Federalism and Products Liability Reform
Michael I. Krauss, Professor of Law, George Mason University
EARLY DRAFT—please do not cite without author’s permission
Life in America is safer and less risky than it has ever been. Yet the business of tort law, which forcibly re-allocates risks which one person has wrongfully imposed on another, is thriving as never before. Reliable data is hard to come by, in part because many tort suits settle before they become public. Nonetheless, one thorough study indicates that between 1930 and 1994 the total cost of tort liability in America grew almost four times faster than the rate of growth of the American economy. From 1984 to 1994 alone the increase in total tort liability was 125%. In one state, Alabama, punitive damage awards increased from 1980 through 1985 at a rate 440 times the rate of job growth. Tort awards (including the costs of tort litigation) now consume upwards of 2.6% of gross production in some areas, according to another study.
Record tort awards in individual cases are set each year. The year 2000 gave rise to a staggering $145 billion award in the Engle Florida class-action lawsuit against tobacco companies. Enormous awards remain the exception rather than the rule, but they are growing more common. Lawyers’ Weekly’s “top ten verdicts of 1999” reported two verdicts topping $1 billion; and all of LW’s “top ten” verdicts for 1999 exceeded $100 million. In fact, the total dollar value of 1999’s top ten awards was twelve times the 1997-dollar amount. Only the nation’s largest award from 1997 would have been large enough to be included in the 1999 “top ten” list. Topping the 1999 list was a $4.9 billion jury verdict against General Motors, in a case where a GM vehicle burned after being rear-ended by a drunk driver traveling at 70 mph.
A solvent defendant must pay the tort award as well as his own attorney’s fees and costs. Only about half the cost of tort adjudication makes its way into any victim’s pockets. The remainder goes to pay court costs, parties’ lawyers’ fees, expert witnesses, and the like. For corporate defendant these costs are always ultimately borne by its employees, shareholders, and (in cases where the corporation is held liable despite having behaved appropriately) by consumers of its products. 
In some cases tort liability will change a corporation’s behavior. Part of this change may be for the better. Surely it is good that manufacturers engage in cost-effective quality and design control, for example; if it requires contemplated tort liability to provide this incentive, so be it. Nevertheless, some behavior modification accomplished by modern tort law is arguably contrary to our national interest. Fear of substantial liability for certain risks, if these risks are not easily calculated, has undoubtedly led firms to avoid activity which might well have accomplished much good. Thus, a recent article in Science reports that liability concerns have led at least two companies to delay research on an AIDS vaccine, while a third abandoned its HIV project altogether. Bendectin, the only treatment proven effective against “morning sickness,” is no longer produced because the feared cost of defending against groundless tort suits was greater than the expected profits for the drug’s patent holder.
Products that are never developed may not be “missed” by consumers, though of course the latter remain impoverished even if unaware of the gains they might have procured. In any event, at the “retail” level of tort liability, power tools now carry ridiculous warnings that no purchaser reads but that all pay for. Safe new extension ladders incorporate such a significant “tort insurance” premium that many consumers continue to use their rickety old model. The premium for what one might call “transfer-based products liability” will vary from product to product. In some instances, it may be no greater than the state sales tax. In other cases, it may represent a substantial percentage of what would otherwise be the product’s market-clearing price.
Tort liability extends beyond responsibility for products, of course. It is widely claimed that fear of excessive medical malpractice liability has caused doctors to order redundant and expensive diagnostic tests, and even operations which would not be justifiable on medical grounds alone. But mass product liability suits, unlike medical malpractice and other areas of tort liability, are subject to an intrinsic structural bias that exacerbates the effects of current trends. That bias is, I think, to a great extent an anomalous function of current choice-of-law rules.
An expansion of class action litigation, as well as more “creative” individual lawsuits have resulted in an explosion of liability claims associated with mass-produced products. From automobiles to asbestos to intrauterine devices to heart valves to breast implants to, most recently of course, cigarettes and firearms, manufacturers of products have been exposed to crushing liability. The Rand Corporation has concluded that product claims constitute an increasingly large percentage (currently sixteen percent) of all federal tort litigation. Fully 31% of product claims result in mammoth awards (over $1 million), which remain exceedingly rare for non-product tort claims. Punitive damage awards are vastly more likely in products cases than in other tort cases.
Many tort suits (for instance, nuisance proceedings, medical malpractice cases and automobile collisions) pit what might be called an indigenous local plaintiff against an indigenous local defendant, and concern an incident that occurred locally. When this type of suit is litigated before a jury, there is no geographic bias against either party. Nor is there what is sometimes termed a “public choice” problem: neither party can expressly or impliedly maintain that a victory will bring money into the locality, or will keep money from leaving the locality. Other kinds of bias (class, race, etc.) are of course possible in these “indigenous” tort cases. Parties can attempt to guard against these kinds of bias through challenges to the composition of the jury venire.
Other tort suits (especially those invoking the doctrine of respondeat superior, which holds an employer liable for the negligent behavior of an employee on the job) set local indigenous plaintiffs against local firms. The jury is after all composed only of individuals -- so the local firm does suffer some structural bias here. Such a jury may be tempted to transfer wealth to the individual plaintiff. In many cases, however, this redistributive temptation may be offset by the jury’s desire to maintain employment and economic activity in the locality. It is hard to know how these offsetting incentives ultimately unfold – they might conceivably result in a rough structural neutrality between plaintiff and defendant.
Early products liability suits tended to oppose individual plaintiffs and local firms, as most products were manufactured near their place of consumption. With the advent of successive “paradigm shifters” such as assembly-line production, the interstate highway system and mass electronic marketing, markets for goods have become national. Today, products liability suits are very likely to pit an in-state individual against a corporate defendant with few if any local ties.
To check this postulate, Chris DeMuth of the American Enterprise Institute counted published New Jersey products liability cases in 1900 and at twenty-year intervals up to 1980. New Jersey is an “active” products liability state, but it is also a heavy manufacturing state. DeMuth found a distinct tendency to sue out-of-State defendants, which has likely increased greatly since his last datum point of 1980:
Manufacturing Facilities for New Jersey Products Cases
Though products today are typically manufactured in one location for national distribution, they are subject to more than fifty separate bodies of law when and if a consumer complains that she has been injured by it. This is because products liability law, a subset of tort law, is not one of the named areas of federal competence under our Constitution. Products liability law in particular, like tort law in general, has always been a component of state law. State tort law is of course at its base common law, that is, it is developed not ab nihilo by legislatures but incrementally by state courts.
The use of state law is mandated even if a products liability suit is tried in federal court. Lawsuits may be initiated in, or removed to, federal court even if they involve only ‘state’ products liability questions, because of federal “diversity” jurisdiction if a case pits citizens of different states against each other. The United States Supreme Court held in the famous Erie Railroad v Tompkins case that state common law, not any then-developing federal common law, must prevail in diversity cases. The Supreme Court recognized that common federal “diversity” law would quickly swamp state law, and would in effect nationalize areas of jurisdiction left to the several states under our Constitutional division of powers. In other words, federal diversity jurisdiction is a matter of procedural protection, not of substantive uniformity of legal rules, the court held. Diversity jurisdiction is meant to assure the out-of-state litigant (typically the defendant) that his state citizenship would not bias the court against him. It was not designed to authorize federal imposition of substantive solutions to legal problems.
National manufacturers consequently must reckon with more than fifty bodies of law. Though on any given subject matter states’ rules may be quite similar, the multiplicity of laws is not without costs. National manufacturers must, for instance, retain attorneys aware of tort rules and precedents in each of the states. In some cases these administrative costs may create inefficiencies, as was noted first by Judge Posner, then by others. This is especially true if ‘choice of law’ rules operate to exacerbate substantive inefficiencies, as will be made clear below. On the other hand, national corporations cope rather well with various state highway codes (UPS trucks run in every state) and contract rules (Mobil Corp.’s outside counsel are presumably familiar with the contract law of their jurisdiction).
Retaining state law has distinct advantages from the perspective of comparative law. Justice Brandeis, among others, recognized that states offer competing laboratories in which solutions to problems can be tried and tested: “It is one of the happy incidents of the federal system that a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.”
In the typical products liability case, a consumer has purchased a product, has been injured while using the product, and sues to recover damages resulting from that injury. Most purchases still take place close to home, almost all product use takes place near the home, and no state hosts a plurality of manufacturers’ head offices or factories. Thus, injured consumers are by and large injured in their home state, which is also the state in which the product was purchased. The product, however, more often than not was designed and manufactured in another state.
Assume for the moment that the victim sues in her home state (call the state of trial “the forum state”), that that state’s court finds it has personal jurisdiction over the suit, and that that court concludes that its own substantive products liability law applies to resolve the dispute. Such a suit, then, pits an indigenous individual plaintiff against an out-of-state corporate defendant, and the plaintiff’s state’s laws apply.
So far we have no reason to believe that any systematic injustice is about to be committed. After all, the laws of the forum must be generally applicable, i.e., they must apply even whether the defendant manufacturer is located in-state or out-of-state. Imagine, though, that the forum state’s products liability rules are ambiguous in some way that bears on the dispute at hand. Imagine, for instance, that the defendant offers a legal argument that is powerful, but not clearly dispositive, and that if the forum court agreed with this argument it would be forced to take the case from the local jury and declare the local plaintiff’s suit groundless. At the margin, will a state judge be tempted to transfer wealth in-state by nudging PL law to a slightly more pro-plaintiff stance? Of course, the previous state of the law may have been ‘optimal’ in some sense, and the incremental move now made would therefore be inefficient. But the cost of this inefficiency would be borne by shareholders, workers and consumers across the nation, while all the benefits of the move would accrue inside the state.
Lest the reader think this is academic speculation, Justice Richard Neely of the West Virginia Supreme Court disclosed, in a 1988 book, that he felt motivated to adjust product liability rules away from their efficient level whenever this brought money into West Virginia. Justice Neely implemented his ideas, while beseeching the United States Supreme Court to stop him before he trespassed again, in Blankenship v General Motors. In Blankenship, a so-called “crashworthiness” case, a plaintiff who had negligently caused his GM vehicle to crash alleged that its poor design exacerbated the injuries he suffered in the collision. In most “crashworthiness” cases one difficulty is demonstrating that the allegedly defective design actually worsened plaintiff’s injury. Since no crash can be replicated exactly, how can one know how a “well-designed” car would have stood up to this collision? And what should the court do if the “well-designed” car appealed to by the plaintiff is not made by any manufacturer? On these points, two general trends had emerged in the case law:
· The first requires the plaintiff to “offer proof of an alternative, safer design, practicable under the circumstances, … of what injuries, if any, would have resulted had the alternative, safer design been used.” The costs of producing such evidence are typically prohibitive, especially if no existing automobile had the design features the plaintiff desires.
· The second approach shifts to the defendant the burden of proving that the appealed-to design was not feasible, or would not have reduced injuries, as long as the plaintiff had offered evidence that injuries were, in fact, enhanced under the defendant’s design choice as opposed to the invoked alternate design. Clearly, proving a negative is difficult, in part because the defendant cannot easily replicate the crash.
In other words, the party with the burden of proof here would likely lose.
The West Virginia court, speaking through Justice Neely, ruled that the first approach was more rational, but decided to adopt the second rule anyway. It did this, it said, because West Virginia consumers were already paying markups caused by those states that had adopted the second rule every time a car was purchased in the state. Justice Neely felt that West Virginians might as well benefit from the rule themselves. Indeed, the Court announced that henceforth, “in any crashworthiness case where there is a split of authority among states on any issue, … we [will] adopt the rule that is most liberal to the plaintiff.”
The dilemma sketched by Justice Neely, and provisionally corroborated by Alex Tabarrok can be usefully conceptualized as the classic “Prisoners’ Dilemma” of game theory. A “Prisoners’ Dilemma” is a predicament in which individuals, acting independently, are impelled to bring about a result that is undesirable from each of their points of view. This particular Prisoners’ Dilemma springs from the reality that products liability suits typically involve local plaintiffs and out-of-state defendants. The most desireable outcome would obtain if each state’s judicial system crafted products liability rules that treated out-of-state parties (mostly defendants) on even terms to in-state parties (most plaintiffs). But if all states had such even-handed rules, an individual state could extract large profits for local residents by “defecting”, i.e., by adopting rules that exploit defendants, most of whom will turn out to be located out-of-state. This temptation could yield sub-optimal outcomes nationally if it were replicated, as is illustrated in the following table which imagines a simple scenario with two states:
PRISONER’S DILEMMA AS CONCEPTUALIZED BY JUSTICE NEELY
«STATE B (PAYOFF IS SECOND NUMBER)
The matrix (the numbers are arbitrary – other numbers would equally illustrate the dilemma) illustrates that state A would optimize its return (i.e., the first number in the pair will be maximized), regardless of state B’s behavior, if state A adopts a rule which exploits out of state defendants. If state B has a neutral rule, state A’s payoff increases from 40 to 55 by adopting a discriminatory rule. If state B has a discriminatory rule, state A increases its payoff from -10 to 5 by discriminating in turn. This is so even though an exploitative rule by A causes a net loss of 35, socially. State B has similar incentives. The expected outcome is therefore the lower right-hand corner, i.e., the situation wherein both states have adopted exploitative, non-efficient rules. This result is “Pareto-inferior” to the upper left-hand corner outcome, which maximizes social benefits.
Note that products liability Prisoners’ Dilemma is unlikely to be neutralized by industrialization. Even in highly industrialized states like New Jersey, the majority of products consumed are produced outside the jurisdiction. As will be shown below, under such conditions courts will have an incentive to provide “Equal Protection” by exploiting in-state and out-of-state manufacturers, without in any way jeopardizing the attractiveness of the state as a site for manufacturing. The fact that New Jersey has arguably led the way in products liability “innovations” is in no way incompatible with the existence of the Prisoners’ Dilemma.
Nor is the dilemma likely to be resolved by state legislative action. For reasons analogous to those explaining the judicial “Prisoners’ Dilemma”, state legislatures are at lease as unlikely as courts to adopt rules clarifying products liability doctrines in ways favorable to out-of-state interests. Indeed, the bulk of legislative tort reforms have concerned such matters as automobile accidents and medical malpractice, both of which tend to involve in-state defendants. Legislation dealing with situations usually involving out-of-state defendants tends to expropriate their assets whenever possible. Even when supreme coordination (as opposed to the lack of coordination inherent in the Prisoners’ Dilemma) is required to do so, out-of-state defendants will tend to be targeted – as the multistate settlement with tobacco manufacturers shows.
Current products liability law would not be the result of a “Prisoners’ Dilemma” if manufacturers could price their products differently from state to state, as a function of the costs (including liability costs) of doing business in that state. If individual state pricing were possible, each state would arguably behave as if in an “autarky”, i.e., as if it alone bore the consequences of its legal decisions. The matrices of Table II would be quite different in an autarky: in brief, any inefficient legal change by state A would result in increased costs for consumers in that jurisdiction.  Because the costs of A’s legal change could not generally be externalized to other states, there would be no strategic advantage to State A making the change.
Unfortunately, current “choice of law” rules preclude the “autarky” solution.
Consider the traditional choice of law rule for torts. That rule, styled lex loci delictus, provides that the law of the state in which an alleged tort occurs governs any lawsuit arising from the tort. Since a tort cannot exist absent injury, lex loci delictus essentially applies the law of the place of injury. Thus, say a consumer in state A travels to state B to purchase a product manufactured in state C. If that product’s use in state A subsequently results in an injury to the consumer, courts in state A will both assume jurisdiction and apply the products liability law of state A to determine whether the manufacturer is liable for the injury. Courts in other states, if they had jurisdiction for some reason, would also apply the law of state A.
If courts in state A use the laws of state A to govern accidents occurring in state A, while courts in state B use the laws of state B to govern accidents occurring in state B, it seems that the autarkial solution might result. Merchants in state A might be able charge “A prices” for their goods, while merchants in state B charge “B prices”, each set of prices reflecting inter alia a given “liability premium”. Nevertheless, this is not possible because of what economists call “arbitrage” and insurance theorists call “adverse selection.” Say state B has legal rules which favor defendants more than do the rules of state A. Say a manufacturer decides to charge a higher (wholesale) price for goods destined to merchants in state A, in order to cover the premium it must pay for unavoidable liability. If this happens, some consumers in A will simply purchase their products from merchants in B, sans premium. Lex loci delictus provides that the law of state A applies to accidents occurring in state A, regardless of the location of retail sale. Thus, consumers in A derive the same tort “protection” when they purchase their product in state A as when they purchase it in state B. Manufacturers will therefore not be able to discount the price of goods sold in state B, to take account of B’s less stringent products liability rules, because low-priced goods sold in state B may well incur the liability of pro-plaintiff state A.
Would lex loci delictus at least tend to encourage manufacturers to leave State A and relocate in State B, with its more favorable products liability law? If such a tendency existed, then this would alter incentives, perhaps sufficiently to avoid the Prisoners’ Dilemma entirely. Alas, there is no reason to believe such an inducement to relocate would exist. After all, under lex loci delictus a manufacturer’s liability in no way depends on its location. A manufacturer’s exposure to liability in any state will be strictly the same, whether the manufacturer is located in that state or in some other state.. Indeed, if state A believes it is siphoning money from other states through its liability rules, as Justice Neely’s opinion implies, it could well choose to use part of that “rent” to subsidize or ‘bribe’ manufacturers to locate or to remain there. Thus, indirectly, lex loci delictus might discourage plant location in low-liability locations.
In sum, under lex loci delictus the “dominant strategy” in states A and B would be to adopt more stringent liability laws than global efficiency might mandate. Empirically, the impossibility of reacting to products liability rules by changing location or by tailoring prices to state law is borne out: manufacturers report that they consider state product liability rules as a factor to which they are unable to adjust.
As a choice of law rule, lex loci delictus has been superseded by “interest analysis” in a majority of states. Under “interest analysis”, the forum court determines which substantive law to apply to a dispute with multi-jurisdictional elements by ascertaining which state has the greatest “interest” in determining the outcome of the case. Using “interest analysis” allows a state to apply its own law to accidents occurring outside its boundaries. Theoretically (but I have found no instance of this ever happening), “interest analysis” also allows a state to decline to apply its own law to accidents occurring inside its boundaries. In fact, courts typically uses “interest analysis” to conclude that the forum state’s ‘interest’ in protecting its citizens “abroad” exceeds the ‘interests’ of the loci delictus state in determining the legal consequences of that accident. Such “protection” is of course only requied if the loci delictus state has liability rules that would favor the defendant. Thus, “interest analysis” is often seen to exacerbate the “Prisoner’s Dilemma”.
In a provocative article, Professor Bruce Hay maintains, to the contrary, that the rise of “interest analysis” helps neutralize some of the tendencies of lex loci delictus, by encouraging manufacturers to locate in low-liability areas. Briefly, here is Hay’s argument:
1. Contrast state A, whose rules result in relatively extensive manufacturers’ liability, with state B, a less pro-plaintiff jurisdiction. Assume also that, for reasons discussed above, manufacturers are unable to price their product differentially in states A and B to reflect this different liability potential.
2. It has been shown that under lex loci delictus manufacturers have no incentive to relocate their plants. But under “interest analysis”, as long as the state of manufacture creates an additional “interest” in that state’s law being applied, then at the margin companies will have an incentive to locate in states with favorable liability rules. For instance, citizens of state B, injured in state B by a product manufactured in state A, might under “interest analysis” persuade the courts of state B to invoke the pro-plaintiff products liability rules of state A. But if the manufacturer were to relocate to state B, there would be no grounds to apply the other state’s law to this case. This provides an incentive, at the margin, for firms to relocate to low-liability states, and this incentive counteracts the court’s natural tendency to favor a local plaintiff.
3. According to this argument, one might expect states with low manufacturer liability to adopt “interest analysis” rules as their choice of law rule, as a way to lure manufacturers into their jurisdiction. Courts with pro-plaintiff products liability rules could be expected to stick with lex locus delicti.
Alas, Hay’s hypothesis does not correspond to the “facts on the ground.” Hay admits that that the states that have adopted “interest analysis” tend in fact to be high-liability states. Activism in one field breeds activism in others: an empirical analysis shows that states which continue to follow lex loci delictus are not only less ‘innovative’ in ‘choice of law’, but also in modifying products liability and other common law rules generally. States whose substantive rules result in less frequent manufacturer liability do not adopt interest analysis for the same reason they do not change their substantive rules. Given the current Prisoner’s Dilemma, the only satisfactory explanation for some states maintaining lower liability rules is that they have some motivation different from maximizing pecuniary gains for their residents. Otherwise, the only stable equilibrium of the game would be one in which every single state adopted high liability rules. Plaintiffs simply fare better in “interest analysis” states than in lex loci states.
A good illustration of “interest analysis” in action is Duncan v Cessna: In that case, plaintiff’s husband, a resident of Texas, was killed in New Mexico during the crash of a Cessna aircraft in which the husband was taking flying lessons. The decedent had traveled to New Mexico to contract with the flight school, and had signed a release holding the school harmless for the general risks of flying. The release did not name the Cessna Corporation, but did state that it protected “any other corporations or persons whomsoever responsible therefore, whether named herein or not”. Under New Mexico law, this release would have benefited Cessna, which presumably could therefore charge lower aircraft lease rates to New Mexico flight schools. But the plaintiff sued in her home state, Texas. The Texas court held that Texas law, which did not allow Cessna to avail itself of the release, applied to the accident. New Mexico, the Texas court wrote, had little interest in seeing its law applied, because defendant Cessna was a Kansas corporation and New Mexico was therefore presumably indifferent to its welfare. On the other hand, Texas, wrote the Texas court, is acutely interested in the protection and compensation of injured Texas residents. The Texas court conveniently overlooked the fact that New Mexico may have a distinct interest in allowing lease terms in the state to reflect the state’s legal rules. [In other words, New Mexico has an interest in liability autarky.] The refusal to apply New Mexico law to the case prevented New Mexico from achieving this result. Unless it can somehow restrict New Mexico flying lessons sto New Mexico residents, Cessna no longer has a basis for charging lower equipment lease rates to charter firms in New Mexico.
Plaintiffs choose the forum state, which will therefore typically the jurisdiction in which they believe they have the greatest chance of success. The forum state is also the state making the choice of law decision. Plaintiffs will only choose a low-liability state if there really is no other choice. And under those circumstances, where typically both the plaintiff and the place of accident are the same low-liability state, the odds are that the law of that low-liability state will apply regardless of whether it uses lex loci delictus or “interest analysis”. Under these circumstances, “interest analysis” will favor local (typically the plaintiffs’) interests. Only when the plaintiff’s preferred state has for some reason barred the suit would interest analysis ever favor defendants. This tendency to use “interests” to favor the local over the out-of-state party is illustrated by Rutherford v Goodyear Tire & Rubber. In Rutherford, plaintiff was injured in his home state of Indiana following the explosion of a tire mounted in Kentucky, in the assembly plant of an automobile manufacturer, which was one of the defendants. Indiana’s statute of repose barred the suit, so plaintiff sued in Kentucky, whose statute of repose was more favorable. The Kentucky court refused to apply Kentucky law, stating that Kentucky had no interest in applying its product liability law to Kentucky products causing injury in other states. The Kentucky court chose to apply Indiana law, including its statute of repose. Note that this solution would also have resulted from a lex loci delictus rule.
“Facts on the ground” do not favor Professor Hay’s hypothesis. Indeed, pro-plaintiff jurisdictions have occasionally refused to use interest analysis to protect their own firms from out-of-state plaintiffs, thereby demonstrating an ideological commitment incompatible with Professor Hay’s prediction of the reasons for their adoption of the “interest analysis” rule.
The Prisoners’ Dilemma involves more than merely the choice of law. After a law has been chosen, it must be interpreted in an individual case. One expects to see more adverse interpretations of the same law against out-of-state defendants than against in-state defendants. This tendency will likely be exacerbated when the judge is elected and must run regularly for re-election.
To claim that “choice of law” rules have precluded an autarkial solution and exacerbated the prisoners’ dilemma, resulting in excessive products liability, is not to claim that no countervailing tendencies against liability exist. If no such tendencies existed then firms would tend to be held liable without any connection to injuries. Spiraling products liability might result in General Motors being held liable whenever anyone dies in a car crash involving a GM car, for any reason including natural causes. That this has not occurred suggests that countervailing forces preclude a complete degeneration of products liability law and interpretation. Countervailing factors might countervail include:
· Judges and juries are subject to more than political and economic temptation. Religious values and other beliefs about individual responsibility clearly influence determinations of liability.
· State and federal constitutional protections preclude overt takings from out-of-state defendants.
· Some direction of juries comes from appointed judges. Appointed judges will not be subject to “rent-seeking” pressures as intense as those afflicting elected judges.
· Out-of-state manufacturers can lobby state legislatures to enact pro-defendant tort reform, emphasizing the first two factors above -- and political contributions from outside the state are permitted.
C. Is a federal takeover of products liability law the appropriate solution to the Prisoners’ Dilemma?
“Prisoners’ Dilemmas” are quintessential coordination problems. Coordination problems can be resolved, generally speaking, in one of two ways: either through the centralized imposition of a solution from outside the afflicted group or through the prior altering of incentives for decentralized choice making. Our search for an autarkial products liability system might involve the former or the latter.
Nationalization of substantive products liability law represents the first type of solution – autarky imposed by removing decision making to the central level, where all effects are presumably felt. This approach has been advocated by numerous observers, and is regularly proposed by Congress. Because of the national market that now exists for products, and because products liability is closely related to economic transactions, a national products liability statute might be rationalized as a constitutional exercise of the “interstate commerce” power.
This paper will, by and large, gloss over the issue of whether Congress has in fact the constitutional right to nationalize products liability. That is because this paper’s claim is that, even if it were constitutional, a federal takeover of products liability law is not propitious. A federal takeover would create several kinds of problems.
Products liability law is a subset, of recent vintage, of tort law, which of course is state-based. Should a federal statute cover products liability, courts would need to juggle federal and state law whenever lawsuits raise both products liability and other tort issues. Is liability of a manufacturer and, say, an employer “joint and several”? How should a plaintiff’s case against the manufacturer of an allegedly defective car be harmonized with her case against the driver who allegedly failed to control that car correctly? How would common-law state doctrines (developed by courts, and subject to revision upon principled argument to a court) coordinate with federal statutes, which are “frozen” and thus beyond the scope of judicial tinkering? Would the federalization and ‘codification’ of products liability law not lead to an increase in inefficient, rent-procuring rules? How would this federal legislation mesh with state tort reform? Since plaintiffs’ and defendants’ lobbies at the federal level are fairly evenly matched, any federal legislation is likely to be fatally ambiguous about this interface, as recent cases have shown. Even if a federal tort takeover were feasible and unambiguous, would it not lead to a takerover of contract law as well? Is the disappearance of the states as the principal locus of private ordering in our interests? Strong arguments can be mustered to the contrary.
Beyond the problem of coordination, and the risk of massive federal intrusion into local law, uncertainty about the content of “perfect” products liability legislation argues strongly against action by Washington. As is the case for all of private ordering, constant and complicated calculations of the utility functions, risk preferences and, particularly, moral stances of individuals is needed to determine the correct allocation of the risks of products among manufacturers and consumers. There is no particular reason to believe, in fact, that a ‘one-size-fits-all’ solution exists – it is probable that moral views and risk preferences vary across individuals and regions. As Michael Greve has recently written, “topography and climate aside, no one would mistake Texas for New York, or Nebraska for Massachussetts.” Diversity diminishes Washington’s knowledge of “correct” products liability rules.
State legislation is, if properly arranged so that costs and benefits are largely reflected inside the state, conducive to a competition (a second and distinct knowledge problem) that will produce the information needed to determine and reflect collective preferences. If a state’s products liability rules are too generous to plaintiffs, or to manufacturers, and if the costs of these rules are reflected within that state, undesirable characteristics of those rules may eventually lead to modifications. The cost of a product might rise tremendously, for example, if prices in one jurisdiction had to incorporate a high premium to cover accidents that would avoidable if only the victim had used due care. Consumers in that state might if given a choice decide to choose competing rules in jurisdictions whose products laws call for a degree of “self-insurance” that consumers find more acceptable. Competition for good laws among states can in this way serve the same purpose as competition for good products.
To the contrary, however, federal legislation is on a par with coercive imposition of one product on all. When a federal private law rule is imposed on the country, it had better be “the right” one (and there had better be one “right” solution for our diverse population) – for it is much more expensive to opt out of a country’s laws than it is to leave a state. As one perceptive commentator notes, “[t]he choice between state authority and federal authority is the choice between competition and monopoly.” As the federal legislature suffers from less competition, it learns over time much more slowly. Learning over time (a third knowledge problem) clearly favors state action over federal legislation.
Some have averred that no binding allocation of products liability risks, by either federal or state authorities, is desirable. Rather, states could compete to propose optional, disclaimable products liability packages, and a federal statute might specify that people have freedom of contract with respect to product risks. Under such a rule, General Motors might market automobiles with the following legally binding (on purchasers, who would be bound to secure the consent of their passengers) statement:
“Warning: we have determined that, given our production and quality control techniquees, one in each 500,000 vehicles is defectively made and will explode during its first year of use, without any fault by the driver. We hereby disclaim liability for injuries brought about by such explosions.”
Although the doctrine of assumption of risk is appealing, as a general matter, to this author, there are many problems with any federal ‘libertarian rule’. Here are four of them:
· The libertarian rule implies that there is no collectively felt need to establish minimal safety standards which manufacturers may not waive. But it is difficult to know a priori that shifting some risks to manufacturers would not produce a collective good. Local collectivities might be sensitive to such collective goods, and desirous of implementing them.
· The libertarian rule is in direct conflict with products liability law across all fifty states since the 1960’s. There is little reason to believe that legislatures (federal or state), courts or juries would react more favorably to the libertarian rule now than they did forty years ago. In addition to seminal products liability cases in each of the states, widespread adoption of the Uniform Commercial Code (with its claim, quite compatible with visions of private ordering, that manufacturers’ express warranties impliedly contradict their disclaimers, making them unenforceable) is opposed to the federal imposition of the libertarian rule
· The libertarian rule largely assumes away the problem of limited consumers’ rationality, and of intrinsic asymmetries in information which make manufacturers much more efficient bearers of certain subsets of risk than are consumers.
· The libertarian rule is oblivious to injuries suffered by third parties (e.g., pedestrians injured by defectively manufactured vehicles), who assumed no risk of defect.
For all these reasons, federal nullification of state products liability rules seems neither desirable nor politically feasible.
If both substantive federal products liability laws and federal nullification of state products liability rules are inappropriate, is there any role for federal legislation in resolving the current Prisoners’ Dilemma? This essay argues that federal choice-of-law rules would be both legitimate and valuable. Choice-of-law rules could resolve the Prisoners’ Dilemma without shutting down the states’ private ordering laboratory.
Without defending this view at length, I posit that two fundamental principles imply and circumscribe federal authority in matters of ‘choice of law’.
· The principle of equal citizens. States must as a general matter treat citizens of sister states on an equal basis with their own citizens. This implies that states may not exploit choice-of-law rules to favor local citizens over citizens of sister states. As has been stated above, in practice both lex loci delictus and (especially) “interest analysis” rules have contributed to violations of this principle.
· The principle of territorial states. The allocation of authority among states is territorial. This fundamental principle is so obvious that the Constitution assumes it to be so. State constitutions and state admission acts make the territorial basis for state sovereignty quite clear. The implication of this principle is that a state’s claim to regulate behavior or to govern a dispute must be based on something that happened within its territory. A state’s interest in extending the territorial reach of its own law is not sufficient to grant it jurisdiction.
These two principles have important implications. They suggest, first and foremost, that the effect of “interest analysis” on products liability, spurred on by Brainerd Currie, does not pass muster under the second principle. As we have seen, interest analysis was essentially developed to extend the reach of state law, i.e., to apply state law to characterize a transaction which occurred in a different state.
Lex loci delictus does respect the territorial principle: the fact that an accident happened inside the territory of a state is a constitutionally sound reason to use that state’s rules to determine the legal obligations arising from the accident. But as practiced, lex loci delictus has produced a Prisoners’ Dilemma under which states are encouraged to exploit subjects of sister states, in favor of their own residents. This is violative of the first principle of legitimate state action.
From this analysis it follows that the constitutional scheme for allocating products liability authority among the states, given our national marketing structure, is no longer complete without a new federal allocation of authority via choice of law rules. This authority cannot reside in each state, which will tend to make choice of law rules that favor its own citizens over out-of-state opponents. The authority to make choice of law rules compatible with the Privileges and Immunities and Full Faith and Credit Clauses of the Constitution therefore resides in Congress, or (failing Congressional action) in the Supreme Court.
Federal choice of law rules are therefore clearly constitutional; indeed, the current dearth of national choice of law rules arguably might even constitute an abdication of federal constitutional duty. But which choice of law rule reconciles sound principles of federalism with the need to resolve the products liability prisoners’ dilemma? There are several plausible candidates, to be sure. In the end, though, a “state of first retail sale” choice of law rule best reconciles our constitutional structure to the current demands of the national market.
One solution, proposed by Dean Harvey Perlman and inspired by the competition for incorporation of companies, is a federal rule which would allow manufacturers a binding choice among the various states’ products liability régimes.
Corporations can currently choose among states’ corporation statutes by selecting a “state of incorporation”, even if this state is not the one in which their facilities are based. In this way, it is said that states compete to have the most efficient incorporation statutes (since fees are derived from this incorporation). Delaware is to all appearances the current winner of the contest. It derives both incorporation fees and legal “royalties” (court costs, lawyers’ salaries) from its success in this competition.
Dean Perlman hopes to export this régime to products liability. Thus, a manufacturer incorporated in Georgia and producing a widget in, say, Florida, sold at retail in South Carolina, might designate Virginia law as applicable to disputes arising from the use of the widget. Virginia need have no territorial connection with the widget. If adequately designated (perhaps via some marking on the widget, or on its packaging), Virginia law would apply. The widget would presumably be priced to reflect expected liability under Virginia law.
Of course, a manufacturer could under this proposal offer in each state an array of liability packages along with its widgets: consumers would purchase their widget with a selection of Alabama-to-Wyoming liability rules, concomitantly priced to reflect their chosen state’s expected liability cost, much as car buyers can purchase different kinds of extended warranties with their vehicles. [In practice, though, adverse selection problems are such that each manufacturer might select a single state’s law to apply to its products, and thereby save legal costs by specializing in the substantive legislation of that state. Consumers, for their part, would consider the products liability rules attached to the product when deciding whether the “price is right”. Presumably, a poor choice of products liability rules by the manufacturer would lead to financial losses, and consumers would be drawn to the products of competitors offering more efficient products liability packages.
Dean Perlman’s proposal has the advantage of allowing consumers to freely select their liability régime. Although multiple choices (Pontiacs with New Jersey or Virginia protection, etc.) are possible, it is likely that a single state’s law would apply to each product. If they want a widget manufactured by company X, consumers will obtain state Y’s legal rules; if they want another state’s rules they will likely have to select a widget made by a competitor to X. The Perlman proposal would in essence allow manufacturers to ensure that all their products, sold nationwide, are governed by the same liability rules. As long as those rules remain relatively stable, a manufacturer would be able to price its products with confidence that buyers are “purchasing” an optimally chosen package of risks.
Perlman’s rule increases predictability, but it makes a mockery of the territorial basis of state sovereignty. No tie to territory (not even to the locus of the company’s incorporation or “citizenship”) is needed to select a given state’s law. The consumer, a citizen of one state, has no necessary territorial link to the state whose law is applied. This consumer never accepted the sovereignty of that state by traveling to it, or indeed by any action other than the clause in the purchase of the product, which purchase itself likely as not took place in a state other than the one whose law is chosen. Nor is the consumer likely to be conversant with the chosen state’s liability rules, as would be the case if the consumer lived in or traveled to the state in question.
Perlman’s proposed choice of law rule comes closest to the federal nullification of state products liability rules, as discussed above. It is subject to all the weaknesses of the nullification proposal. Thus, asymmetry about the content of liability rules would be more pronounced than is presently the case. It is easier to expect a Marylander to know something about her own state’s laws (she is or can be a participant in Maryland’s political process) than about (say) Hawaii’s. One result of this asymmetry is that the manufacturer might become a much more influential political player on (say) the Hawaii products liability scene than it would normally be. The manufacturer would almost surely be a more active player than would be consumers living out of state. Manufacturers might be likely to choose the products liability law of a state with more pro-defendant rules than are demanded by consumers, especially if they believe that consumers are unable to accurately perceive and measure this when making their purchasing decisions. This could provoke a reverse prisoners’ dilemma, a ‘race to the bottom’, instead of an efficiency competition among rules as is the case for incorporation laws. Put differently, the reason why the “freedom to charter” works well is the presence of fully informed marginal shareholders, whose presence deters the “race to the bottom”. There is no reason to believe the consumer market exhibits these characteristics.
It is questionable whether individual states would have an incentive to ‘get it right’ when enacting their products liability rules. States would essentially be ‘selling’ products liability rules to manufacturers under Dean Perlman’s proposal. Would they receive a percentage of each sale as an incentive to select a popular rule? Would states derive any ‘seignorage’ from the development of widely used rules, analogous to Delaware’s incorporation fees?
Quite apart from all these problems, the Perlman solution is politically improbable. The greater the information asymmetry between manufacturer and consumer, the less likely it is that Congress would adopt a régime or that courts would enforce the resulting contractual allocation of risk.
William Niskanen proposed an alternate federal choice of law rule in 1995. Under Niskanen’s proposal, liability for a manufacturer’s products would be governed by the products liability rules of the jurisdiction in which that manufacturer had the largest number of employees.
This idea comes closest to the “choice of incorporation” régime in place in corporate law. Corporations would presumably choose to locate manufacturing facilities in jurisdictions whose products liability rules were most attractive to them. Niskanen’s proposal, unlike Perlman’s, would likely provide substantial ‘seignorage’ to states when determining their liability rules. “Getting it right” would arguably result in a tremendous increase in manufacturing activity. Indeed, Niskanen’s proposal turns on its head the perverse incentives provided by the lex loci delictus rule, which (as we have seen) makes a corporation indifferent between locating in a high-liability or a low-liability state. Finally, unlike Perlman’s proposal, Niskanen’s does present a tie (albeit minimal) to territoriality and to a state’s political process. After all, manufacturers would have to be “residents”, if not “citizens” of the state whose law is chosen.
As with the Perlman proposal, though, and unlike lex loci delictus, Niskanen’s plan requires no overt consumer act of submission to a state’s sovereignty. In no sense is the purchaser “responsible” for the legal rule chosen. Again, this makes the plan questionable from the point of view of participatory democracy, and less likely to be adopted by legislatures or enforced by courts.
In addition, the Niskanen proposal may have “Public Choice” problems. Whereas Perlman would allow both defendant and plaintiff to be ‘strangers’ to the state whose law is being applied, Niskanen requires the defendant to be likely more influential than are purchasers in the determination of the chosen liability rules. An auto manufacturer in Michigan, for example, would likely be very persuasive if it argued that that state should adopt more pro-defendant rules. As Niskanen would apply these rules to sales by the manufacturer throughout the country, the plan would in effect violate the first Laycock principle sketched out above: it would prefer in-state to out-of state interests.
In addition, the Niskanen proposal hypothesizes that manufacturing states will not be tempted by a ‘race to the bottom’ – presumably because consumers would at the margin turn down products from inefficiently pro-defendant states, preferring to pay more for products made in more ‘plaintiff-friendly’ states. However, this seems difficult to reconcile with the existing corpus of regulations and intellectual property rules which are barriers to the free entry Niskanen needs for his plan to work.. Thus, for example, pharmaceutical companies may have a patent monopoly on a certain medication. Such companies might have a distinct interest in choosing a pro-defendant state for their manufacturing. Even if that state’s rules were inefficiently pro-defendant, the patent monopoly would preclude competitors from manufacturing the same product in a different state.
Professor John Kozyris has proposed that choice of law rules specify that the law of the “intended place of use” of a product be applicable to all products liability suits. The most important benefit of this plan is that it would be easy to adopt and enforce. The vast majority of people injured by products are consumers or those in privity with consumers -- by definition these people have a connection to the jurisdiction whose law is applied. Thus, Kozyris’s plan applies legal rules that are in a meaningful way chosen by the plaintiff. By choosing to use the product in a jurisdiction the plaintiff has in essence assented to that jurisdiction’s exercise of sovereignty over the accident. Indeed, it is probable, though not certain, that the consumer is a citizen of that state, and in that capacity he has a constitutional “input” on the jurisdiction’s legal rules. In this sense, Kozyris’s proposal is more ‘legitimate’ than Perlman’s or Niskanen’s. It has a strong territorial link. And while Kozyris’s proposal (unlike Perlman’s or Niskanen’s) would not result in uniform pricing for each manufacturer’s products, it is not clear that the lack of uniform pricing is a real problem. Again, the nature of federalism is that companies know they are subject to different rules in different jurisdictions.
Unfortunately, Kozyris’s proposal does have a very severe defect. Quite simply, it does not appear to resolve the Prisoner’s Dilemma. Since two widgets sold at the same location might be intended for use in two different states, a vendor could not charge different prices (to reflect different ex ante liability outcomes) without conducting a rather expensive investigation into the purchaser’s intent. Higher up the chain, at the manufacturer’s level, it would be even more difficult to price each unit of production. As a result, differential pricing of widgets to reflect the different liability rules is unlikely. Thus, a retailer in a given state is likely to charge all purchasers the same price, even though different legal rules will apply to different purchasers. Put another way, a purchaser in a high liability state is likely to cross state lines to purchase his widget, then claim the benefit of his home state’s law if an injury involving the widget arises. This arbitrage (or adverse selection) will in effect prevent the differential pricing – a manufacturer will understand that expected liability from sales in a state are not reflected by that state’s liability rules. Thus the high liability state’s law will be applied at the expense of the low-liability state, as currently occurs. This is in violation of the first of Laycock’s rules for choice of law.
This choice of law rule, first proposed by Michael McConnell, would apply to each widget the product liability rule of the state of that widget’s first retail sale. Thus, if a Virginian traveled to Maryland to purchase a lawn mower, products liability for that mower would be determined by Maryland law, even if an eventual accident occurred in Virginia.
The key characteristic of this rule is that it allows mfg to effectively calculate expected liability for each retailer’s product, given that the state of retail sale of that product is known in advance. This overcomes the Prisoners’ Dilemma, because no arbitrage is possible. Every single widget sold in, say, Virginia will be subject to Virginia products liability law, regardless of where the purchaser lives or intends to use it. If the purchaser’s home state has a more pro-plaintiff products liability rule, and if the purchaser desires the greater protection his own state’s rule provides, he will have the incentive to purchase the product in-state. However under this choice of law rule, unlike either lex loci delictus or “interest analysis”, a purchaser seeking high-liability protection will have to pay for that protection as part of the purchase price. He will not be able to externalize the cost of this protection to consumers nationwide, as is currently the case. In addition, the purchaser will have endorsed the application of Virginia law by traveling to that state to purchase the product. This satisfies the territorial requirement.
This rule would create a more autarkial market for products liability protection, thereby satisfying the second choice of law requirement. Consumers could choose the amount of liability protection they wanted and pay for that level accordingly. Asymmetric access to knowledge of the applicable liability rule would diminish greatly due to retail competition. Retailers in high-liability states would have an incentive to explain to consumers that they are receiving greater “protection” as part of their purchase price (much as retailers of name-brand products have an incentive to explain to consumers that their product is more than worth the brand premium). Of course, the “protection” being offered may not be desired by consumers. Say, for example, that the price includes an insurance premium to pay for state products liability that includes manufacturer responsibility for gross misuse of a product by the consumer. In that case, careful consumers might presumably prefer to pay less for the product in a neighboring state where such “protection” is not “offered”. Home state retailers would lose sales to careful consumers under such a scenario. Note, though, that these retailers are well placed to make political representations intended to modify the liability rule to better reflect true consumer preferences. This rule is thus respectful of the local political process, and allows for input by local residents, not merely by out-of-state corporations. This makes the proposal politically palatable.
Nor would consumers be “held captive” by their own state’s products liability rules. They could escape these rules, and their cost, through “voice” (by joining local retailers and lobbying for a change in liability rules) and through inexpensive “exit” (by purchasing their widgets in another state). Through mail order or online purchase, consumers across the country could choose the level of protection they wanted best. If a state provides more, or less, protection than residents of any given state want, there will be fewer retail sales in that state.
If the ‘law of first retail sale’ does serve to resolve the prisoners’ dilemma, one of two things are likely to occur. As the national unraveling process described by Justice Neely ceases, products liability law might assume varying content across the country. As is the case for welfare and tax policy, there is no particular reason to believe that the population of New York has the same attitude toward collective sharing of risks as does the population of, say, Wyoming. Indeed, it seems likely that different communities would be freer to enact their collective preferences under the ‘law of first retail sale’ rule than they are currently. Recall that the current choice of law rules make West Virginians pay the same premium as New Jerseyites for the rules New Jersey’s court has fashioned – and this has led West Virginia to abandon the rules its institutions had preferred.
If laws and interpretations do vary across the country, in an autarkial fashion, will producers, wholesalers and retailers consider these variations when pricing their products? It is hard to see why they would not. Providers of goods and services already consider risks shaped by state law whenever autarky reigns. There is no reason to believe consideration would not be given to differential products liability rules.
On the other hand, it is theoretically possible that the content of all fifty states’ products liability rules will change in the same direction. Efficient assumption of risks by consumers might be authorized nationwide, for example. Consumer misuse might be sanctioned more severely. Other changes might be made nationwide. If this occurred, products would not have different state “liability premiums”. But this would not be a flaw in the ‘‘law of first retail sale’ rule. Rather, it would indicate that preferences for products liability rules are in fact rather uniform across the land, and were previously distorted (in a rather uniform fashion) by the Prisoners’ Dilemma created by the old choice of law rules.
Bruce Hay argues that the ‘Law of place of first retail sale’ rule “might produce national liability levels that are lower than most states prefer”. Hay asserts that consumers systematically underestimate the risk of products when they purchase them. Thus, they would tend to irrationally decide to save a little money now, and choose to purchase in low-liability states, when they “really” want to pay the extra premium and avail themselves of higher liability rules.
This is a difficult critique to rebut, relying as it does on the existence of consumer irrationalities that cannot be verified by consumers’ purchasing choices. If people are irrational in the way Hay states, however, they would certainly be more likely to understand their “true” preferences thanks to the “wake-up call” of differential pricing, which this plan promotes. For retailers have a strong commercial incentive, under this plan, to educate consumers about the true risks and benefits of the varying levels of protection they are purchasing with their product.
This notion of irrationality may be based on the often-verified belief that people tend to underestimate the probability of low probability risks. However, in the case of a large national market, statistics would be available and retailers would have an incentive to provide consumers with information about just how likely an injury from certain products is. Assuming no differential production methods across states, the only differences in accident rates will result from moral hazard problems: reduced consumer caution in states with stricter rules of liability. States with more risk averse residents will opt for higher liability rules, and this will increase the probability of accidents. Thus, residents will face the correct trade-off: increased coverage in case of an accident, in exchange for a higher probability of accidents, and therefore a higher price for insurance. It is entirely possible that with different levels of underlying aversion to risk, different states will choose different coverage-insurance premium combinations. At the moment, risk averse states purchase too much insurance, from the social point of view, because their “purchase” of greater coverage comes at an “insurance premium” that underestimates the overall cost to the national market.
Risk aversion does appear to differ among people and across areas. Poll after poll indicates that Canadians are, in the main, content with a level of socialized health care protection that Americans are by and large keen to assume privately. Under the ‘first retail sale’ rule (unlike, say, the Perlman plan), states have a great incentive to promote their products liability protection levels, because they at the same time promote retail activity in the state. The resulting competition would arguably go a long way to overcome the problem of ignorance complained of by Hay, if it does exist.
When a consumer travels to another state to purchase a product, she assents to that state’s jurisdiction in a meaningful way. What if she merely picks up a phone, mails a letter, or uses her computer to buy a product from an out of state reseller? Is the assent to the foreign jurisdiction as transparent?
This is admittedly a hard question to answer. Consumers who purchase by mail order arguably know they are in some way outside the protection of their state’s law – their insistence that their purchase is not subject to their home state’s sales tax is a nice illustration of their awareness of this issue. Presumably, regulations could oblige mail order and Internet vendors to prominently display the name of their host state, with (in Internet cases) hyperlinks to federally approved summaries of that state’s product liability rules. This would make acquiescence to the retailer’s state’s law just as informed as is the case for many service contracts.
This solution does, admittedly, resemble the Perlman plan of allowing manufacturers to freely designate any state’s liability rules as applying to a particular sale – and the same objections made above would apply. Alternatively, federal law could mandate that the state of first retail sale is the state to which the product is shipped. Each consumer would then need only be familiar with the PL rules governing the location of her mailing address, which would generally be her residence. This would oblige residents of high liability states to physically travel to (or open a postal address in) another state if they wish to avail themselves of that other state’s liability rules. Whether that solution is an advantage or not is debatable: “exit” is surely made more expensive, and mail order firms are obliged to charge different prices nationally -- all of this seems bad. On the other hand, affirmative acceptance of the pro-defendant state’s rules is easier to legitimately infer when a conscious act of travel is undertaken; and making exit a bit more difficult also makes it more likely at the margin that a consumer will choose political “voice”, which might be good.
My own view is that simplicity is best served by applying the liability rules of the state of delivery to mail-order and internet purchases. Many American consumers can, if they wish, easily travel to a neighboring state. On the other hand, for telephone purchases from a “brick-and-mortar” establishment to be delivered by that establishment to a purahaser in a contiguous state, the state of sale should prevail. This would provide states contiguous to pro-plaintiff jurisdictions a powerful motivation to gauge the satisfaction those citizens have with their liability rules.
Most products (from soap to food) are only sold
to a consumer once. Other goods, from lawn mowers to automobiles, are
often resold. The ‘law
of first retail sale’ rule would continue to apply the first state
of sale’s laws regardless of the place of resale.
This might perhaps take some buyers of second-hand products by
surprise. On the other hand, products sold at the retail level more
than once are typically easy to engrave with some marking (“VA”,
“MD”, etc.) which identifies applicable law, just as they are easy
to mark with serial numbers or the Underwriters’ Laboratories logo.
Second-hand purchasers are already used to acquiring the
original purchaser’s remaining warranty coverage and (in the case of
automobiles) emissions eligibility.
In both these cases the resale buyer takes his product with the
attributes given to it at the first retail sale.
It is true that the second-hand purchaser has made less of a
commitment to the state of first retail sale.
But the prominence of the marking does create a meaningful
consent to the foreign state’s jurisdiction.
And retailers’ publicity, which will presumably have
emphasized the costs and benefits of various states’ products
liability rules will arguably have an impact on the resale market
price of differently engraved items.
Thus, an item governed by the law of a state with a “statute
of repose” might have a different resale price than an item governed
by the products liability rule of a state without this limitation.
Original purchasers under the ‘law of first retail sale’ plan voluntarily assent to the sovereignty of the state of first sale. So would, vicariously, those in privity with these purchasers – subsequent purchasers, borrowers of the product, etc. What, though, of injured strangers? What legal régime should be applied to the New Jersey child, injured in New Jersey by a stone thrown from his next-door neighbor’s lawn mower, if that neighbor had traveled to Pennsylvania to purchase the mower (say, to benefit from lower prices caused by Pennsylvania’s hypothetically more pro-defendant products liability rules)?
This is a conceptually important problem, even if it is not practically widespread – the overwhelming majority of products liability plaintiffs consist of purchasers and people in privity with purchasers. Nevertheless, in those cases when a true stranger is injured, some limitation on the ‘law of first retail sale’ is required. It appears impermissible, under the criteria advanced by Laycock, to apply to a New Jersey resident a foreign law to which neither the resident nor his agents have assented. The one exception to the “state of first retail sale”, then, would be for these true strangers. For them, lex loci delictus should apply. As these third parties constitute an extremely small percentage of products liability plaintiffs, this exception would not prevent meaningful differential pricing of products by manufacturers according to the state of first retail sale. While statistically a very small component of products liability, true third parties may be a politically important factor in the adoption of any legislative plan. Excluding them from the rule of “first retail purchase” diffuses objections along those lines.
The ‘state of first retail sale’ plan is perfectly compatible with federalism. It fulfils what is arguably a federal constitutional duty, while imposing no substantive federal rule of liability. It allows a state’s products liability and general tort rules to network well. Yet it resolves the ‘Prisoners’ Dilemma’ that is arguably at the base of current products liability law.
Some federal regulation will be required to make this choice of law system operational. The following areas must surely be addressed:
The federal government might mandate labeling requirements for products, establishing a consistent way to communicate the state of first retail sale to subsequent consumers. Of course, there is some chance that such a requirement would spawn another large federal bureaucracy, as with food labeling requirements. Conceivably, the federal statute that introduces the rule of first retail sale for choice of law could limit the application of such a rule to products that are clearly and unmistakably labeled. Some common law would develop to allow courts to figure out what constitutes a clear label, but this would allow the manufacturers of different products some freedom in choosing the most efficient way to label their products. Manufacturers would have strong incentives to label products in such a way as to trigger the federal statute, because the capacity to price differentially based on PL rules benefits them greatly.
Whose law applies to goods purchased abroad? Federal law would have to answer this question. Possibilities include the state of residence of the first purchaser, the state of the importer, and the country of first retail sale. If the foreign manufacturer comes into the US to sell its products, then it should become acquainted with the state liability rules. If on the other hand an American consumer goes to Scotland to buy whiskey, we can assume that he has a chance to observe that one cannot succeed in Scotland on a claim that whiskey manufacturers deliberately cause addiction to their product.
In the case of goods brought into the US, the most common procedure today is for the foreign company to establish a subsidiary within the US. This company is a separate corporate entity and can be considered a US corporation, incorporated in a particular state, for all legal purposes. (In fact, some important NAFTA benefits only accrue to subsidiaries, not to branches, and this has further encouraged the practice of establishing subsidiaries). Few American products liability suits concern goods truly purchased abroad, so this is not presently an acute practical problem. But it is another important conceptual issue for the federal government to address.
Most importantly, the adoption of a “state of first retail sale” choice of law rule is not sufficient to guarantee its enforcement. As was indicated earlier, interpretation of law is as important as choice of law.
Say a New Jersey court is to apply Pennsylvania law to a products liability suit (because the New Jersey plaintiff traveled to Pennsylvania to purchase the product). If Pennsylvania law differs from the New Jersey forum’s law in a significant way, the forum court may be tempted to “misread” applicable law. As noted above, the prisoners’ dilemma applies to more than just the choice of law – it concerns the interpretation of this law as well.
This moral hazard must be constrained; otherwise, manufacturers will have no confidence that the choice of law system will be genuine. If this confidence is lacking, autarkial pricing will not be possible.
Of course, a lawsuit by a New Jersey plaintiff against an out-of-state manufacturer would in principle be removable to federal court, which would be charged with applying the applicable state’s law under the Erie Railroad doctrine. Unfortunately, case law has required complete diversity for removal to be an option. As a result, plaintiffs have been able to assure the presence of a state judge by joining an in-state defendant (typically, a retailer) to their products liability suits, even if they do not intend to pursue the local lawsuit. This case law must be modified, by the Supreme Court or by federal statute, to spell out that federal jurisdiction exists at the option of any out-of-state defendant if a case is filed in state court. Similar modifications have been recently proposed as part of federal tort reform packages. Unlike much other recent federal tort reform, the incomplete diversity amendment, standing alone, is not hostile to federalism principles. To the contrary, by ensuring that a state’s law will not create a national “Prisoners’ Dilemma”, the incomplete diversity amendment cements true federalism.
Products liability law is a subset of state tort law. As such, it must be allowed to evolve as an expression of each state’s considered view of the protection it wishes to offer to consumers who use potentially dangerous products. Currently, those responses are skewed toward more liability than each state might consider optimal, were the consequences of this liability felt inside the state. Products liability law is today a classic “Prisoners’ Dilemma”.
Some critics have proposed federalizing products liability law to resolve this dilemma. However, in addition to the harm a federal takeover would inflict on our constitutional division of powers, uniformity of products liability law is undesirable for substantive reasons. Our invincible ignorance about the desires of consumers, and the comparative advantage of expressing moral beliefs in decentralized forums, makes the laboratory of states the ideal setting for torts in general, and for products liability law in particular.
Through adoption of a federal “choice of law” régime, it is possible to resolve the “Prisoners’ Dilemma” while respecting substantive federalism. This paper has sketched out the reasons for such a régime, the variations such a régime could have, and the best way to make it operational.
With a new federal administration and a Congress interested in both tort reform and renewed federalism, it may now be possible to reconcile the two principles through federal legislation. Choice of law, a federal duty long neglected, is worth a serious look now.
 Thanks to George Priest and Joyce Sadka for their comments on previous drafts.
 Cite A. Wildhavsky
 Only about 3% of tort suits ever make it to trial. Many settlements are confidential and therefore never reported.
 Tillinghaust-Towers Perrin, Tort Costs Trends: An International Perspective, New York, (1995).
 Punitive damages are court-ordered financial payments made by the tortfeasor to the victim that go above and beyond compensation for the victim’s loss. Punitive damages are frequently justified on the public policy considerations of punishing the victim and/or setting an example for other potential wrongdoers. They are discussed infra in the section titled “Punitive Damages.”
 Keith J. Ward & Ass., A Study to Address Relationships Between Economic Development and the Need for Tort Reform, Birmingham, (Mar. 15, 1993).
 The Economics of Civil Justice Reform in Massachusetts, Beacon Hill Institute, 1997.
 The Engle damages award is the largest in U.S. history. This class –action lawsuit, in a Florida state court was brought on behalf of all Florida smokers (estimated at 700,000) against tobacco companies. The case has been harshly criticized for allowing the claims of such a diverse group to go forward, and for the amount of the award. Defense lawyers, following the award have removed the case to a federal district court. In the meantime, the plaintiffs are said to be losing an estimated $39,726,027 per day in interest on the award. There is a good likelihood that the award will be reduced as Florida law bars court awards so large that they would bankrupt companies. See e.g. Marianne Lavelle and Angie Cannon, Chewing Big Tobacco: Will the Record $145 Billion Verdict Stand? U.S. News & World Rep., Jul. 24, 2000; Richard Hubbard, Reporter’s Notebook: Engle Tobacco Case Chills in U.S. Court, Dow Jones News Serv., Aug. 21, 2000.
 See e.g. Alexander Tabarrok and Eric Helland, Court Politics: The Political Economy of Tort Awards, J.L. & Econ.157 (1999) This article also provides statistical evidence that the costs of losing will be higher out-of state defendants.
 Corporations, which are legal fictions consisting of a ‘nexus’ of contracts among shareholders, employees and suppliers, don’t “pay” for tort awards in the same sense that individuals do. Depending on the elasticity of supply and demand, factors of production (capital and labor) will bear in varying proportions the costs of tort awards resulting from inefficient corporate behavior.
 Excessive liability rules do not increase consumer safety and raise retail prices across the board. Many people seem to think that there is a linear relationship between liability rules and safety – that increased liability always produces more safety, and so the only question a society has to answer is how much safety it wants. But this is not the case. In a competitive industry, if a corporation is held liable for damages caused by misbehavior, i.e., for design or manufacturing practices it could have easily (efficiently) avoided, it will not be able to factor the amount of a tort award into the price consumers pay for the corporation’s goods or services. It the corporation attempted to raise prices in such circumstances, efficiently organized firms would be able to under-price the goods and services of the negligent corporation. However, if liability is imposed on a corporation despite its having taken all cost-efficient courses of action, the tort award becomes a “cost of doing business”. If this cost is imposed “across the board” on all corporations in an industry, the consumer price of goods and services will be affected. See Michael Krauss, Restoring the Boundary: Tort Law and The Right to Contract, Cato Institute, Policy Analysis #347, 1999.
 The profit motive, even absent tort liability, will furnish appropriate incentives unless consumers are ignorant or irrational, or unless some other important market imperfection exists.
 Is Liability Slowing AIDS Vaccines?, Science, Apr. 10,1992, at 168-169.
e.g. Joseph Sanders, The
Bendectin Litigation: A Case Study in the Life Cycle of Mass Torts,
42 Hastings L.J. 301
 See Krauss, “Loosening the Food and Drug Administration’s Drug Certification Monopoly: Implications for Tort Law and Consumer Welfare”, 4 George Mason Law Review 457 (1996) for a generalized study of the effects of this consumer ignorance. The technical problem is that consumer welfare losses are estimated using demand curves, but demand curves can really only be estimated for existing products. However, some work has been done on trying to estimate optimal levels of variety, both horizontal and vertical (quality). In this type of research the assumption is that although a certain quality of the good does not exist, consumers know about all qualities and varieties.
 See Krauss, "Tort Law, Moral Accountability and Efficiency", 2 Journal of Markets and Morality 114 (1999)
 That is, liability which in no way reflects on misbehavior by the manufacturer or retailer, but has as its sole purpose to transfer money from a producer to an injured consumer. As discussed above, in note 11, transfer-based liability will lead to price increases or inefficient design changes.
 In the case of many vaccines the transfer premium is clearly over 100%. Cite studies.
 What is, by the way, “responsibility for products” as discussed in this paper? In general, it is legal recourse when an alleged defect in a tangible product causes property damage or personal injury.
 Baldwin et al., Defensive Medicine and Obstetrics, 274 J. Am. Med. Ass’n 1606-1610 (1995); See also Jonathan J. Frankel, Medical Malpractice Law and Health Care Cost Containment: Lessons for Reformers from the Clash of Cultures, 103 Yale L.J. 1297 (1994).
 L. Dubay, et al., The Impact of Malpractice Fears on Cesarean Section Rates, 18 J. Health Econ. 491-522 (1999)
 Thus, illnesses and accidents which in the past would have been seen as the result of assumption of risk (e.g., smoking), or of contributory negligence by consumers (e.g., driving while inebriated and without buckling one’s safety belt), today result in the filing of lawsuits against the manufacturer who provided the cigarette, or who allowed the car to be driven without an automatic seat belt.
 On the latter, see Krauss, Fire and Smoke: Government Lawsuits and the Rule of Law, Independent Institute, 2000.
 Cite Rand study.
 31.9% of all state tort trials in the nation’s 75 largest counties involved automobile accidents, according to the Rand study (p. 2).
 E.g., Osborne v McMasters (first Products liability suit, opposing victim and local apothecary, who had mislabeled a poison, resulting in poisoning of the victim).
 New Jersey has pioneered many shifts in favor of individual plaintiffs against defendant corporations. See, e.g., Henningsen v Bloomfield Motors, 161 A.2d 69 (1960) (holding that a waiver of the right to sue for a manufacturing defect in a new product was void); Beshada v Johns-Manville, 447 A.2d 593 (1982) (holding that a manufacturer could be liable for a “defective” design even if no one in the world had a better design to offer); O’Brien v Muskin, 463 A.2d 298 (1984) (holding that a trespasser who dived into a four-foot deep above-ground swimming pool could sue the manufacturer of that pool, and that a court could declare that above-ground pools were all intrinsically defective.
 “Out-of-state” refers to the presence of plants and employees, not to legal abstractions such as state of chartering.
 C. DeMuth, “Should Product-Liability Law be Nationalized?”, Revised draft, unpublished, 1985, at p. 50. As will be seen below, in many products liability cases multiple defendants are sued, but the intention is to seek recovery from the out-of-state firm (typically the manufacturer).
 Fifty states, District of Columbia, Puerto Rico, Guam, etc.
 Brief note about Louisiana, Puerto Rico.
 Cite relevant provisions.
 Erie R.R. v Tompkins, 304 U.S. 64 (1938).
 This procedural protection has been substantially diminished by judicial interpretation of federal statutes, allowing state courts to retain jurisdiction unless diversity is complete. A case study of this problem may be found in “The American Torts Crisis meets NAFTA: The Loewen Case”, 8 George Mason Law Review 1 (2000).
 R. Posner, The Federal Courts, 1985, at 179.
 E.g., see Bruce Kobayashi & Larry Ribstein, “An Economic Analysis of Uniform State Laws”, 25 J. of Leg. Stud. 131 (1996)
 in New State Ice Co. v Liebmann, 285 U.S. 262, at 311 (Brandeis, J., dissenting)
 Cite data on percentage of purchases through internet (WSJ) and through mail order (get data from debate on sales tax in Congress) to show that vast majority of purchases are made close to home.
 With the exception of automobile use, of course -- but most driving still does take place in one’s state of residence.
 It matters not, here, whether or not the defendant manufacturer removes the case to Federal court on diversity grounds. In any case the Federal court will apply the domestic state’s substantive law to the dispute. And as has been stated above, removal can currently be prevented with ease, merely by joining a local defendant (say, the retailer) to the lawsuit.
 The forum is the state whose court hears the case. This is also usually the state whose laws are applied.
 In reality the nature of the legal system is such that many litigated cases are ‘interstitial’, raising issues in a slightly different way than they have ever been raised before. Easy cases never make it to trial. Those cases that do make it to trial typically involve either an acute difference of view of the facts or a “gap” in the state’s law.
 For instance, the previous state of the law might have incorporated a “consumer misuse” defense, according to which consumers injured by products they misuse may not recover from manufacturers. This defense is needed, inter alia, to impart appropriate incentives to consumers. But how clear must the misuse be? Should consumer misuse be fatal to the plaintiff’s case if the misuse was “foreseeable” by the manufacturer (drunk driving, or travelling at 100 m.p.h.,, or declining to wear a seat belt, are all foreseeable)? A plaintiff may be highly motivated to nudge the consumer misuse rule toward this arguably non-optimal result in an individual case.
 See note 11.
 R. Neely, The Product Liability Mess, 1988
 406 S.E.2d 781 (W. Va., 1991)
 Following Huddell v. Levin , 537 F.2d 726 (3rd Circuit, 1976)
 Following Fox v. Ford Motor Co. and Mitchell v. Volkswagenwerk, A.G.,669 F.2D 1199 (8th Circuit, 1982)
 406 S.E.2d 781, at 785-786.
 cite to Regulation article on this subject.
 See Robert Axelord, The Evolution of Cooperation, 1984.
 Investments are less costly, commerce less risky, etc.
 In reality, “state A” might be the “forum state”, while “state B” might be “all other states.
 Total payoff in top left box: 80 (40 + 40). Total payoff in bottom left box: 45 (55 -10).
 I.e., it it worse from each party’s subjective perspective (absent weird envy or some similar pathology).
 Legislatures are virtually always free to modify the common law of their state. [But see recent Ohio case, where the court denied that the legislature could enact substantial tort reform.]
 R. Epstein, “The Political Economy of Product Liability Reform”, 78 Am. Econ. Rev. 311 (May 1988 special issue).
 Cite to Levy policy analysis for CATO, pointing out antitrust implications.
 Michael W. McConnell was perhaps the first to conceptualize this solution to the products liability prisoners’ dilemma, in “A Choice of Law Approach to Products-Liability Reform”, in W. Olson (ed.) New Directions in Liability Law, 1988 at 88. This article constitutes in large part a detailed elaboration of Professor McConnell’s thesis.
 If the costs of
discriminating against out-of-state firms were internalized, and
given the figures used in Table II above, the result of adopting
an inefficient pro-local plaintiff rule would be reflected in the
B (SECOND NUMBER)
can be seen, in this scenario neither State A nor State B has an
incentive to adopt a rule hostile to out-of-state defendants. [And
to the extent that each state would also see its own shareholders
and employees affected (see notes 9 and 10, supra),
the disincentive would be exacerbated.
 A’s legal rules might still have an economic impact in other states. For instance, depending on the elasticity of demand for products, increased local liability in State A could result in decreased supply, increased unemployment in the state of manufacture, decreased tax base in that state, etc. Alternatively, if State A had a large population, a liability rule change might result in a design modification instead of a price increase. The design modification might affect consumers in all jurisdictions, if there are significant economies of scale. This kind of externality is both commonplace (every time we purchase something we increase demand for it slightly, and thus increase price marginally for others) and morally unobjectitonable -- for none of the additional costs imposed on the nation confer any subsidy to consumers in state A.
 Note here, rather long in the final draft, will distinguish ‘personal jurisdiction’ rules (when can a court hear a case) from ‘choice of law rules’ (what law to apply to the case).
 This premium may be a literal insurance premium, or it may be a reserve set aside by a self-insuring manufacturer, to cover expected liability costs.
 Remember that prices will
increase to reflect liability costs only if those costs do not
represent cost-efficient design or manufacturing changes.
For efficiently avoidable liability, i.e., negligence, a
manufacturer will simply not be able to pass costs on in a
competitive market. See
note 11, supra.
 G. Eads & P. Reuter, Designing Safer Products: Corporate Responses to Product Liability Law and Regulation 107 (1983)
 It is very difficult to classify states’ choice of law rules with precision, as many states have “combination” rules of great complexity. Upwards of 40 states have apparently adopted one form of ‘interest analysis’ or another. See Russell Weintraub, “Choice of Law for Products Liability”, 52 Ark. L. Rev. 157, 163-164 (1999)
 Bruce Hay, “Conflicts of Law and State Competition In The Product Liability System”, 80 Geo. L. J. 617 (1992)
 Courts in state B, it is hypothesized, will find their desire to transfer wealth from A exceeds their desire to apply the pro-defendant rule they believe to be appropriate.
 Hay, at 649.
 Michael Solimine, “An Economic and Empirical Analysis of Choice of Law”, 24 Ga. L. Rev. 49 (1989)
 665 S.W.2d 414 (TX, 1984)
 943 F.Supp.789 (W.D. Ky, 1996)
 A statute of repose quiets any litigation after a given period – here, a given period after the assembly of the automobile. Statutes of repose are similar to statutes of limitation, except that the latter (but not the former) may be suspended (or “tolled”) by such factors as the age of the victim.
 Indeed, it seems that the better tactic might have been to sue in Indiana and persuade the Indiana court to apply Kentucky law via “interest analysis”. Indiana apparently uses lex loci delictus, however. This is another illustration of the fact that adopting “interest analysis” (in Indiana) would benefit plaintiffs, not defendants.
 Cite Priest article on the history of the development of products liability rules.
 See, e.g., Gantes v Kason Corporation, 679 A.2d 106 (N.J., 1996). In Gantes, the New Jersey Supreme Court tackled a case involving an accident in Georgia, which killed a Georgia woman and resulted in a wrongful death suit against the New Jersey manufacturer of the machine that allegedly killed her. Georgia’s statute of repose precluded the suit, so the plaintiff sued in New Jersey. The court applied New Jersey law to hold the manufacturer liable, stating that its concern for injured victims overrode its fear of discouraging manufacturing in the Garden State. Clearly, at the margin, firms have more of an incentive to locate in “interest analysis” states (like Kentucky: see Rutherford, supra) that discriminate against out-of-state residents than in “interest analysis” states (like New Jersey) that tend to maintain their pro-plaintiff stance in all cases. [This incentive is quite minimal, however -- it is most surely drowned by other factors influencing firm location.]
 Cite Tabarrok data.
 At least everywhere but in Michigan…
 Cite dicta remark in Gore v BMW to this effect; but see Krauss article on Loewen.
 Allude to studies by Tabarrok.
 G. Schwartz, “Considering the Proper Federal Role in Tort Law”, 38 Ariz. L. Rev. 917, at 936-7. (1996)
 cite to game theory; explanation of what a coordination problem is.
 Imagine each prisoner refusing to learn the identity of his co-conspirators, trusting this information to a third party who will therefore control every decision to “rat.”
 See Schwartz & Mahshigian, “A Permanent Solution for the Product Liability Crisis: Uniform Tort Law Standards”, 64 Den. U. L. Rev. 685 (1988); Reed & Watkins, “Product Liability Tort Reform: The Case for Federal Action”, 63 Neb. L. Rev. 389 (1984).
 Cite recent federal P.L. efforts.
 Sen. Spence Abraham’s piece in Policy Review, cited here.
 Cite to Michael Greve’s paper in the forthcoming Hoover book.
 Briefly describe the creation of modern products liability law, from tort law, in the 1960’s via seminal California cases which renounced the use of res ipsa loquitur, etc.
 Two or more defendants are “jointly and severally” liable when any one of them can be called on to pay the entire tort awards. Some states, like California, have greatly extended the concept of joint and several liability. Others, like New Mexico, have virtually abolished joint and several liability.
 See William Powers, “Some Pitfalls of Federal Tort Reform Legislation”, 38 Ariz. L. Rev. 909 (1996).
 One explanation for Congress’ constant revisiting of the products liability reform issue is that keeping the issue potentially “on the table” makes it easier for members of both federal political parties to engage in fundraising from interested groups. See, e.g., J. Abramson, “Product-Liability Bill Provides Opportunity for Long-Term Milking of PAC’s by Congress”, Wall St. Journal, June 21, 1990, at A16.
 Geier v Honda, FDA preemption cases, etc.
 See Gary Schwartz, “Considering the Proper Federal Role in American Tort Law”, 38 Ariz. L. Rev. 917 (1996)
 Cite Krauss, Tort Law and Private Ordering.
 See Krauss, “Tort Law and Private Ordering”, 35 St. Louis U. L. J. 623 (1992).
 Moral and political outlooks are directly related to views about products liability. What ‘causes’ a product to injure a consumer – is it consumer misuse, insufficient redundancy of safety devices, third party negligence, or manufacturer cost-cutting? As readers of products liability cases know, in virtually every case accidents involving products implicate some combination of all these factors. Deciding which factor constitutes the cause of an injury is ultimately a moral issue. If one believes that individuals are primarily responsible for their own fate, the knowing misuse of a product by a consumer assumes greater importance. If on the contrary one sees consumers as lacking free will, and acting as involuntary pawns in a game played by powerful commercial interests, corporate cost-cutting in determining design and manufacturing processes will be more important.
 Greve, “Federalism after the Election”, AEI Federalism Outlook #4, December 2000, at 2.
 Thus, in Obrien v Muskin (see above), the New Jersey Supreme Court intimated that above-ground pools, in toto may be socially inappropriate products. The court implied that the manufacturers of such pools should essentially insure users against the hazards of swimming. If price differentiation were possible, such pools might cost much more in New Jersey than elsewhere. This will create powerful lobbies to modify the New Jersey rule, unless Garden State residents truly have an aversion to above-ground pools.
 Cite here to Austrian teachings, Hayek, Mises, etc.
 Harvey Perlman, “Products Liability Reform in Congress: An Issue of Federalism”, 48 Ohio St. L. J. 503 (1987), at 507.
 Cite Peter Huber, Liability: The Legal Revolution and its Consequences, 1988, at 195; Paul Rubin, Tort Reform By Contract, 1993 at 24.
 cite Henningsen, etc.
 Cite UCC provisions, explain how they are compatible with laizzez-faire, etc.
 Cite Kahneman & Tversky, etc.
 Croley & Hanson, “Rescuing the Revolution: the Revived Case for Enterprise Liability”, 91 Mich. L. Rev. 683 (1991), at 766-767.
 It has been ably defended elsewhere. See Douglas Laycock, “Equal Citizens of Equal and Territorial States: the Constitutional Foundations of Choice of Law”, 92 Colum. L. Rev. 249 (1992)
 This is one of the corollaries of Article IV’s “Privileges and Immunities” and “Full Faith and Credit” Clauses.
 To be more precise, lex loci delictus and interest analysis have provided incentives to adopt substantive products liability rules that disregard the interests of out of state individuals and corporations, in order to favor state residents. The rules themselves havenot violated the “no discrimination” rule. Lex loci delictus will apply the law of the place of the accident, regardless of who the parties to the accident are. In other words, the choice of law rule is neutral on its face regarding the residency status of the parties. If a state declined to decide a case involving an accident inside its borders, involving two non-citizens, then the differential treatment of the non-resident would result from rules of personal jurisdiction, not from the choice of law rule. Interest analysis seems more prone to result in direct discrimination against out of state parties because it is more malleable. True discrimination against outsiders is due to the substantive rules resulting from the (often selective) application of the “interest analysis”, however.
 Consider for example the Constitution’s restrictions on new states: “No new State shall be formed or erected within the Jurisdiction of any other State; nor any State be formed by the Junction of two or more states, or Parts of States, without the Consent of the Legislatures of the States Concerned as well as of the Congress.” U. S. Const. Art. IV, § 3, cl. 1. the word “jurisdiction” is a synonum for territory. A state’s authority to govern – its ‘jurisdiction’ – is a place within which no new state can be formed. When the Constitution states that no new state shall b formed “within the jurisdiction” of another, it does not mean “within the reach of the interests of another”, for then every state’s creation would be in breach of this rule. It can only mean “within the territory” of another. See Laycock, 92 Colum. L. Rev. 249 (1992) at 317.
 The territorial definitions of states are specified in their organic acts – for examples, see Laycock, 92 Colum. L. Rev. 249 (1992) at 318. as the Supreme Court held early on, “Title, jurisdiction, sovereignty, are therefore dependent questions, necessarily settled when boundary is ascertained….”, Rhode Island v Massachussetts, 37 U.S. (12 Pet.) 657 (1838), at 733.
 Currie, Selected Essays on the Conflict of Laws, 1963.
 The Full Faith and Credit Clause includes the following: “Congress may b general Laws prescribe the Manner in which such Acts, Records and Proceedings shall be proved, and the Effect thereof.” U. S. Const., art. 4, § 1. Congress may, it is generally agreed, thus specify which state’s law gets full effect in different classes of cases. See Walter Cook, “The Powers of Congress Under the Full Faith and Credit Clause”, 28 Yale L. J. 421 (1919), at 425-426; Laycock, 92 Colum. L. Rev. 249 (1992) at 331.
 Laycock, 92 Colum. L. Rev. 249 (1992) at 331. I am not, here, adopting Laycock’s view as my own. If choice of law rules are mandated by the Full Faith and Credit clause of the Constitution, they are presumably required in the field of Contract as much as in Tort or Products Liability. It is enough, for my purposes, to claim that it is consitutionally permissible to have a federal choice of law rule, under the commerce clause.
 Harvey Perlman, “Products Liability Reform in Congress: An Issue of Federalism”, 48 Ohio St. L. J. 503 (1987)
 Roberta Romano, The Genius of American Corporate Law, 1993.
 Cite Ribstein & O’Hara, 2000 Chicago Law Review article.
 Cite Posner article on related point concerning waivers.
 Manufacturers would presumably avoid choosing the products liability law of a state the law of which is unstable and arbitrary. To the extent that states derive “rents” from having their legal rules selected, this is an additional incentive for stability of legal rules.
 This makes the Perlman solution less legitimate than the current system regarding incorporation.
 See supra text corresponding to notes 100-104.
 Schwartz, 38 Ariz. L. Rev. 917, at 938. (1996)
 Shareholders choosing a state of incorporation have an incentive to choose the state offering the most efficient rules for corporate governance (such as rules that help to mitigate the agency costs prevalent within corporations). There is no reason to believe that manufacturers have similar incentives to choose efficient products liability rule states. Of course, if consumers knew perfectly what the liability rules for each product cost them implicitly, liability rules that are too generous to manufacturers would be penalized by consumer demand and would sell at lower prices, and this would counteract manufacturer incentives to race to the bottom. However, this would require an unrealistically high level of knowledge on the part of consumers, of technology, of firm production processes and internal decisions, and of underlying risks. Accord, cite Roberta Romano piece.
 Cite to Ribstein paper, which claims that the selected state’s attorneys will tend to dominate litigation, receiving seignorage, and would therefore be proxies for adoption of efficient rules. With the advent of interstate practice of law, it is not obvious that states would be able to procure true royalties, however. Court fees rarely if ever absorb the true social costs of dispute resolution. And bar admission fees do not appear to be a significant revenue item.
 Perhaps flesh this out in a paragraph or two, showing how the information asymmetry was at the base of the refusal to enforce contracts such as in Henningsen.
 Niskanen, “Do Not Federalize Tort Law” 4 Regulation 34 (1995)
 Unless the state of incorporation is chosen instead of the state of manufacturing activity.
 Schwartz, 38 Ariz. L. Rev. 917, at 938. (1996)
 Here, though, the in-state interests would be those of manufacturers, not consumers.
 Kozyris, “Choice of Law for Products Liability: Whither Ohio?”, 48 Ohio St. L. J. 377 (1987). Note that Kozyris was proposing that a state adopt this rule. However, there is no reason not to analyze his proposal as a potential federal solution.
The investigation would not be expensive for certain products,
e.g., automobiles, where the state of intended use corresponds
roughly to the state of registration.
See on this point the discussion infra.
 “A Choice of Law Approach to Products-Liability Reform”, in W. Olson (ed.) New Directions in Liability Law, 1988 at 88.
 Assuming here that the applicable law is the law of the place of the seller’s retail location. This location would have to be prominently displayed on the retailer’s publicity or web page. Failure to publicize this location might result in the plaintiff’s home state being chosen.
 One important caveat should be mentioned. If manufacturers respond to liability awards by inefficiently increasing the number of quality and safety inspections, placing superfluous additional warnings on the product, etc., then it may be prohibitively costly for manufacturers to adjust such practices depending on the state of sale. In other words, the natural laboratory referred to before may simply not work as well for mass produced goods, because returns to scale may make it more efficient for manufacturers to standardize all their production processes, regardless of differences in levels of liability. To the extent that manufacturers merely adjust their prices to incorporate “insurance” for purchasers, however, the autarkial outcome seems likely.
 E.g.: life insurance (different laws on suicide by state, taken into account by actuaries in determining likelihood of claims); auto insurance (different liability rules helping to determine premiums); apartment rent prices (multi-state developers use different pricing policies to take account of each state’s rent control and other related laws); firearm sales prices (Maryland requirement of ballistics test with each firearm surely factored into sales price of handguns in Maryland). In each of these cases, autarky is possible (one’s zip code is used to determine insurance premiums; handguns may not be purchased outside one’s state of residence; etc.)
 Hay, at 646.
 Cite to Greve’s op-ed piece on sales tax.
 Cite to Carnival cruise case.
 A second-hand buyer of a car in California, Mexico, Canada, etc. acquires all rights to register the vehicle. California emissions requirements follow the car, not the owners.
 E.g., the forum law may allow full recovery of damages when the injured consumer has misused the product; the law of first retail sale may deny or reduce recovery in such cases.
 Cite relevant provision of Constitution and of statute.
 Cite seminal case. Mention that complete diversity is at most a statutory, not a constitutional, requirement.
 See Krauss, “The American Torts Crisis meets NAFTA: The Loewen Case”, 8 Geo. Mason L. Rev.1 (2000).
 Cite to recent bills relaxing complete diversity rule.
 Cite to Michael Greve book on Federalism.