Product Safety Regulation and the Law of Torts
Despite many convincing anecdotes, tort law's impact on technological innovation is inconclusive. The best studies suggest that the relationship, if one exists, is complex and multidimensional.1 Although real-world anecdotes are suggestive, they are unreliable as a basis for making policy. One hears many stories, for example, about the negative impact of tort suits on the practice of medicine. However, research has shown that only one in 10 injuries attributable to medical malpractice results in a lawsuit.2 This finding undercuts the contention that an avalanche of frivolous malpractice suits has made doctors excessively cautious. Similarly, the more extreme claims linking product liability suits to lowered levels of research and development appear to have no basis in fact. One study has even found that at low to moderate levels of expected liability, increased liability costs encourage research and development.3 Only at very high liability levels is the effect negative.4 This result is quite plausible since increases in expected liability costs could induce spending on innovations to improve safety.5
Nevertheless, even if the alarm expressed by some observers is overstated, tort law may still be creating inefficient incentives for product innovation. If we believe that the market does not create efficient incentives to produce safe and healthy products, perhaps direct regulation is superior to tort law. If tort law will remain a fixture of the American legal landscape, perhaps it can be redesigned to complement the regulatory system rather than work in opposition.
TORTS VERSUS STATUTES
Both tort law and statutory law have regulatory effects. This paper will address the question of how tort law and regulation by statute should fit together. While some critics contend that the size of jury awards implies that the tort system should be entirely abandoned as a way of regulating product quality, that is an unrealistic proposal. Nevertheless, one can distinguish the relative efficiency of tort law and statutory law in alternative settings. Different situations may require reliance on one system over the other, as the following examples suggest.6
Statutory law works best in the following circumstances:
If the harm is very diffuse, with many people harmed in a small way. No one has much of an incentive to sue individually, and even though class action suits are an option, they are not always effective.
If damages are imposed on large numbers of people. In these cases, the tort system, which operates on an ex post and case-by-case basis, may be less effective than an ex ante regulatory system. The latter approach facilitates economies of scale and conserves on information in cases of individualized but similar harms. Such problems are most effectively controlled through a standard-setting process at the government level.
When the damages cannot be tied to a single, identifiable source. For example, if polluting smokestacks are causing many people to suffer, there is nothing to be gained by reducing the problem to a set of disputes between particular individuals and particular smokestacks. Pollution damages are a statistical problem, and it is a waste of resources to try to control damages through a set of individualized decisions in the tort system.
If the companies causing the injuries are too poor to pay for the harm they cause.
In contrast, the tort system is more effective for regulating very low probability events because it may be administratively cheaper. Developing, enacting, and enforcing regulations is a time-consuming and expensive process, and one does not want to burden the system by over-regulation.
In theory, the relative importance of these factors should help one choose between regulating through the incentives provided by the tort system or through an ex ante statutory system. However, in practice, in areas like toxic torts, product liability, and medical malpractice, the line between the two systems has blurred. The courts have innovated by making themselves into little regulatory agencies and setting up institutions to administer their judgments. This is a trend that has little to recommend it.
Courts are not very good at acting like agencies. When one sees this happening, that is an argument for establishing regulatory institutions.
TORTS AND STATUTES AS COMPLEMENTS
Given the existing interconnections and overlaps, is there some way to make the tort system and the regulatory system complementary, rather than competitive? Three possibilities exist.
First, when various dangers exist that have not been regulated, the tort system provides a stop-gap measure pending the passage of legislation. Thus in Larsen v. General Motors (391 F.2d 495, 506 [8th Cir. 1968]) a federal court of appeals holds that "[t]he common law standard of [reasonableness] … can at least serve the needs of our society until the legislature imposes higher standards." The court recognizes that a problem exists and acknowledges that there has not been a systematic approach from the relevant regulatory body. Thus, the courts are filling in the gaps, albeit in an imperfect way.
Second, statutes are often treated as baselines in tort suits. The federal courts have taken this approach in product and occupational health and safety cases. They have ruled, for example, that the Food, Drug, and Cosmetic Act sets only minima and does not preempt tort suits for damages (Abbot v. American Cyanamid Co., 844 F.2d 1108 [4th Cir. 1988]). Similarly, tort suits involving automobile design (15 U.S.C. § 1397(c) and e.g. Sours v. General Motors Corp. 717 F.2d 1511, 1516-1517 [6th Cir. 1983]) or exposure to nuclear materials (Silkwood v. Kerr-McGee, 464 U.S. 238 ) are not preempted by regulatory statutes.7
In the occupational safety and health area, a baseline statute could designate the most obvious risks, the risks that everyone would agree to eliminate. Then bargaining between workers and management could establish higher standards in particular workplaces or industries. The tort system, then, would make it possible for people to argue that the standards should be higher in particular cases. In the context of a statute that was explicitly a baseline, the tort system would accommodate special cases, given the fact that the world is more diffuse than the regulation indicates. By the same token, a manufacturer should be able to argue that it should not be held to as high a standard as others because of the particular way its product is used. The court system could provide for these exceptions.
The third possibility for complementarity is a tort system that acts as a supplemental enforcement device. This is possible when the negligence rule in torts is equal to or less stringent than the regulatory standard. It is also possible in a pure, strict (or absolute) liability system that imposes liability whenever a product-linked injury occurs. When the tort system acts as such a compensation system, or when the negligence standard is no
higher than the regulatory standard, a company that does not meet either the regulatory or the negligence standard can be held liable and required to pay damages. If the company meets the regulatory standard, it would not be held liable.8
Obviously, problems will arise if the tort system, in practice, imposes standards that are higher than those of the regulatory agency. By doing so, it undermines the notion that the regulatory system sets standards where benefits are balanced against costs. Under such conditions the two systems would work at cross-purposes. The conflict between torts and regulation would be exacerbated by the existence of punitive damages.9
Labeling law provides an example of how tort law can complement direct regulation. In 1993 two circuit courts ruled on federal preemption of tort claims, one under the pesticide statute and the other under the hazardous substance law.10 The statutes contain very similar language. Both clearly preempt state statutes that try to impose different labeling requirements. The laws do not directly address the issue of tort suits.
The cases were decided by the fourth and the eleventh circuits. Under the principles articulated here, the fourth circuit made the correct ruling, and the eleventh circuit did not. The eleventh circuit, in a case involving the pesticide law, simply concluded that lawsuits challenging the adequacy of labels were preempted.11 The fourth circuit took a more nuanced approach. In a case dealing with an exploding paint thinner, the court found that the labels conformed with the law. It then went on to argue that preemption only applies to suits claiming that a label, which complies with the federal standard, should have been stronger. A tort suit is permitted, however, if one can demonstrate that the label did not conform to federal standards.12 In this way, the tort system operates as an enforcement device. It supplements the limited resources of public regulatory agencies without undermining statutory goals.
Commentators have long urged legislators and regulatory agencies to charge fees set to reflect the risks created by regulated firms and to establish performance-based standards. Incentive-based reforms allocate regulatory costs to those who can bear them most efficiently, encourage firms to search for innovative ways to reduce harms, and force producers to reflect the risks they impose on society.
Such reforms, however, could be undermined by a poorly informed judiciary. If courts equate regulation with standard setting, then they may treat only command-and-control regulation as behaviorally significant. In a recent case, for example, the Superfund law was described as ''not a regulatory standard-setting statute" because polluters pay for the cost of
abating hazardous wastes "through tax and reimbursement liability" (State of New York v. Shore Realty, 759 F.2d 1032, 1041 [2d Cir. 1985]).
Incentive schemes require a fundamental rethinking of the relationship between tort law and statutory law. Following the conventional wisdom of economists and policy analysts, regulators have begun to use incentives and subsidies to affect behavior in lieu of command-and-control standards. The Environmental Protection Agency has experimented with "bubbles," "offsets," and ''banking." The 1990 Amendments to the Clean Air Act seek to control acid rain through a system of tradeable pollution rights (Clean Air Act of 1990, §§ 403–405). Similar proposals exist to pay workers to use protective devices under the Occupational Safety and Health Act and to establish marketable rights for water pollution.
How should courts handle claims by defendants that incentive-based regulatory statutes preempt tort actions? Judges who view regulation as confined to standard setting might allow tort actions on the ground that these statutes are not "regulatory" because they do not establish uniform standards but "only" create incentives. Yet the argument for preemption of tort law is even stronger in the case of incentive-based regulations than in the case of command-and-control regulation. With standard setting based on either technology or performance, tort actions can complement regulatory agency activity if agency enforcement is not comprehensive or if the fines levied bear little relationship to damages. In contrast, a well-designed incentive system signals to a firm the social costs of its activities. A fee system resembles a tort liability system: No fixed standards are set, but firms respond to the cost of damages. The regulated entity must purchase the right to impose social costs in the same way that a tort judgment requires payment for harms. The main difference is the comprehensiveness of a fee schedule, which the state sets so that all firms are covered. A firm's liability does not depend on the contingency of private litigation and jury damage awards.
If fee schedules are set to reflect the social costs of the regulated firm's activities, then tort actions would be redundant at best and counterproductive at worst. Tort judgments would undermine such a regulatory scheme, especially if courts applied a strict liability standard, the type of standard that some judges have found least "regulatory" (Silkwood v. Kerr-McGee, at 276 n.3). Thus, incentive-based statutes should include a provision clearly preempting tort actions. For example, if the Environmental Protection Agency charges effluent fees, those damaged by the discharges that occur should not be able to sue since this would create inefficient care-taking incentives on the margin.
The only role for lawsuits by private individuals would be to force the agency to enforce its own rules; such suits might permit private recovery of damages for harm caused by lax enforcement. For example, the Consumer
Product Safety Act permits suits for damages against firms that violate agency rules. In situations where the damages are too diffuse to motivate private litigation, the recovery could be some multiple of fees that the agency could have exacted and could be paid to the Treasury with the public interest litigant recovering legal fees. Thus, although ordinary tort actions would be preempted, certain specialized private remedies might supplement agency enforcement just as tort actions do which use regulatory standards as the standard of negligence.
Tort law provides more than a set of regulatory incentives; behavior modification is not its only legitimate function. It is also a compensation system triggered by victims' complaints. If a regulatory statute bars private tort actions, those who were previously able to sue for damages will be disadvantaged, a result courts seem reluctant to permit. In finding that Karen Silkwood could sue for punitive damages in state court despite a federal statute that preempted state regulation of the nuclear industry, the Supreme Court noted that the statute did not provide for compensation and stated that "It is difficult to believe that Congress would, without comment, remove all means of judicial recourse for those injured by illegal conduct" (Silkwood v. Kerr-McGee at 251). If compensation of victims is not addressed by a purely regulatory statute yet remains a policy goal, conflict may arise between the statute and tort law. Compensation-oriented courts may apply conventional tort doctrines that are at cross purposes with regulatory policies.
We need to focus on situations where regulatory policies conflict with a compensation-oriented tort law. Where truly innocent victims exist, denying compensation to those who formerly could bring damage actions may be unjust and unwise. Yet retaining conventional tort actions in the face of regulatory statutes can undermine the behavioral impact of statutes. Other solutions must be found to the problem of providing compensation.
If the victims are numerous and their losses fall into broad, easily identified categories, such as lost limbs or particular types of cancers, then the compensation goal could be served by direct subsidy programs similar to workers' compensation or the black lung compensation program. In contrast, if the victims are few in number and their problems are idiosyncratic, the law should either permit private rights of action for damages analogous to those permitted under the Consumer Product Safety Act and the Comprehensive Environmental Response, Compensation, and Liability Act (Superfund),13 or it should allow tort actions under strict liability principles solely as a means of achieving compensation.
The tort system deals inadequately with problems that do not fit easily into traditional tort categories, problems such as latent cancer risks and harms with attenuated chains of causation. The innovations that the courts have developed to manage class actions and consolidate cases are transforming the courts into quasi-regulatory agencies. Real agencies are likely to perform better than awkward judicial hybrids that have many of the disadvantages of both forms.
If Congress reforms the regulatory system to rely more heavily on incentive schemes, the judicial role should become even more modest. Under incentive schemes that require firms to pay for the damage they cause, statutes should preempt tort actions in order to avoid overdeterrence. For programs affecting many people, compensation should be effected through a separate system of social insurance. Private lawsuits would be permitted under the statute only to compel regulated entities to comply with existing regulatory standards.
But in policy areas that have not yet been reformed, a limited role remains for tort law or, at least, for private causes of action embedded in statutory schemes. Negligence law can be complementary to command-and-control regulation if it adopts the agency's standard not just as a minimum but as the measure of due care. Conversely, a true, strict (or absolute) liability regime would obviate a judicial risk-benefit calculation; only a determination of causation would be required. The choice between negligence and strict liability should then depend on how society evaluates the importance of giving victims an incentive to take care versus the distributive effects of initially shifting all losses to injurers.
An efficiently operating system of tort and regulatory law might indeed affect the research choices of business firms, but that would be a result to applaud, not condemn. If manufacturers are induced to take into account the costs imposed by their products on society, this will give them an incentive to make appropriate research and development choices. Innovation will not be discouraged, but it may be redirected. The current mixture of tort law and direct statutory regulation, however, does not appear to conform to the economic ideal. Policymakers should, however, seek a more consistent system, not impose artificial limits on either tort judgments or regulatory initiatives.
Portions of these remarks are derived from Susan Rose-Ackerman, "Regulation and the Law of Torts," American Economic Review-Papers and Proceedings , 81:54-58 (May 1991). Amore extended discussion can be found
in Susan Rose-Ackerman, Rethinking the Progressive Agenda: The Reform of the American Regulatory State New York: Free Press, 1992, Chapter 8, pp. 118–131.