Contracting over liability: Medical malpractice and the cost of choice

University of Pennsylvania Law Review

Volumn 158, Issue 4, 2010, Pages 957-1023

  Arlen, Jennifer ^a  

^a NEW YORK UNIVERSITY SCHOOL OF LAW   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 77951795404     PISSN: 00419907     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Review

References (314)

1. See Patricia M. Danzon, Liability for Medical Malpractice (discussing evidence that patients suffer more than 150,000 iatrogenic fatalities annually, more than half of which were caused by medical negligence), in 1 HANDBOOK OF HEALTH ECONOMICS 1339, 1351 (Anthony J. Culyer & Joseph P. Newhouse eds., 2000);
2. see also infra Section I.A (discussing the rate of medical error and its causes).
3. Id. at 1351-1352
4. See Jennifer Arlen & W. Bentley MacLeod, Torts, Expertise, and Authority: Liability of Physicians and Managed Care Organizations, 36 RAND J. ECON. 494, 507-11 (2005) (arguing that malpractice liability can be used to induce medical providers to invest optimally to reduce medical error);
5. infra note 45 (discussing empirical evidence).
6. See infra note 19.
7. See, e.g., RICHARD H. THALER & CASS R. SUNSTEIN, NUDGE ch. 14 (2008) (asserting that patients should be permitted to contract over liability);
8. Richard A. Epstein, Contractual Principle Versus Legislative Fixes: Coming to Closure on the Unending Travails of Medical Malpractice, 54 DEPAUL L. REV. 503, 505 (2005) [hereinafter Epstein, Contractual Principle] (asserting that it is a fundamental error to treat malpractice reform as a tort problem because "designing a governance regime calls for a contractual response");
9. Richard A. Epstein, Medical Malpractice, Imperfect Information, and the Contractual Foundation for Medical Services, 49 LAW & CONTEMP. PROBS. 201, 202 (1986) (arguing in favor of contractual malpractice liability);
10. Richard A. Epstein, Medical Malpractice: The Case for Contract, 1 AM. B. FOUND. RES. J. 87, 94-95, 149 (1976) (contending that contractual physician liability is the superior approach to medical liability);
11. Clark C. Havighurst, Private Reform of Tort-Law Dogma: Market Opportunities and Legal Obstacles, 49 LAW & CONTEMP. PROBS. 143, 161-62 (1986) (claiming that a larger role for private agreements in the determination of physician liability would benefit all parues involved);
12. Glen O. Robinson, Rethinking the Allocation of Medical Malpractice Risks Between Patients and Providers, 49 LAW & CONTEMP. PROBS. 173, 198 (1986) (explaining why medical malpractice liability should be governed by contract);
13. see also Keith N. Hylton, Agreements to Waive or to Arbitrate Legal Claims: An Economic Analysis, 8 SUP. CT. ECON. REV. 209, 263 (2000) (insisting that parties to consensual relationships should be allowed to contract over liability).
14. See, e.g., CLARK C. HAVIGHURST, HEALTH CARE CHOICES 265-302 (1995) (advocating contractual MCO liability for physician negligence);
15. Patricia M. Danzon, Tort Liability: A Minefield for Managed Care?, 26 J. LEGAL STUD. 491, 493-94 (1997) (supporting limited contractual liability for MCOs);
16. Clark C. Havighurst, Vicarious Liability: Relocating Responsibility for the Quality of Medical Care, 26 AM. J.L. & MED. 7, 8 (2000) [hereinafter Havighurst, Vicarious Liability] (asserting that MCOs should be liable for all medical malpractice committed by healthcare providers but should be allowed to alter or eliminate this liability by contract);
17. William M. Sage, Enterprise Liability and the Emerging Managed Health Care System, 60 LAW& CONTEMP. PROBS. 159, 208 (1997) (arguing that Congress should impose contractual liability for medical negligence on MCOs);
18. see also PAUL H. RUBIN, TORT REFORM BY CONTRACT 75-77 (1993) (advocating patient contracting over malpractice damage awards through health insurers);
19. THALER & SUNTEIN, supra note 5, at 212-13 (suggesting that insurers should be allowed to contract with patients over physician liability for negligence);
20. Richard A. Epstein & Alan O. Sykes, The Assault on Managed Care: Vicarious Liability, ERISA Preemption, and Class Actions, 30 J. LEGAL STUD. 625, 644-48 (2001) (arguing mat states should allow MCOs to limit liability by contract).
21. Throughout this Article, the term "informed patients" is used to refer to patients who are informed about the costs and benefits of imposing liability by contract. They are not informed about each individual medical provider's risk of medical error, however. This Article assumes that patients are informed about the costs and benefits of malpractice liability in order to evaluate proponents' claim that contractual liability enhances patients' welfare when patients make rational, informed, contract choices. This Article shows that even in this favorable situation, contractual liability is not efficient.
22. For a summary of the conventional economic case for contract, see PAUL C. WEILER, MEDICAL MALPRACTICE ON TRIAL 96 (1991). For an exposition of the claim that contracting over liability would allow patients to adopt rules that achieve the balance between cost and quality for which they are willing to pay,
23. see Epstein, Contractual Principle, supra note 5, at 507-509
24. See infra Section II.B.
25. See, e.g., Tom Baker & Timothy Lytton, Allowing Patients to Waive the Right to Sue for Medical Malpractice: A Response to Thaler and Sunstein, 104 Nw. U. L. REV. (forthcoming 2010) (manuscript at 5-12) (observing that behavioral biases may undermine patients' ability to contract effectively on their own behalf).
26. See, e.g., Epstein & Sykes, supra note 6, at 647-48 (claiming that MCO contractual liability would not be plagued by serious information problems);
27. see also J. Mark Ramseyer, Products Liability Through Private Ordering: Notes on a Japanese Experiment, 144 U. PA. L. REV. 1823, 1825 (1996) (arguing that the claim that consumers are not sufficiently informed to contract over products liability is empirically "questionable");
28. Alan Schwartz, Proposals for Products Liability Reform: A Theoretical Synthesis, 97 YALE L.J. 353, 378-84 (1988).
29. Recent scholars to embrace contractual malpractice liability include Cass Sunstein. See THALER & SUNSTEIN, supra note 5, at 212-13. Courts also now enforce clauses requiring mandatory arbitration of medical claims; mese clauses affect expected liability.
30. See Carol A. Crocca, Annotation, Arbitration of Medical Malpractice Claims, 24 A.L.R.5TH 1, 47-52 (1994).
31. See, e.g., Epstein, Contractual Principle, supra note 5, at 509.
32. This Article defines an MCO as any insurer that attempts to influence the quality of care selected either directly, through utilization review, or indirectly, through financial incentives provided to physicians to cut costs (e.g., capitation).
33. See sources cited supra note 6.
34. See supra note 7.
35. See Robinson, supra note 5, at 183-84 (positing that the case for contract depends on whether "in general, private parties are likely to achieve results that are at least as good and fair for themselves as would be achieved by paternalistic intervention");
36. see also HAVIGHURST, supra note 6, at 266 (arguing that there seems "to be a strong inducement for consumers and providers jointly to discover and to agree on rules and procedures that reduce the uncertainty, complexity, and volume of litigation"). For a discussion of adverse selection,
37. see infra Section III.C and Part IV.
38. There have been many proposals for medical-entity liability. See, e.g., Kenneth S. Abraham & Paul C. Weiler, Enterprise Medical Liability and the Evolution of the American Health Care System, 108 HARV. L. REV. 381, 398-414 (1994) (proposing hospital enterprise liability for medical error);
39. Jennifer Arlen & W. Bentley MacLeod, Malpractice Liability for Physicians and Managed Care Organizations, 78 N.Y.U. L. REV. 1929, 1961-79 (2003) (favoring MCO liability for both their own negligence in making treatment decisions and physician negligence);
40. Michelle M. Mello & Troyen A. Brennan, Deterrence of Medical Errors: Theory and Evidence for Malpractice Reform, 80 TEX. L. REV. 1595, 162326 (2002) (arguing in favor of hospital enterprise liability created by requiring hospitals to pay for and provide malpractice liability insurance to their affiliated physicians);
41. and Sage, supra note 6, at 164-69, 206-209 (discussing the benefits of MCO liability for medical malpractice).
42. See Arlen and MacLeod, supra note 19, at 1982-1983.
43. See id. at 2003-2004.
44. Infra Section III.B;
45. see also Jennifer Arlen, Private Contractual Alternatives to Malpractice Liability, in MEDICAL MALPRACTICE AND THE U.S. HEALTH CARE SYSTEM 245, 257 (William M. Sage & Rogan Kersh eds., 2006). Renegotiation also undermines patients' ability to use contractual liability to regulate the moral hazard problem.
46. See Arlen & MacLeod, supra note 19, at 2002 (explaining that patients and providers cannot use renegotiable contractual liability to induce physicians to make efficient nonverifiable investments in postcontractual care because providers know that patients will waive liability after providers decide how much to invest);
47. Abraham Wickelgren, The Inefficiency of Contractually-Based Liability with Rational Consumers, 22 J. L. ECON. & ORG. 168, 173-75 (2006) (showing that contracting over products liability is inefficient if consumers are able to waive liability after the producer has invested in quality, because consumers' incentives to waive once investments are fixed will result in producers underinvesting in product quality in anticipation of liability being waived);
48. see also Drew Fudenberg & Jean Tirole, Moral Hazard and Renegotiation in Agency Contracts, 58 ECONOMETRICA 1279, 1280 (1990) (finding that the ability to renegotiate a contract undermines the principal's ability to use incentive contracts to induce agents to invest in effort postcontract).
49. See infra Section III.C (discussing adverse selection).
50. This Article does not separately consider standard form contracting between patients and physicians because this type of contracting would face all the problems identified in Part IV, as well as additional problems. See infra note 123.
51. See Danzon, supranote 1, at 1351.
52. See INST, OF MEDIQNE, To ERR Is HUMAN: BUILDING A SAFER HEALTH SYSTEM 27 (Linda T. Kohn et al. eds., 2000) [hereinafter To ERR IS HUMAN] .
53. Michelle M. Mello et al., Who Pays for Medical Errors? An Analysis of Adverse Event Costs, the Medical Liability System, and Incentives for Patient Safety Improvement, 4 J. EMPIRICAL LEGAL STUD. 835, 847 (2007);
54. see also Chunliu Zhan & Marlene R. Miller, Excess Length of Stay, Charges, and Mortality Attributable to Medical Injuries During Hospitalization, 290 JAMA 1868, 1872 (2003) (finding that hospital errors result in excess national health care costs of almost $5 billion per year).
55. See PAUL C. WEILER ET AL., A MEASURE OF MEDICAL MALPRACTICE: MEDICAL INJURY, MALPRACTICE LITIGATION, AND PATIENT COMPENSATION 42-14, 137-39 (1993) (identifying both inadequate physician expertise and institutional problems in hospitals as leading causes of medical error);
56. Lori Andrews, Studying Medical Error in Situ; Implications for Malpractice Law and Policy, 54 DEPAUL L. REV. 357, 363 (2005) (finding that many medical errors are attributable to individuals, as well as to factors under the control of hospitals);
57. Michelle M. Mello & David M. Studdert, Deconstructing Negligence: The Role of Individual and System Factors in Causing Medical Injuries, 96 GEO. L.J. 599, 620 (2008) (same);
58. infra Section I.A.
59. See Arlen & MacLeod, supra note 3. 30 At one point, people assumed that most medical errors were caused by a few bad doctors. Empirical analyses consistendy refute this "bad apple" view of medical error.
60. See, e.g., John E. Rolph et al., Identifying Malpractice-Prone Physicians, 4 J. EMPIRICAL LEGAL STUD. 125, 150 (2007). Accordingly, even if the medical community were able to police its members adequately, this regulation would not eliminate the primary causes of medical error.
61. See Andrews, supra note 28, at 390 fig.3 (showing that almost 38% of medical errors in a Chicago hospital were at least partially attributable to individual error resulting from poor technical performance, poor judgment, or the failure to obtain or act on information, and that a substantial percentage of mese errors were also caused by inadequate training or supervision);
62. Mello & Studdert, supra note 28, at 606 tbl.1 (finding 48% of medical errors producing a claim were attributable, at least in part, to the physician's lack of technical competence or knowledge and that 57% were attributable to a failure of vigilance or memory);
63. see also ATUL GAWANDE, COMPLICATIONS 197-98 (2002) (describing the findings of three studies where autopsies of patients who died in the hospital revealed that 40% were misdiagnosed and that one-third of these patients would have been expected to live if properly diagnosed and treated) ;
64. JEROME GROOPMAN, HOW DOCTORS THINK 24 (2007) (noting evidence suggesting that 10-15% of physicians' diagnoses are wrong);
65. Elizabem A. McGlynn et al., The Quality of Health Care Delivered to Adults in the United States, 348 NEW ENG. J. MED. 2635, 2641 (2003) (showing that patients on average receive only about 55% of recommended care);
66. Mark A. Schuster et al., How Good Is the Quality of Health Care in the United States?, 76 MILBANK Q. 517, 521 (1998) (finding that, for chronic conditions, only "60 percent [of patients] received recommended care and 20 percent received contraindicated care").
67. See TO ERR IS HUMAN, supra note 26, at 4;
68. WEILER ET AL., supra note 28, at 9-10;
69. Andrews, supra note 28, at 387;
70. Mello & Studdert, supra note 28, at 617.
71. See WEILER ET AL., supra note 28, at 43 (finding that 3.7% of the hospitalized patients studied were injured by the medical treatment they received); infra note 75 (discussing this study). A subsequent study employing onsite observation of hospital error found that almost 18% of hospital patients were the victims of at least one error sufficiently serious to prolong their hospital stay. Many of these errors were not recorded in the hospital's written records.
72. See Andrews, supra note 28, at 362;
73. see also Robert H. Brook et al., Effectiveness of Nonemergency Care Via an Emergency Room, 78 ANNALS INTERNAL MED. 333, 337 (1973) (finding that only 25% of patients seeking care in an emergency room received at least "minimally adequate medical care");
74. Knight Steel et al., Iatrogenic Illness on a General Medical Service at a University Hospital, 304 NEW ENG. J. MED. 638, 639 (1981) (finding that 9% of 815 patients were harmed by medical error that was either life threatening or produced disability).
75. See Mello & Studdert, supra note 28, at 605 (showing that 56% of errors where claims were filed had systemic causes, usually in addition to individual causes);
76. see also Andrews, supra note 28, at 362-363 (finding that hospital practices, including administrative systems, are important contributing causes of medical error).
77. See Mello & Studdert, supra note 28, at 606 tbl.1 (finding that inadequate supervision was a contributing cause of 20% of the medical errors that resulted in claims).
78. Hospitals also can decrease error by investing in health care technology, such as computerized physician order entry system that reduce drug errors. See GAWANDE, supra note 31, at 63. In addition, U.S. hospitals could substantially reduce deadly hospital-induced infections by adopting better procedures akin to those used in Europe, where hospitals have nearly eliminated the risk of hospital-acquired, antibiotic-resistant staphylococcus.
79. See Kevin Sack, Swabs in Hand, Hospital Cuts Deadly Infections, N.Y. TIMES, July 27, 2007, at A1.
80. Physician exhaustion produced by long hospital shifts is a known cause of medical error. See Drew Dawson & Kathryn Reid, Fatigue, Alcohol and Performance Impairment, 388 NATURE 235, 235 (1997) (finding that people kept awake for 24 hours suffered a deficit in cognitive psychomotor performance "equivalent to the performance deficit observed at a blood alcohol concentration of roughly 0.10%");
81. Christopher P. Landrigan et al., Effect of Reducing Interns' Work Hours on Serious Medical Errors in Intensive Care Units, 351 NEW ENG. J. MED. 1838, 1842 (2004) (finding that interns on the traditional extended schedule made 35.9% more "serious medical errors" than did interns on a lighter schedule).
82. See Mello et al., supra note 27, at 847.
83. Id.
84. See id. at 836 (noting the "strong business [rationale]" for investing in safety improvements);
85. see also Danzon, supra note 1, at 1353-54 (describing a study finding that hospitals vary widely in their error rates, with some having a negligent-error rate of only 1% and others having a negligent-error rate of 60%-a disparity that could not be entirely explained by differences in illnesses and patient populations).
86. See Mello et al., supra note 27, at 837-838 (finding that hospitals do not bear the cost of the extra medical care required by medical errors attributable to them).
87. Most patients believe that physicians and hospitals do not differ significantly in the quality of care they provide because they believe that health care regulation has eliminated any significant differences in physician or hospital quality. Jacquelyn J. Jewett &Judith H. Hibbard, Comprehension of Quality Care Indicators: Differences Among Privately Insured, Publicly Insured, and Uninsured, 18 HEALTH CARE FINANCING REV. 75, 90 (1996) ;
88. see also KAISER FAMILY FOUND. & AGENCY FOR HEALTH CARE RESEARCH & QUALITY, NATIONAL SURVEY ON AMERICANS AS HEALTH CARE CONSUMERS: AN UPDATE ON THE ROLE OF QUALITY INFORMATION, at Summary §1 (2000) (finding that only 47% of patients believed that there were big differences in the quality of care between local hospitals);
89. cf. Mark R. Chassin et al., Benefits and Hazards of Reporting Medical Outcomes Publicly, 334 NEW ENG. J. MED. 394, 394-95 (1996) (showing that public reporting of evidence identifying higher-risk physicians and hospitals did not affect patients' willingness to seek services from these providers). Patients' belief that provider quality is constant stands in contrast with empirical evidence that licensed medical providers differ significandy in the quality of care they deliver.
90. See supra text accompanying notes 38-40 (discussing the variation in error rates across providers).
91. Arlen & MacLeod, supra note 3, at 510-512 (showing that, absent liability, physicians underinvest in postcontractual care).
92. See id. at 507-08 (showing that, absent liability, even compassionate physicians have an incentive to underinvest in patient safety); see also Michael Spence, Consumer Misperceptions, Product Failure and Producer Liability, 44 REV. ECON. STUD. 561, 563-564 (1977) (showing that liability is needed to induce producers to invest optimally in product quality if consumers underestimate product risks). Malpractice liability is needed notwithstanding recent efforts to provide patients with better information about providers' quality because current disclosure policies tend to be quite limited. For example, most disclosure policies only inform patients whether a provider (e.g., a hospital) follows specific safety practices and do not disclose patient outcomes.
93. See, e.g., Leslie P. Kernisan et al., Association Between HospitalReported Leapfrog Safe Practices Scores and Inpatient Mortality, 301 JAMA 1341, 1348 (2009) (finding that higher Leapfrog safe practice scores were not associated with significant risk-adjusted decreases in inpatient mortality in the 1075 hospitals that completed Leapfrog's 2006 Safe Practices Survey);
94. see also Meredith B. Rosenthal & R. Adams Dudley, Pay-for-Performance: Will the Latest Payment Trend Improve Care?, 297 JAMA 740, 741 tbl. (2007) (showing that 91% of pay-for-performance "programs target clinical quality measures," while only 37% "include patient satisfaction measures"). This limited disclosure does not provide incentives for providers to invest in safety measures that are not subject to disclosure requirements.
95. See Peter K. Lindenauer et al., Public Reporting and Pay for Performance in Hospital Quality Improvement, 356 NEW ENG. J. MED. 486, 486 (2007) (finding that patient outcomes are better when hospitals have direct financial incentives to avoid poor outcomes man they do when hospitals only voluntarily report information about quality).
96. See Arlen & MacLeod, supra note 3, at 503-04 (showing that physicians will invest optimally in patient safety and MCOs will make optimal decisions regarding utilization review if each is subject to optimal medical malpractice liability);
97. see also Abraham & Weiler, supra note 19, at 410-11 (arguing that hospital enterprise liability could provide effective incentives for hospitals and physicians to reduce medical error);
98. Havighurst, supra note 6, at 27 (claiming that MCO tort liability will improve incentives to take care by shifting focus from cost control to overall efficiency). Empirical evidence suggests that liability and other financial sanctions for bad outcomes can improve the quality of medical care patients receive.
99. See, e.g., Janet Currie & W. Bendey MacLeod, First Do No Harm? Tort Reform and Birth Outcomes, 123 Q.J. ECON. 795, 797 (2008) (finding that caps on noneconomic damages increase complications in labor and delivery, as would be expected if liability deters procedures that increase risk); Lindenauer et al., supra note 44, at 486 (finding that hospitals provide superior quality care when given direct financial incentives to avoid bad outcomes);
100. see also TOM BAKER, THE MEDICAL MALPRACTICE MYTH 108-10 (2005) (explaining how the high malpractice liability costs of the 1980s induced the American Society of Anesthesiologists to adopt reforms that reduced death rates from anesthesia to four per one million patients - lowering malpractice liability insurance rates substantially);
101. Danzon, supra note 1, at 1341 ("The limited empirical evidence of provider response to liability and the deterrent effect of claims suggests - but cannot prove - that the net benefits of the malpractice system may plausibly be positive.");
102. cf. Mello & Brennan, supra note 19, at 1604-1605 (presenting evidence that malpractice liability may improve hospitals' safety records, but noting that the evidence on liability's deterrent effect is mixed).
103. Evidence suggests that malpractice liability is sufficiently accurate to achieve its deterrence goals if properly designed. See, e.g., David M. Studdert et al., Claims, Errors, and Compensation Payments in Medical Malpractice Litigation, 354 NEW ENG. J. MED. 2024, 2028 fig.1 (2006) (providing evidence that 82% (653/798) of the malpractice claims in which patients received a payment involved injuries attributable to medical error).
104. See generally BAKER, supra note 45, at 83-87 (discussing empirical evidence on litigation accuracy);
105. Arlen & MacLeod, supra note 19, at 1933 (examining the optimal scope of physician and MCO negligence liability for medical malpractice);
106. Philip G. Peters, Jr., What We Know About Malpractice Settlements, 92 IOWA L. REV. 1783, 1803-05 (2007) (presenting empirical evidence suggesting that setdement outcomes are driven by the strength of a plaintiff s case);
107. infra note 75 (discounting claims that malpractice liability is random).
108. See Danzon, supra note 1, at 1343 ("Malpractice [liability insurance] premiums account for roughly 1% of total health care spending, hence are not a significant contributor to the level or growth of health care costs.");
109. BAKER, supra note 45, at 40 (same);
110. Paul C. Weiler, The Case for No-Fault Medical Liability, 52 MD. L. REV. 908, 909 (1993) '("[T]he idea that containing medical liability costs will make any appreciable dent in health care costs is absurd."). The amount spent on medical malpractice liability premiums is far less (one-fifth to one-tenth the size) than the amount spent on either workers' compensation or automobile insurance, even though substantially more people are killed by medical malpractice each year than by workplace and automobile accidents combined.
111. See BAKER, supra note 45, at 63.
112. See infra Section IIA (discussing the problems plaguing the current system).
113. See, e.g, Epstein, Contractual Principle, supra note 5, at 505-06;
114. see also Hylton, supra note 5, at 209-210
115. See Epstein, Contractual Principle, supra note 5, at 505 (arguing that malpractice liability reform could be best achieved by allowing patients to contract with providers over liability);
116. see also Danzon, supra note 6, at 517-18 (criticizing state legislative proposals that would extend liability to all MCOs rather than deferring to contract terms);
117. Epstein & Sykes, supra note 6, at 625-27 (arguing against reforms seeking to impose mandatory liability on MCOs on the grounds that any such liability should be contractual);
118. Robinson, supra note 5, at 198-199 (claiming that state-imposed liability may not be necessary if private parties are free to determine liability by contract).
119. See Epstein, Contractual Principle, supra note 5, at 505-506
120. Proposals for states to adopt entity-level liability while permitting patients and providers to contract around it fall into this category. See supra note 6.
121. For a summary of the conventional economic case for contract, see, for example, PAUL WEILER, MEDICAL MALPRACTICE ON TRIAL 96 (1991);
122. Epstein, Contractual Principle, supra note 5, at 507-08;
123. and Robinson, supra note 5, at 183-184
124. Cass Sunstein and Richard Thaler focus their defense of contractual liability on the compensatory role of tort liability and largely ignore deterrence concerns. THALER & SUNSTEIN, supra note 5. Yet one cannot assert, as they do, that patients would be better off if allowed to contract over liability without considering the effect of contractual liability on patients' ability to use liability to induce providers to invest in patient safety. Deterrence is particularly important since leading reforms would make malpractice more effective.
125. See sources cited supra note 45;
126. infra Section IIA;
127. see also Abraham & Weiler, supra note 19, at 410-11 (explaining how hospital liability would create better incentives to reduce medical error);
128. Arlen & MacLeod, supra note 19, at 1993-95 (showing how MCO liability would reduce medical error). Thus, we should not embrace contractual liability unless proponents can show that it would not hurt patients who would benefit from the deterrence that state-imposed malpractice liability could be designed to provide. Beyond this, contractual liability is hard to justify solely on compensation grounds. First, for many patients the expected cost of using liability solely to obtain compensation exceeds the expected benefit. Providers subject to liability must charge an additional fee equal to the expected cost to the provider of bearing liability. Were compensation the only effect of liability, the amount that providers would charge for liability would exceed the benefit of liability for the vast majority of patients. The provider's premium would reflect the expected damage award plus her litigation costs, whereas the patient would receive the damage award net of her litigation costs, producing a total load of more man 50%. Moreover, risk averse patients could insure more cheaply through first-party insurance. Thus, states interested only in compensating injured patients could achieve this goal more effectively by adopting no-fault or social insurance for medical error, because liability takes too long and is too expensive to provide patients with effective insurance.
129. See, e.g. Danzon, supra note 1, at 1369 (observing that 60 cents of each dollar spent on medical malpractice liability insurance is lost to either litigation costs or administrative expenses, and only 40 cents reaches the patient as compensation; by contrast, 90 cents of each dollar spent on first-party insurance is available for compensation);
130. Mark Geistfeld, Should Enterprise Liability Re- place the Rule of Strict Liability for Abnormally Dangerous Activities?, 45 UCLA L. REV. 611, 627-32 (1998) (arguing that third-party insurance provided through the tort system is more expensive than first-party insurance);
131. cf. Studdert et al., supra note 46, at 2026-2027 tbl.1 (finding that median defense-side costs in paid claims were $27,954, which is about 13.5% of the median payment of $206,400).
132. See, e.g., Sanford Grossman, The Informational Role of Warranties and Private Disclosure About Product Quality, 24J.L. & ECON. 461, 463 (1981) (showing that nonnegotiable contracting over warranties can induce producers to invest optimally in product quality, assuming that consumers independendy value liability for insurance reasons);
133. George L. Priest, A Theory of Consumer Product Warranty, 90 YALE L.J. 1297, 1347 (1981) (showing that market forces will pressure producers to offer full warranties, thereby providing incentives for them to invest optimally in postcontractual quality);
134. Spence, supra note 44, at 569-70 (showing that contracting over liability will lead to efficient product markets when producers can use voluntary liability to signal quality);
135. see also Hylton, supra note 5, at 222 (presenting a model of contracting over waivers that implicidy assumes that accidents depend only on an injurer's postcontractual decision to take care to benefit one person).
136. Patients pay these costs either direcdy, in the form of higher medical bills, or indirectly, through higher health insurance premiums.
137. See Priest, supra note 55;
138. Spence, supra note 44;
139. see also Hylton, supra note 5.
140. See sources cited infra note 63.
141. Postcontractual expertise is important because best medical practices change so quickly that physicians cannot reliably select the right treatment without continually investing in expertise. Cf. Annetine C. Gelijns et al., Uncertainty and Technological Change in Medicine, 26 J. HEALTH POL. POL'Y& L. 913, 914 (2001) (noting that "approximately 35% of the 200 largest selling prescription drugs are new each year," and in one year alone the Food and Drug Administration approved approximately 5000 new and modified medical devices).
142. See, e.g., Priest, supra note 55 (describing the investment theory of contractual liability, in which producers contract to bear liability when they are better able than consumers to reduce product risks);
143. see also Hylton, supra note 5 (modeling contractual liability as a solution to producer moral hazard).
144. See supra Section I.A.
145. Providers will not make durable investments that are cost-effective in light of their benefit to all patients if the providers do not benefit from the effect of durable investments on future patients; they will only make investments that are cost-effective given their benefit to current patients.
146. See Spence, supra note 44, at 569-70 (showing that producers can use voluntary nonnegotiable liability to signal product quality, thereby enabling consumers to purchase me quality of product they prefer);
147. see also Grossman, supra note 55, at 474-477 (showing that a monopolist will offer a full warranty, even when consumers cannot observe product quality, in order to avoid the negative price effect of signaling poor quality by offering an incomplete warranty).
148. A related objection to contractual liability is that contracting would be undermined by behavioral biases that may cause patients to act contrary to their own interests. See Baker & Lytton, supra note 10, at 5-12.
149. See, e.g., Epstein & Sykes, supra note 6 (positing that information problems associated with MCO contractual liability are not significant);
150. see also Ramseyer, supra note 11, at 1825 (asserting that claims that consumers are not sufficiently informed to contract over products liability are empirically "questionable");
151. Schwartz, supra note 11, at 380 (concluding that the evidence on consumers' approach to product risks "more strongly supports the view that consumers are informed than the view that they are ignorant");
152. cf. Mark Geistfeld, Note, Imperfect Information, the Pricing Mechanism, and Products Liability, 88 COLUM. L. REV. 1057, 1058 (1988) (discussing how pricing mechanisms for warranties can be modified to reduce information problems).
153. See sources cited supra note 12.
154. See supra note 17.
155. The proposition that we must directly consider whether contractual malpractice liability is efficient, rather than relying on analyses of products liability, is strengthened by economic analyses of health care that have identified numerous inefficiencies plaguing other types of health care contracts. See generally Joseph P. Newhouse, The Economics of Health Insurance (finding that contracting in health care markets is distorted by moral hazard, adverse selection, and other problems), in 3 THE NEW PALGRAVE DICTIONARY OF ECONOMICS 872-875 (Steven N. Durlauf & Lawrence E. Blume eds., 2d ed. 2008). As a result of these inefficiencies, certain health care contracts, specifically health insurance policies, are regulated. Accordingly, before embracing contractual liability it is important to determine whether the inefficiencies that plague other health care contracts also would afflict malpractice liability contracts.
156. See supra note 24 and accompanying text.
157. This Article assumes for argument's sake that consumers are informed about the costs and benefits of imposing liability in order to evaluate contractual liability in the most favorable light. Nevertheless, patients may not be sufficiently informed to contract effectively. See, e.g., Arlen, supra note 22, at 253-54, 263-64 (discussing information problems afflicting contractual malpractice liability);
158. see also Mark Geistfeld, The Political Economy of Neocontractual Proposals for Products Liability Reform, 72 TEX. L. REV. 803 (1994) (same);
159. Neil D. Weinstein, Unrealistic Optimism About Susceptibility to Health Problems: Conclusions from a Community-Wide Sample, 10 J. BEHAV. MED. 481, 49496 (1987) (finding that people underestimate the probability that they will fall ill, and thus underestimate the probability that they will need medical care).
160. See, e.g., Epstein, Contractual Principle, supra note 5, at 509 (asserting that legislative solutions to the liability problem are inferior to solutions that allow patients to contract over liability). The separate claim that patients would benefit from the ability to contract over the standard of care is considered in Part V.
161. This Article focuses on negotiable individual contracting because it is the most choice-enhancing form of contracting. Also, standard form contracting with individual providers would be rendered inefficient by information costs, search costs, and medical duress if not publicly disclosed in advance of the patient seeking care. Should providers disclose their standard form contracts on the web, then individual contracting would be plagued by adverse selection, in addition to the problems discussed above. See discussion infra note 123.
162. This Article focuses on medical-entity liability because this reform has garnered broad support from medical malpractice liability experts. See sources cited supra notes 6, 19.
163. This Article shows that contracting over liability would dramatically reduce the amount of malpractice liability, but only by hurting patients who would have benefited from state-imposed liability.
164. See Abraham & Weiler, supra note 19, at 408;
165. Mello & Brennan, supra note 19, at 1616-17. Another problem plaguing malpractice liability is that neither the standard of care nor damages are set optimally.
166. See Jennifer Arien, Tort Damages (showing that under current legal rules governing damages for serious personal injury and death, injurers' expected damages are too low to induce them to take due care and engage in optimal activity levels), in 2 ENCYCLOPEDIA OF LAWAND ECONOMICS 682, 718 (Boudewijn Bouckaert & Gerrit De Geest eds., 2000);
167. Danzon, supra note 1, at 1347-48 (arguing that the current practice of predicting negligence determinations based on medical custom biases the standard of care in favor of excessive care when the custom was established by physicians treating patients with fee-for-service health insurance);
168. Paul C. Weiler, Reforming Medical Malpractice in a Radically Moderate - and Ethical - Fashion, 54 DEPAUL L. REV. 205, 217-218 (2005) (advocating for damage reforms that include a schedule for noneconomic damages). By contrast, the evidence does not support the oft-repeated claim that malpractice liability is random, falling on physicians regardless of whether they erred. Instead, studies find that the vast majority of malpractice claimants who receive payments were injured by medical error.
169. See supra note 46. Scholars who argue that malpractice liability is random generally cite the Harvard Medical Practice Study (HMPS) to support this claim. Yet the HMPS cannot be used to assess litigation accuracy because it included too few malpractice lawsuits to produce any statistically significant results about the quality of litigation outcomes.
170. Patricia M. Danzon, Medical Malpractice, in 2 THE NEW PALGRAVE DICTIONARY OF ECONOMICS AND THE LAW 624, 626 (Peter Newman ed., 1998). In addition, the HMPS should not be used to analyze litigation accuracy because it was designed to provide a conservative estimate of medical error and not to assess litigation accuracy. It thus employed a protocol for identifying medical negligence that underestimates the amount of negligence. First, the study relied solely on written hospital records and research, which often fail to document medical errors.
171. See Andrews, supra note 28, at 370 (finding that more than 60% of the patients seriously injured by medical error were "not brought to the hospital's attention" through either written occurrence reports or potential claims files). Second, the HMPS coded the injury as not attributable to error if either one of the two physician reviewers concluded that no error occurred, even if the other reviewer concluded that the patient's injury was caused by medical error.
172. See BAKER, supra note 45, at 28-29 (discussing the HMPS);
173. cf. Michelle M. Mello & Kadiryn Zeiler, Empirical Health Law Scholarship: The State of the Field, 96 GEO. L.J. 649, 682-84 (2008) (concluding that the debate over malpractice liability reform has not been adequately informed by controlled empirical studies, but has instead been overly influenced by preconceptions supported by uncontrolled descriptive studies).
174. Indeed, on average a provider who errs is sued only 10% to 12% of the time. PATRICIA M. DANZON, MEDICAL MALPRACTICE: THEORY, EVIDENCE, AND PUBLIC POLICY 23-24 (1985) (describing a study showing that only 10% of patients injured by a negligent adverse event filed a medical malpractice case);
175. see also Andrews, supra note 28, at 370 (finding that although 185 patients in a Chicago hospital (17.7%) were the victims of a medical error producing a serious injury, only thirteen patients (1.2%) brought a claim);
176. Danzon, supra note 1, at 1354 (noting that in the HMPS, less than 15% of negligent injuries produced a malpractice claim);
177. David M. Studdert et al., Negligent Care and Malpractice Claiming Behavior in Utah and Colorado, 38 MED. CARE 250, 255 (2000) (finding that 97% of the patients who suffered negligent injury did not sue).
178. Studdert et al., supra note 46, at 2028 (finding that 27% of the victims of medical error who filed claims did not recover);
179. see also Lee D. Cranberg et al., Do the Claims Hold Up? A Study of Medical Negligence Claims Against Neurologists, 4 J. EMPIRICAL LEGAL STUD. 155, 161 (2007) (finding that a majority of the patients who filed valid claims against neurologists did not recover);
180. David M. Studdert Sc Michelle M. Mello, When Tort Resolutions Are "Wrong": Predictors of Discordant Outcomes in Medical Malpractice Litigation, 36 J. LEGAL STUD. 547, 565 (2007) (finding that patients with valid claims are particularly likely to be denied recovery if they proceed to trial).
181. David A. Hyman et al., Do Defendants Pay What Juries Award? Post-Verdict Haircuts in Texas Medical Malpractice Cases, 1988-2003, 4 J. EMPIRICAL LEGAL STUD. 3, 7 (2007);
182. Kathryn Zeiler et al., Physicians' Insurance Limits and Malpractice Payments: Evidence from Texas Closed Claims, 1990-2003, 36 J. LEGAL STUD. S9, S10-11 (2007).
183. See Mark Geistfeld, Malpractice Insurance and the (Il)Legitimate Interests of the Medical Profession in Tort Reform, 54 DEPAUL L. REV. 439, 444 (2005) (concluding that malpractice liability premiums are lower than they would be if physicians paid the total cost of all medical negligence);
184. supra note 78.
185. Individual physician malpractice liability insurance premiums reflect a physician's specialty and time in practice and location, but do not adjust to reflect a provider's individual claims experience. See Danzon, supra note 1, at 1360-62;
186. Mello & Brennan, supra note 19, at 1616. By contrast, hospital malpractice liability insurance is experience-rated.
187. See generally Geistfeld, supra note 79, at 444 (noting that individual physician malpractice insurance premiums are not experience-rated);
188. William M. Sage, Medical Malpractice Insurance and the Emperor's Clothes, 54 DEPAUL L. REV. 463, 473 (2005) (noting that physicians are charged premiums based on average loss experience in their specialty, not individual experience).
189. See Mello & Brennan, supra note 19, at 1620-1621
190. See Aetna Health Inc. v. Dávila, 542 U.S. 200, 221 (2004) (holding that ERISA preempts patients' state tort actions against their health insurers for alleged failures to exercise reasonable care in handling insurance coverage decisions). For a justification of MCO liability,
191. see Arlen & MacLeod, supra note 19, at 1942-1944
192. Hospitals often escape liability for negligence by hiring surgeons, anesthesiologists, and radiologists as independent contractors. MCOs also generally hire physicians as independent contractors. See Arlen & MacLeod, supra note 19, at 1990 (arguing that optimal tort liability includes MCO liability for physician negligence).
193. See generally Jennifer H. Arlen & W. Bendey MacLeod, Beyond Master-Servant: A Critique of Vicarious Liability (analyzing how the independent contractor rule distorts principle-agent relationships in ways that create excessive risk), in EXPLORING TORT LAW 111, 117-140 (M. Stuart Madden ed., 2005).
194. See sources cited supra note 19;
195. cf. Stephen M. Shortell & Sara J. Singer, Improving Patient Safety by Taking Systems Seriously, 299 JAMA 446-447 (2008) (arguing that health care organizations should be given financial incentives to protect patient safety).
196. See sources cited supra note 6.
197. MCOs and hospitals would be particularly responsive to malpractice liability because, unlike physicians, these entities either self-insure or purchase liability insurance that is experience-rated. See supra note 80 and accompanying text (discussing experience-rating);
198. see also Abraham & Weiler, supra note 19, at 410 (observing that one advantage of hospital liability is that insurers can charge medical entities liability insurance premiums that "reliably reflect the prior claims experience of the particular institution");
199. Arlen & MacLeod, supra note 19, at 1990 n.205 (observing that one advantage of MCO liability is that MCOs would either self-insure or purchase experiencerated malpractice liability insurance).
200. Entities could use their monitoring information to induce physicians to reduce their error rates more effectively than is possible through the use of individual liability alone because entities possess direct evidence about providers' per-patient outcomes (and direct evidence on error) even when no claim is filed. Thus, entities could sanction providers who deliver inadequate care even if the patient did not file a claim. Moreover, entities can sanction physicians in ways not insulated by physician liability insurance. Furthermore, entities could facilitate experience-rating of individual liability insurance by sharing their information about physicians' patient-health-adjusted outcomes. Finally, entity-level liability would reduce the underclaiming problem because patients injured by medical care in a hospital would no longer be deterred from claiming by the difficulty of having to identify the individual responsible for their injuries See Arlen & MacLeod, supra note 19, at 1990-1995 (discussing the benefits of MCO liability);
201. see also Abraham & Weiler, supra note 19, at 398-400, 407-14 (explaining why hospital enterprise liability is a more effective mechanism for medical error than individual physician liability);
202. Mello & Brennan, supra note 19, at 1626 (asserting that one advantage of channeling liability costs through hospitals is that it facilitates experiencerating of liability insurance).
203. The expected liability imposed by any individual patient is not sufficiently high to justify a hospital making substantial expenditures based on liability to one patient. This is true even though the median award for valid claims is about $200,000. Studdert et al., supra note 46, at 2027. A profit-maximizing entity weighs the cost of a safety investment against its expected benefit, which depends on its expected potential liability to the patient This amount is the per-patient expected damage award if the patient is injured, sued, and recovers, adjusted to reflect the probability that (i) the patient imposing liability will be the victim of medical error, (ii) the injury will be serious, (iii) the victim will sue, and (iv) the plaintiff will be successful. If we make these adjustments, a median award of about $200,000 translates into an expected liability cost of only $328, assuming that 9% of patients are the victims of medical error,
204. see discussion supra note 33, 25% produce serious injuries, and that only 10% of these serious injuries result in malpractice claims,
205. see discussion supra note 76, of which only 73% lead to a victory for the patient,
206. see Studdert et al., supra note 46, at 2028. Similarly, the average damage award of $485,000 translates into a per-patient expected cost of liability of only $796. These costs are too low to induce a medical entity to make the substantial expenditures in administrative practices, personnel, and equipment purchases needed to reduce medical error based on the threat of liability to a single patient.
207. Hospitals benefit all of their patients when they reform administrative practices to improve hand-off procedures, the supervision of residents and interns, and the monitoring of physicians. Cf. infra note 96 and accompanying text (discussing empirical evidence that providers standardize care across patients).
208. To the extent that hospitals standardize administrative systems and other safety measures, see infra note 96, then all hospital patients are governed by the same administrative systems and safety measures, whemer they impose liability or not. Given mis, individual patients are better off not imposing liability because patients who impose liability must pay more for the same quality of care than patients who do not impose liability. Patients would pay more if they imposed liability because hospitals would charge them for the expected cost to the hospital of the patients' right to receive compensation for injuries resulting from medical negligence. But most patients do not want to pay for the right to obtain compensation through the tort system. First, the amount they must pay for this right - which is based on the damages that providers must pay plus their litigation costs - exceeds its expected benefit, which is the damage award minus plaintiff-side litigation costs (and exclusive of those of the defendant). In addition, patients can often obtain insurance against many losses resulting from medical negligence at lower cost by purchasing first-party insurance.
209. See supra note 54 and accompanying text.
210. The finding that the deterrence benefit of contractual liability is less than the deterrence benefit of state-imposed liability establishes that contractual liability is not an efficient reform mechanism even if patients also value the right to compensation. Patients only have efficient incentives to contract into liability if they can replicate the benefit of optimal state-imposed liability by contract. This necessarily includes the deterrence benefits of liability. Moreover, patients would be unlikely to value contractual liability solely as a mechanism for obtaining insurance against the risk of medical negligence because malpractice liability is not an effective mechanism for compensating patients. See supra note 54 and accompanying text.
211. See supra Section I A.
212. See supra note 54 and accompanying text;
213. see also supra note 91 and accompanying text This Article focuses on inefficiencies that arise from the structural differences between contractual and state-imposed malpractice liability. In addition, patients will not have optimal incentives to contract into liability with individual providers if they are insured against some of the losses arising from medical error, since insurance reduces the ex post (post-insurance) cost to them of error.
214. See Geistfeld, supra note 70, at 830-31 (arguing that consumers will not contract efficiendy over liability for product defects if they are insured against their losses, assuming that insurance premiums do not adjust to reflect the risk of the products that each consumer purchases);
215. see also Jon D. Hanson & Kyle D. Logue, The First-Party Insurance Externality: An Economic Justification for Enterprise Liability, 76 CORNELL L. REV. 129, 132, 148-53 (1990) (observing that first-party insurance can blunt potential victims' incentives to invest in deterrence since insurers do not classify insureds according to the riskiness of their product purchases);
216. see also discussion infra note 123.
217. See supra Section I.A.
218. Stuthes of medical error reveal that patients particularly benefit from liability when it induces medical providers to invest more in expertise, administrative systems, and health information technology pre- and postcontract, since deficient expertise and poor systems are leading causes of medical error. See supra text accompanying notes 3040 (discussing the causes of medical error).
219. For example, once a physician develops expertise in a particular illness or treatment choice, she usually uses this expertise to benefit all affected patients. She will not knowingly deny some patients the right to that expertise because they have not paid for premium care. Similarly, physicians who develop administrative systems or surgical procedures to reduce error generally apply them to all relevant patients and do not limit them to a select few. See Richard G. Frank & Richard Zeckhauser, CustomMade Versus Ready-to-Wear Treatments: Behavioral Propensities in Physicians' Choices, 26 J. HEALTH EGON. 1101, 1102 (2007) (finding that physicians tend to follow norms ramer than customizing care for individual patients). The claim that providers tend to standardize collective aspects of care - such as expertise, treatment assessments, and systems - even when patients have paid for care of differing quality - also is supported by evidence that the quality of care a patient receives from her physician depends not only on whether that patient herself has decided to insure through an MCO (which will place pressure on the physician to reduce costs) but also on whether a high proportion of the physician's other patients are insured through an MCO.
220. Sherry Glied & Joshua G. Zivin, How Do Doctors Behave When Some (But Not All) of Their Patients Are in Managed Care?, 21J. HEALTH ECON. 337, 352-53 (2002) (concluding that the quality of care a patient receives from her physician depends not only on whether the patient is enrolled in an MCO but also on whether the physician's other patients are predominately managed care patients);
221. see also Paul A. Heidenreich et al., The Relation Between Managed Care Market Share and the Treatment of Elderly Fee-for-Service Patients with Myocardial Infarction, 112 AM. J. MED. 176, 179-80 (2002) (providing evidence that a patient's expected outcome from treatment by a particular physician depends on the portion of the patients enrolled in managed care in the local market and not just on that patient's choice of insurer).
222. See Arlen & MacLeod, supra note 19, at 2003 (finding that contracting over liability to govern physician expertise is not efficient because expertise is a collective good). The interdependence of patients' liability choices is heightened by the fact that many important medical care investments are "lumpy" - the provider faces a discrete choice between making a substantial investment in a particular type of care or no investment at all. For example, a hospital seeking to protect patients by purchasing and implementing a physician-order-entry system (to computerize prescriptions) must incur the full cost of purchasing the system, installing it, and implementing it. It cannot purchase a marginal (one dollar) increment of the system. Similarly, it is difficult for a hospital to change policies regarding supervision or patient hand-off procedures in marginal (one dollar) in- crements. Hospitals generally need to change procedures in a uniform way across an entire ward (and often the entire hospital), which entails a substantial expenditure.
223. See supra note 96 and accompanying text (discussing the standardization of health care practices). The magnitude of the costs involved in the hospital reforms needed to protect patient welfare are sufficiently great that no one patient would impose enough expected liability to induce these expenditures because a provider's expected liability to one patient is not particularly high.
224. See supra note 88 and accompanying text. Liability only induces such investments if the physician faces liability to a sufficiently large number of patients to justify a substantial investment in error reduction.
225. Patients waiving liability would lower their expected costs because providers would charge higher prices to patients imposing liability than to those who do not, even if the quality of care is the same, because providers serving liability-imposing patients must recapture the expected cost of the liability-imposing patients' right to compensation. Given this, patients would benefit from waiving contractual liability, because the premium providers must charge in order to accept liability exceeds the compensation patients generally expect to receive. Providers will charge a premium for liability that reflects their expected gross damage payments plus their expected defense costs, whereas patients will receive compensation net of bom their own litigation costs and defense-side costs. See supra note 54 and accompanying text (explaining mis point and noting that patients can more cheaply insure through first-party insurance than through liability);
226. see also Weiler, supra note 47, at 926 (observing that "malpractice litigation uses up approximately 55% to 60% of the claims dollar - not even counting the business costs incurred by all forms of private insurance").
227. We can demonstrate this claim through the following example. Assume that the state imposed optimal liability reform. Asume there are 100 patients. Asume that if liability is imposed for all patients, each physician would spend $3000 to adopt optimal care, which would reduce each patient's risk of suffering a $1000 injury due to medical negligence from 1/10 to 1/100. Asume litigation costs are $300 per claim. The total net social benefit of using liability to reduce the risk of error across all patients is [(1/10)(1000)-(1/100)(1000)]100-3000-(1/100) (300) (100) = 5700, which yields a benefit of $57 per patient. (Figure Presented) Notwithstanding the fact that each patient is better off when liability is imposed, each patient offered the right to contract out of liability will do so because she gains nothing from her individual decision to impose liability. To see this, assume the patient believes that the other 99 patients will impose liability. In this case, the threat of liability to these other patients is sufficient to induce the provider to invest $3000 in care. This will lower the risk of injury to the remaining contracting patient to 1/100, whether or not she imposes liability. Given this, the patient will waive liability because doing so enables her to obtain care at a lower cost If the patient imposes liability, the provider will charge her an extra $10 to reflect the providers' expected payments to the patient. Yet the patient only expects to receive $7 in expected compensation, because she must pay expected attorney's fees of $3 (300(1/100) = 3). By contrast, if a patient assumes that none of the others will impose liability, then she also will not because imposing expected liability on the physician of $100[(1/10)(1,000)] will not induce the provider to invest $3000 in care. Of course, if each patient reasons the same way, then they will all waive liability and the physician will not invest $3000 in care - to the detriment of every patient.
228. See Alen & MacLeod, supra note 19, at 2003-2004. The free-rider problem is particularly great when liability is needed to induce providers to make substantial, discrete, collective investments that cannot be undertaken incrementally, such as health care technology, improvements in administrative systems or support staff, and substantial investments in expertise. In this situation, each patient knows that her medical provider will make the investment only if she faces liability for a substantial number of her patients. If she does face liability, she will make the investment regardless of what any particular patient does. If she does not, she will not make it, regardless of what any one patient does. Given this, each patient will rationally waive liability because she knows that her decision will have no effect on the provider's expected investment in discrete care.
229. See supra note 54 and accompanying text.
230. Individual providers could not eliminate mese inefficiencies by using standard form contracts because standard form contracting with individual providers would be plagued by a host of additional problems. See infra Section IVA & note 123 (discussing the numerous problems associated with standard form contracting between individual providers and patients at the point of service).
231. See supra subsection I.C.2. When signaling through voluntary liability enables consumers to distinguish high-quality providers from low-quality ones, then voluntary liability also provides medical caregivers with an incentive to become high quality by investing in precontractual quality.
232. See supra text accompanying note 62.
233. In the classic signaling model, high-quality physicians are able to use liability to signal quality because their lower risk of error translates into lower expected liability costs. Thus, they can offer to bear liability at a lower price than can low-quality physicians. This result does not hold if physicians obtain malpractice liability insurance which is not experience-rated, however, because now low- and high-quality physicians bear the same expected liability costs in the form of standardized premiums.
234. Moreover, the patient's expectation about the provider's quality is also fixed the patient having obtained all available information about the provider's quality from her offer to bear liability.
235. The conventional result that contracting over liability will result in a "separating equilibrium" in which only high quality producers offer to bear liability - thereby signaling their superior quality - is based on a model in which producers make take-itor-leave-it liability offers. In addition, in these models, consumers are risk averse and bear no costs if they enforce their right to compensation. Given this, in these models, consumers value warranties bodi for the signal they provide about the producer and for the insurance against the risk of product defects, and thus they accept liability. See Grossman, supra note 55, at 474-77 (showing that a monopolist will offer a full warranty using a model in which consumers are risk averse and warranty offers are not renegotiable);
236. Spence, supra note 44, at 569-570 (showing that producers can use voluntary contracting over liability to signal quality when consumers are risk averse, injuries involve purely monetary losses, and other conditions are met).
237. See, e.g., Danzon, supra note 1, at 1369 (observing that the load burden on firstparty insurance is only 10 cents of every premium dollar whereas the load burden on malpractice liability insurance is 60 cents of every premium dollar);
238. see also Richard A. Epstein, Products Liability as an Insurance Market, 14 J. LEGAL STUD. 645, 668-669 (1985) (arguing that products liability does not confer a net benefit when viewed purely as an insurance system).
239. Consider two physicians negotiating over liability with a risk neutral patient: a low-quality physician who has an error risk of 0.1 and a high-quality physician with an error risk of 0.01 as a result of unobservable precontractual investments that increased her marginal treatment costs by $30 per patient. The patient suffers a $1000 loss if her physician errs. Litigation costs are $300. Asume low-quality care is cosdess. The patient would prefer to be treated by the high-quality physician. In return for $30 in additional cost, she obtains $90 in expected benefit resulting from a reduction in her expected losses from (0.1)1000 to (0.01)1000. This produces an expected net benefit of $60 - which translates into a net benefit of $57 if liability is used to signal quality. (Figure Presented) The high-quality provider cannot use liability to signal quality if the patient can negotiate for a waiver after selecting the physician because patients who believe the signal will negotiate to eliminate liability. Once a patient has selected the physician, expected quality is fixed. At this point, liability only operates as a mechanism for compensating the patient. Given this, each patient should ask the doctor to waive liability in return for a price reduction, because the amount the physician must charge her for the right to compensation ($10) exceeds her expected recovery, which is only $7 after litigation costs. The patient could leave both parties better off by, for example, offering to waive liability in return for a price reduction of $9. Ex post, the high-quality physician will accept this offer. The expectation of renegotiation precludes the highquality physician from using liability to signal quality, however, because low-quality providers also will offer to bear liability, confident in the expectation that patients who select them believing the signal will ask to waive liability.
240. Alen, supra note 75. Negotiation also undermines the use of contractual liability to regulate the moral-hazard problem.
241. See Alen & MacLeod, supra note 19, at 200203 (explaining how renegotiation undermines the ability of patients and physicians to use contractual liability to address moral hazard because providers know that patients have an incentive to waive after providers invest in care);
242. Abraham L. Wickelgren, The Inefficiency of Contractually-Based Liability with Rational Consumers, 22 J.L. ECON. & ORG. 168, 170 (2006) (showing that contracting over products liability by homogenous consumers is not efficient if consumers can waive liability after producers select product quality);
243. see also Drew Fudenberg & Jean Tirole, Moral Hazard and Renegotiation in Agency Contracts, 58 ECONOMETRICA 1279, 1280-81 (1990) (showing that principals cannot use incentive contracts to solve the moral hazard problem if they can renegotiate the contract after a risk-averse agent has invested in postcontractual effort because, once effort is fixed, both the principal and the risk-averse agent benefit from renegotiating the contract to eliminate the agent's risk of being penalized for bad outcomes in return for the principal paying a lower salary).
244. Some proponents of contractual liability recognize that state-imposed liability pools customers into a single insurance pool, a pooling that would be undone by contractual liability. Proponents assume, however, that using contractual liability to undo this pooling would necessarily increase consumers' welfare. See, e.g., RUBIN, supra note 6, at 40-44;
245. Epstein, supra note 107, at 652. Yet economists have shown that both lowand high-cost patients may be better off with pooling because pooling reduces the cost of obtaining the benefits of liability even to low-cost patients.
246. See generally Charles Wilson, A Model of Insurance Markets with Incomplete Information, 16 J. ECON. THEORY 167, 200-02 (1977) (showing that, when adverse selection is present, total social welfare may be higher when insurance is mandatory and pooled);
247. infra Section IV.B (arguing that MCO contractual liability would be harmful to consumers because it would increase the adverse selection problem).
248. Moreover, patients with complex conditions may impose additional costs on providers to the extent that treating them requires additional effort and resources for which the provider is not fully compensated by insurance.
249. In other words, contracting over malpractice liability would be distorted by a problem economists call "adverse selection," which can occur when the cost or benefit to one party to a market transaction depends on an unobservable characteristic of the other party. George A. Akerlof, The Market for "Lemons": Quality Uncertainty and the Market Mechanism, 84 Q.J. ECON. 488, 490-91 (1970). In this situation, contracts will deviate from optimality as a result of some people's efforts to signal that they are a good type by bearing costs that other (less desirable) people would find too costly to bear.
250. See Michael Rothschild & Joseph Stiglitz, Equilibrium in Competitive Insurance Markets: An Essay on the Economics of Imperfect Information, 90 Q.J. ECON. 629, 648 (1976). For a useful survey of the literature on adverse selection,
251. see Charles Wilson, Adverse Selection, in 1 THE NEW PALGRAVE DICTIONARY OF ECONOMICS, supra note 63, at 25, 25-31,
252. and Joseph P. Newhouse, The Economics of Health Insurance, in 3 THE NEW PALGRAVE DICTIONARY OF ECONOMICS, supra note 63, at 872, 872-875 This Section only briefly discusses adverse selection. Section IV.B discusses it more thoroughly.
253. This is true so long as the expected cost of offering liability to the patients likely to contract into it exceeds the average per-patient cost of offering liability to the population at large.
254. Moreover, to the extent that lower-cost patients elect to waive liability, this would only increase the price charged for liability, leading more patients to waive lia- bility even though the benefit to them of liability exceeds the cost to the provider of bearing it.
255. See also infra text accompanying note 127 (discussing how pooling can redress an incompleteness in insurance markets that precludes healthy patients from purchasing insurance today to protect against the risk of becoming ill in the future).
256. See infra note 123 and accompanying text (explaining why proponents favor MCO standard form contracting over standard form contracting with individual physicians).
257. In considering contracting through medical entities, this Aticle focuses on contracting with MCOs because MCO contracting is superior to contracting with either hospitals or large physician groups because patients are better to coordinate care, financing, and incentives across their various medical providers when they contract with MCOs than when contracting individually with hospitals or provider groups. See, e.g., HAVIGHURST, supra note 6, at 171-73 (arguing that enterprise contractual liability imposed on MCOs is superior to hospital liability for medical error);
258. Alen & MacLeod, supra note 19, at 1995-1996 n.221 (asserting that MCOs are the superior entities to bear entity-level liability);
259. Danzón, supra note 1, at 1382 (concluding that contracting through managed care plans is superior to patient contracting with individual physicians);
260. Sherry Glied, Managed Care (hypothesizing that managed care plans may be able to effectively improve performance and information provisions), in 1 HANDBOOK OF HEALTH ECONOMICS supra note 1, at 707, 725 (Anthony J. Culyer & Joseph P. Newhouse eds., 2000);
261. Havighurst, supra note 6, at 8-9 (arguing in favor of MCO contractual liability); Sage, supra note 6, at 163-64 (concluding that MCO liability is superior to hospital liability for medical error).
262. But see Abraham & Weiler, supra note 19, at 39394 (arguing for enterprise liability for hospitals).
263. But cf. Joseph P. Newhouse, Reimbursing Health Plans and Health Providers: Efficiency in Production Versus Selection, 34 J. ECON. LITERATURE 1236, 1240 tbl.1 (1996) (observing that 48% of employees provided with insurance by their employers are only offered one health plan).
264. This form of MCO liability differs from that preferred by proponents of MCO contracting who advocate unfettered contracting between MCOs and patients over liability. This Section examines collective contracting with commitment because unfettered MCO contracting would be subject to the collective action and time-inconsistency problems identified in Part III. To evaluate MCO liability in the most favorable light, this Aticle focuses on the form of MCO contracting that avoids these problems.
265. Some have argued that MCO contracting ameliorates information problems because health insurance contracts are the product of negotiations between informed employers and insurers. See, e.g., Epstein & Sykes, supra note 6, at 647-48. There are two problems with this argument First many employers, especially small employers, may not be fully informed about the benefits of an effective malpractice liability regime. Second, even if informed, employers will not use their superior information to select the liability provisions that maximize their employees' welfare unless their employees also know the costs and benefits of liability. When employees are uninformed, employers will select the clauses that their uninformed employees think benefit them, because these clauses will be valued on labor markets. If employees care about price but underestimate the effect of liability on quality, employers will negotiate to reduce both quality and cost, even if the cost savings are less than the cost to the employees of the resulting reduction in the quality of care.
266. See Sherry A Glied, The Employer-Based Health Insurance System: Mistake or Cornerstone? (explaining that employers generally mimic the behavior of most workers when they select health plans, and thus place a lot of weight on price considerations and pay less attention to quality), in POLICY CHALLENGES IN MODERN HEALTH CARE 37, 46-47 (David Mechanic et al. eds., 2005);
267. see also David Dranove & Mark A. Satterthwaite, Monopolistic Competition When Price and Quality Are Imperfectly Observable, 23 RANDJ. ECON. 518, 519-26 (1992) (showing that competition in markets may lower consumer welfare in situations where consumers can easily compare producers' prices but cannot easily assess quality differences; this will lead producers to focus on lowering price even if it means lowering quality below optimal levels). 121 Should individual providers publicly disclose their liability clauses (e.g., on the internet), the search-cost problem would be reduced, but only by introducing adverse selection. Disclosure of providers' liability clauses would exacerbate adverse selection because MCOs contracting with providers would design their insurance plans and contracts based on an expectation that liability providers who offer their patients liability contracts will attract higher-cost patients. MCOs would put liability-assuming providers into one insurance plan and no-liability providers into another, and price these plans accordingly.
268. Indeed, to the extent that reformed malpractice liability operates effectively to induce hospitals to make systemic investments in care, the patients most likely to benefit are those who expect to be hospitalized during the contract.
269. MCO contracting over liability is superior to individual provider contracting liability for other reasons. First, individual providers cannot easily give insured patients adequate financial incentives to waive liability because insured patients do not pay providers directly for the cost of the care they receive, and thus may not obtain the full benefit of any price breaks offered by physicians. MCOs, by contrast, can ensure that patients benefit when they waive liability by lowering the price for no-liability insurance. HAVIGHURST, supra note 6, at 171. Standard form contracting over liability also is better accomplished through MCOs because individual-provider standard form contracting generally is rendered inefficient by imperfect information, search costs, and bounded rationality. Individual physicians do not face adequate incentives to offer optimal liability terms when contracting with patients seeking immediate medical care. Even when patients would prefer and be willing to pay to retain liability, physicians can reduce costs by offering standard form contracts that include liability waivers. Physicians offering such terms will not lose many patients because patients seeking medical care are unlikely to know the waiver clause exists, as almost no one reads standard form contracts.
270. See generally Yannis Bakos, Florencia Marotta-Wurgler & David R. Trossen, Does Anyone Read the Fine Print? Testing a Law and Economics Approach to Standard Form Contracts 3 (N.Y. Univ. Law & Econ. Research Paper Series, Paper No. 09-40, 2009), available at http://lsr.nellco.org/cgi/viewcontent.cgi? article= 1199&context=nyu-lewp (providing evidence that only a small percentage of consumers read standard form contract terms);
271. see also Danzon, supra note 1, at 1382 (noting that patients actively seeking medical care often are in no condition to consider properly the issue of contractual liability). Moreover, those who do read their contracts often will continue to seek treatment - even if the liability clause is welfare reducing because it would be too costly for them to delay care in the hope of finding a provider with a contract they prefer. Patients seeking immediate necessary medical treatment are especially unlikely to search for a new provider because the cost of delaying care in the hope of finding a provider with better terms is likely to exceed the expected benefits. Search costs are particularly high when patients cannot easily learn alternative providers' liability clauses in advance or would face a delay in obtaining care from another provider. The expected benefits of searching also may be low for patients whose choice of alternative providers is limited by their insurance or by the dearth of qualified providers in their geographic area.
272. See generally Alan Schwartz & Louis L. Wilde, Imperfect Information in Markets for Contract Terms: The Examples of Warranties and Security Interests, 69 VA. L. REV. 1387, 1409-15 (1983). MCO contractual liability is less afflicted by information problems and medical duress because MCOs would incorporate liability clauses into the health insurance contracts they present to their subscribers each year. Patients normally enter into such contracts when they are not under medical duress. Patients also usually have more time to deliberate over health insurance plans than over contracts presented to them by medical providers when they seek treatment. Moreover, employers and unions often bring important clauses to their employees' attention during the open-enrollment process.
273. See Danzon, supra note 1, at 1382;
274. see also HAMGHURST, supra note 6 at 171;
275. Havighurst, supra note 6, at 16.
276. In other words, MCOs would need to insist that any provider who contracts to assume liability for its subscribers is also subject to liability to the subscribers of other insurers.
277. Most privately insured patients are insured through an entity that operates as an MCO. Nevertheless, in considering the costs and benefits of MCO contractual liability, it is important to recognize that more than 17% of the nonelderly population (under age sixty-five), and 15% of the entire U.S. population (including the elderly) are uninsured. Paul Fronstin, Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 2009 Current Population Survey, 344 EBRI ISSUE BRIEF 1, 4 (2009), available at www.ebri.org/pdf/briefspdf/ebri-13-9-2009-No334-HI-Cvgl.pdf (providing insurance figures for 2008). In states such as Florida, Texas and New Mexico, more than 20% of the population is uninsured. The Federalist Prescription, ECONOMIST, Jan. 13, 2007, at 28. Beyond this, more than 19% of the nonelderly population and 29% of the total U.S. population (including the elderly) are covered by public insurance.
278. Fronstin, supra, at 4.
279. See Marcel Kahan & Michael Klausner, Standardization and Innovation in Corporate Contracting (Or "The Economics of Boilerplate"), 83 VA. L. REV 713, 729-35 (1997) (discussing how learning and network externalities can lead contracting parties to adopt contracts with inefficient terms).
280. See generally Lewis Kornhauser, Comment, Unconscionability in Standard Forms, 64 CAL. L. REV 1151, 1183 (1976) (discussing inefficiencies plaguing standard form contracting).
281. Proponents of contractual liability recognize that insurance markets are affected by adverse selection but incorrectiy assume that selection effects unambiguously favor contractual liability. They assume contractual liability is better because mandatory liability in effect requires all patients to purchase the same liability at the same price, thereby requiring low-risk patients to subsidize high-risk ones. See, e.g., RUBIN, supra note 6, at 40-44 (describing the adverse selection problem and concluding that "elimination of the forced transfer from the tort system would conform with general notions of fairness [and would be] efficient");
282. Epstein, supra note 107, at 660 ("If the skillful and the incompetent, the watchful and the careless must be treated as falling within a single risk classification, then the former must subsidize the latter."). Yet economic analysis has shown that patients do not necessarily benefit from efforts to reduce pooling. Indeed, low-risk patients who value liability may well be better off when the state imposes liability (thereby pooling them with higher-cost patients whom they must cross-subsidize). This is because if liability is not mandatory then these patients could not obtain liability without actively seeking it and would thereby join a pool that is disproportionately composed of high-cost patients. The cost to lowrisk patients of obtaining liability voluntarily thus would be much higher than the cost to them of obtaining liability imposed by the state. Thus, low-risk patients who want to impose liability benefit from pooling because they can purchase the insurance they desire more cheaply under pooled liability than they can under contractual liability.
283. See generally Wilson, supra note 110, at 173-76 (showing that purchasers of health insurance may be better off when required to purchase pooled insurance because low-risk insurers may be better off subsidizing high-risk ones, as they would have to with mandatory insurance, than they would be under voluntary insurance, because adverse selection is likely to lead them to be underinsured). Moreover, even low-risk patients who do not benefit from liability in the short run may be better off when liability is mandatory if they benefit over the long run from the right to obtain lower-cost liability protection later when they are ill and need it.
284. See infra note 129 and accompanying text.
285. Insurers are allowed to charge a different amount to insure an individual or a family but often are precluded from adjusting premiums to reflect differences in subscribers' health status. Of particular importance, federal law precludes group health insurers from charging individuals different premiums based on their health status. Group insurers also cannot exclude subscribers based on health status. See Mary Crossley, Discrimination Against the Unhealthy in Health Insurance, 54 U. KAN. L. REV. 73, 113-16 (2005) (discussing the Health Insurance Portability and Accountability Act's nondiscrimination provisions). Some states have gone even further, requiring insurers to charge the same premium for a given benefit package to all subscribers within a given geographic location, even if the different subscriber pools have different risk characteristics.
286. See id. at 112 (describing the "community rating" systems some states have employed). Issuers are allowed to charge different premiums for different plans, however and can take the expected average health status of expected subscribers into account when setting these premiums.
287. See Glied, supra note 120, at 41 (arguing that job-based insurance coverage through large firms offers a form of long-term private health coverage against future risk).
288. David M. Cutler & Richard J. Zeckhauser, The Anatomy of Health Insurance, in 1 HANDBOOK OF HEALTH ECONOMICS, supra note 1, at 563, 572.
289. Id. at 571-572
290. See Newhouse, supra note 118, at 1253-1256 (discussing evidence that MCOs use plan design to select for healthier patients).
291. For example, one study of family health policies found that a patient seeking to switch from a low-quality plan (charging a premium at the bottom tenth percentile of the premium distribution) to a plan offering no more than a 40% increase in health benefits (as measured by spending) would have to incur a three-fold increase in premium costs for this relatively small increase in coverage. This three-fold increase in costs to obtain a mere 40% increase in coverage reflects the fact that the better plan attracts a disproportionate percentage of patients who need medical care. Thus, the patients attracted to this plan have higher expected average costs than the patients insured by the low-quality plan. Id. at 1253-55. The Federal Employees Health Benefits Plan also appears to exhibit a pricing structure distorted by selection effect.
292. Id. The plans differed only in the amount of cost sharing, not in the coverage provided. Id. Athough the actuarial difference in the two plans was only 10%, the price charged for the high-option plan was almost double that charged for the low-option plan, as a result of the expectation that this plan would attract people with greater health care needs.
293. Id.
294. See supra text accompanying note 129.
295. For example, largely to combat adverse selection, states impose certain minimum-quality constraints on the coverage that plans can offer, such as requiring coverage of certain medical procedures. See Russell Korobkin, The Efficiency of Managed Care "Patient Protection " Laws: Incomplete Contracts, Bounded Rationality, and Market Failure, 85 CORNELL L. REV. 1, 60-62 (1999) (contending that market forces will not induce MCOs to offer optimal coverage because MCOs know they can profitably restrict coverage for certain diseases because healthy consumers are unlikely to notice the restriction since they rarely read health plan coverage provisions carefully, and MCOs can profit from discouraging subscription by unhealthy, sophisticated consumers who do read plan terms and also are expensive to serve).
296. See supra Section III.C.
297. Patients who anticipate being healthy might want liability but could live without it in return for a sufficiently large discount. By contrast, patients who need serious care would find the no liability plans unattractive. Consequently, MCOs would use their right to contract over liability to help them segment the market.
298. See Cutler & Zeckhauser, supra note 130, at 607 ("Generally, the sick are drawn to more generous plans than the healthy.").
299. It might appear that the adverse selection problem is muted by risk averse patients, who are likely both to disproportionately prefer liability and to be healthier on average. Yet, contrary to this expectation, empirical analysis reveals that risk averse people impose lower costs on life insurers but do not impose sufficiently lower costs on acute health insurers to counteract the exponentially higher cost of insuring the unhealthy people who value this insurance. Thus, adverse selection seriously plagues the market for acute health insurance. See David M. Cutler et al., Preference Heterogeneity and Insurance Markets: Explaining a Puzzle of Insurance, 98 AM. ECON. REV. 157, 161 (2008) ("For acute health insurance, the lack of any systematic offsetting effect of risk tolerance may explain why... this market is, on net, adversely selected."). Similarly, it might seem that insurers would be less inclined to penalize liability subscribers since liability deters errors that increase medical expenses. But this effect will not seriously mute the adverse selection problem if the increased medical costs associated with the patients who prefer liability exceeds the cost savings to the MCO associated with any resulting reduction in medical error. This is likely since MCOs pay all the added health care costs of the very ill patients who value liability and these costs are exponentially higher for the very ill. By contrast, MCOs only bear higher medical costs as a result of error for the smaller portion of patients injured by error, and then only when the error harms the patient without killing her immediately. Thus, the perpatient adverse selection effect on plans that attract the ill can be expected to dwarf the per-patient deterrence benefits from liability.
300. See generally Cutler & Zeckhauser, supra note 130, at 606-624 (discussing the effect of adverse selection on the health insurance choices of higher- and lower-risk patients).
301. See id. at 616 ("The disappearance of generous plans as a result of dynamic processes of adverse selection is termed a 'death spiral.'");
302. Newhouse, supra note 118, at 1255 (explaining that limits on insurance pooling can lead to a "premium death spiral").
303. Contracting over the standard of care would permit patients to agree that doctors should treat them less intensively than other patients. One way patients could do this is to agree that a physician would provide only cost-effective treatments (and not provide some effective but excessively expensive treatments). Moreover, patients could agree that in determining whether a treatment is appropriate - in the sense of costeffective - the evaluation could be made using a value of life that is lower than the valuation used for other patients.
304. See Epstein, Contractual Principle, supra note 5, at 509 (describing the bargaining that would take place if patients could negotiate their standard of care).
305. This Part focuses on the claim that contracting would produce optimal variation in the standard of care because Part III reveals that patients would not use their right to contract to select the level of care that maximizes their welfare. Patients would not have optimal incentives to contract over the level of care provided because collective goods and renegotiation problems would preclude them from obtaining the full deterrence benefit of any standard of care imposed by contract In addition, they would be overcharged for contracting into higher quality of care as a result of adverse selection. Thus, for the reasons given in the preceding Part, contracting would not provide patients with optimal incentives to contract into the care levels that benefit them.
306. See, e.g., Robinson, supra note 5, at 182-183 (suggesting that, because patients bear the costs of medical care and obtain its benefits, from an efficiency perspective contractual arrangements would allow the parties to establish the rules that they prefer).
307. See id. at 194 (arguing that informed patients would employ contract to demand the level of care that is optimal for them).
308. See Hoangmai H. Pham et al., Care Patterns in Medicare and Their Implications for Pay for Performance, 356 NEW ENG. J. MED. 1130, 1132-1133 (2007) (noting that most Medicare beneficiaries are under the care of at least two primary care physicians and five specialists working in four different practices).
309. It also is helpful for all providers to be ostensibly governed by a common standard of care because patients often are cared for by many providers, practicing out of many different institutions. Standardizing the duty of care helps facilitate the coordination of care across these providers. In turn, coordination would be more difficult if each different provider treating a patient were governed by a dramatically different standard of care.
310. See Kahan & Klausner, supra note 126, at 733 ("A long as the value of customized terms exceeds the value of the potential standard term to the initial user, [parties] may customize even if standardization would be socially optimal.").
311. See, e.g., id. at 729-36 (describing the effects of learning and network externalities on the efficiency of standard terms in contracts);
312. Michael Klausner, Corporations, Corporate Law, and Networks of Contracts, 81 VA. L. REV. 757, 805-08 (1995) (arguing that standard terms adopted by heterogeneous firms may be inefficient because of network externalities);
313. Kornhauser, supra note 126, at 1174-75, 1177-1179 (stating that inefficient standard terms can occur even when mere are no bargaining defects because of search costs, consumer irrationality, and bounded memory).
314. Convergence on a uniform term would present a challenge for contractual liability because it would raise the question of why we need to achieve this uniformity through costly individual contracts instead of through the state. After all, if patients are as informed about the costs and benefits of liability - and as attentive to this issue as contractual liability proponents assume them to be, it would seem that they should be sufficiently informed and attentive to induce legislatures to adopt optimal reforms. Many of the pathologies of interest-group politics (e.g., rent-seeking) arise because citizens do not already know the costs and benefits of a legal rule and do not attend to the issue. These problems with state action would be muted if citizens - and thus, presumably, the state - are perfectly informed about the effects of a legal rule, attend to the issue, and develop a uniform solution which in theory could even provide for the socially optimal level of variation across consumers. See generally DENNIS C. MUELLER, PUBLIC CHOICE III 343-348 (2003) (discussing rent-seeking behavior and the political process).