February 23, 2000
Regulation Through Litigation: Assessing the Role of Bounty Hunters and Bureaucrats in the American Regulatory Regime
In modern life, there are two obvious ways to address threats to human health and safety that may be presented by private commercial activity. One is by litigation, whereby an individual who has been harmed by private conduct sues a business for compensation under common law (or perhaps even statutory) tort liability principles, with the additional objective of creating incentives for businesses to behave differently in the future. The other possibility is through the regulatory state, whereby the government promulgates rules ex ante and an agency then investigates, judges, prohibits or requires certain conduct in the future, and sometimes awards compensation for harm. Both regimes can serve as regulators who set standards for reasonable conduct. But, how do we prioritize, coordinate, or integrate these two institutions to achieve the desired end? How well suited to the task of standard setting and technical evaluation is the tort system? And, under what circumstances do marketplace forces provide all the incentives necessary to alert the public to possible harms and deter businesses from undesirable activities?
To respond to these questions the Manhattan Institute, The Federalist Society and the U.S. Chamber of Commerce held a conference on the subject in February 2000.
This conference was the second in a series of conferences on the newest wave of legal reform issues. The first, Regulation by Litigation: The New Wave of Government Sponsored Litigation was held in June 1999. The transcript of that conference is available on request.
The conference sponsors wish to thank Kim Kosman, once again, for her superb editing services. We also wish to thank the panelists and moderators for their thoughtful, insightful dialogue.
Judyth W. Pendell, Director, Center for Legal Policy, The Manhattan Institute
Leonard Leo, Director, Lawyers Division, The Federalist Society
James Wootton, President, Institute for Legal Reform, U.S. Chamber of Commerce
Panel One: Litigation or Government Regulation?
MR. JAMES WOOTTON: Most of us are familiar with former Labor Secretary Robert Reich’s comment that the era of big government is over and now the era of regulation through litigation has begun. Many assumed that he was an unqualified supporter of that development. He has recently indicated otherwise, however, in describing government litigation as a threat to democracy and warning that we should be careful about using it as a substitute for legislation. The Chamber also is concerned with this development. We recognize that, at the end of the day, there has to be litigation to challenge enforcement actions. Litigation is legitimately used, to some extent, to enforce regulations. But the new breed of government litigation that exists is something we find quite chilling.
Regulation through litigation concerns us for three reasons. First, there is an anti-democratic aspect of the new government suits. If we are not careful about which issues we take away from the democratic process, the constituencies that feel strongly about these issues are going to feel disenfranchised. Court and jury ordered remedies make people feel cut off from government. When issues of great importance are settled by undemocratic means, people feel that they have been shut out of the decision-making process.
Second, when you make available to private parties the power of the state, conflicts of interests arise. State attorneys general give contingency fee lawyers the power to sue. These attorneys are often encouraged to make political contributions to the attorney general or are given suing rights to reward past generosity. When the dollars are so large and the power is so great, there is an invitation to corruption. To give a recent example, the FBI agent in charge of the Morales Five investigation—which is ongoing in Texas—has left the FBI and joined one of the target firms.1 This is the kind of impropriety, or appearance of impropriety, that could become more common if we accept that the right to sue is for sale.
Finally, there is a growing and dangerous view that we are no longer a country subject to the rule of law. If the government can change the rules to prosecute an unpopular industry or to assure a certain outcome, then there is really no limit to how rules are applied to anyone in the process.
MR. TODD BUCHHOLZ: We are here to assess the role of “bounty hunters” and of bureaucrats engaged in the quest for consumer protection, health, and safety. This panel is entitled “Litigation or Government Regulation,” and I wonder whether I’m the right moderator. The way Congress is run, the way agencies are run, and the way the courts are run, the moderator ought to be Regis Philbin—these processes take on more of the tone of a game show than serious government business. The prizes, however, are not contributed by corporate sponsors; they are swiped from corporate shareholders.
We’ve all heard stories about out of control litigation. The McDonald’s cup of coffee and the two and a half million dollar verdict resulting from a spill. The nearly four million dollar settlement from Hooters, the restaurant chain, because two men who applied for the job of waitress couldn’t fill the required outfit. The Supreme Court recently considered whether a high school football coach could pat his players on the backside without constituting sexual harassment. The list goes on and on.
Conservatives applaud various types of deregulation: airlines, trucking, communication, and trade. Many of us are convinced that deregulation has been a major factor behind the new wealth in America, making our economy richer and more flexible. But we are now seeing re-regulation in society, regulation coming from the courtroom.
JUDGE GUIDO CALABRESI: I speak to this topic today not as a judge, but as a legal scholar. As a judge, I am perfectly happy to apply regulations, and I’m perfectly happy to apply the laws that allow litigation. In diversity cases, I will apply the laws of various states, no matter how absurd they are. Obstructing rules that I disagree with, isn’t my job as a judge. Expressing my strong opinions on the subject is, however, my job as a teacher.
In a way, today’s topic is misstated. Instead of choosing between regulation or litigation, we should be more precise and describe the debate as one between regulation and incentives. Should safety be pursued by the imposition of bureaucratic rules? Or should safety be pursued by giving incentives to those who are capable of administering risks?
Courts aren’t capable of administering risks. Agencies aren’t capable of administering risks in any reliable way. The only people, in a free enterprise society, who can administer risk and choose the amount of safety that is worthwhile are entrepreneurs. What encourages entrepreneurs to administer risks correctly? Is it best done by a system of incentives—do it right and make money; do it wrong, and lose money? Or is it better done by setting up a bureaucratic system of regulation that tells entrepreneurs what is and what is not acceptable? I would suggest that the presumption should be in favor of incentives.
Of course, a system of incentives has its problems. Just look at the system of incentives we have at the moment, which is litigation. But for every horrible that exists in litigation, you can have the identical horrible in a system of regulation. The disadvantage of regulation, however, is that its problems are more entrenched. One example is agency capture, which I don’t think is all that important. Much more important is stultification or obsolescence.
Regulation is always out of date and usually hinders the true entrepreneur. Building codes, enacted for safety reasons, regularly get in the way of innovation in the construction industry. Similarly, consider Ralph Nader’s so-called reforms with respect to automobiles. They determine what kinds of automobiles will be accepted, and, in the process, exclude innovators who might find improved safety another way.
But there is something else even more important to think about. To the extent that we are willing to let risk be regulated—to let somebody else decide what is okay and what is not—we are abolishing the entrepreneur’s very reason for being. We like entrepreneurs and enterprise because we do not trust the government to decide what is worthwhile. We do not trust it to decide who will get into the shoe industry or any other type of activity. This is why we create incentives.
To the extent that we regulate, or decide what is worth having, we get rid of any reason for having profit. Long ago, the Chicago economist Frank Knight defined profit as being the return one gets for taking an uninsurable risk, one that can not be quantified ahead of time. Profit is not monopoly rents. It is not return on capital. It is not payment for one’s labor. Those things are available to many people. The person who makes a profit is a person who handles more competently something that other people can not handle as well: risk.
With respect to accidents, we need the same kind of system if we’re going to justify profits. Does that mean having the current litigation system? No. There is much in that system that is wasteful, silly, and ripe for reform. But if the choice is between litigation incentives or regulatory mandates, the preference for anyone who is devoted to a libertarian enterprise system must be for litigation. We have to change our current litigation system, but with caution. Accident costs are real costs; they are costs that are part of the production, sale, and use of goods and services. We want to reduce these costs, but any such reform has real dangers. Under the guise of reducing these costs, we may actually remove them from the only people who can make appropriate decisions. Regulation follows and—at first—most people will be happy because they can live with the regulatory rules. In the end, however, if we do this we will take a long step towards destroying the free enterprise system.
Who can administer accident costs best? Who can decide what is worthwhile and what is not? Sometimes it’s the victim; sometimes it’s the injurer. More often than not, though, it is the entrepreneur who can best make safety decisions. To the extent that victims don’t adequately watch out for themselves, the best way to make them do so often is through regulation. People can be punished if they don’t behave well, or if they are unduly careless. But the entrepreneur usually should not be similarly punished. The better way to encourage the entrepreneur is to say, “if you produce a safer product, you will pay less in compensation for injuries.” Now that can be done by a structured system of costs rather than through our litigation system. If, however, the choice is offered between a system that creates incentives through litigation or a system that imposes regulation, anyone who’s devoted to enterprise and the entrepreneurial system should opt for litigation, despite its many problems.
PROFESSOR DAN KESSLER: I’d like to talk about a current issue in the litigation versus regulation debate: the patients’ bill of rights. What are the merits of using the tort system instead of regulation to promote quality care in HMO’s? Since the patients’ bill of rights means different things to different people, I will consider one aspect of the several proposals for reform, which is a change to federal law that lifts the current federal preemption against suing HMO’s and other health plans for medical malpractice.
The findings here are not definitive. On the whole, however, the evidence indicates that expanding the reach of the tort system to HMO’s—which is what most of these proposals would do—is not likely to be an effective or efficient way to address real and valid concerns about the quality of health care. More generally, my research shows that expanding the liability system actually does just the opposite.
Some suggest that the experiences of states that have adopted state patients’ bills of rights—Texas, Michigan, Missouri, California—demonstrate what the likely effects of federal reform would be. Whatever you think of these state experiments—and there is some feedback from what has happened in Texas—these state bills of rights are largely preempted by existing federal law. The effects of these state law reforms, both good and bad, are going to be much less than the effects of any federal reform. Because of existing federal law, there is really no direct evidence of how effective lifting the preemption on suing HMO’s would be at insuring the cost or quality of care.
There is, however, evidence on the consequences of expanding malpractice liability for the cost and quality of care. My research with a colleague at Stanford, Mark McClellan, shows that incremental increases in malpractice liability lead to more defensive medical practices—precautionary treatments with minimal medical benefit administered out of fear of legal liability.2 We found that an increase in liability adds to what we call the “malpractice pressure” that doctors face: the time, effort, expense, and hassle of defending against malpractice claims and lawsuits. In turn, increasing malpractice pressure leads to more defensive medicine.
If expanding liability to HMO’s decreases the malpractice pressure on doctors, then it might be a way to improve cost effectiveness and quality of care. Alternatively, if expanding liability to HMO’s increases malpractice pressure on doctors—say by leading to more frequent and more complex suits in which doctors are inevitably involved, even if they’re not the defendant of record—then it will lead to more defensive practices and more treatment intensity that is very costly but provides few benefits to patients.
The problem with using litigation to get appropriate health care is that nobody pays for care at the point of decision. Doctors don’t pay for care. Patients don’t pay for care because of health insurance. This is what economists call “moral hazard.” It is very hard to get the right level of care with a system of tort-based incentives when there is moral hazard.
So what do we do? I will suggest two alternatives, although I have no evidence on whether they will be effective or useful. First, we want to get HMO’s to make appropriate benefits decisions on average. Second, we want to provide patients with a kind of insurance against bad HMO behavior, a guarantee against arbitrary and capricious denials of benefits at a time when people are worst suited to handle them.
As for appropriate benefits decisions, we might enact mandatory, potentially publicly financed, HMO quality measures, which would facilitate competition among HMO’s on the basis of the care they offered. These measures could be based on patient surveys, or on outcomes-based studies of the quality of care. In fact, many HMO’s, through private trade associations, are already reporting quality measures in response to consumers’ demands. To protect patients against capricious decisions, mandatory second opinions on benefits decisions could be made available, at a patient’s request, by an outside panel not financed by the HMO’s. In California, at least, managed care plans have already implemented this sort of thing in response to patients’ requests.
To summarize, evidence suggests that more litigation is not a good vehicle for quality control. A patients’ bill of rights that expands the reach of the liability system is likely to lead to more malpractice pressure on doctors, which in turn is likely to lead to more defensive medicine that offers very little in the way of patient benefits. More research into regulatory alternatives that directly address patients’ concerns, like the two I suggested, would be a better solution.
PROFESSOR W. KIP VISCUSI: I’m going to restrict myself to three points. First, regulation works pretty well in terms of creating safety incentives. In fact, when regulation does exist, the safety incentives are generally too great—regulation is too stringent. EPA and OSHA, for example, have restrictive legislative mandates that prohibit benefit costs tests, at least in the interpretation of the agencies. As a result these regulations, as well as other regulations, are too stringent. In terms of FDA reviews, year after year, people complain that there is too much weight on errors of commission, as opposed to errors of omission. Drugs don’t reach the market as fast as they should. The main exception to this trend would be the Department of Transportation, which tends toward under-regulation.
If you look at studies of regulatory impacts, you can actually find significant effects, particularly for the EPA. The air is cleaner. The water is cleaner. To the extent that a problem exists, it is excessive stringency. For this reason, in situations where regulation exists, I’ve advocated a regulatory compliance defense against regulation. The objective is to get firms to provide an efficient level of safety. Once you’ve complied with the regulation that should be sufficient. These regulations create incentives beyond what you would want from the standpoint of economic efficiency.
My second point concerns the relationship between litigation and safety incentives. My colleagues and I have examined data regarding product safety levels and product liability costs by industry, as well as changes in products in response to product liability costs. We find the following patterns. First, incentives work. Economists, such as Judge Calabresi, who are great believers in incentives, have not misplaced their beliefs. Low and moderate penalties on firms increase innovation; in fact, as you increase the level of product liability costs over sales at the low and moderate level, it increases every measure of innovation. You have more patents. You have new products, new product introductions, and more changes in product design. We don’t know that all these changes are good, but at least they are in line with the usual story about how economic incentives operate.
If you have extremely high penalties for industries relative to industry sales, however, that tends to shut down in regard to innovation. This is not an anecdotal hypothesis—it is based on real data across firms and statistical estimations. The companies in this scenario essentially shut down in regards to innovation. We found this, for example, in the pharmaceutical industry in the 1980’s.3 Companies withdrew from innovation with respect to anything to do with contraception or the birth process.
I’ve been doing work recently on the effects of punitive damages. Do punitive damages, which tend to be random and capricious, provide a deterrent effect to careless or irresponsible behavior? The answer is a clear no. I have a paper in the Georgetown Law Journal in 1998 that tried to find a link between punitive damages and any beneficial safety incentive.4 Wherever I looked—toxic chemical accidents, the number of facilities reporting reduced toxic discharges, surface water discharges, accidental deaths, medical malpractice misadventure deaths—I could find no beneficial response to punitive damages.
In terms of what role the courts play, then, low and moderate levels of incentives seem to play a constructive role. Very high penalties and/or random and capricious punitive damages do not. The incentives that work tend to be those present in the more routine types of court cases, not in the heavy hitters that we often read about in the paper.
My third point: jurors do a lousy job of risk-cost balancing. I have a paper coming out this month in the Stanford Law Review that describes an experiment I conducted.5 I gave a series of cases to approximately 500 mock jurors. I asked them to decide whether punitive damages were warranted in these cases, and if so, how high they should be. The cases they considered concerned automobile accidents generally modeled after the GM and Ford Pinto–type cases where the companies did risk analysis. The imaginary companies assessed the benefits and costs of particular safety innovations, and then decided on that basis that costs exceeded benefits.
The results are as follows: if a company does not do any analysis at all, jurors are a little unhappy. If a company does regulatory analysis, assessing benefits and costs, jurors are very unhappy. If a company uses a compensatory damage amount of, say, $800,000 to value life, the juries hit the companies harder than if they had done no analysis at all. Regulatory analysis is, apparently, more offensive then just putting a product out on the market.
The story gets worse. You might think that if a company does an analysis the way the government does its analysis—if it, for example, follows the procedures and value of life used by the National Highway Traffic Safety Administration—that the jurors would be pleased. In fact, they actually hit the companies harder. Jurors hit companies with a higher punitive damage amount when life is valued at three million dollars than when life is valued at $800,000. Jurors, it seems, are trying to send companies a message. They focus on the value that the companies themselves use. If a company values life at three million dollars, jurors assume a higher award is necessary to get their attention.
Companies, if they think systematically about risk and assign a high value to life, get hammered if this information is given to a jury. Economically, at least, they are far better off doing no analysis at all and simply dumping risky products on the market. This is not a good message.
We found other systematic errors as well with respect to hindsight bias. We gave our juries a case involving a railroad accident and a choice whether to fix the railroad track now or later. Most people opted to fix it later. Then the same case was presented with a tragic consequence. The track was broken and not fixed. A train fell off the track. Not surprisingly, everybody wanted to levy punitive damages. Before the fact, jurors would not fix the track. After the fact, they levied punitive damages. I should add that I’ve run the same scenario on a sample of state judges, including one who’s in this room. I’m pleased to report that they did much better. Whereas two-thirds of the jurors wanted to award punitive damages, two-thirds of the judges did not. This result suggests that more authority over the setting of punitive damages should be taken away from jurors and given to judges if we’re to have punitive damages at all.
There are other studies that suggest that jurors do a lousy job of mapping their punishment intent into a dollar value in terms of damage numbers. In short, jurors are not well equipped to do the kind of risk/cost balancing that we would like them to do. They should not be in charge of setting safety levels for society.
Why is my view different than Judge Calabresi’s? The judge is one of the inventors of law and economics, and I claim to be a law and economics person. I think the main reason behind our differences, perhaps, is that we are focusing on different issues. For some areas, such as automobile accidents, the courts do a great job; juries can determine whether or not a driver was being careful. For other things, like trying to set an overall product safety level on a marketwise basis, I’d rather take my chances with government regulation.
MR. EDWARD WARREN: I’m speaking as a practicing lawyer this morning. Over the course of the last three decades, I have been trying regulatory issues. In the last ten years or so, I have practiced tort law, that alternative regulatory system. I have firm opinions about which of these two systems makes the most sense and in which cases. Kip’s remark about distinguishing between problems is important—circumstances dictate which regulatory mechanism is best.
Government-sponsored tort litigation—cigarettes, guns, lead paint—represent the culmination of more than three decades of shift away from the administrative/regulatory approach. Our courts have moved away from traditional tort law concepts and expanded their role considerably. This is true in the context of awarding both equitable relief and damages. Before we even look at the economic questions involved, this trend raises fundamental questions about the respective roles of the three branches of government. Is it wise, let alone democratic, to turn away from legislation and executive branch regulation as the preferred means of addressing health, safety, and consumer protection? How much does this trend call upon the courts to depart from their traditional role of allocating responsibility retrospectively on a case-by-case basis? Finally, what limits on the courts’ equitable jurisdiction remain if they are willing to oversee consent decrees which serve the purpose of statutes or regulations?
There is no time today to compare the recent evolutions of the tort and administrative systems. Suffice it to say, for institutional reasons and reasons of democratic legitimacy, I favor regulation through legislation and delegated authority to the executive branches of government. Further administrative regulation is preferable to litigation from an economic efficiency standpoint.
The first question to ask when considering any particular regulation, whether it’s by litigation or by administrative agency, is, is it warranted? I’m very conventional here. I think that the need for regulation is determined by market failures, such as a breakdown of competition, or identifiable social costs that are not being internalized in the price of goods or services. It is also important to ask whether administrative costs permit firms to internalize the social costs by private ordering. The framework for deciding this question was provided by Ronald Coase. If the answer to the regulatory question is “yes,” then the issue is how the social cost question should be solved. Even when the private ordering is not superior, however, we have to ascertain whether regulation will do more good than the social harm of not regulating.
Having set these hurdles for regulation, I still think that much, maybe most, of our health, safety and environmental regulation over the last thirty years—at least the statutes—pass these tests. This is not to say that specific regulations always do. Indeed, I think many regulations don’t pass the rudimentary cost benefit test, especially at the margin. Often they ignore risk-risk trade-offs, which are very important.
If these are the questions that need to be asked, it seems that the courts are ill suited to answer them. There is no reason why courts should defer to self-interested private attorneys—for the plaintiffs or the defense—in trying to address these issues. I don’t think government attorneys, including state attorneys general, are going to be any more effective, objective, or neutral in answering these questions. Public choice concerns and all the problems with legislation notwithstanding, democratically elected legislatures enacting statutes administered by executive branch officials are much more likely than the courts to get these questions right.
Despite my concerns about institutional competency and economic efficiency, I do not expect that judges are going to be excusing themselves soon from the regulatory scene. There are some things we can do, however, to mitigate the excesses of the present system. First, we have to recognize that administrative regulatory statutes and the tort system often overlap directly. They regulate the same problem, sometimes in an inconsistent manner. This overlap can be reduced by a more generous recognition of the scope of federal preemption in appropriate cases and by broader acceptance of compliance with federal standards as a defense in tort actions.
Second, courts can be encouraged to follow the Supreme Court’s Amchem case. They can also refuse to exercise continuing jurisdiction when consent decrees extend far beyond the underlying case or controversy. This kind of activity transforms the court into a de facto administrative agency. During the AT&T breakup in 1984, for example, Judge Greene became a virtual alternative to the FCC. This is a growing problem.
Finally, as Kip has said, we need to reassess the use of punitive damages. There is no relationship, it seems, to the economic basis for regulation. That concept needs to be abandoned, or at least reevaluated, and reformed to a substantial degree.
MR. BUCHHOLZ: The UK, Canada and Australia don’t have juries handling personal injury cases. My question for Judge Calabresi: does his enthusiasm for the court system extend to issues of personal injury and other liability cases?
JUDGE CALABRESI: The question of juries is what people focus on in this debate—as if a system of incentives requires juries. It does not. My impression is that juries are not bad when it comes to determining liability: who ought to pay, where the loss ought to lie. Juries are problematic, however, when it comes to determining the size of damages. I don’t think, though, that it does any good to refer to the lack of juries in England or Australia. The reason that juries do what they do in America is clear enough. They are concerned with something more than appropriate levels of safety. Juries in the United States understand that, in many accidents, the only way the injured party will get medical coverage is through a huge verdict. Decisions are made on this basis, whether or not it is fair to the defendant. It is silly to charge juries with doing something wrong when they see their role differently than we do. If I could redesign the system, I would have a system of automatic incentives. And juries would not determine the size of damages. But unless we move to such an automatic liability system we are not going to get very far politically with a plan to get rid of juries.
I agree with the other panelists that one has to look at many different areas to see whether regulation works better than incentives. And I specifically agree with Dan Kessler that incentives in the medical malpractice area work terribly. People in this industry respond to incentives but they respond the wrong way. The incentives we have created here are wrong; but we do not seem to be capable of figuring out a system of incentives that works in this area. Incidentally, we have also yet to figure out a system of regulation that works in the health care insurance context. This area is so dynamic that all the worst faults of regulation emerge there, too.
The problem with health care reform is this: we would like things to be better than they can be. As a result, we look for ways to control the industry, and this simply makes things worse, whether it’s done by regulation or litigation. It would be far better to get out of the business entirely—we should figure out what level of medical insurance is needed, establish that, and beyond that realize that the true horribles will be taken care of regardless of what we do. Unless we create such a compensation system, however, juries are going to continue to respond to the fact that they think people need to be compensated.
PROFESSOR VISCUSI: Let me speak to the jury issue. My view is not that we need to abolish juries. Indeed, I think there is value in citizen participation in the legal system. Having said that, I think the right question—which is the question Judge Calabresi posed—is what is appropriate to expect from lay jurors when they decide complex cases?
EDWARD WARREN: About a hundred years ago, Learned Hand wrote a piece in the Albany Medical Annals on expert testimony and what we ought to do about it.6 For Judge Hand, expert testimony is simply a form of opinion testimony—in the common law, strongly disfavored. What we are asking jurors to do is to decide between competing experts who are presenting testimony that is beyond their comprehension. With respect to these kinds of scientific questions, Hand argued that the court should have its own outside expert or panel. Judges should empanel a series of experts to address the questions, and then instruct the jury on questions of science just as they are instructed on questions of law.
It has taken 100 years, and we still haven’t come around completely to Judge Hand’s position. The Supreme Court’s decisions in Daubert,7 Joiner,8 and Kumho Tire,9 however, are moving in the right direction. Justice Breyer has been a very thoughtful participant in this debate.
Setting damages—especially punitive damages—is exceedingly difficult for lay jurors. They do not understand that health, safety, and consumer protection issues are major economic regulatory questions. They treat these cases simply as judgment calls about the fault of the defendant before the court in a particular instance. The jury is asked, “Is a punitive award justified? Is the conduct willful and wanton?” and little more than that. The charge which is given to the civil jury in the punitive damage context is terrifically open-ended. With little or no guidance or instruction from the judge, it’s hardly surprising that wildly varying and capricious awards exist. Even assuming, for argument’s sake, that the rules that govern punitive damages make sense, they are impossible for jurors to carry out in any realistic way.
In sum: when considering punitive damages, just as when considering expert testimony, we have to be realistic about what we can fairly expect of lay jurors. We need to develop a system, if we’re going to retain punitive damages, where the courts play a much more substantial role in guiding and governing the process.
JUDGE CALABRESI: On the topic of punitive damages, I think the courts should be more directive on damages in general. Judges should use remittitur and things of that sort, and that applies to punitive damages as well.
When we talk about punitive damages, though, we often confuse two things. There is the way punitive damages are actually awarded, which has all of the flaws that the various speakers have referred to. This is quite different from the question of whether there is a role for damages that are more than what is necessary to compensate a particular victim in a particular case. On this latter issue, there is a fine article by Mitchell Polinski10 that explains that one of the functions of punitive damages is not all to be punitive. These damages are there to insure that the cost allocated isn’t merely the compensatory damages awarded in the relatively few cases that go to court and in which recovery is given. The true cost also includes those damages suffered in all the cases that do not go to trial. If the fine that is levied on a thief were the value of the property stolen, thieves would steal all the time because they do not get caught that often.
Similarly, people do not pay damages all the time that they do harm. Accordingly, punitive damages are used as a kind of multiplier—a company must think beyond particular cases to make a correct cost benefit analysis. But, though correct, this conception of damages is also highly problematic. This is because it, in effect, gives juries the freedom to create a monster, and this is so even though the awarding of some punitive damages does have an underlying reason.
When you call that “punitive,” and you give it to the jury, and then you say that these go to the particular victim, you’ve created a monster, a monster to do something that, in an another context, would make sense.
As I stated earlier, the issue isn’t litigation versus regulation, but rather what kind of incentive system is preferable to the bureaucratic system of regulation. For that reason, there should be a role for damages that are not punitive, as we call them today, but that perform a similar function.
There is another important consideration in this context. When a defendant in a tort case, like Ford or General Motors, tells the court that it made a cost benefit analysis and, on that basis, decided that a safety feature was not worthwhile, a jury is going to award huge punitive damages. No jury wants to be told that safety was not the primary concern. The public does not want to know that, in fact, we are always deciding between life and the cost of saving life. That’s the nature of the human condition. The solution is a system where costs are assessed and paid and nobody speaks a word about cost benefit analysis. There is a tragic cost to being too truthful.
PANELIST: As Plato observed, there is a place for noble lies in any kind of government. Another related problem is that the jury, by the time it is making an assessment about damages—especially punitive damages—has obvious sympathy for the plaintiff. Perhaps damages should be assessed by jurors who are under a veil of ignorance regarding the particular plaintiff. Imagine if a general or a president, in sending young men into war, was forced to choose soldiers on a personal basis and explain his decision to each man’s parents. There are times when objectivity is crucial and sentimentality utterly distracting.
PROFESSOR KESSLER: I agree with Judge Calabresi: regulation versus litigation is really a question of regulation versus incentives. I too favor using incentives through the tort system. The problem is, in some sectors, like health care, we can’t get the incentives right. Because of this limitation, I have a slightly rosier view of the regulatory approach. With regulation, you can aim directly at the social problems that are causing particular losses. The system of incentives, when it is fouled up, is almost impossible to fix.
In the context of health care, the two main problems are getting information to people so that they can make appropriate decisions and dealing with people’s fear of uncertainty, or natural risk aversion. This is what the patients’ bill of rights debate is about. Something that gets people information and that reduces the possibility of having a bad outcome is more appropriate than trying to work with a terrible system of incentives.
PROFESSOR VISCUSI: Let me clarify my argument about juries. I agree that punitive damages are a “monster,” as Judge Calabresi calls them. I don’t advocate getting rid of juries; I advocate eliminating punitive damages. As for the regulatory compliance defense, I would like to take out of the courts all the cases where the companies have complied with government regulations. Again, this is not a question of getting rid of juries. It is a question of keeping things out of the courts that involve market-wide decisions, choices of hazard warnings, and choices of overall levels of product safety that hinge on judgments regarding the market. This is different from whether some particular individual chose not to read the warning or some particular individual chose not to take the precautions.
We might as well talk about a series of cases that are really the motivation for this panel. We have had the state tobacco cases. We currently have gun and lead paint litigation. We have many new lines of litigation emerging in which the products involved are highly regulated already. The question is: should we be using the courts to achieve additional regulation? Consider the state tobacco settlement. As a part of the agreement, in addition to transferring billions of dollars to the states and their attorney friends, there are substantial new regulations in effect that affect the cigarette industry and the selling of cigarette products. Cigarette advertising, for example, is controlled by the court agreement. None of these regulations had their benefits assessed. Nobody assessed the costs of implementation. None of these regulations were offered up for public comment and evaluation. If a government agency had proceeded in this way, it would have been clobbered by OMB. Unsubstantiated assumptions do not survive the federal regulatory process.
My point is simple. If we’re going to start promulgating new and far-reaching regulations, we should do it the right way. Apply procedures that are set up for federal regulations, and don’t hedge when back room deals are made and people begin to barter. I would favor relying on regulation rather than government suits because it is a surer way to establish sound rules and accountability in the system.
MR. WARREN: We should focus on the cigarette litigation and discuss what really happened. The companies settled with the states pursuant to a consent decree that was presented to a judge. That decree was, in essence, a book of regulations. This has two pernicious aspects. First, this is regulation that did not go through any of the economic efficiency evaluations we all recognize as important. Second, and even more disturbing, this process thrust the courts into a new and improper role.
Let’s put aside for a minute the fact that these things were done by state rather than federal courts. It is worth recalling that suits have a case or controversy requirement. The exercise of equitable jurisdiction by the court thereafter, as a matter of continuing jurisdiction, ought to bear some relationship to the original controversy. The idea that lawyers can come into court with a consent decree—a bounded set of federal register regulations—and have the court take responsibility for carrying out those regulations for the indefinite future strikes me as wrong. It is a transformation of the role that the judiciary was intended to play.
JUDGE CALABRESI: I have a couple of correctives. We should not be too guided by the cigarette situation. If we decide on the basis of what happened there what is an appropriate system of regulation versus what is an appropriate system of incentives for safety, we will be highly misled. I happen to dislike the cigarette settlement for many reasons, not the least of which is because it puts the courts in the position of regulating. And that is a mistake. This is, however, not what courts normally do in tort cases.
I also think that the statement Kip made about the benefits of a regulatory defense is unduly influenced by a couple of almost unique instances. All of us would agree that a certain amount of regulation in automobile situations is worthwhile. As a society, we do not say, “go through red lights, and if you have an accident, pay the cost of it,” because we all agree that it is cheaper to have a regulation that tells people they can go through green lights, but can not go through red lights. We all think that it is too dangerous for someone to drive at the age of seven and so we have a regulation prohibiting it. We do not allow a seven-year-old to drive even though she may be able to pay her costs.
That said, none of us would think that because there is permission to drive through a green light or that because somebody over sixteen is allowed to drive that is the end of the story. No one thinks a seventeen-year-old can drive as badly as he wishes and not be subject to paying damages.
Because we have agreed that there are certain minimum levels that everybody should live up to, we have not also agreed that those are the maximum levels. Let me suggest that it is often possible for bureaucrats to determine what the minimum level of behavior should be, but it is particularly difficult for them to determine what is the highest level of safety that is worthwhile. This higher level is exactly where innovation changes things. Once you have set both a minimum and a maximum level, you’ve taken away important incentives. Traditionally, under tort law, the fact that there was a prohibition did not mean that living up to the prohibition protected you from a tort suit. Now, through federal preemption, we are getting just that. This should concern anyone who considers himself a federalist; it is yet another way in which the federal government will destroy state authority.
One further thought about democracy. Yes, courts are less democratic than legislators. Whether they are as democratic as administrative agencies—which I happen to think are not very democratic—is another matter. There is, however, one factor that everyone forgets: who pressures legislators to make changes? Essentially it is repeat players, or those who have a continuing interest in an issue. Those who pay damages will lobby for regulations that they think are right; so will the plaintiffs’ lawyers. The people who are not there lobbying are people who are actually injured. They are not there because, at any given moment, there are relatively few people who are injured. Further, government action is generally prospective only, so the injured won’t necessarily gain from legislative changes. The one thing to be said for courts—perhaps not much, but something—is that they are a place where the individual victim comes in and has the chance to be heard. That makes the discussion of democracy in this matter more complex than Mr. Warren suggested.
MR. BUCHHOLZ: This is why, at the outset, I posed the choice between the ignorant jurors or the conniving legislators and their lobbyist friends. Kip, you suggest that satisfying a federal regulation should be a defense against liability claims. In an age where we applaud the deregulation of many industries, would that be an invitation for re-regulation? Wouldn’t industries want more rules, and thus more defenses available at trial?
PROFESSOR VISCUSI: Most of the deregulation efforts have been applied to things like trucking, banking, and airline fares. There has not been a big cry for deregulation in health, safety, or environmental areas. We have had many attempts—mostly unsuccessful—to pass regulatory reform bills to reform what regulators do and put rules on a sounder basis. No one, however, has advocated getting rid of health and safety regulations because there are legitimate market failures with things like pollution. You can not leave people to their own devices. OMB has long advocated efficient regulations for health, safety and the environment, as opposed to deregulation. I personally would restrict deregulation to the area of economic regulation.
That’s not to say we don’t want better regulations, like the kind that would retain safety incentives. If you have specification standards as opposed to performance standards, you are not providing enough incentives for safety. Nevertheless, you still have companies like Volvo which will find a way to make safer cars than other car companies. We want to have regulations that retain, as much as possible, incentives for safety. We want to augment those incentives that are already present. The most powerful incentive for safety is not going to come from the regulation. It’s not going to come from the courts. It’s going to come from the markets. Safety incentives are transmitted through the market, through the price that firms face when they’re making these safety decisions. This should be the case, at least for products and jobs involving market transactions.
MR. WARREN: I would like to distinguish between two concepts that appear in the intersection between the tort system and the federal regulatory system. On one hand, there is federal preemption. It is not easy to say what rules should be applied in regard to field preemption and implied preemption. As a general rule, however, there ought to a greater use of preemption in the appropriate cases. This reduces overlap considerably.
On the other hand, there is the question of whether compliance with federal standards can be asserted as a defense in a tort suit. This is governed by state tort law which is informed by the Third Restatement. You can easily find situations where a company has not only complied with a federal standard but has considered and documented alternatives which would be highly costly and achieve very little. The company then gets punitive damages assessed against it because it engaged in the very documentation that is worthy of commendation.
I think Kip is whistling in the dark if he thinks that you can have regulation with respect to safety and not have it encourage other regulation. That isn’t the way regulators work. Once regulators are allowed to decide that a certain level of safety is appropriate, it is a very small step to allow them to determine that one kind of car is better than another kind of car or that a certain number of businesses should be engaged in a certain business. Of course, everyone wants the regulation that they can control or the regulation that does what they think is right. This is wishful thinking.
In 1984, a rather conservative Supreme Court told courts that they had to give deference to regulators in Chevron.11 This was at a time when the regulators were, by and large, non-aggressive. We now have another group of regulators who are considerably more active. Many courts are very uncomfortable because they have to show them similar deference.
We have to adopt a longer vision of what’s going to happen. Think about the enterprise system as a whole. Do we say that people can go into a business and go bankrupt only if a reasonable person would not have entered that business? Do we go into a business only if people who know best about it, i.e. regulators, grant a certificate of necessity? There are a few regulated industries, and by and large, they are not very efficient precisely because they work that way. We allow people in our system to make risky decisions, to bear the costs of those decisions, and to win if they decide correctly.
The ideal is an administratively efficient way to encourage safety. You want a system where, wherever possible, costs are imposed instead of regulation. Those who are best at innovation, i.e. entrepreneurs, should be encouraged to figure out a way of reducing those expenses. Now, there is no guarantee that if a safer product is made, people will actually buy it. Hudson tried to do that a long time ago, and we learned that safety doesn’t necessarily sell. The people at large—those ignorant jurors—don’t understand that a car is safer. They do, however, understand that a car costs less. Cars cost less because they are safer; manufacturers have less to pay. The market works better than a direct safety appeal.
AUDIENCE MEMBER: My concern is that we are overlooking the issue of notice of what the law requires. Consider the Honda case now pending before the Supreme Court. There were elaborate federal regulations dictating when airbags were required in cars. The regulations set dates by which certain percentages of fleets had to be equipped with airbags. Honda followed the regulations closely. Honda, however, was sued by a woman who was seriously injured in a crash because her 1987 Accord did not have an airbag, despite the fact that when the accident occurred, airbags were not required.12
The market would not have rewarded a car company for putting airbags in vehicles prior to federal regulation. Doesn’t the whole issue of notice of what the law requires come into play when a business is sued for failure to do more than what the regulation required? Does this come down to an issue of federal law preempting state tort law?
JUDGE CALABRESI: The traditional libertarian says if it turns out that something you did was costly and did harm, you should pay. If you guessed right, you’ll make a lot of money. That person will have been able to charge more (indeed everybody will have been able to charge more), a free marketeer says, precisely because of taking a prior risk. It’s the same kind of risk as the risk that the market will turn down. It’s the same kind of risk that one takes when he or she makes an innovation. It’s the kind of risk that gives rise to profits because, if the entrepreneur guessed right and developed, say, the airbag/seatbelt combination that gave the greatest safety for its cost, that entrepreneur would make a lot of money. Are we willing to go this far? If we don’t believe we can stand that degree of free enterprise, then we are going to have regulation and control. But such regulation and control calls into question the whole market idea.
AUDIENCE MEMBER: Judge, you yourself say that safety does not sell. Had Honda put airbags in before anyone else, they would not have been rewarded in the market.
JUDGE CALABRESI: Let me explain this with a simple scenario. Take a situation where people have automobile accidents. The automobile manufacturers, let us assume, pay all such accident costs. Imagine that these costs are properly evaluated, not by juries, but through a rational system. In this situation, automobile manufacturers would be able to charge an amount for cars that would cover the costs of accident risks. That cost would become part of the price of cars, as is the price of steel. The automobile manufacturer who came up with a seatbelt/airbag combination that reduced the number of accidents would keep more of that higher price because that manufacturer would have to pay fewer damages. The manufacturer who came up with another combination that resulted in more accidents would end up broke. Safety in this scenario would not be a matter of selling. The consumer would never face the question of safety directly.
MR. WARREN: The Honda case is an illustration of the point that I’ve been trying to make. It is the worst of both worlds to have two systems operating on the same problem. We’d be better off with the tort system that Judge Calabresi advocates, or the regulatory system that Kip advocates. Either one would be better than both. The point about notice gets to the very heart of this matter. Even the definition of a rule in the Administrative Procedure Act is a matter of prospective application. When we have rules coming out of the Department of Transportation, if there is not fair notice, those rules will get reversed by the court of appeals. And yet, here we have a situation where there is the complete absence of any kind of notice as to what the rule is because the rule is established by the court in a particular case. Since the courts’ role is to decide individual cases using retrospective information, they cannot possibly provide fair warning.
A new kind of tort law has been in vogue for the last fifty or seventy five years. The new understanding is that the tort law is a system for establishing incentives—like regulation, it is forward looking. This is a highly questionable proposition, especially in a world where you have dual regulatory systems dealing with the same problems.
AUDIENCE MEMBER: A question for Judge Calabresi. You argue that some form of incentive system through the tort process is better than a kind of mixed up regulatory system. It’s hard to argue with that proposition. If the choice is between the common law tort regime and the kind of modern unreformed regulatory system that exists, I would agree. Unfortunately, this is not the choice we have. Our choice is between the flawed, unreformed regulatory system and a completely unworkable and unpredictable tort system. Certainly the current condition of the tort system makes a tremendous difference. It is hard to defend a system where a coin toss determines liability and a roulette wheel determines damages. If you want to ignore the tobacco case, that’s one thing. But the gun cases illustrate an increasing trend: the litigation system is completely random and unresponsive to old common law rules, which I think generally were efficient. How can you defend a system that continues to move away from that ideal?
JUDGE CALABRESI: I confess to be something of an idealist. As a judge, I make decisions every day trying to make the system that we have a better one, within the limited authority I am given. But when we’re talking here, in the presence of people who can move and shake the system, I think we ought to think broadly about what kind of system we want. If we don’t set our goals high enough, we will get a system that may work for a little while but will ultimately fail.
Many years ago, I worked on a plan to reform automobile accident law in the state of Hawaii. It made no sense to have a different system of law in Connecticut from that in New York or Rhode Island because Connecticut people drive across state boundaries all the time. But Hawaii, because of its location, presented the opportunity of actually accomplishing something. I came up with a system that would have created an adequate system of incentives without regulation. It was not a system that put the liability on the automobile manufacturers, it was a form of first-party system. It was going through the Hawaii legislature until the trial lawyers bought a half time spot during the Super Bowl and managed to kill it off.
The advertisement showed how much money was at stake. Even in those days, half time during the Super Bowl was an expensive proposition. The real problem, however, was that the Chamber of Commerce was asleep. They simply weren’t present. If you think the existing choice between litigation and regulation is a Hobson’s choice—which it probably is—then you ought to support the right people in the right situations and then wait for the political moment when something can actually happen. Senator Danforth tried to operate in this way; he viewed things in the long term unlike many of his colleagues. I would urge you to think in those terms. If you want us to focus on how we can improve the current system, then I believe we should think at a deeper level.
AUDIENCE MEMBER: I’d like to ask a question about rule implementation. In many instances, because of things like capture by regulated industries, agencies have not implemented statutes to the extent that many think they should. The EPA, for example, has essentially given up on many of the businesses it is trying to regulate. Agencies say, in effect, “here is what the law is. Violate it at your peril.” Perhaps I’ve caricatured this position to an unfair degree, but this seems to be a classic example of regulation by litigation.
MR. WARREN: I know something about this issue because I’m dealing with a related case for an automobile manufacturer. Agencies increasingly feel—especially the EPA—that they are unfairly constricted. Agencies do not feel like they can impose additional requirements by regulation; nor do they feel they can push the frontiers of law through enforcement and requirement interpretation. There is, I think, a fundamental fair notice problem presented in these cases. There’s a couple of important D.C. Circuit cases that speak to this issue, specifically the GE and Chrysler cases where that sort of thing happened.
Announcing in a policy statement that rules are changing from X to X plus ten doesn’t satisfy the requirements of the Administrative Procedure Act. Accordingly, these issues are going to be fought on a fair notice basis. Having said that, the phenomenon you identify is one of increasing importance. The best defenses that exist right now are basically arguments about notice and the lack of deference that ought to be given to the prosecutorial staff, as distinct from the rulemakers, which certainly they should get deference. These are important issues, but the solution is for judges, not lawyers to decide.
AUDIENCE MEMBER: I think what prompted the previous discussion of juries was Mr. Buchholz’s observation that in Britain, Canada and Australia juries are not used to decide certain kinds of civil cases. Beyond the role of juries, do any of the panelists think that the experience of other economically advanced democracies has anything to teach us about the proper balance between regulation and litigation? The United States surely isn’t the only place in the world that’s engaged in the quest for consumer protection and human health and safety.
PROFESSOR VISCUSI: To some extent, it may be. I routinely go to conferences organized by groups such as the OECD and talk about regulation. What I’m thinking about is health, safety and environmental regulation; they’re thinking about things like taxi medallions. Regulations, of course, exist in other countries, but our health, safety and environmental regulations dwarf those of any other country. We are way ahead of everyone else, which is not necessarily a good thing. Other countries don’t know what we’re talking about when we refer to the problems of punitive damages. They don’t know what we’re talking about when we say that hazardous waste site cleanups are inordinately expensive. There are certain excesses of the United States system that have yet to emerge in other countries.
JUDGE CALABRESI: It is true: most advanced countries in the world don’t have our systems of regulation. It may be one reason why they’re advanced. Unfortunately, many of these countries are beginning to copy the worse aspects of our system. Instead of moving towards a system of incentives that would move regulation out of the courts—linking damages to real costs and so on—they are moving in the direction of copying our tort system. While they don’t have things like punitive damages yet, this trend still causes me great concern. When I visit other countries, I find myself sounding very different from what I’m sounding like here. I tell my non-American colleagues to be careful and not assume that these things work that well in the United States. I tell them that if they just buy the American tort system, they are buying something that has plenty of problems.
AUDIENCE MEMBER: Isn’t it the case that, until very recently, Australia, Canada and the UK strongly discouraged contingency fees for plaintiffs’ attorneys? Weren’t they considered unethical?
JUDGE CALABRESI: That is true. Most countries did not have contingency fees until recently. Now many do. But what is happening with contingency fees in Europe is, nonetheless, very different from contingency fees in the United States. Contingency fees in the United States are a percentage of the take. What that does is to create incentives for people to bring suits that have a low probability of success as long as they have a very high yield. Accordingly, in the United States one sees many risky, high paying suits being brought. The English version allows a lawyer to recover nothing if she or he loses. But if the lawyer wins, the take is three times, maybe four times, maybe two times, (depending on the contract) what the normal fee would have been under a traditional fee arrangement. That means that a suit that has a very low probability of success is not going to be taken on a contingency fee basis regardless of the probable size of damages. On the other hand, a suit that has a moderate or low payout, but a pretty good chance of success, would be attractive to a lawyer. If we thought about this practically, and not in broad or ideological terms, we could ask if this is better or worse than the American method.
AUDIENCE MEMBER: And if we did ask, how would you respond?
JUDGE CALABRESI: One has to consider this question in the context of the society. In a society where many costs, like medical care, are covered outside of torts, the English way is surely better. For that reason, I think one of the greatest tort reforms we could enact, actually, would be to provide most people in this country with medical insurance. If that were done, we would find that many of the silly things we do in torts today would lose their appeal.