admissible to mitigate the award, it is hard to see why other post-spill facts relating to Exxon’s blameworthiness for damaging the environment were not admissible to at least maintain the jury’s punitive damages award.

Pre-spill facts are particularly relevant to assessing the reprehensibility of Exxon’s conduct. For example, oil companies, including Exxon, promised Alaskans that tankers operating in Prince William Sound would have double-bottomed hulls, but broke this promise almost as soon as they received their operating permits. 223 Ott, supra note 14, at 186–87. Reinforced hulls could have reduced the volume of the spill by as much as sixty percent. 224Keith E. Sealing, Civil Procedure in Substantive Context: The Exxon-Valdez Cases, 47 St. Louis U. L.J. 63, 68 n.29 (2003). Exxon representatives promised victims of the spill that if “you show that your motel goes out of business, that we can take care of. . . . If you can show that you have a loss as a result of this spill, we will compensate it,” 225 Ott, supra note 14, at 47 (alteration in original). but then fought to have all the claims dismissed and delayed paying the judgment for nearly twenty years. 226 Id. at 113.

Those who vested their hopes in the payment of the jury’s original verdict must have endured torments similar to those of poor Carstone in Dickens’s Bleak House. 227 See Charles Dickens, Bleak House 936 (Nicola Bradbury ed., Penguin Books 1996) (1853) (“Richard [Carstone], more worn and haggard, haunted the court day after day, listlessly sat there the whole day long when he knew there was no remote chance of the suit being mentioned, and became one of the stock sights of the place. I wonder whether any of the gentlemen remembered him as he was when he first went there.”); Ott, supra note 14, at 272 (“Closure for individuals is a different story. Those who lean heavily on the Supreme Court’s decision may close this chapter of their lives soon after the court makes a ruling. Those who wait for the herring to rebound may be waiting years or decades. Those who died—over 6,000—while their cases churned through the judiciary system will never have had the peace of closure.”). In 2007, Exxon (now ExxonMobil) earned $40.6 billion in profits, breaking its own previous record and once again establishing itself as the most profitable corporation in history. 228Jad Mouawad, Exxon Sets Profit Record: $40.6 Billion Last Year, N.Y. Times, Feb. 2, 2008, at C3. In light of this fact, and considering the Supreme Court’s reduction of the company’s liability for punitive damages to roughly one-tenth of the jury’s original $5 billion verdict, it is not difficult to understand what lies behind the popular impression that the Court promotes the interests of corporations over individuals. 229 See Jeffrey Rosen, Supreme Court, Inc., N.Y. Times, Mar. 16, 2008, (Magazine) at MM38 (discussing the pro-business trend in Supreme Court decisions). In conclusion, the Court thoroughly undermined its own statement that retribution is a justification for punitive damages both by its choice of a one-to-one cap and its treatment of the facts in Exxon Shipping.

V. An Outlier Case

The Court in Exxon Shipping emphasized that the one-to-one ratio applies only in maritime cases. 230Exxon Shipping Co. v. Baker, 128 S. Ct. 2605, 2631–32 (2008). For punitive damages litigation outside maritime law, the extent to which Exxon Shipping will prove influential remains an open question. 231As the Court made clear in Exxon Shipping, the case is not binding outside of admiralty. Id. at 2626–27. Several courts have already cited the case as persuasive precedent for limiting punitive damages more than the Supreme Court’s constitutional punitive damages cases require. 232 See, e.g., Leavey v. Unum Provident Corp., 295 F. App’x 255, 259 n.1 (9th Cir. 2008) (“While the Supreme Court’s recent decision in Exxon Shipping Co. v. Baker . . . review[ed] a jury award for conformity with maritime law, rather than the outer limit allowed by due process, . . . the Court’s statements in that case support the district court’s decision to reduce the award here.” (citations omitted) (internal quotation marks omitted)); Hayduk v. City of Johnstown, 580 F. Supp. 2d 429, 484 n.46 (W.D. Pa. 2008) (“Although Exxon is a maritime law case, it is clear that the Supreme Court intends that its holding have a much broader application.”). Since, as this Comment argues, the rule of Exxon Shipping excessively limits punitive damages, it will be important to distinguish it in non-maritime punitive damages cases.

Historically, limitations of liability has been an important principle of maritime law. 233 Thomas J. Schoenbaum, Admiralty & Maritime Law § 15-1 (4th ed. 2009). These limits are justified by the inherent risks of shipping, the need to promote maritime commerce, and the unfairness of holding shipowners liable for circumstances outside their control. 234 Id. According to Thomas Schoenbaum’s treatise, Admiralty and Maritime Law, “the principle of limitation of liability remains vital to those involved in the shipping industry and even to those who seek its services.” 235 Id. The Limitation of Shipowners’ Liability Act, enacted by Congress in 1851, remains the binding legal source of the limited liability principle in U.S. maritime law. 236 Id. In British Transport Commission v. United States, the Supreme Court described the Act as intended to limit shipowners’ liability to the value of their ships and thus to encourage maritime commerce and shipbuilding. 237354 U.S. 129, 133 (1957). Clearly, maritime law limited liability long before Exxon Shipping.

Furthermore, before Exxon Shipping, the Court’s decision in Miles v. Apex Marine 238498 U.S. 19 (1990). had the practical effect of limiting punitive damages in maritime cases. 239John W. DeGravelles, Uncertain Seas for Maritime Punitive Damages, Trial, Jan. 2004, at 50, 51 (“Although Miles did not address punitive damages specifically, it became the launching pad for much of the case law eliminating them from maritime law.”). Thus, maritime law historically imposed special limits on both punitive damages and general liability that do not apply to non-maritime. 240Somewhat tautologically, maritime law principles are based on distinctly maritime concerns. See Schoenbaum, supra note 234, § 3-5 (“Thus the test for admiralty tort jurisdiction requires that an incident (1) occur on navigable waters; (2) bear a substantial relationship to traditional maritime activity; and (3) have a potentially disruptive impact on maritime commerce.”). While Exxon Shipping makes no explicit reference to the liability-limiting principles of maritime law (or to the punitive damages implications of Miles), the history of maritime law reveals that applying the holding of a liability-limiting case like Exxon Shipping is simply inappropriate in the context of non-maritime law. In sum, Exxon Shipping is a maritime case, and its holding should be strictly confined within the limits of maritime jurisdiction.

Conclusion

The Supreme Court laid out its constitutional punitive damages jurisprudence in BMW of North America Inc. v. Gore 241517 U.S. 559 (1996). and State Farm Mutual Automobile Insurance Co. v. Campbell. 242538 U.S. 408 (2003). These cases addressed genuine concerns with the punitive damages system and provided reasonable guidelines for courts to follow when reviewing punitive damages awards. 243 See Posner, supra note 80, at 207. The Court’s most recent decision in Exxon Shipping breaks with these earlier cases in two significant ways: it is not well-reasoned, and the rule it establishes is inflexible and unsupported by any but the most extreme views of punitive damages. Because it is a maritime case and because punitive damages promote fairness and efficiency, Exxon Shipping should neither be cited nor relied upon in non-maritime cases.

Footnotes

1128 S. Ct. 2605, 2633 (2008).

2538 U.S. 408 (2003).

3517 U.S. 559 (1996).

4 Id. at 562 (“The Due Process Clause of the Fourteenth Amendment prohibits a State from imposing a ‘grossly excessive’ punishment on a tortfeasor.” (quoting TXO Prod. Corp. v. Alliance Res. Corp., 509 U.S. 443, 454 (1993))).

5 See Exxon Shipping, 128 S. Ct. at 2619–34.

6 See Joanna M. Shepherd, Tort Reforms’ Winners and Losers: The Competing Effects of Care and Activity Levels, 55 UCLA L. Rev. 905, 910 (2008) (“Deterrence is the function of tort law by which the law creates incentives that induce people to avoid inappropriately dangerous activities.”).

7 Exxon Shipping, 128 S. Ct. at 2633 (“[T]he unpredictable outlier cases . . . call the fairness of the system into question.”).

8Alex Sienkiewicz, Towards a Legal Land Ethic: Punitive Damages, Natural Value, and the Ecological Commons, 15 Penn St. Envtl. L. Rev. 91, 99 (2006).

9 See Racich v. Celotex Corp., 887 F.2d 393, 396 (2d Cir. 1989).

10 Exxon Shipping, 128 S. Ct. at 2621 (“Regardless of the alternative rationales over the years, the consensus today is that punitives are aimed . . . at retribution and deterring harmful conduct.”); Marc Galanter & David Luban, Poetic Justice: Punitive Damages and Legal Pluralism, 42 Am. U. L. Rev. 1393, 1428 (1993).

11Mark A. Drumbl, Collective Violence and Individual Punishment: The Criminality of Mass Atrocity, 99 Nw. U. L. Rev. 539, 577 (2005) (“[T]he goals of deterrence . . . may well conflict with those of retribution . . . .”).

12Galanter & Luban, supra note 10, at 1396 (“[I]f punitive damages are pared back too drastically, civil law may be underenforced.”).

13 James W. Dearing & Everett M. Rogers, Agenda-Setting 39 (1996) (noting the “massive media attention given to the Exxon Valdez incident”); Hannah Lendon & Brian Martin, Environmental Disasters, in Brian Martin et al., Justice Ignited 99, 103 (2006) (“Of all spills, the Exxon Valdez is most well-known.”).

14 See Riki Ott, Not One Drop: Betrayal and Courage in the Wake of the Exxon Valdez Oil Spill 19, 66–67 (2008) (“Exxon launched an aggressive public relations campaign to quiet ‘rumors’ of extensive damage from its spill.”).

15 Exxon Shipping, 128 S. Ct. at 2640 (Breyer, J., concurring and dissenting) (“The jury could reasonably have believed that Exxon knowingly allowed a relapsed alcoholic repeatedly to pilot a vessel filled with millions of gallons of oil through waters that provided the livelihood for the many plaintiffs in this case.”); see also id. at 2612 (majority opinion) (detailing Captain Hazelwood’s consumption of alcohol).

16 Charles Perrow, Normal Accidents 374–75 (1999) (describing the contingency plans that oil companies were required to submit before being allowed to operate in Prince William Sound as “fantasy documents”).

17 Ott, supra note 14, at 273.

18 U.S. Gen. Accounting Office, Alaskan North Slope Oil: Limiting Effects of Lifting Export Ban on Oil and Shipping Industries and Consumers 12 (1999), available at http://www.gao.gov/archive/1999/rc99191.pdf.

19 Id.

20Brief of the Alaska Legislative Council et al. as Amici Curiae Supporting Respondents at 10, Exxon Shipping Co. v. Baker, 128 S. Ct. 2605 (2008) (No. 07-219), available at http://www.abanet.org/publiced/preview/briefs/pdfs/07-08/07-219_RespondentAmCuAlaskaLegCouncil.pdf.

21 Ott, supra note 14, at 24.

22Brief of the Alaska Legislative Council et al., supra note 20, at 10; Ott, supra note 14, at 273 (“[In 1972] oilmen and/or [the] Nixon administration repeatedly promised that state-of-the-art construction, tankers, navigational procedures, and oil spill response equipment will make ‘operations in Port Valdez and Prince William Sound the safest in the world.’”).

23 U.S. Gen. Accounting Office, supra note 18, at 12.

24 Ott, supra note 14, at 2; see also U.S. Coast Guard, Oil Spill Response Research & Development Program: A Decade of Achievement 37 (2003).

25 Samuel K. Skinner & William K. Reilly, Nat’l Response Team, The Exxon Valdez Oil Spill: A Report to the President 8 (1989) [hereinafter Nat’l Response Team Report], available at www.uscg.mil/History/.../ExxonValdezNRT1989Report.pdf.

26 Id.

27Lendon & Martin, supra note 13, at 103. However, even if Alyeska Pipeline Service Company, Exxon, and the Coast Guard had been more prepared, an effective cleanup would have been unlikely. See Eliot Marshall, Valdez: The Predicted Oil Spill, Science, April 1989, at 20.

28 Nat’l Response Team Report, supra note 25, at 3.

29 In re Exxon Valdez, 270 F.3d 1215, 1236 (9th Cir. 2001). Exxon officials were aware of the captain’s alcohol problems. Robert M. Schoch, Case Studies in Environmental Science 44 (1996).

30 Nat’l Response Team Report, supra note 25, at 13.

31 Id. at 26–27, 31.

32 Id. at 25.

33 Id. at 27.

34 Id. at 28.

35 Id. at 31.

36 Id.

37Brief of Sociologists, Psychologists & Law & Economics Scholars as Amici Curiae Supporting Respondents passim, Exxon Shipping Co. v. Baker, 128 S. Ct. 2605 (2008) (No. 07-219), available at http://www.abanet.org/publiced/preview/briefs/pdfs/07-08/07-219_RespondentAmCuSocPsychLawEcoScholars.pdf.

381 Linda L. Schlueter, Punitive Damages 5 (5th ed. 2005).

39 See, e.g., Lucinda M. Finley, The Hidden Victims of Tort Reform: Children, Women, and the Elderly, 53 Emory L.J. 1263, 1307 (2004) (“[C]aps on punitive damages that tie them to the amount of economic loss only can have a disparate impact on injured women.”).

40Mark C. Weber, Mass Jury Trials in Mass Tort Cases: Some Preliminary Issues, 48 DePaul L. Rev. 463, 473 (1998).

41 Schlueter, supra note 38, at 5.

42 Wilkes v. Wood, (1763) 98 Eng. Rep. 489, 490 (KB).

43 Accord Mathias v. Accor Econ. Lodging, Inc., 347 F.3d 672, 677 (7th Cir. 2003) (“[Compensatory damages] would be too slight to give the victim an incentive to sue . . . . [T]o limit the plaintiff to compensatory damages would enable the defendant to commit the offensive act with impunity provided he was willing to pay . . . .”).

44 Schlueter, supra note 38, at 9.

45 Id. at 10.

46 Merest v. Harvey, (1814) 128 Eng. Rep. 761, 761 (Heath, J.).

47 Schlueter, supra note 38, at 6.

48David Fink, Note, Best v. Taylor Machine Works, The Remittitur Doctrine, and the Implications for Tort Reform, 94 Nw. U. L. Rev. 227, 228 (1999) (“The remittitur doctrine exists in virtually every jurisdiction nationwide . . . .”).

49 Schlueter, supra note 38, at 6. However, in certain cases a grand jury would review the jury’s findings, and if the grand jury overturned these findings, members of the jury would be subject to severe penalties. Id. at 6 n.31.

50 Id. at 8.

51 Id.

52 2 Mason 119, 121 (1820).

53 Id. at 16.

54Day v. Woodworth, 54 U.S. (13 How.) 363, 371 (1851).

55John T. Nockleby & Shannon Curreri, 100 Years of Conflict: The Past and Future of Tort Retrenchment, 38 Loy. L.A. L. Rev. 1021, 1021 (2005) (“In the 1970s insurance companies, tobacco interests, and large industry launched a political campaign . . . . Unlike previous reform efforts that sought to change rules of law through case-by-case adjudication in the courts, the self-styled tort ‘reform’ movement pursued a much grander vision: transforming the cultural understanding of civil litigation . . . by attacking the system itself . . . . [A]dvocates seek to persuade the public through advertising and lobbying that the civil justice system is corrupted . . . .” (footnote omitted)).

56John A. Albers, Note, State of Confusion: Substantive and Procedural Due Process with Regard to Punitive Damages After TXO Production Corp. v. Alliance Res. Corp., 26 U. Tol. L. Rev. 159, 175 (1994) (“[S]ince 1986 many states enacted statutes that either placed caps upon the recovery of punitive damages or stiffened the burden of proof applicable to such awards.”).

57 See BMW of N. Am., Inc. v. Gore, 517 U.S. 559, 615–19 (1996) (Ginsburg, J. & Rehnquist, C.J., dissenting) (providing comprehensive survey of state legislation affecting punitive damages).

58 Id. at 613–14 (“[T]he reexamination prominent in state courts and in legislative arenas . . . serves to underscore why the Court’s enterprise is undue.” (footnote omitted)).

59 Percival E. Jackson, Dissent in the Supreme Court 521 (1969) (“Ordinary legislation, not disguised as constitutional interpretation, is flexible and subject to ready change in response to public opinion without the tug-of-war which rules of stare decisis generate. Nor need it overcome the obduracy of men with life tenure who, like most men, are not given to confess error.”).

60517 U.S. 559.

61538 U.S. 408 (2003).

62 BMW, 517 U.S. at 568; State Farm, 538 U.S. at 416.

63Coniston Corp. v. Vill. of Hoffman Estates, 844 F.2d 461, 465 (7th Cir. 1988) (“The other objection to the due process route in a case such as the present one is that it depends on the idea of ‘substantive’ due process. This is the idea that depriving a person of life, liberty, or property can violate the due process clause of the Fifth and Fourteenth Amendments even if there are no procedural irregularities . . . .”).

64 See BMW, 517 U.S. at 582 (limiting punitive damages despite right to a jury trial and appeals); State Farm, 538 U.S. at 416 (limiting punitive damages).

65 State Farm, 538 U.S. at 416.

66 Id.

67 Id.

68 Id. at 419.

69 BMW, 517 U.S. at 575.

70 State Farm, 538 U.S. at 425.

71 E.g., A. Mitchell Polinsky & Steven Shavell, Punitive Damages: An Economic Analysis, 111 Harv. L. Rev. 869, 874 (1998).

72 State Farm, 538 U.S. at 425.

73 Id.

74 Id. (citing Pac. Mut. Life Ins. Co. v. Haslip, 499 U.S. 1, 23–24 (1991)).

75 Id. The Court has never defined the amount of compensatory damages that should be considered “substantial,” however.

76 Id.

77 Id. (“Our jurisprudence and the principles it has now established demonstrate, however, that, in practice, few awards exceeding a single-digit ratio between punitive and compensatory damages, to a significant degree, will satisfy due process.”).

78 See, e.g., Action Marine, Inc. v. Cont’l Carbon Inc., 481 F.3d 1302 (11th Cir. 2007) (upholding a punitive damages award); Mathias v. Accor Econ. Lodging, Inc., 347 F.3d 672 (7th Cir. 2003) (upholding a punitive damages award in excess of nine-to-one ratio).

79 See Marshall, supra note 27, at 20.

80 See, e.g., David Friedman, Law’s Order: What Economics Has to Do with Law and Why It Matters 206–11 (2001) (discussing economic rationales for applying punitive damages in various intentional tort scenarios); Richard A. Posner, Economic Analysis of Law 206–08 (7th ed. 2007) (discussing various rationales for punitive damages).

81 See Polinsky & Shavell, supra note 71, at 874.

82This is a departure from Judge Posner’s theory, which holds that punitive damages would be inefficient in both negligence and strict liability cases. Posner, supra note 80, at 206. Additionally my argument that the analysis of punitive damages differs when the legal standard is negligence sets it apart from Polinsky and Shavell’s study. See Polinsky & Shavell, supra note 71, at 886 (“Because damages should equal harm under the strict liability rule, and because we assume that damages should equal harm under the negligence rule for the reasons given, we generally will not distinguish between the rules in our subsequent discussion.”).

83This conclusion is in partial agreement with existing literature but is based on different reasoning. See Posner, supra note 80, at 207.

84Browning-Ferris Indus. of Vt., Inc. v. Kelco Disposal, Inc., 492 U.S. 257, 275 (1989) (“[P]unitive damages advance the interests of punishment and deterrence . . . .”); see Mathias v. Accor Econ. Lodging, Inc., 347 F.3d 672, 676–77 (7th Cir. 2003) (“[O]ne function of punitive-damages awards is to relieve the pressures on an overloaded system of criminal justice by providing a civil alternative to criminal prosecution of minor crimes. . . . [An] award of punitive damages . . . serves the additional purpose of limiting the defendant’s ability to profit from its fraud by escaping detection . . . .”).

85 See Posner, supra note 80, at 206–07.

86Shepherd, supra note 6, at 911.

87Gideon Parchomovsky & Alex Stein, Torts and Innovation, 107 Mich. L. Rev. 285, 314 (2008) (“Tort law is commonly thought of as a mechanism of assigning liability to wrongdoers and thereby forcing them to internalize the costs they impose on others. This, indeed, is its primary effect.”).

88Russell B. Korobkin & Thomas S. Ulen, Law and Behavioral Science: Removing the Rationality Assumption from Law and Economics, 88 Cal. L. Rev. 1051, 1096 (2000) (“Traditional law and economics suggests that precaution is efficient when its benefits outweigh its costs . . . .”).

89 See United States v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947) (“[I]n algebraic terms: if the probability be called P; the injury, L; and the burden, B; liability depends upon whether B is less than L multiplied by P: i.e., whether B [is less than] PL.”).

90 See Posner, supra note 80, at 168 n.1.

91Michelle J. White, The Economics of Accidents, 86 Mich. L. Rev. 1217, 1219 (1988) (reviewing Steven Shavell, Economic Analysis of Accident Law (1987)) (“Economists assume that accidents occur less often and involve less damage when potential injurers and victims use higher levels of care.”).

92 Cf. Jennifer H. Arlen, Compensation Systems and Efficient Deterrence, 52 Md. L. Rev. 1093, 1096–97 (1993) (“Standard economic analysis has shown that in the unilateral risk context, strict liability rules can be used to induce both efficient caretaking and efficient activity levels, because strict liability can be employed to force injurers to bear the full social cost of any risks they create.”).

93 Id. at 1096 (describing the “level at which the total social cost of accidents is minimized—that is, the level that minimizes the cost of reducing (or eliminating) the risk in question, plus the expected cost to the members of society of the resulting injuries”).

94The model used here is not intended to reflect the real-world probability of accidents in the chemical transportation industry, nor does it realistically depict the amount of money spent on precautionary measures. The model does, however, accurately illustrate the relationship between the amount spent on safety precautions and the probability of a harmful accident occurring.

95The first $10,000 spent will decrease the probability by 0.0357. The second $10,000 spent will decrease the probability by 0.0060. And the third $10,000 spent will decrease the probability by 0.0025.

96 Posner, supra note 80, at 168 (“[I]nputs of care are scarce and therefore their price rises as more and more are bought.”). Consequently, the amount of care per dollar decreases as more “care” is purchased.

97This relationship is captured in a function that takes as its argument the dollar amount spent on precautions and whose output is the probability of harm. The equation used for this example is f(x) = 200 / (x + 4000) + 0.001. It was designed to have a y-intercept of ~ 0.05, a slope that is initially negative but that approaches 0, and a value for y that approaches 0.001 as x approaches infinity.

98 Posner, supra note 80, at 168.

99The single line in Graph 1 represents both private and social costs, which are equal in this case.

100Paul G. Mahoney, Precaution Costs and the Law of Fraud in Impersonal Markets, 78 Va. L. Rev. 623, 628 (1992) (“The optimal deterrence framework holds that liability rules minimize social cost by forcing defendants to choose levels of activity and care that reflect social costs and benefits rather than the defendants’ own private costs and benefits.”).

101Polinsky & Shavell, supra note 71, at 887–89.

102 See, e.g., Kirk Johnson, Ex-Grace Officials on Trial in Asbestos Poisoning, N.Y. Times, Feb. 18, 2009, at A15 (“Charlie Welch, 55, who worked as a guard at the mine for a time and says he too suffers from asbestosis . . . remembers the trains carrying vermiculite in open rail cars, billowing dust plumes through town as they rumbled out to the wider world.”). The example of shipping asbestos-laden material in open rail cars proves that tortious conduct may be both difficult to detect and slow to manifest. Id.

103 See Paul H. Rubin, John E. Calfee & Mark F. Grady, BMW v Gore: Mitigating The Punitive Economics of Punitive Damages, 5 Sup. Ct. Econ. Rev. 179, 189 (1997).

104 Posner, supra note 80, at 168 (“Hand wrote that a potential injurer is negligent if but only if B < PL . . .”). The “marginal Hand Formula” described by Judge Posner is similar to the formula used in these examples. Id. at 168 n.2 (explaining how the “marginal Hand Formula” is derived).

105 See United States. v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947). The formulation “If B < PL then the actor is negligent” is in fact an oversimplification. See infra note for a discussion of a more accurate method for assessing negligence based on the “marginal Hand Formula.”

106 W. Page Keeton et al., Prosser and Keeton on Torts § 30, at 164, § 31, at 169 (W. Page Keeton ed., 5th ed. 1984).

107The model presented in this Part reveals that this may often be the case.

108Fed. Deposit Ins. Corp. v. W.R. Grace & Co., 877 F.2d 614, 623 (7th Cir. 1989) (“The most straightforward rationale for punitive damages, as for fines and other criminal punishments that exceed the actual injury done by (or profit obtained by) the tortfeasor or criminal, is that they are necessary to deter torts or crimes that are concealable.”).

109This determination could require some rather complex calculations, but this Comment maintains that only a case-by-case method for assessing punitive damages in negligence cases will achieve efficient results.

110For example, a divergence of optimums occurs when social and private costs vary in such a way that the potential injurer is better off taking a socially inefficient level of precaution.

111Polinsky & Shavell, supra note 71, at 874. Interestingly, in some cases a punitive award along these lines will be justified even if the tortfeasor’s conduct implies no culpability. In such situations, there is a genuine disconnect between the rationales of retribution and deterrence. See infra Part III.C.

112 Cf. Barbara Pressley Noble, At Work; The Legacy of Jack McGann, N.Y. Times, Nov. 15, 1992, § 3, at 27 (reporting that a single case filed against an employer for discriminatory termination of insurance benefits helped turn the issue into a cause célèbre).

113 Friedman, supra note 80, at 208 (“Part of what makes us describe a tort as reckless is the failure of the tortfeasor to take even the simplest and most obvious precautions.”); Posner, supra note 80, at 207. Non-economic formulations of recklessness are less precise. See Keeton et al., supra note 106, § 34, at 214 (“[T]here is often no clear distinction at all between such [willful, wanton, or reckless] conduct and ‘gross’ negligence, and the two have tended to merge and take on the same meaning, of an aggravated form of negligence, differing in quality rather than in degree from ordinary lack of care.”).

114 Posner, supra note 80, at 207–09 (“B is positive but extremely low, while P and L are both extremely high.”). Judge Posner’s assertion that both P and L are extremely high does not make much sense. Even in his example, it seems unlikely that P would be “extremely” high. It seems more reasonable to judge recklessness based on the percentage difference between B and P x L.

115Craig S. Lerner & Moin A. Yahya, “Left Behind” After Sarbanes-Oxley, 44 Am. Crim. L. Rev. 1383, 1386 (2007) (contrasting “bean counters” with “swashbucklers—that is, people who are risk-neutral with those who are possibly risk-loving with respect to business matters and legal compliance”).

116 Cf. Avishalom Tor, The Fable of Entry: Bounded Rationality, Market Discipline, and Legal Policy, 101 Mich. L. Rev. 482, 514 (2002) (“Profitability and growth are frequently the focus of entrants’ attention, the very factors that lead many of them to embark on new ventures. Although entrants will not analyze these factors perfectly, given their bounded rationality, entrants will nevertheless focus much attention on them. Entrants are likely to ignore other background variables . . . especially if these variables do not affect entrants’ ability to embark on a new venture but ‘only’ the venture’s prospects. The analysis of such variables, if done at all, would therefore be more likely to fall prey to the processes of overconfidence, leading entrants to exhibit a relative insensitivity to their presence.”).

117 See supra note 42 and accompanying text.

118 James G. March, A Primer on Decision Making: How Decisions Happen 10 (1994) (“Time and capabilities for attention are limited. Not everything can be attended to at once. Too many signals are received. Too many things are relevant to a decision. Because of these limitations, theories of decision making are often better described as theories of attention or search than as theories of choice. They are concerned with the way in which scarce attention is allocated.”).

119 Friedman, supra note 80, at 208 (“Part of what makes us describe a tort as reckless is the failure of the tortfeasor to take even the simplest and most obvious precautions. That suggests that his behavior was clearly inefficient, so we need not worry about over deterring it.”).

120 Posner, supra note 80, at 207 (“Take the case of recklessness. I decide to rest my eyes while driving, and plow at high speed into a flock of pedestrians. B is positive but extremely low, while P and L are both extremely high.”).

121 See Moin A. Yahya, Deterring Roper’s Juveniles: Using a Law and Economics Approach to Show that the Logic of Roper Implies that Juveniles Require the Death Penalty More than Adults, 111 Penn St. L. Rev. 53, 69 (2006) (“Now, suppose that the individual is risk-loving. . . . [T]here exists a penalty that will deter a risk-loving individual, but the penalty required to deter [is] about 70% higher than what [is] needed to deter the risk-averse individual.”).

122 E.g., W. Kip Viscusi, The Social Costs of Punitive Damages Against Corporations in Environmental and Safety Torts, 87 Geo. L.J. 285, 335 (1998) (arguing that punitive damages fail to incentivize risk-reducing behavior in corporations).

123Polinsky & Shavell, supra note 71, at 949.

124 Id. (“We find this conception of the punishment goal unappealing . . . because it necessitates believing that people would, after reflecting on the matter, want to impose a penalty on what ultimately is an artificial legal construct. The notion that individuals would want to punish firms per se strikes us as not entirely different from the idea that individuals would want to punish inanimate objects for causing harm . . . .”).

125Viscusi, supra note 123, at 288.

126Polinsky & Shavell, supra note 71, at 949.

127 See Bryan Mercurio, Resolving the Public Health Crisis in the Developing World: Problems and Barriers of Access to Essential Medicines, 5 Nw. U. J. Int’l Hum. Rts. 1, 27 (2006) (“[P]harmaceutical companies routinely sell pharmaceuticals at heavily reduced prices to developing countries in order to promote goodwill and, in all probability, to counter the negative publicity the industry has received in recent years.”).

128In fact, some have noted that shareholder voting power is on the rise. See, e.g., Lisa M. Fairfax, Shareholder Democracy on Trial: International Perspective on the Effectiveness of Increased Shareholder Power, 3 Va. L. & Bus. Rev. 1, 5 (2008) (“This examination reveals not only that shareholders have become more active within recent years, but also that their activism has had an impact on corporate affairs.”).

129Moreover, allowing punitive damages to be imposed against corporations that act reprehensibly could encourage ethical investing. Cf. Einer Elhauge, Sacrificing Corporate Profits in the Public Interest, 80 N.Y.U. L. Rev. 733, 784 (2005) (“An increasing number of investors now put their money in funds committed to avoid investments in corporations that create environmental harms . . . .”).

130 Cf. Stephanie E. Tanger, Enforcing Corporate Responsibility for Violations of Workplace Immigration Laws: The Case of Meatpacking, 9 Harv. Latino L. Rev. 59, 79 (2006) (“In response to the lawsuit, Tyson fired the managers who pled guilty and claimed to redouble its efforts . . . .”).

131Often, low-level employees will be judgment-proof. This in turn provides part of the economic justification for the rule of respondeat superior. See Posner, supra note 80, at 188 (“[M]ost employees lack the resources to pay a judgment . . . .”).

132For instance, Exxon fired Captain Hazelwood soon after the Valdez oil spill, and he also faced criminal prosecution. Exxon Valdez Trial to Be Held in Anchorage, N.Y. Times, Oct. 29, 1989, § 1, at 27.

133Polinsky & Shavell, supra note 71, at 949.

134 Oliver Wendell Holmes, Jr., The Common Law 8–38 (Dover 1991) (1881). As Justice Holmes explained:We have now followed the development of the chief forms of liability in modern law for anything other than the immediate and manifest consequences of a man’s own acts. . . . We have seen a single germ multiplying and branching into products as different from each other as the flower from the root. It hardly remains to ask what that germ was. We have seen that it was the desire of retaliation against the offending thing itself. . . . A consideration of the earliest instances will show, as might have been expected, that vengeance, not compensation, and vengeance on the offending thing, was the original object. The ox in Exodus was to be stoned. The axe in the Athenian law was to be banished. The tree, in Mr. Tylor’s instance, was to be chopped to pieces.Id. at 34.

135Business law also gives corporations many of the legal rights enjoyed by natural persons. See Santa Clara County v. S. Pac. R.R. Co., 118 U.S. 394, 396 (1886) (“The court does not wish to hear argument on the question whether the provision in the Fourteenth Amendment to the Constitution, which forbids a State to deny to any person within its jurisdiction the equal protection of the laws, applies to these corporations. We are all of opinion that it does.” (quoting Chief Justice Waite)) (syllabus). This adds strength to the argument that corporate “persons”—and not only natural persons—should face the possibility of economic punishment through punitive damages.

136 See Bert Swart, International Trends Towards Establishing Some Form of Punishment for Corporations, 6 J. Int’l Crim. Just. 947, 951–52 (2008) (“[C]orporations have their own institutional memories. They can remember things that happened in the past and learn from their experiences. They can correct their policies if they have made mistakes, or if circumstances make it desirable for them to change them. So far as their relationship with the law is concerned, they can make free choices about whether or not to comply with the law. Their freedom of choice in this respect makes it both possible and justifiable to hold them accountable for their choices. Experience also shows that sanctions imposed on corporations for having violated the law often induce them to adjust their goals or policies in order to prevent repetition.”).

137Viscusi, supra note 123, at 288.

138 Id. at 335.

139 Id. at 286–87.

140 Id. at 288.

141 Id. at 307.

142 Id. at 310.

143 Id. at 315.

144 Id. at 322.

145 See id. at 285. Viscusi, however, argues that punitive damages are assessed more or less at random, making it impossible for a corporation to respond to the threat of punitive damages liability by spending more on precautions. Id. at 309.

146 See, e.g., David A. Super, Privatization, Policy Paralysis, and the Poor, 96 Cal. L. Rev. 393, 408 (2008) (describing the “laser focus on the bottom line” that characterizes private business).

147Eric A. Posner & Cass R. Sunstein, Dollars and Death, 72 U. Chi. L. Rev. 537, 554 (2005) (“To understand how deterrence works, one must take an ex ante perspective. . . . As is well known, people can be given optimal incentives to take care if they are required to pay damages for any financial losses that they cause . . . .” (footnotes omitted)).

148 See Keeton et al., supra note 106, § 2, at 9–10 (“Something more than the mere commission of a tort is always required for punitive damages. There must be circumstances of aggravation or outrage . . . or such a conscious and deliberate disregard of the interests of others that the conduct may be called willful or wanton.” (footnotes omitted)); id. § 35, at 212–13 (explaining that conduct labeled as “willful,” “wanton,” or “reckless” remains, “at essence, negligent, rather than actually intended to do harm,” yet “is held to justify an award of punitive damages” (footnote omitted)).

149Viscusi, supra note 123, at 285.

150 See Sandra F. Gavin, Stealth Tort Reform, 42 Val. U. L. Rev. 431, 438 (2008) (“In 1983, the Rand Institute studied the ‘serious public policy problem, namely the manufacture of products that may have been unreasonably dangerous to their users[,]’ to determine what external pressures had the greatest influence on promoting products safety and concluded, ‘[o]f all the various external social pressures, product liability has the greatest influence on product design decisions.’” (alteration in original) (quoting George Eads & Peter Reuter, Designing Safer Products, Corporate Responses to Product Liability Law and Regulation, at iii–viii (1983))).

151Polinsky & Shavell, supra note 71, at 887 n.44.

152 Id. at 886–87.

153 See, e.g., Galanter & Luban, supra note 10; Dan Markel, Retributive Damages: A Theory of Punitive Damages as Intermediate Sanction, 94 Cornell L. Rev. 239 (2009); Catherine M. Sharkey, Punitive Damages as Societal Damages, 113 Yale L.J. 347 (2003); Benjamin C. Zipursky, A Theory of Punitive Damages, 84 Tex. L. Rev. 105 (2005).

154George P. Fletcher, Paradoxes in Legal Thought, 85 Colum. L. Rev. 1263, 1264 (1985) (“The criteria of crime, criminal responsibility, and punishment have yet to receive an adequate account in the literature of law and economics.”).

155 Posner, supra note 80, at 216 (“But the fact that any sort of rape license is even thinkable within the framework of the wealth-maximization theory that guides so much of the analysis in this book is a limitation on the usefulness of that theory. What generates the possibility of a rape license is that fact that the rapist’s utility is weighted the same as his victim’s utility. If it were given a zero weight in the calculus of costs and benefits, a rape license could not be efficient. The only persuasive basis for such a weighting, however, would be a moral principle different from efficiency.” (emphasis added)). For other examples of moral atrocities that a theory based purely on economic efficiency would recommend, see id. at 11, 12, 27.

156 Galanter & Luban, supra note 10.

157 Id. at 1428.

158 Id.

159 Id. at 1461.

160 Id.

161 Id. at 1428 (emphasis added).

162 Id. at 1432 (“A retributivist scales punishment to the heinousness of the offense, and that is not measured by the magnitude of harm.”).

163 Id.

164 Id.

165 Id. at 1461 (“In our view, the limits of punitive damages have to do entirely with the heinousness of the wrongful act; they have nothing to do with the size of compensatory awards. Thus, we oppose proposals to cap punitive damages at some small multiple of compensatory damages.”).

166Jeffrey R. White, State Farm and Punitive Damages: Call the Jury Back, 5 J. High Tech. L. 79, 79 (2005) (“Nowhere does the civil jury speak louder than when it awards punitive damages against a defendant who has violated our common understanding of acceptable behavior. The jury verdict speaks as the conscience of the community.”).

167David F. Partlett, The Republican Model and Punitive Damages, 41 San Diego L. Rev. 1409, 1427 (2004).

168 Id.

169 Id.

170 See Cass R. Sunstein, On the Psychology of Punishment, 11 Sup. Ct. Econ. Rev. 171, 171 (2004) (“Jurors are intuitive retributivists . . . .”).

171Drumbl, supra note 11, at 577 (“[T]he goals of deterrence . . . may well conflict with those of retribution . . . .”).

172 See Keeton et al., supra note 106, § 2, at 9 (“Something more than the mere commission of a tort is always required for punitive damages.”).

173 See supra Part III.A.3.

174 See, e.g., Reeves v. Carlson, 969 P.2d 252, 256 (Kan. 1998) (“To be reckless, conduct must be such as to show disregard of or indifference to consequences, under circumstances involving danger to life or safety of others.”).

175 See supra Part III.

176Polinsky & Shavell, supra note 71, at 955 (“It is evident that the best level of punitive damages should be a compromise between the levels that are optimal when each objective is considered independently.”).

177Galanter & Luban, supra note 10, at 1432.

178 Id. at 1433.

179Exxon Shipping Co. v. Baker, 128 S. Ct. 2605, 2613 (2008).

180 Id. at 2614.

181 Id. The jury awarded $5,000 in punitive damages against the Valdez’s captain. Id.

182 Id.

183 Id.

184 Id. at 2611.

185 Id. at 2619.

186 Id.

187 Id., passim.

188 Id. at 2620–25.

189 Id. at 2625.

190 Id. at 2626. The Court did not apply the Gore guideposts—reprehensibility, disparity, comparison to civil penalties—as required by the due process punitive damages cases. See BMW of N. Am., Inc. v. Gore, 517 U.S. 559, 575 (1996) (setting forth the guideposts).

191 Exxon Shipping, 128 S. Ct. at 2627.

192 Id.

193O.W. Holmes, The Path of the Law, 10 Harv. L. Rev. 457, 459 (1897). Justice Holmes was then serving on the Supreme Judicial Court of Massachusetts. Id. at 457.

194 Exxon Shipping, 128 S. Ct. at 2627. This argument is not universally accepted. According to Jeffrey R. White:Business practices and corporate decisions require cost planning and consideration of potential liability exposure. Unpredictable punitive damage awards are an unwelcome wild card.This argument received a cool reception in state courts in the 1980s. Many took the view expressed by the Maine Supreme Court [in Tuttle v. Raymond, 494 A.2d 1353, 1359 (Me. 1985),] that “the lack of any precise formula by which punitive damages can be calculated is one of the important assets of the doctrine.” As the New Jersey Supreme Court explained [in Fischer v. Johns-Manville Corp., 512 A.2d 466, 477 (N.J. 1986)]:Anticipation of these damages will allow potential defendants, aware of dangers of a product, to factor those anticipated damages into a cost-benefit analysis and to decide whether to market a particular product. The risk and amount of such damages can, and in some cases will, be reflected in the cost of a product, in which event the product will be marketed in its dangerous condition.Thus, [according to Palmer v. A.H. Robins Co., 684 P.2d 187, 218 (Colo. 1984)], ‘[i]f punitive damages are predictably certain, they become just another item in the cost of doing business.’ That result defeats the purpose of punitive damages in deterring misconduct and turns them into a user fee that permits defendants to continue their misconduct for a price.White, supra note 167, at 88 (footnotes omitted).

195 Exxon Shipping, 128 S. Ct. at 2628–29.

196 Id. at 2629.

197 Id. at 2629–33.

198 Id. at 2624.

199 Id. at 2624–25.

200 Id. at 2625.

201Brian J. Ostrom, et al., A Step Above Anecdote: A Profile of the Civil Jury in the 1990s, 79 Judicature 233, 238 (1996).

202 Exxon Shipping, 128 S. Ct. at 2625.

203 Id. However, the facts in two cases will never be exactly the same. See George Priest, Introduction to Cass R. Sunstein et al., Punitive Damages: How Juries Decide 1, 2 (2002) (“The magnitude of punitive damages verdicts appears to vary substantially across juries. But this judgment, too, is problematic. In some sense, no two cases are alike. Thus, there is an inherent difficulty in evaluating one verdict against another . . . .”).

204 Exxon Shipping, 128 S. Ct. at 2625–26.

205 Id. at 2626. The case involving the $4 million award is BMW of North America Inc. v. Gore, 517 U.S. 559 (1996), the first case in which the Supreme Court struck down a state court award of punitive damages. The Supreme Court, 1995 Term—Leading Cases, 110 Harv. L. Rev. 135, 145, 150 (1996). Not only is it a weak form of argumentation to rely on anecdotal evidence—much less a single anecdote—but the Court avoided mentioning that the Alabama Supreme Court remitted the damages to $2 million, and then, after the U.S. Supreme Court struck down the award, the Alabama Supreme Court again remitted the award, this time to just $50,000. Thus, the Court exaggerated the true disparity between awards in the two cases. In addition, selecting a case decided before the Court clarified its punitive damages jurisprudence in State Farm Mutual Automobile Insurance Co v. Campbell, 538 U.S. 408 (2003), is misleading. In Exxon Shipping, the Court assessed the supposed inadequacy of the law in 2008, and BMW likely would have been decided differently if it had been heard after State Farm. See W. Kip Viscusi, The Blockbuster Punitive Damages Awards, 53 Emory L.J. 1405, 1420–26 (2004) (discussing the effect of State Farm on large punitive damages awards).

206 Exxon Shipping, 128 S. Ct. at 2626.

207 Id. at 2625–34.

208 See Jonathan M. Karpoff & John R. Lott, Jr., On the Determinants and Importance of Punitive Damage Awards, 42 J.L. & Econ. 527, 529 (1999) (“An additional concern is not just whether punitive damages are predictable, but whether they are predictable for the right reasons.”).

209The Supreme Court has frequently relied on statistical studies. See Adam Liptak, From One Footnote, a Debate over the Tangles of Law, Science and Money, N.Y. Times, Nov. 25, 2008, at A16 (“The Supreme Court has often considered academic studies in its decisions, starting with Louis D. Brandeis’s famous 1908 brief collecting medical and other evidence to support laws limiting work hours. Lawyers still call such submissions ‘Brandeis briefs.’ The court’s signal triumph, Brown v. Board of Education in 1954, cited studies from psychologists and others, and citations to empirical work are commonplace these days.”). However, the Court’s use of the studies it cited in Exxon Shipping has already drawn heavy criticism:“The opinion reads like a bad joke,” said Jeffrey J. Rachlinski, a law professor at Cornell. “They say they know of no study showing punitive damages are orderly in any way, and yet they cite” a study by Theodore Eisenberg, a prominent empirical legal studies scholar at Cornell, “showing punitive damages are pretty orderly.”Professor Eisenberg struggled to stay respectful about the [C]ourt’s approach to his work . . . . He finally settled on this phrase: “I believe the [C]ourt went seriously astray” in concluding that his work supported a reduced award.Id.

210 Exxon Shipping, 128 S. Ct. at 2624–25 (emphasis added).

211 Id. at 2633.

212 Sentencing-Guideline Study Finds Continuing Disparities, N.Y. Times, Nov. 27, 2004, at A11 (describing study by U.S. Sentencing Commission finding that the average federal prison inmate’s sentence is fifty months).

213A one-to-one cap will only allow total damages to be twice the amount of compensatory damages. As explained in Part III, when the probability of the defendant paying damages is only 0.5, then (assuming the reward for negligence does not already provide the defendant an adequate incentive to take efficient precautions) compensatory damages must be multiplied by two. If the odds of paying damages were any less than 0.5, then a one-to-one cap would prevent the tort system from charging the defendant enough to ensure that private and social costs do not continue to diverge.

214The one-to-one cap will be especially harmful when compensatory damages are low or when the defendant’s taste for risk warrants significantly higher damages. See supra Part III.

215 Exxon Shipping, 128 S. Ct. at 2621 (“[T]he consensus today is that punitives are aimed not at compensation but principally at retribution and deterring harmful conduct.”).

216Some felt that the Court hardly acknowledged Exxon’s blameworthiness. See Adam Liptak, Damages Cut Against Exxon in Valdez Case, N.Y Times, June 26, 2008, at A1 (“Jeffrey L. Fisher, a lawyer for the plaintiffs, said there was ‘a great deal of sadness’ among his clients. ‘What is painful,’ Mr. Fisher said, ‘is that there seems to have been some disagreement between the dissenters and the majority on how reprehensible Exxon’s conduct was.’”). Some scholars have argued that the Court was already ignoring the retributivist rationale for punitive damages before Exxon Shipping was decided. See Paul J. Zwier, The Utility of a Nonconsequentialist Rationale for Civil Jury-Awarded Punitive Damages, 54 U. Kan. L. Rev. 403, 404 (2006) (describing “the harm the Court inflicts by minimizing the retributive justification and usurping the discretion of the jury” and arguing that “the Court backhandedly endorsed the law and economics, or deterrence, model of punishment and ignored the broader effects on social norms and values that result from taking the retribution analysis out of the hands of a common law jury”).

217 Exxon Shipping, 128 S. Ct. at 2613.

218See supra Part I.

219 See Denise E. Antolini, Punitive Damages in Rhetoric and Reality: An Integrated Empirical Analysis of Punitive Damages Judgments in Hawaii, 1985–2001, 20 J.L. & Pol. 143, 148 (2004) (“[C]onsumer advocate Ralph Nader decried Exxon’s failure to pay the award more than a decade after the verdict, calling the company’s response to the award ‘a legal war of attrition, while thousands of Alaskans and others suffered.’” (quoting Ralph Nader, Justice Delayed Is Justice Denied, CommonDreams.org, Oct. 9, 2000, http://www.commondreams.org/views/101000-109.htm)).

220 See, e.g., Ott, supra note 14, at 23 (reporting that in the effort to get the Trans-Alaska Pipeline System permit, “Exxon’s Ken Fountain told lobbyists, ‘I don’t care if every goddamn fish dies, get that [authorization] permit . . . .’”). Some of Exxon’s methods for cleaning up the spill oil were harmful to the environment and to the health of cleanup workers, and an Exxon official admitted that the actual cleanup was secondary to “its public image aspects.” Id. at 52 (describing how the high-pressure “hot-water wash” used to clean the shoreline “cooked clams, mussels, and other intertidal animals that had survived the spill and destroyed sea plants and animals that supported the entire coastal ecosystem”).

221Exxon Shipping, 128 S. Ct. at 2613 (“In the aftermath of the disaster, Exxon spent around $2.1 billion in cleanup efforts.”).

222Some consider Exxon’s cleanup efforts to have been ineffective and mishandled, and claim the cleanup imposed serious health risks on the cleanup workers. Ott, supra note 14, at 52. Some critics think that Exxon’s expenditure of $2 billion on cleanup may have been more harmful to Prince William Sound than doing nothing. Schoch, supra note 29, at 44–45. If so, it is a valid question whether the cleanup should have been considered as a mitigating factor at all.

223 Ott, supra note 14, at 186–87.

224Keith E. Sealing, Civil Procedure in Substantive Context: The Exxon-Valdez Cases, 47 St. Louis U. L.J. 63, 68 n.29 (2003).

225 Ott, supra note 14, at 47 (alteration in original).

226 Id. at 113.

227 See Charles Dickens, Bleak House 936 (Nicola Bradbury ed., Penguin Books 1996) (1853) (“Richard [Carstone], more worn and haggard, haunted the court day after day, listlessly sat there the whole day long when he knew there was no remote chance of the suit being mentioned, and became one of the stock sights of the place. I wonder whether any of the gentlemen remembered him as he was when he first went there.”); Ott, supra note 14, at 272 (“Closure for individuals is a different story. Those who lean heavily on the Supreme Court’s decision may close this chapter of their lives soon after the court makes a ruling. Those who wait for the herring to rebound may be waiting years or decades. Those who died—over 6,000—while their cases churned through the judiciary system will never have had the peace of closure.”).

228Jad Mouawad, Exxon Sets Profit Record: $40.6 Billion Last Year, N.Y. Times, Feb. 2, 2008, at C3.

229 See Jeffrey Rosen, Supreme Court, Inc., N.Y. Times, Mar. 16, 2008, (Magazine) at MM38 (discussing the pro-business trend in Supreme Court decisions).

230Exxon Shipping Co. v. Baker, 128 S. Ct. 2605, 2631–32 (2008).

231As the Court made clear in Exxon Shipping, the case is not binding outside of admiralty. Id. at 2626–27.

232 See, e.g., Leavey v. Unum Provident Corp., 295 F. App’x 255, 259 n.1 (9th Cir. 2008) (“While the Supreme Court’s recent decision in Exxon Shipping Co. v. Baker . . . review[ed] a jury award for conformity with maritime law, rather than the outer limit allowed by due process, . . . the Court’s statements in that case support the district court’s decision to reduce the award here.” (citations omitted) (internal quotation marks omitted)); Hayduk v. City of Johnstown, 580 F. Supp. 2d 429, 484 n.46 (W.D. Pa. 2008) (“Although Exxon is a maritime law case, it is clear that the Supreme Court intends that its holding have a much broader application.”).

233 Thomas J. Schoenbaum, Admiralty & Maritime Law § 15-1 (4th ed. 2009).

234 Id.

235 Id.

236 Id.

237354 U.S. 129, 133 (1957).

238498 U.S. 19 (1990).

239John W. DeGravelles, Uncertain Seas for Maritime Punitive Damages, Trial, Jan. 2004, at 50, 51 (“Although Miles did not address punitive damages specifically, it became the launching pad for much of the case law eliminating them from maritime law.”).

240Somewhat tautologically, maritime law principles are based on distinctly maritime concerns. See Schoenbaum, supra note 234, § 3-5 (“Thus the test for admiralty tort jurisdiction requires that an incident (1) occur on navigable waters; (2) bear a substantial relationship to traditional maritime activity; and (3) have a potentially disruptive impact on maritime commerce.”).

241517 U.S. 559 (1996).

242538 U.S. 408 (2003).

243 See Posner, supra note 80, at 207.

J.D. Candidate, Emory University School of Law (2010). I am especially grateful to Professors William Buzbee, Joanna Shepherd, and Alexander “Sasha” Volokh for their guidance and insightful feedback throughout the writing process. I am also grateful to the editing team at the Emory Law Journal. And finally, I would like to thank my wife Lindy.