Tort Liability in the United States
The U.S. tort system is not centralized, which makes collecting comprehensive data about it difficult. Roughly 95 percent of lawsuits over torts are filed in state courts, rather than federal courts, the Congressional Budget Office (CBO) estimates. Moreover, in the vast majority of cases, plaintiffs and defendants reach out-of-court settlements, whose terms typically remain private. (For example, 97 percent of tort cases that "terminated" in federal district courts in fiscal year 2000 were disposed of before a verdict was reached.)
Data from the Bureau of Justice Statistics that cover 45 of the nation's 75 largest counties indicate that plaintiffs won 48 percent of the cases that reached a verdict in state courts in 1996 (the latest year for which that information is available).(2) In those cases, the average time between filing and completion was 22 months. Automobile-related torts accounted for 49 percent of the cases, followed by premises liability (22 percent) and medical malpractice (12 percent). The median award to successful plaintiffs was $31,000 for all cases, but it varied widely for different categories of torts: from $18,000 in automobile-related cases to $286,000 for medical malpractice and $309,000 in asbestos cases.
Looking at trends over time, data from 16 states tracked consistently by the National Center for State Courts show that the number of tort cases filed each year rose by 70 percent between 1975 and 1990 (its peak) and then fell by 19 percent by 2000. Relative to population, the rate of filings was 8 percent lower in 2000 than in 1975--212 cases per 100,000 residents compared with 230 cases.(3)
Figures that suggest an overall decline in tort cases, however, mask continuing growth in the number or impact of some important categories of torts. For example, the Physician Insurers Association of America reports that median court judgments for medical malpractice rose from $100,000 in 1990 to more than $300,000 in 2001--an increase of 138 percent after correcting for inflation. And researchers at RAND report that the number of claims filed for asbestos exposure nearly tripled in just two years, between 1999 and 2001.
 
The Basic Economics of Tort Liability
From the economic point of view, the efficiency of the tort system is measured by how well it minimizes the sum of several types of costs:
· The costs of injuries (including medical costs, lost productivity, and pain and suffering);
· The costs of efforts to prevent or avoid injuries (including efforts to make products safer, which tend to raise consumer prices) and the opportunity costs of goods and services that are not provided (such as potential medical drugs that do not reach the market or municipal pools that are closed for fear of lawsuits) or goods and services that are provided but forgone by some risk-averse consumers (such as air travel);
· The costs of administration and implementation (particularly attorneys' fees and the administrative costs of insurance that potential injurers and victims buy to redistribute the risks they face); and
· Indirect costs to the economy (such as the disruption costs of plant closings and bankruptcies).
What constitutes equity in relation to the tort system is ultimately subjective, but there is consensus that compensating victims for their injuries--at least in some cases and to some degree--is equitable.
Tort liability is only one means by which society addresses the efficiency and equity issues posed by injuries; other means include market forces, regulation, and public insurance funds. Market forces can help control injury costs in several ways. Under conditions of competition and good information, producers of goods and services respond to consumers' desires for safer products, employers respond to employees' desires for safer workplaces, and insurance companies offer policies to respond to potential victims' desires to reduce the uncertainty they face.
One efficiency rationale for supplementing market forces with some form of government involvement is simply that many injuries--automobile accidents, releases of toxic chemicals, and so forth--are unrelated to any economic transaction. Indeed, some academic economists favor restricting the scope of tort liability to such "stranger" injuries. For other types of injuries, making an efficiency argument for government intervention requires the existence of some market imperfection: perhaps potential victims lack good information about the risks they face, suffer from biases that limit their ability to use the information, or have few choices because of monopoly or collusion in the market.(4) Of course, government actions have their own weaknesses and thus may not improve efficiency in practice. For example, regulation requires centralized information about costs and benefits, and regulators may be co-opted by the parties they regulate.
Tort liability supplements the market in a more decentralized way. The basic idea is that making injurers pay for the harm they cause not only compensates victims but also gives injurers (if not victims) appropriate incentives to reduce the frequency and severity of that harm. The different liability standards used by the courts aim to achieve those goals in different ways: in particular, under the doctrine of strict liability, injurers are responsible regardless of how much care they exercise in trying to minimize injuries, whereas under the doctrine of negligence, they are responsible only if their actions fail to meet a standard of due care.(5)
The tort system is no panacea, however, even in principle--it is difficult if not impossible to craft liability rules that can consistently achieve the desired levels of both efficiency (taking into account all of the relevant costs) and equity. For example, because the expected level of compensation may affect the degree of care that potential victims exercise, the efficiency objective of cost-effective deterrence can conflict with the equity objective of compensation. Moreover, because the terms of that trade-off can vary, a single rule may not achieve the desired balance between efficiency and equity in all cases.
In practice, tort liability is further limited because information--particularly the information needed to determine the cause of an injury--is incomplete and costly. The transaction costs of the tort system derive from information problems: lack of complete information is what allows plaintiffs and defendants to hold divergent views and encourages them to devote resources to proving their respective cases. Information problems are also the root cause of courtroom errors, and they can make it hard to set standards for due care at efficient levels.
 
The Costs and Benefits of Tort Liability
Analyzing the policy questions that surround tort liability is difficult because of incomplete data not only on tort cases themselves but also on the indirect costs and benefits of the tort system. Indeed, from the standpoint of economic efficiency, the actions that the system encourages potential injurers and victims to take (or refrain from taking) to avoid injuries can be more important than some of the direct "costs" associated with individual cases.
In efficiency terms, the primary benefits of the tort system are measured not by payments to victims--which represent transfers of wealth but not gains or losses to society as a whole--but by reductions in injury costs. Those benefits arise indirectly, through precautions taken by potential injurers (for example, efforts to design safer products or reduce production defects).(6) Thus, they are not observable in data on trials or settlements.
Several important types of costs are also indirect, including the costs of specific actions that firms take to reduce the injury risks associated with their products (such as including air bags in automobiles), the opportunity costs of goods and services not offered because of liability concerns or not purchased because of liability-related price increases, and the disruption costs of layoffs and bankruptcies.
Arguments About the Effectiveness of Tort Liability's Incentives
Indirect benefits and costs are very difficult to measure. In general, data do not exist to show how liability affects the degree of care that potential injurers take--let alone how injury costs change as a result of that care. Moreover, theoretical analysis alone cannot answer the key questions, because the extent to which the potential efficiency benefits of tort liability are realized depends on the relationship between the true costs of injuries and the expected costs to injurers. If potential injurers expect to pay one dollar more for each additional dollar of injuries they cause, they will have the optimal incentive to take all (and only) cost-effective precautionary actions. But they might anticipate paying more than one dollar per dollar of additional injury (for example, because of excessive punitive damages) or less than that (for example, if some of their torts go undetected or if their liability costs are insured and their premiums do not rise commensurately). For potential injurers whose actions are thought at the time to be harmless--such as the firms that manufactured or used asbestos before its health risks were identified--there is no expectation of increased liability costs and hence no specific incentive for precaution.(7)
Controversy over both the efficiency and equity effects of liability has particularly focused on nonpecuniary damages (punitive damages and compensatory damages for pain and suffering). Critics argue that large nonpecuniary damages are awarded arbitrarily and unpredictably, with little connection to the actual harm or to the character of the injurer's conduct. In that view, such damages are not only inequitable but also inefficient: arbitrary and unpredictable awards do not provide incentives for precaution but do raise costs, thereby distorting price signals. Critics further argue that nonpecuniary damages, whether arbitrary or not, have a separate adverse effect on the distribution of risk--in particular, that liability for pain and suffering implicitly provides consumers with a form of inefficient overinsurance.(8)
In contrast, supporters of the liability system argue that large punitive damages can serve equity by expressing society's disapproval of behavior that reflects wanton disregard or contempt for potential victims. Such damages can also promote efficiency, they say, by providing proper incentives for the prevention of injuries that have a significant probability of going undetected. (For example, if bolt manufacturers expect the role of defective bolts to go unrecognized in four out of five accidents that their products cause, they will have inefficiently low incentives to prevent defects unless they expect to pay five times the actual damage on those occasions when they are penalized.) Supporters further argue that pain and suffering represent real losses that should be reflected in the prices of products (to send consumers efficient signals) and that limiting awards for such losses might undercompensate some injury victims.
Evidence About the Effects of Tort Liability
Without direct data or clear theoretical predictions about the incentive effects of tort liability, analysts have tried to tease out the truth statistically. However, their most detailed efforts to date, which have focused mainly on punitive damages, have not yielded conclusive results. The best available study of the effects of punitive damages in the United States found no evidence that the 46 states that allow such damages have fewer environmental or safety torts than the four states that do not allow them. However, that lack of evidence may simply reflect the limitations of the data.(9)
Given the scarcity of data on the benefits and many of the costs of the current tort system, economists cannot judge the system's efficiency. But they can answer a narrower question about its cost-effectiveness as a means of compensating injury victims. The best available data on the direct costs of tort cases suggest that victims who file claims receive an average of 46 cents from each direct dollar spent on the system (with the other 54 cents going to attorneys' fees and insurance expenses).(10) The best available data show that such transaction costs are proportionately much smaller in public insurance programs--20 percent nationwide in state workers' compensation programs (though that figure excludes spending on claimants' attorneys, which is reportedly rising) and 15 percent in the federal Vaccine Injury Compensation Program. A no-fault public insurance program for other torts would probably not lower transaction costs to those levels, in part because of the costs of establishing which injurers were responsible for particular injuries. Nonetheless, it is safe to say that the existing tort system is a relatively costly way to compensate victims and, thus, that any justifications for it must rest on its effects on deterrence, equity, or both.
 
Policy Options for Changing the Tort System
The controversies over the costs and benefits of the tort system have led to numerous proposals for change at the federal level. This report discusses the potential advantages and disadvantages of various policy options in qualitative terms. Those options--which were chosen to illustrate the trade-offs between efficiency and equity that lawmakers face--fall into three main categories.
Policies for reducing the scope of tort liability include options that would eliminate liability for all injuries or all "nonstranger" injuries, exempt products certified as safe by a federal regulatory body (such as the Food and Drug Administration or the Consumer Product Safety Commission), or replace tort liability with a federal compensation system for injury victims, like the present state-level workers' compensation system and the federal fund for vaccine victims.
Policies that are more incremental in nature but that could be applied broadly to all types of torts include options that would restrict compensation for pain and suffering or punitive damages, limit fees charged by plaintiffs' attorneys, reduce the use of joint-and-several liability (under which one or a few injurers can be held responsible for paying all of the damages caused by a number of injurers), or modify the "collateral-source rule" (under which the amount of damages owed by a defendant does not take into account any benefits that an injured plaintiff has received from an insurance policy or other independent source).
Policies targeted toward particular types of tort cases include options that would create specialized courts to hear medical malpractice cases, establish minimum medical criteria for asbestos claims and perhaps set up a victims' compensation fund, tie the fees received by plaintiffs' attorneys in class-action suits more closely to benefits actually received by the class members, or allow defendants to shift more class-action cases from state courts to federal courts.
In most cases, data limitations make it impossible for CBO to determine whether a particular option would be likely to improve or reduce economic efficiency. Nonetheless, the economic perspective leads to some general conclusions that decisionmakers may wish to keep in mind as they consider proposed changes to the liability system.
· The impact on efficiency of using tort liability to try to improve on market outcomes may be either positive or negative--depending on the incentives that liability provides for potential injurers and potential victims and on how those incentives interact with incentives and constraints from other sources, such as government regulations and private insurance policies.
· Most, if not all, options for changing the tort system involve some trade-offs. In particular, policies that seem desirable on efficiency grounds may be problematic from the equity perspective, or conversely.
· Federal involvement in an area governed predominantly by state law may be justified by its benefits for interstate commerce, but it limits state innovation and experimentation (as well as the ability of U.S. residents to "vote with their feet" by choosing to live under one state's liability regime rather than another's).
· Because the efficient solution is the one that minimizes the sum of several different costs, which may vary in their relative importance, different policies may be appropriate for different types of tort cases.
 
1.  Bryan A. Garner, ed., Black's Law Dictionary, 7th ed. (St. Paul, Minn.: West Group, 1999), pp. 1496-1497.
2.  Department of Justice, Bureau of Justice Statistics, Tort Trials and Verdicts in Large Counties, 1996, NCJ 179769 (August 2000).
3.  See Brian J. Ostrom, Neal B. Kauder, and Robert C. LaFountain, eds., Examining the Work of State Courts, 2001: A National Perspective from the Court Statistics Project (Williamsburg, Va.: National Center for State Courts, 2001), with accompanying spreadsheets available at www.ncsconline.org/D_Research/csp/2001_Files/2001_Tort-Contract_Tables.xls.
4.  There is no presumption that market forces tend to produce equitable outcomes; hence, arguments for government intervention can also be made on equity grounds.
5.  Even in a case judged under strict liability, the injurer may not be held responsible if the victim's own behavior contributed too much to the occurrence of the harm.
6.  Some indirect benefits may also arise from better distribution of risk. In principle, risk-averse consumers who expect to be compensated for injuries more fully through the liability system than they would be through their own insurance may be more willing to buy certain goods or services (space heaters, perhaps).
7.  However, the mere possibility that seemingly harmless activity may later produce tort claims increases uncertainty and gives potential injurers general incentives to buy insurance, investigate possible risks, and take generic prevention or avoidance measures (such as not researching or developing new products), which may be efficient or inefficient.
8.  The argument is that consumers benefit by insuring themselves against pecuniary losses, such as lost income or increased medical costs, but not against pain and suffering (as illustrated by the fact that people generally do not purchase life insurance policies for their young children). Thus, when producers expect to pay nonpecuniary damages and build the costs of those damages into the prices they charge for goods and services, consumers implicitly pay a kind of insurance premium for coverage they would not otherwise choose to buy. The effect of that implicit premium and coverage is to shift wealth inefficiently--raising it in the event of an injury, but not by enough to justify the reduction in wealth in the case of no injury.
9.  One key limitation is that the control group includes only four states; another is that punitive damages would not be expected to deter typical torts but only those rare ones that were egregious enough to be the subject of such damages. See W. Kip Viscusi, "The Social Costs of Punitive Damages Against Corporations in Environmental and Safety Torts," pp. 285-345, Theodore Eisenberg, "Measuring the Deterrent Effect of Punitive Damages," pp. 347-357, and David Luban, "A Flawed Case Against Punitive Damages," pp. 359-380, all in Georgetown Law Journal, vol. 87, no. 2 (November 1998).
10.  Tillinghast-Towers Perrin, U.S. Tort Costs: 2002 Update--Trends and Findings on the Costs of the U.S. Tort System (2003).
1
Introduction
Injuries have many causes, including economic activities: consumers are injured or killed by defective products, workers are hurt on the job, train passengers are injured by derailments, and patients are harmed by medical errors. Markets provide broad incentives to control the number and costs of such injuries. For example, employers can save on wage costs by making jobs less hazardous; drivers with good safety records pay lower insurance premiums; and enhanced safety features can give a product a marketing advantage over its competitors. In addition, the insurance market responds to people's desire to reduce the financial uncertainty associated with potential injuries.
Society uses three tools to augment the safety incentives and insurance opportunities provided by the market: regulation, public compensation programs, and tort liability. In particular, the U.S. tort liability system is intended to reduce the number of injuries--by providing incentives for individuals and firms to take appropriate care--and to compensate those who are harmed.(1)
"Tort" is defined very broadly in law as an injury to "one's person, reputation or feelings" or damage to "real or personal property."(2) Tort liability is the court-enforced obligation of a "tortfeasor" (injurer) to pay for a victim's losses. The concept of tort liability evolved as a generalization of various specific types of injuries--including trespass, deceit, slander, and assault and battery--some of which generally occur outside the context of economic activity. Even today, tort law's diverse origins are reflected in a complex and heterogeneous body of common law. One example of that complexity is the boundary line between tort law and contract law. Notwithstanding the broad definition of a tort, injuries caused by a breach of contract are generally addressed under contract law--with the exception of injuries involving medical malpractice or defective or dangerous products, which are addressed as torts even when the parties have an explicit contract.
Tort law is almost exclusively contained in state law, and the large majority of tort cases are filed in state courts. Not surprisingly, therefore, most past efforts to reform the tort liability system in the United States have taken place at the state level. In particular, most states have adopted one or more reforms favoring defendants during the past 30 years--especially in 1986, when a perceived insurance crisis led to 41 new state laws.(3) The courts have also taken action at various times: recently, for example, the U.S. Supreme Court reiterated an earlier ruling that the Due Process Clause of the Constitution establishes limits on punitive damages.(4)
Still, many critics of the current tort system say that additional federal action is needed for several reasons. At the general level, they argue that the system's costs are too high, particularly because of excessive "transaction costs" (mainly compensation to plaintiffs' and defendants' attorneys) as well as excessive and arbitrary awards for noneconomic losses ("pain and suffering") and for punitive damages. Such high costs sometimes have perverse negative effects on safety, they argue--for example, by discouraging firms from conducting safety research that could create a legal "paper trail" or by raising the prices of risk-reducing goods and services, such as medical care. Critics also contend that plaintiffs frequently bring frivolous lawsuits when they know that the defendant is inclined to settle out of court to avoid the costs of litigation.
The tort system's critics also take issue with specific types of cases. They argue that medical malpractice claims are contributing to a crisis in the cost and availability of certain health care services, that claims for exposure to asbestos by people who show no evidence of illness are burdening the courts and pushing firms into bankruptcy, and that misuse of the class-action mechanism is allowing local judges and juries who are biased against distant corporate defendants to bring verdicts that have damaging national implications.
Supporters of the current tort system question the factual basis of some of those criticisms. They note that the number of tort cases filed nationwide has been falling since 1996. Moreover, they say, large awards for punitive damages are rare and are often reduced before payment is made. They further argue that the costs of the tort system are worthwhile given the system's contributions to the social goals of compensating victims, holding injurers responsible for their actions, and improving safety. Supporters of the present system also maintain that proposed reforms are generally too broad and that fewer negative consequences would occur if the Congress allowed the states and the judiciary to address any real problems that exist.
The Congress has modified tort law several times in the past, although its actions have generally focused on torts that spring from a particular cause or affect specific industries. For example, the General Aviation Revitalization Act of 1994 exempted manufacturers of small planes from liability for crashes if the planes are more than 18 years old and not used in scheduled service. In addition, federal compensation programs that have been created for victims of vaccine injuries and the September 11 terrorist attacks limit the ability of people receiving compensation to sue for damages.(5)
This primer looks at the tort system from an economic perspective: it discusses the factors that influence whether tort liability improves or reduces economic efficiency and analyzes, in qualitative terms, the likely effects of some reform initiatives. In principle, tort liability can provide incentives for potential injurers and victims to take precautions, thus reducing injury rates. In practice, however, implementation problems can reduce the strength and value of those incentives to the point that liability decreases efficiency and may even have net negative effects on safety. Although available evidence about the tort system is too limited to support many firm conclusions, it does indicate that the system is more costly than other methods of compensating victims. Thus, an economic perspective suggests that changes to the system should focus on improving the incentives it provides, lowering its costs, or both.
 
1.  The aims of the tort system are sometimes also said to include providing a forum for the less privileged to be heard and punishing egregious behavior.
2.  Bryan A. Garner, ed., Black's Law Dictionary, 7th ed. (St. Paul, Minn.: West Group, 1999), pp. 1496-1497.
3.  Those reforms included limits on nonpecuniary damages, reductions in the scope of joint-and-several liability, and offsets for benefits from collateral sources (which are discussed in Chapter 5).
4.  The Court held that punitive damages must be "both reasonable and proportionate to the amount of harm to the plaintiff and to the general damages recovered." It also alluded to rough limits on the ratio of punitive damages to compensatory damages in cases involving only economic losses. State Farm Mut. Automobile Ins. Co. v. Campbell, 538 U.S. ___ (2003).
5.  For a summary of some federal tort reform laws, see Henry Cohen, Federal Tort Reform Legislation: Constitutionality and Summaries of Selected Statutes, CRS Report for Congress 95-797A (Congressional Research Service, updated May 2, 2003).
An Overview of the U.S. Tort System
Most legal scholars agree that the scope of liability in the U.S. legal system has expanded dramatically in the past three decades. That expansion began early in the 20th century with a growing acceptance of the notion that more extensive tort liability would serve to compensate injured parties and reduce the level of accidents. Although no complete set of data is available, limited data call attention to several prominent features of the current tort litigation landscape--that the vast majority of tort claims are settled out of court, that state courts handle the bulk of torts, and that different types of torts have different impacts on litigants and on the tort system.   The Expansion of Tort Liability in the United StatesU.S. tort law is based primarily on common law--in which judicial rules are developed on a case-by-case basis by trial judges--rather than on legislation. Tort liability is assigned using two basic standards: strict liability and negligence. Under strict liability, injurers are held fully liable for their victims' losses without regard for whether they were actually negligent or intended to harm anyone.(1) Under a negligence standard, by contrast, injurers are held liable only if they failed to meet a certain standard of care. According to legal scholars, a number of important developments have increased the scope of liability for torts in the United States. Early English tort law, the antecedent of U.S. tort law, was chiefly concerned with making injurers pay for the losses of their victims, with little emphasis on fault or negligence.(2) That standard was used in the United States until the 19th century, when U.S. common law established negligence as the basis for tort liability. However, strict liability continued to apply in certain cases, such as injuries caused by wild animals kept as pets or damage to crops caused by trespass of domestic animals.(3) Some scholars argue that the requirement for plaintiffs to show that defendants had been negligent effectively limited the scope of the U.S. tort system.(4) The turn of the 20th century saw public policy increasingly emphasize victim compensation and accident reduction. The enactment of workers' compensation laws--which established a public insurance system aimed at lowering employers' payments while making workers' recovery of damages automatic--played an important role in the evolution of tort law and policy.(5) Before workers' compensation programs, the only remedy that injured workers had was to prove their employers negligent through the tort system. Workers favored legislation instead because they often had been unable to recover damages or had experienced delays or high costs when they had been successful. For their part, employers favored legislation because it limited their liability and made payments predictable.(6) That shift away from tort law to a public compensation system led to more thought about how tort liability could be improved or better applied in other types of cases. By the 1940s, legal scholars had begun to think about two ways in which the tort system could serve the wider goal of enhancing social welfare. First, they saw the economic concept of "cost internalization" as a tool for reducing accident rates: if potential injurers know they will be held liable for accidents, they will take appropriate action to avoid liability. In that view, by awarding damages to compensate victims, tort law would serve as a mechanism to ensure that potential injurers faced the appropriate future costs of their actions.(7) Second, some scholars argued that the tort system could provide a kind of accident insurance for victims. They did not focus on the possibility that an expanded liability system could increase carelessness on the part of potential victims, nor did they adopt any of the methods that traditional insurance policies use to deal with that problem.(8) Rather, they focused exclusively on the distributional goal of relieving victims of the burden of accident losses and spreading that burden across a broader population. One area in which those concepts proved appealing in practice was product liability.(9) Historically, product liability was dealt with either as a breach of warranty under contract law or as a tort subject to the negligence standard. Under contract law, recovery in such cases was limited to repair and replacement of the product; under tort law, recovery was limited by the difficulty of proving negligence. In the 1960s, the courts moved rapidly toward a standard of strict liability for defective products; in 1964, that standard was accepted and recommended by the American Law Institute in its second Restatement of the Law volume on torts. By the mid-1970s, most states had adopted provisions that were either identical or similar to those in the Restatement.(10) In addition, the concept of negligence has undergone significant reinterpretation over time, according to legal scholars. The law now takes into account the fact that manufacturers often have more ability than consumers to avoid accidents; thus, it is more likely to view failure to take inexpensive action as negligence or to attach liability to indirect or partial contribution to an injury.   Characteristics of the Tort System TodayGetting a complete picture of the state of the U.S. tort system is difficult because no data are available that cover all of the tort cases brought in the various jurisdictions across the country. However, the National Center for State Courts (NCSC) provides some data on trends in civil filings in general-jurisdiction courts in several states.(11) It also conducts periodic surveys of civil trials in the nation's 75 largest counties for the Bureau of Justice Statistics (BJS). In addition, data about cases disposed of in federal court are available from the Administrative Office of the U.S. Courts. In 16 states consistently tracked by the NCSC, tort filings in general-jurisdiction courts grew from 189,520 in 1975 to 260,745 in 2000, which appears to support the common view that the number of tort cases is rising. But controlling for population growth in those states indicates that tort filings relative to population declined by 8 percent over that period--from 230 per 100,000 residents in 1975 to 212 in 2000.(12) Additionally, total tort filings in those 16 states were relatively constant from 1986 to 1996 and have shown a downward trend since then, falling from 320,976 filings in 1996 to 260,745 in 2000. In drawing inferences about the tort system as a whole, however, it is important to note several limitations of the available information. · Data do not exist for those tort disputes that do not go to trial, because the details of settlements are usually private.(13)· Collecting consistent data between the various jurisdictions is difficult. The overwhelming majority of tort filings occur at the state level, and the structure of state courts and the laws under which they operate differ from state to state. Moreover, those courts have not tended to view keeping records on the details of case outcomes as being central to their mission.· Both anecdotal and statistical evidence about damage awards can be misleading because the amount of damages actually paid can be reduced after a trial.(14)· Overall trends can be misleading because various categories of torts have different economic impacts, and the timing and disposition of mass torts (cases involving large numbers of people) can significantly skew the numbers. Settlement Versus TrialThe majority of tort disputes never reach a trial verdict. For example, of the 41,696 tort cases that were terminated in U.S. district courts in fiscal year 2000, only 3 percent were decided in trials.(15) The NCSC similarly reports that "[t]he vast majority of all [state] tort cases are disposed through some form of settlement, with only 3 percent of all tort matters resulting in a jury trial."(16) Litigants have mutual incentives to save on litigation costs by settling out of court. They avoid uncertain trial outcomes and delays and can agree to keep settlements confidential.(17) In some cases, settlements may be reached through alternative methods of dispute resolution, such as voluntary arbitration or mediation. Generally, details of civil disputes settled before a trial are not reported to the courts and hence are not included in publicly available data. Those data therefore show only part of the picture--there may be important differences between cases that go to trial and cases that settle out of court.(18) For example, cases that go to trial probably involve larger dollar amounts, on average. Nevertheless, trial verdicts set precedents for all cases and thus affect the incentive to settle by signaling the value and probability of success to future litigants.(19) Where Are Tort Cases Heard?The vast majority of tort filings occur in state courts. In 2000, more than 700,000 torts were filed in state general-jurisdiction courts, compared with only about 37,000 in federal courts, the Congressional Budget Office (CBO) estimates.(20) Liability standards are not uniform among the various jurisdictions. For example, the extent to which damages may be reduced if the injured party contributed to the accident differs among states. In addition, a small number of local courts have been described as "class-action magnet courts" and criticized for being biased toward plaintiffs.(21) U.S. district courts have jurisdiction in civil cases when a case deals with a federal question, the federal government is either a defendant or plaintiff, or the case involves "diversity of citizenship."(22) Of the tort cases that were terminated by trial in federal courts in fiscal year 2000, 72 percent involved diversity of citizenship, 18 percent involved a federal question, and 11 percent involved the U.S. government as a defendant or plaintiff. Many of those tort cases did not originate in federal courts: 28 percent were removed from state courts.(23) Categories of Tort CasesDifferent types of torts pose different challenges for the goals of the liability system, so it is useful to track the trends in filings for important categories of torts. For example, overall statistics on torts can be substantially driven by developments in mass torts. The General Accounting Office found that asbestos litigation accounted for half of the growth in tort filings that occurred in federal courts between 1974 and 1986.(24) Although tort filings as a whole have fallen significantly since 1996, that overall trend masks important developments in key categories of torts. The major areas of tort litigation, based on their share of total tort trials completed in the general-jurisdiction courts of the 75 largest U.S. counties in 1996, are automobile-related torts (49 percent), premises liability (22 percent), and medical malpractice (12 percent).(25) Many of the common categories of torts pose few policy problems. For example, automobile torts often have low awards: a median jury award of $18,000 for winning plaintiffs in state courts in the 75 largest counties in 1996, compared with a median award of $31,000 for all torts (see Table 1).                 
Table 1. 
Characteristics of Tort Cases Decided by Trial in State and Federal Courts, 1996
 
  Percentage of CasesWon by Plaintiff Median Award(Dollars) Percentage of Awards 
   $250,000or More $1 Millionor More
 
All Tort Cases 
  State 48.2 31,000   16.9   5.8 
  Federal 45.8 139,000   38.1   14.6 
 
Automobile Accidents 
  State 57.5 18,000   8.7   3.4 
  Federal 59.7 100,000   37.4   11.6 
 
Medical Malpractice 
  State 23.4 286,000   51.0   20.2 
  Federal 39.8 252,000   54.3   22.9 
 
Asbestos 
  State 55.6 309,000   50.6   12.1 
  Federala 40.0 465,000   50.0   0 
 
Product Liability Other Than Asbestos 
  State 37.1 177,000   41.2   16.3 
  Federal 26.6 368,500   62.0   24.0 
 
Source: Congressional Budget Office based on Department of Justice, Bureau of Justice Statistics, Tort Trials and Verdicts in Large Counties, 1996, NCJ 179769 (August 2000), and "Civil Terminations, 1996 and 1997," Federal Court Cases: Integrated Data Base (Federal Judicial Center, Washington, D.C.), dataset nos. 103 and 104, available from the Inter-university Consortium for Political and Social Research at www.icpsr.umich.edu/cgi/archive.prl?study=8429.
Notes: State numbers are based on trials in general-jurisdiction courts in the 75 largest counties in the United States in calendar year 1996 (calculated from data for fiscal years 1996 and 1997). Federal numbers are based on trials in U.S. district courts in calendar year 1996.
Data on awards exclude cases won by defendants and those with awards of zero.
a. There were only two federal asbestos cases with monetary awards in 1996.
 
In contrast, torts that have received the most public attention--such as product liability cases (including asbestos litigation) and medical malpractice cases--often involve larger stakes and have more significant effects on courts' resources, victims' compensation, the viability of businesses, and insurance premiums. Winning plaintiffs in state courts received a median award of $309,000 in asbestos cases and $286,000 in medical malpractice cases in the 75 largest counties in 1996. Moreover, about 20 percent of medical malpractice awards and 16 percent of product liability awards (other than in asbestos-related cases) were at least $1 million, compared with only 6 percent of state courts' awards for all torts. (Although punitive damages were infrequently awarded in medical malpractice and asbestos cases, they were also higher than the average for all torts: median amounts of $250,000 and $110,000, respectively, compared with $38,000 for tort cases overall.)(26) At the federal level, medical malpractice cases terminated by trial in U.S. district courts in 1996 had a median final award of $252,000. Policymakers and the business community have been concerned about the increasing costs of asbestos litigation. Researchers at RAND estimate that claims for asbestos-related compensation totaled $54 billion through 2000, with estimated future costs ranging from another $145 billion to $210 billion.(27) They also identified 67 bankruptcies related to asbestos litigation through 2002, up from three through 1982. Adding to that concern is the latency that occurs in the onset of asbestos-related disease, which makes it difficult to predict firms' exposure to liability and to ensure adequate compensation for victims who have yet to file claims.(28) Asbestos cases also differ from other torts in the number of firms affected. The median number of defendants in asbestos litigation was 18, compared with a median of one defendant in tort cases overall.(29) Over time, the targets of asbestos suits have expanded from the original manufacturers of asbestos-related products to include customers who may have used those products in their facilities. According to RAND, the total number of defendants in asbestos litigation rose from 300 in 1982 to more than 6,000 in 2000.(30) Recent data show no growth in the total number of medical malpractice tort claims, but the size of awards has increased. Median payments for medical malpractice claims at trial rose from about $100,000 in 1990 to over $300,000 in 2001, according to the Physician Insurers Association of America.(31) (BJS data from the nation's largest counties show an increase in median awards in state courts from $201,000 in 1992 to $286,000 in 1996.)(32) That rise coincides with increasing malpractice premiums for doctors, which critics blame for leading to a reduction in the availability of health care in some parts of the country. Another set of tort claims that concerns policymakers is cases filed using the class-action procedure, in which a small number of plaintiffs represent a larger group of people who were similarly affected by the same product or tort. Class-action cases are designed to address relatively small but numerous losses for which individual suits would be impractical. However, when the class is large enough, even claims that are trivial individually can have a significant effect on particular firms and even whole industries. Limited data are available on class actions, but those cases are more likely than other torts to be filed in federal court. For example, a RAND study estimated that during the 1995-1996 period, 40 percent of reported class-action decisions arose in federal court, whereas CBO estimates that less than 5 percent of all torts are filed in federal court.(33)  1.  Strict liability--sometimes called "liability without fault"--is based on the breach of an absolute duty to make something safe. See Bryan A. Garner, ed., Black's Law Dictionary, 7th ed. (St. Paul, Minn.: West Group, 1999), p. 926.
2.  See Stuart Speiser and others, The American Law of Torts, vol. 1 (Rochester, N.Y.: Lawyers Cooperative Publishing Co., 1983), pp. 144-145.
3.  Ibid., pp. 144-145, note 96.
4.  See, for example, George L. Priest, "The Modern Expansion of Tort Liability: Its Sources, Its Effects, and Its Reform," Journal of Economic Perspectives, vol. 5, no. 3 (Summer 1991).
5.  The first workers' compensation law was enacted in 1908 to cover certain federal civilian workers. By 1920, most states had workers' compensation laws; today, all states and the District of Columbia have their own programs. See Cecili Thompson Williams, Virginia P. Reno, and John F. Burton Jr., Workers' Compensation: Benefits, Coverage, and Costs, 2001 (Washington, D.C.: National Academy of Social Insurance, July 2003), p. 6.
6.  Depending on the state, workers' compensation requires employers to purchase insurance from either private sources or a public insurance fund, unless they can prove the ability to self-insure. Employers are responsible for benefit payments specified by state statute for on-the-job injuries regardless of who is at fault.
7.  From society's perspective, the optimal level of accident avoidance is the point at which the total cost of additional avoidance is equal to the total additional benefits obtained. However, without cost internalization, an individual firm does not take losses to accident victims into account and thus underprovides accident avoidance, because the costs outweigh the private benefits. Tort liability works by aligning the private benefits of accident avoidance with the social benefits. Markets for products may also lead firms to internalize the costs of accidents when consumers are well informed about the products and the risks of injuries do not fall on third parties.
8.  See the discussion in Priest, "The Modern Expansion of Tort Liability," pp. 31-50. Those scholars also did not note that using the tort system to provide insurance could adversely affect the supply of goods and services by raising companies' costs.
9.  Early attempts to figure out how the tort system could be used to internalize costs in the case of automobile accidents failed because identifying which of the parties should be held liable for such an accident is generally difficult.
10.  Strict liability for product defects can be seen as an evolution of negligence and warranty law. Gary T. Schwartz, in "New Products, Old Products, Evolving Law, Retroactive Law," New York University Law Review, vol. 58, no. 4 (October 1983), pp. 796-852, points to the fundamental "high correlation between product defect and manufacturer negligence, making the issue of negligence not worth the costs and uncertainties of litigation." Judging defective products under strict liability rather than negligence reduces plaintiffs' burden of proof. In particular, a plaintiff need not show exactly what happened inside the factory, only that the product is defective.
11.  Although all states have at least one court of general jurisdiction, 44 states have limited-jurisdiction courts that hear certain types of cases, such as small claims, traffic, or probate cases. See Neal Kauder, Examining the Work of State Courts, vol. 1, no.1 (Williamsburg, Va.: National Center for State Courts, August 1995).
12.  Data for the early years of that period may not be complete, but that would only sharpen the finding of a decline. See Brian J. Ostrom, Neal B. Kauder, and Robert C. LaFountain, eds., Examining the Work of State Courts, 2001: A National Perspective from the Court Statistics Project (Williamsburg, Va.: National Center for State Courts, 2001), with accompanying spreadsheets available at www.ncsconline.org/D_Research/csp/2001_Files/2001_ Tort-Contract_Tables.xls.
13.  The NCSC data are limited to the number of court filings; they do not give details about cases. They also cover only a subset of states. Another possible source of data on the outcomes of tort cases is insurance company records--since a large percentage of tort awards are paid by defendants' insurers--but companies do not regularly make those records available.
14.  Many, if not most, court awards are negotiated among the parties after trial, are reduced by the judge, or are subject to statutory reductions, such as caps on damages. For example, in the widely reported case involving a person scalded by a cup of McDonald's coffee, the trial court reduced the plaintiff's $2.7 million punitive damage award to $480,000 (three times the compensatory damages).
15.  Congressional Budget Office calculation based on "Civil Terminations, 2000," Federal Court Cases: Integrated Data Base (Federal Judicial Center, Washington, D.C.), dataset no. 117, available from the Inter-university Consortium for Political and Social Research at www.icpsr.umich.edu/cgi/archive.prl?study=8429.
16.  Kauder, Examining the Work of State Courts, p. 2.
17.  Also, in some circumstances, compensatory damages are not taxable but punitive damages are. Therefore, plaintiffs who anticipate that the net value of a trial award will be reduced because of taxes have an incentive to settle for an amount below the expected trial award. See A. Mitchell Polinsky, "Are Punitive Damages Really Insignificant, Predictable, and Rational? A Comment on Eisenberg et al.," Journal of Legal Studies, vol. 26, no. 2 (June 1997).
18.  George L. Priest and Benjamin Klein argue, on the basis of a model of the determinants of settlement and litigation, that disputes selected for litigation are neither random nor representative of all disputes. See Priest and Klein, "The Selection of Disputes for Litigation," Journal of Legal Studies, vol. 13, no. 1 (January 1984), pp. 1-55.
19.  Firms such as Jury Verdict Research Inc. provide data to attorneys to help them value claims on the basis of verdicts for similar cases. Using their own private information about the probability of success, the litigants can calculate the expected return from going to trial. Thus, if a plaintiff thinks she has a 30 percent chance of winning at trial, and successful cases like hers average a trial award of $1,000, she should be willing to settle the case for $300 or more--or even a bit less if she is averse to risk.
20.  The state figure is based on the information that 537,000 tort filings occurred in 2000 in general-jurisdiction courts in 30 states tracked by the National Center for State Courts and that those states contain 72 percent of the U.S. population. According to the Administrative Office of the U.S. Courts, 36,586 tort cases were filed in fiscal year 2000; see Administrative Office of the United States Courts, 2002 Annual Report of the Director: Judicial Business of the United States Courts (2003), Table C-2A, p. 132, available at www.uscourts.gov/judbus2002/appendices/c02asep02.pdf. Those statistics represent the number of cases filed, not the number of plaintiffs or defendants.
21.  John H. Beisner and Jessica Davidson Miller, Class Action Magnet Courts: The Allure Intensifies, Civil Justice Report No. 5 (New York: Center for Legal Policy, Manhattan Institute, July 2002). Some observers argue that state courts in general are biased against large corporate defendants; see, for example, Robert J. MacCoun, "Differential Treatment of Corporate Defendants by Juries: An Examination of the "Deep-Pockets" Hypothesis," Law & Society Review, vol. 30, no. 1 (1996).
22.  Federal questions arise from interpretation and application of the U.S. Constitution, treaties, or acts of Congress. Diversity of citizenship cases are those in which no plaintiff and no defendant are citizens of the same state and at least one plaintiff seeks $75,000 or more in damages.
23.  Congressional Budget Office calculations based on "Civil Terminations, 2000," Federal Court Cases: Integrated Data Base (Federal Judicial Center, Washington, D.C.), dataset no. 117, available from the Inter-university Consortium for Political and Social Research at www.icpsr.umich.edu/cgi/archive.prl?study=8429.
24.  General Accounting Office, Extent of "Litigation Explosion" in Federal Courts Questioned, GAO/HRD-88-36BR (January 1988).
25.  Department of Justice, Bureau of Justice Statistics, Tort Trials and Verdicts in Large Counties, 1996, NCJ 179769 (August 2000). The same data also show that tort cases typically involve individuals suing individuals (42 percent of all torts) or individuals suing businesses (39 percent).
26.  Those medians exclude cases with awards of zero.
27.  Presentation by Deborah Hensler at the U.S. Chamber of Commerce seminar "Understanding Asbestos Litigation: The Genesis, Scope, and Impact," Washington, D.C., January 23, 2003, available at www.legalreformnow.com/resources/012303.pdf. See also Stephen J. Carroll and others, Asbestos Litigation Costs and Compensation: An Interim Report (Santa Monica, Calif.: RAND Institute for Civil Justice, 2002), available at www.rand.org/publications/DB/DB397/DB397.pdf.
28.  Another factor contributing to the concern has been a recent surge in new asbestos cases. New filings in U.S. district court for asbestos cases involving personal injury or product liability rose from 5,041 in fiscal year 2000 to 26,818 in 2002 (see Administrative Office of the United States Courts, 2002 Annual Report of the Director, Table C-2A). But federal cases are only a portion of total claims (which also include state and trust fund claims), and the number of new cases (which represent old exposures now coming to light rather than new exposures) fluctuates greatly from year to year.
29.  Department of Justice, Tort Trials and Verdicts in Large Counties, 1996.
30.  Carroll and others, Asbestos Litigation Costs and Compensation.
31.  Data supplied to the Congressional Budget Office by the Physician Insurers Association of America in spring 2003.
32.  See Department of Justice, Bureau of Justice Statistics, Tort Trials and Verdicts in Large Counties, 1996, and Civil Jury Cases and Verdicts in Large Counties, NCJ 154346 (July 1995).
33.  Deborah R. Hensler and others, Class Action Dilemmas: Pursuing Public Goals for Private Gain (Santa Monica, Calif.: RAND Institute for Civil Justice, March 1999).