assembled. It is perhaps for this reason that policy initiatives and enactments in this area have been limited in scope. Maryland legislation instructs the insurance commissioner to specifically determine whether an insurer’s surplus is excessive before permitting a rate increase.33 New Jersey legislation now requires prior approval of rate increases that exceed 25 percent.34 An initiative based on California’s Proposition 103 was placed on the ballot in Nevada but garnered only a 34.7 percent “yes” vote, the poorest performance of any of the eight ballot measures voted on in that state’s 2004 election.35 Oklahoma actually loosened its insurance rate regulation in 2004, in keeping with a general national trend over the past decade.36 What remains unknown is how much more money is retained by or returned to physicians under stricter regulatory scrutiny of rates. 7 confirm one of the theories behind such panels—that they give parties accurate guidance as to the settlement value of a claim38—but others suggest that this process can add, rather than reduce, costs and delay.39 While there is little evidence that arbitration or mediation reduces litigation cost or delay,40 there are many committed activists in the legal community who continue to advocate ADR as a less adversarial means of handling disputes. Some even suggest that such programs could increase patient safety by giving more feedback to providers.41 Nonbinding ADR is also thought to improve parties’ satisfaction with the legal process.42 In any event, because these measures do not deprive litigants of any substantive rights, they provoke much less controversy than tort reform approaches. D. Improving accuracy and predictability of injury compensation The legal system’s handling of malpractice suits is suspect in the eyes of many because it brings few certainties. In every trial, jurors must choose to believe one of two sets of hired expert witnesses who offer diametrically opposed testimony. Jurors are given no guidelines for awarding non-pecuniary damages, which are, by nature, subjective. And the legal standard applied by juries—whether a doctor provided the standard of care customary in the medical community—has been criticized as “unreasonable”43 and “unreliable.”44 There are several proposals for addressing these dilemmas: • Stricter requirements for expert witnesses may prevent unqualified doctors from testifying but can also be made so restrictive as to make it harder or more expensive for plaintiffs to prove meritorious cases. • Some states have recently clamped down on “forum shopping”—the practice of filing suits in jurisdictions that are considered more favorable to plaintiffs. In most states, venue laws require suits to be filed in the county where the doctor lives or where the alleged malpractice occurred. • In recent years, judges have retained independent experts to testify in important product liability cases. To some extent, the screening panels described above have brought independent expertise to bear in malpractice cases. It has been suggested that judges might obtain independent experts for malpractice trials, but it is unclear who would bear the cost. • Some observers have advocated the creation of guidelines, schedules, or ranges (e.g., graduated caps) for awards of nonpecuniary damages, based on the severity of a plaintiff’s injury. • A jury’s discretion to award such damages could also be curbed by a statute, such as New York’s, that gives judges authority to modify awards that “deviate materially from reasonable compensation.”45 This empowers the judge to compare a verdict with others in similar cases and reduce “outlier” awards. • Researchers at Harvard Medical School and The Urban Institute urge adoption of an administrative, no-fault compensation system based on occurrence of a preventable adverse event, a 8 Figure 5 Average Payment to Malpractice Claimant by Injury Severity, $ 12 3 ,9 17 $ 5 6 ,8 11 $ 10 ,5 4 5 $ 48,366 $ 0 $ 200,000 $ 400,000 $ 600,000 $ 800,000 Death P ermanent, Grave P ermanent, Majo r P ermanent, Significant P ermanent, Mino r Tempo rary, Majo r Tempo rary, Mino r Tempo rary, Ins ignificant Tempo rary, Emo tio nal SOURCE: Physician Insurers Association of America. proposal endorsed by the Institute of Medicine. • A proposal for special health courts would combine an administrative no-fault system for clear-cut medical errors with panels of expert judges to promulgate a new standard of care based on “scientific evidence on the safety and effectiveness of health care.”46 To date, there has been no empirical research evaluating these proposals (although no-fault compensation systems limited to birth injuries that have been implemented in Florida and Virginia have been given considerable scrutiny). Much of the criticism leveled at the civil justice system portrays it as unreliable and prone to awarding extravagant sums of money. But while the Harvard Medical Practice Study found a mismatch between medical errors and claims,47 other empirical research has found that very few non-meritorious claims are made or result in payment,48 and that damage awards are generally proportionate to injury severity49 (see Figure 5). While an ideal legal system should be “reliable” in the sense that it