Saving lives through punitive damages

Southern California Law Review

Volumn 83, Issue 2, 2010, Pages 229-262

  Hersch, Joni ^a   Kip Viscusit, W ^a  

^a VANDERBILT UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 77953309348     PISSN: 00383910     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Review

References (129)

1. Punitive damages are designed to punish and serve as a deterrent. A typical example of punitive damages instructions indicating the punishment and deterrence functions of punitive damages is that of New Mexico: "Punitive damages are awarded for the limited purposes of punishment and to deter others from the commission of like offenses." N.M. Uniform Jury Instructions for Civil Cases § 13-1827 (West 2009). In this Article, we focus on the deterrence role. In Exxon Shipping Co. v. Baker, the U.S. Supreme Court cited the following cases as examples of the consensus that deterrence is the rationale underlying punitive damages: Moskovitz v. Mount Sinai Medical Center, 635 N.E.2d 331, 343 (Ohio 1994) ("The purpose of punitive damages is not to compensate a plaintiff, but to punish and deter certain conduct.")
2. Hamilton Development Co. v. Broad Rock Club, Inc., 445 S.E.2d 140,143 (Va. 1994) (same);
3. Loitz v. Remington Arms Co., 563 N.E.2d 397, 401 (111. 1990) (same);
4. Green Oil Co. v. Hornsby, 539 So. 2d 218,222 (Ala. 1989) (same);
5. and Masah v. General Motors Corp., 780 P.2d 566, 570 (Haw. 1989) (same). Exxon Shipping Co. v. Baker, 128 S. Ct. 2605, 2621 n.9 (2008). The Court also cited Cooper Industries, Inc. v. Leatherman Tool Group, Inc., 532 U.S. 424, 432 (2001) (stating that punitive damages are "intended to punish the defendant and to deter future wrongdoing");
6. State Farm Mutual Automobile Insurance Co. v. Campbell, 538 U.S. 408, 416 (2003) ("[P]unitive damages... are aimed at deterrence and retribution.");
7. and Restatement (Second) of Torts § 908 cmt. a (1979). Id.
8. W. Kip Viscusi, 77ie Devaluation of Life, 3 Reg. & Governance 103,105 (2009).
9. Id. at 103,107. See, e.g., Office of Info. & Regulatory Affairs, U.S. Office of Mgmt. & Budget, Informing Regulatory Decisions: 2003 Report to Congress on the Costs and Benefits of Federal Regulations and Unfunded Mandates on State, Local, and Tribal Entities app. A, at 94 (2003) (noting agencies' use of the VSL in calculating costs and benefits of regulations)
10. Memorandum from Tyler D. Duvall, Assistant Sec'y for Transp. Policy, & D. J. Gribbin, Gen. Counsel, to Secretarial Officers, Modal Administrators, U.S. Dep't of Transp., Treatment of the Economic Value of Statistical Life in Departmental Analyses, available at http://ostpxweb.dot.gov/ policy/reports/080205.htm [hereinafter Duvall & Gribbin Memo] (suggesting how to improve the use of the VSL)
11. National Center for Environmental Economics, Frequently Asked Questions on Mortality Risk Valuation, http://yosemite.epa.gov/ee/epa/eed.nsf/webpages/ Mortdity%20risk%20valuation.html (select "Why Does EPA Use a Value of Statistical Life?") (last visited Nov. 3,2009).
12. Thomas R. Ireland, Recent Legal Decisions Regarding Hedonic Damages: An Update, 13 J. Forensic Econ. 189,190-192 (2000).
13. Eric A. Posner & Cass R. Sunstein, Dollars and Death, 72 U. Chi. L. REV. 537, 587-590 (2005).
14. A. Mitchell Polinsky & Steven Shavell, Punitive Damages: An Economic Analysis, 111 Harv. L. Rev. 869, 889-90, 941-42 (1998). Under Polinsky and Shavell's proposal, if the probability of detection is 1.0, then in a wrongful death case there is no rationale for punitive damages from a deterrence standpoint because their proposal always sets compensatory damages to equal the VSL. Id. at 891.
15. There may, of course, be some bequest motive, but these values have not been found to be extremely large relative to the VSL. In particular, the bequest value for workers in hazardous jobs is "equivalent to a 2.4% chance at another year of life." Michael J. Moore & W. Kip Viscusi, Models for Estimating Discount Rates for Long-Term Health Risks Using Labor Market Data, 3 J. risk & Uncertainty 381,399 (1990).
16. The key aspect of severe personal injury cases is that the injury reduces the marginal utility that the victim can derive from income. W. Kip Viscusi & William N. Evans, Utility Functions That Depend on Health Status: Estimates and Economic Implications, 80 AM. ECON. REV. 353, 370 (1990).
17. See also W. Kip Viscusi, Pain and Suffering: Damages in Search of a Sounder Rationale, 1 MICH. L. & POL'YREV. 141,147-56(1996).
18. State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408,425 (2003).
19. Exxon Shipping Co. v. Baker, 128 S. Ct. 2605,2633 (2008).
20. Bureau of Justice Statistics, U.S. Dep't of Justice, Civil Justice Survey of State Courts, 2005: Codebook (Jan. 14, 2009), available at http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/23862?archive=ICPSR&q= 23862 [hereinafter CJSSC 2005: Codebook].
21. See generally W. Kip Viscusi, Fatal Tradeoffs: Public and Private Responsibilities for Risk (1992) (using the terminology "value of life"). Recent studies use the terminology "value of statistical life." See supra note 3.
22. See supra note 3.
23. This example is intended to illustrate the VSL based on a thought experiment rather than indicating the kinds of statistical studies used to generate the VSL estimates.
24. See viscusi, supra note 12, at 34-42 (providing a survey of VSL literature);
25. W. Kip Viscusi & Joseph E. Aldy, The Value of a Statistical Life: A Critical Review of Market Estimates Throughout the World, 27 J. Risk & Uncertainty 5 (2003) (same).
26. See Viscusi, supra note 2, at 113-118 (reviewing the different estimates used by the EPA and the different studies the EPA took into account).
27. Duvall & Gribbin Memo, supra note 3 (citing Viscusi & Aldy, supra note 15).
28. These differences are reviewed in Viscusi & Aldy, supra note 15.
29. Office of Info. & Regulatory Affairs, supra note 3, app. D, at 146-148 (Office of Management and Budget circular to executive agency heads).
30. As noted above, the Department of Transportation now uses a figure of $5.8 million. Duvall & Gribbin Memo, supra note 3. The VSL numbers used by agencies have risen over time, as the EPA uses values in the $7-$9 million range. Viscusi, supra note 2, at 115. Other values in 2008 dollars used by agencies for certain regulations are $6.2 million, by the Consumer Product Safety Commission; S5.3-S6.8 million, by the Food and Drug Administration; and S3.1-$6.2 million, by the Department of Homeland Security. Id. at 108 tbl.l.
31. The $9 million figure is also consistent with labor-market studies. Viscusi and Aldy presented a meta-analysis of the different wage-risk studies and found a median VSL of $7 million. Viscusi & Aldy, supra note 15, at 63. This value is approximately $8.75 million in 2009 dollars. According to the Consumer Price Index ("CPI") calculator, $1.00 in the year 2000 is equivalent to $1.25 in 2009. U.S. Dep't of Labor, Bureau of Labor Statistics, CPI Inflation Calculator, http://www.bls.gov/data/inflation-calculator.htm (last visited Nov. 3, 2009). For purposes of our discussion, we will use a VSL of $9 million.
32. News Release, Bureau of Labor Statistics, U.S. Dep't of Labor, Usual Weekly Earnings of Wage and Salary Workers: First Quarter 2009, at 1 (Apr. 16, 2009), available at http://www.bls.gov/news.release/archives/wkyeng-04162009. pdf.
33. See Kenneth R. Feinbero et al., U.S. Dep't of Justice, Final Report of the Special Master for the September 11th Victim Compensation Fund of 2001, at 33 (2004), available at http://www.usdoj.gov/final-report.pdf
34. basing work-life estimates on the tables presented in James Ciecka, Thomas Donley & Jerry Goldman, A Markov Process Model of Work-Life Expectancies Based on Labor Market Activity in 1997-1998, J. legal econ., Winter 1999-2000, at 33
35. Conditional on being a worker at age thirty-five, the average expected remaining work-life is twenty-five years from age thirty-five until the worker's death. Work-life estimates are not restricted to consecutive years in the labor force but also take into account interruptions such as periods of unemployment, disability, and the risk of death. See Ciecka et al., supra.
36. We assume that income would be paid at the beginning of each year for twenty-five consecutive years. This assumption is made to simplify the calculations. Recognizing periods of interruptions in work will defer some income beyond twenty-five years, reducing its present value.
37. W. Kip Viscusi, The Value of Risks to Life and Health, 31 J. ECON. LITERATURE 1912,1941-42tbl.7 (1993).
38. See W. Kip Viscusi, The Flawed Hedonic Damages Measure of Compensation for Wrongful Death and Personal Injury, 20 J. Forensic Econ. 113, 118-21 (2007).
39. Noneconomic pain and suffering damages comprise a substantial percentage of compensatory damages awards. 2 Am. Law Inst., Enterprise Responsibility for Personal Injury 199-201 (1991) [hereinafter Enterprise Responsibility].
40. See Viscusi, supra note 26, at 118-121
41. See id. at 120-121
42. Duvall & Gribbin Memo, supra note 3 (citing income elasticities of VSL based on Viscusi & Aldy, supra note 15).
43. Viscusi & Aldy, supra note 15, at 40.
44. See Viscusi, supra note 2, at 109. Based on the authors' knowledge, the Department of Transportation is the only agency to take such income heterogeneity into account.
45. Tort law generally provides for recovery of lost earnings. See 2 Dan B. dobbs, Dobbs Law of Remedies: Damages-Equtty-Restitution § 8.3(1), at 423 (2d ed. 1993).
46. See feinberg et al., supra note 23, at 30-43 (detailing the factors considered in calculating economic and noneconomic damages).
47. This is the standard economic theory for the effect of tort liability costs on prices. See Paul H. Rubin, John E. Calfee & Mark F. Grady, BMW v. Gore: Mitigating the Punitive Economics of Punitive Damages, in 5 SUP. Cr. Econ. Rev. 179, 188 (Harold Demsetz, Ernest Gellhorn & Nelson Lund eds., 1997) (discussing the effect of punitive damages on product prices).
48. See supra notes 7-8 and accompanying text.
49. Posner & Sunstein, supra note 5, at 587-590 Courts have, however, widely rejected use of hedonic damages for compensation.
50. See Ireland, supra note 4, at 190. Below we will show that Polinsky and Shavell adopt a fairly similar approach in their punitive damages formulation for personal injury cases.
51. See infra notes 54-55 and accompanying text.
52. Posner & Sunstein, supra note 5, at 587.
53. Id.
54. Id. at 590.
55. Id. at 584-92 (outlining their proposal, which omits mention of pain and suffering except indirectly through the hedonic loss of the victims, or VSL). For a discussion on the role of pain and suffering compensation, see enterprise responsibility, supra note 26, at 199-230.
56. Posner & Sunstein, supra note 5, at 587-590
57. Id. at 587.
58. Id.
59. See supra notes 16-20 and accompanying text.
60. See Viscusi, supra note 2, at 108 tbl.l, 114 tbl.2.
61. See supra notes 22-23 and accompanying text.
62. Posner & Sunstein, supra note 5, at 588.
63. Polinsky & Shavell, supra note 6.
64. See id. at 941-942
65. Id. at 873-874
66. Exactly how the punishment objective should enter and whether it should be added in some way to the deterrence value is unclear. For a discussion of the punishment objective, see id. at 948-56, and id. at 955 (indicating that the punitive damages amount should be a number between the punishment value and the deterrence value). For a discussion of model jury instructions, see id. app.
67. The classic cite for this theory is Jeremy Bentham, Principles of Penal Law, in 1 the Works of Jeremy Bentham 365,401-402 (John Bowring ed., 1962) (1838-1843).
68. Extensive modem developments exist as well. See. e.g., Robert D. Cooter, Punitive Damages for Deterrence: When and How Much?, 40 ala. L. Rev. 1143,1149-1153 (1989)
69. Polinsky & Shavell, supra note 6, at 887-896
70. For example, let the probability of detection be less than 1.0. Assume for this case that the value of the harm equals Economic Damages + Noneconomic Damages so that all harms can be fully compensated through monetary payments. Penalties then establish optimal deterrence if Probability of Detection x (Punitive Damages + Economic Damages + Noneconomic Damages) = Harm. After some rearrangement, one finds that the punitive damages amount can be expressed in terms of the punitive damages ratio, or (Punitive Damages / (Economic Damages + Noneconomic Damages)) (1 / (Probability of Detection -1)).
71. Polinsky & Shavell, supra note 6, at 889.
72. Id. at 941-942 (footnotes omitted).
73. Michael L. Rustad, The Closing of Punitive Damages' Iron Cage, 38 loy. l.a. l. Rev. 1297,1338-1346(2005).
74. Id. at 1308.
75. Id. at 1350.
76. BMW of N. Am., Inc. v. Gore, 517 U.S. 559 (1996).
77. Id. at 580.
78. Pac. Mut. Life Ins. Co. v. Haslip, 499 U.S. 1 (1991).
79. TXO Prod. Corp. v. Alliance Res. Corp., 509 U.S. 443,462 (1993).
80. State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408,425 (2003).
81. Id.
82. Id. (quoting Gore, 517 U.S. at 582).
83. Exxon Shipping Co. v. Baker, 128 S. Ct. 2605,2625-2626 (2008).
84. Cass R. Sunstein et al., Punitive Damages: How Juries Decide 11-14 (2002).
85. See id. at 212-214 (synthesizing controlled experimental studies on punitive damages that found that mock jurors reached a consensus on the blameworthiness of behavior but there was substantial variability in assigning a punitive damages penalty for such conduct).
86. Exxon Shipping, 128 S. Ct. at 2625-26. As indicated in footnote seventeen of the decision, the Court declined to rely on SUNSTEIN ET al., supra note 67, because that study was funded in part by Exxon. Exxon Shipping, 128 S. Ct. at 2626 n.17.
87. In re Exxon Valdez, 236 F. Supp. 2d 1043, 1049 n.14 (D. Alaska 2002) (quoting Jury Instruction No. 22).
88. Id. at 1049 n.16 (quoting Jury Instruction No. 25).
89. Id. at 1050 n.21 (quoting Jury Instruction No. 35).
90. The U.S. District Court for the District of Alaska, in its decision in In re Exxon Valdez, stressed the importance of adequate instructions: "Without proper instructions, jury verdicts are patently suspect." Id. at 1054.
91. Exxon Shipping, 128 S. Ct. at 2633 (footnote omitted).
92. Id. at 2634 (quoting State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408,425 (2003)).
93. Id. at 2639 (Ginsburg, J., concurring in part, dissenting in part) (citation omitted).
94. Id. at 2633 (majority opinion) ("In a well-functioning system, we would expect that awards at the median or lower would roughly express jurors' sense of reasonable penalties in cases with no earmarks of exceptional blameworthiness within the punishable spectrum (cases like this one, without intentional or malicious conduct, and without behavior driven primarily by desire for gain, for example) and cases (again like this one) without the modest economic harm or odds of detection that have opened the door to higher awards.").
95. Jurinko v. Med. Protective Co., 305 F. App'x 13,27-30 (3d Cir. 2008).
96. Morgan v. N.Y. Life Ins. Co., 559 F.3d 425,441-443 (6th Cir. 2009).
97. The description of the CJSSC data and the data used in the empirical analysis is based on the Department of Justice's description of the study. See CJSSC 2005: Codebook, supra note 11.
98. See Bureau of Justice Statistics, U.S. Dep't of Justice, Civil Justice Survey of State Courts, 1996: Description, at ii (Sept. 18, 2000), available at http://www.icpsr.umich.edu/icpsrweb/IrCPSR/studies/2883?archive=ICPSR [hereinafter CJSSC 1996: Description] (noting that the 1996 study expanded on the 1992 study by including both bench and jury trials).
99. The sampling frame is a two-stage stratified sampling process. For the 1992,1996, and 2001 waves of data, the following sampling frame was used: In the first stage, the seventy-five most populous counties were identified. In the second stage, forty-five or forty-six of these counties were selected for inclusion in the study.
100. See Bureau of Justice Statistics, U.S. Dep't of Justice, Civil Justice Survey of State Courts, 1992: Description, at iii (Oct. 1, 1996), available at http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/6587?archive=ICPSR (sampling forty-five counties)
101. For the 2005 wave of data, the survey preserved the earlier sampling frame by sampling forty-six counties from the seventy-five most populous counties, but also sampled from 110 additional counties outside the seventy-five most populous counties.
102. See CJSSC 2005: Codebook, supra note 11, at 10-12; Bureau of Justice Statistics, U.S. Dep't of Justice, Civil Justice Survey of State Courts, 2005: Description, at ii-iv (Jan. 14, 2009), available at http://www.icpsr.umich.edu/ icpsrweb/ICPSR/studies/23862?archive=ICPSR&q=23862 [hereinafter CJSSC 2005: Description]. The total number of cases included in the 2005 sample was 8872. CJSSC 2005: Codebook, supra note 11, at 6.
103. See CJSSC 2005: Description, supra note 82, at ii-iv.
104. For articles using the first three waves of the CJSSC data, see generally Theodore Eisenberg et al., Juries, Judges, and Punitive Damages: Empirical Analyses Using the Civil Justice Survey of State Courts 1992, 1996, and 2001 Data, 3 J. Empirical Legal Stud. 263 (2006) (analyzing CJSSC 1992,1996, and 2001 studies)
105. Joni Hersch & W. Kip Viscusi, Punitive Damages by Numbers: Exxon Shipping Co. v. Baker, 18 Sup. CT. Econ. REV. (Ilya Somin & Todd J. Zywicki eds., forthcoming 2010) (discussing the Exxon Shipping Court's use of 1992,1996, and 2001 data)
106. Joni Hersch & W. Kip Viscusi, Punitive Damages: How Judges and Juries Perform, 33 J. Legal Stud. 1 (2004) (using CJSSC 1996 data to demonstrate that juries and judges differ in awarding punitive damages);
107. and Thomas H. Cohen, U.S. Dep't of Justice, Punitive Damage Awards in Large Counties, 2001 (Mar. 2005) (analyzing CJSSC 2001 data).
108. See CJSSC 2005: Codebook, supra note II, at 11 (describing variables for the various different damages figures).
109. Cases not involving bodily injury will also be referred to as property damage cases.
110. The authors' calculations come from CJSSC 2005. The sample is comprised of cases in which the plaintiff won, punitive damages were awarded, compensatory damages were greater than zero, and the breakdown of compensatory damages into economic damages and noneconomic damages was reported. The numbers of observations are as follows: 156 (panel A), 44 (panel B), and 112 (panel C). In the 2005 CJSSC data set, case type is reported in twenty-six categories. See CJSSC 2005: Codebook, supra note 11, at 21-22. We group case types into the eleven categories reported in the table. See infra toll.
111. The weights are built into the CJSSC study to account for the sample not being a simple random sample. See CJSSC 2005: Codebook, supra note 11, at 13.
112. Note that the average ratio figures pertain to the average value of the individual punitive damages ratios across all cases, which does not equal the value of the average punitive damages award divided by the average value of compensatory damages.
113. See supra tbl. 1, panel A.
114. See supra tbl. I, panels B & C.
115. See supra tbl. 1, panel C.
116. See supra tbl. 1, panel C.
117. See supra notes 74-77 and accompanying text.
118. The data set includes a total of seven wrongful death cases with positive punitive damages awards, of which three are intentional torts, three are medical malpractice, and one is a motor vehicle tort.
119. See supra Part II.
120. See supra Part III.
121. For example, this does not appear to be a case with extraordinarily large medical expenses or a case in which the present value of the net earnings loss is extremely large. Complications such as these would warrant an adjustment in the target VSL level.
122. The authors' proposed $7,519 million punitive damages award would yield a punitive damages ratio of 5.08:1, well within the single-digit punitive damages ratio guidelines.
123. This relationship follows from the fact that severe bodily injuries lower the marginal utility of income and hence the degree to which money can compensate for the welfare loss. See supra notes 7-8 and accompanying text.
124. Note that based on our proposed theory of punitive damages, if the level of harm is held constant and restricted to wrongful death cases, then higher levels of noneconomic damages should reduce the level of punitive damages that are required for the total damages amount to equal the VSL.
125. Tests for influential observations indicated that the four observations in the sample with compensatory damages greater than $25 million were influential and that inclusion of these observations distorted the relation between damages components and punitive damages. Thus, we drop these four observations from the regression analysis. "An observation is said to be influential if removing the observation substantially changes the estimate of coefficients." UCLA Academic Technology Services, Stata Web Books: Regression with Stata, http://www.ats.ucla.edu/stat/stata/webbooks/reg/chapter2/ statareg2.htm (last visited Nov. 3,2009).
126. The authors' calculations come from CJSSC 2005. The sample is comprised of cases in which the plaintiff won, punitive damages were awarded, compensatory damages were greater than zero and less than $25 million, and the breakdown of compensatory damages into economic damages and noneconomic damages was reported. Robust standard errors are in parentheses and are based on probability weights using the variable "BWGT0."
127. See infra tbl.2.
128. See supra tbl.2.
129. See supra tbl.2.