The Federalism Project

American Enterprise Institute

A Choice-of-Law Approach to Products-Liability Reform

Michael W. McConnell


from New Directions in Liability Law, Proceedings of the Academy of Political Science (New York, 1988)

Volume 37, Number 1[1]


Thanks are due to the Russell Baker Scholars Fund for financial support during the preparation of is essay, to Christopher R. DeMuth and Michael J. Horowitz, who worked with the author on these ideas some years ago at the Office of Management and Budget, and to Lea Brilmayer, David Currie, Richard Epstein, Larry Kramer, Fred McChesney, Bernard Meltzer, Richard Posner, and Alan Sykes for their probing comments on an earlier draft. 

In recent years, interest in products-liability reform has grown. But interest in federal legislation for that purpose has actually declined. Among the reasons are not only the political strength of reform opponents in Congress but .also the difficulties inherent to the project. Congress, unlike the fifty state legislaires, generally and wisely refrains from experimental legislation, since federal mistakes are notoriously difficult to correct. It is hazardous to impose a single national  "solution” to a problem without a reasonable confidence that the solution can be tolerated for years to come.

Reformers have therefore turned to the fifty states, which Justice Louis Brandeis dubbed “laborator[ies]" for conducting "novel social and economic experiments without risk to the rest of the country."[2] Tort law has always been predominantly a state matter, and the prospects for rethinking the relation between tort liability, a the one hand, and innovation, competitiveness, and prosperity, on the other, would seem far more promising at that level. Indeed, more than thirty state legislatures passed some kind of tort reform in 1986 and 1987.

But there is a major structural obstacle to effective state reform that goes beyond the issue of political will. In a unified economy where products move in national markets, no state can ensure that its own consumers, manufacturers, and workers will capture the major benefits of reform legislation. This essay draws on the insights of the "public choice" school of economics and political science to show how the current legal structure militates against products-liability reform at the state level and, more tentatively, to argue that meaningful reform may require new federal rules governing the choice of law in products-liability cases.

The Importance of Choice of Law

In theory, each state is free to set its own products-liability laws. Some states im­pose strict liability for certain hazards, while others use a negligence standard; some place monetary caps on noneconomic losses, while others allow recovery limited only by the sensibilities of the jurors; and some have relatively liberal standards for awarding punitive damages, while others virtually preclude damage awards of that type. But it is a fiction that any state can actually enforce the products-liability laws on its books. Since most products are made in one state and used in another, at least two states are usually involved, and often more. Obviously, they will not all be able to get their way when their laws differ. Accord­ingly, an elaborate branch of legal doctrine has arisen, called "conflict of laws" or "choice of law," whose goal is to devise relatively predictable and fair rules for determining which state's law should govern in a multistate transaction.

There is no consensus, to say the least, on the best way to resolve the choiceoflaw problem in products liability. As one leading scholar in the field has put it: "There is no conflicts magic that can make sense of the underlying bedlam of rules that passes for products liability laws."[3]  Whatever their prescriptions, choiceoflaw theorists almost invariably use what could be called static analysis. They take the substantive law of each state as a given and see as their task to select which of the various fixed state laws would be best to apply in a particular case. Absent from the debate have been the dynamic effects of choiceoflaw rules: how they influence the substantive legal doctrines the various states will adopt.

A number of normative assumptions should be made explicit. This essay does not assume that any particular substantive standard of products liability is desir­able or undesirable in itself. Instead, it assumes that the "right" substantive rules are those a state would choose to adopt if it were an autarky, that is, a self-sufficient economy, all of whose production and consumption took place within its own borders. In such a case, since all manufacturers and all consumers would be citizens of the state, all the costs and benefits of a change in the laws would be in principle taken into account by the state's democratic processes. The legislative outcome reached under those circumstances is akin to what economists would call a market-clearing  price in a market without externalities, that is, where participants could not throw off costs onto unwilling outsiders.

The optimal products-liability law might well vary from state to state, just as market clearing prices vary from place to place, depending on supply-and-demand conditions. But factors that skew this process of political market-clearingleading states to enact beggar-thy-neighbor laws in hopes of exporting costs to citizens of other statesare here assumed to lower the net satisfaction of the national citizenry with the legislative outcome and thus to be inefficient and undesirable. Products-liability reform, therefore, is legislation that moves in the direction of laws that a state would adopt if left to its own devices.

Under the current system of choice of law, changes in a state’s products-liability laws that reduce the size or probability of plaintiff recoveries injure the interests of consumer-plaintiffs (and plaintiffs’ attorneys) within the state, who either have to sue under less-favorable rules or must undergo the cost and inconvenience of suing elsewhere. Yet such reform, at best, is of limited use in protecting the state's own manufacturers. By the standard outlined above this structure of choice of law is inefficient; it artificially inhibits prodefendant reforms in a way that does not reflect the settled views of the citizenry as a whole.  This essay therefore focuses on inhibitions to prodefendant reform. 

The Structural Bias against Reform

Two economic factors combine to prevent states from protecting their manufac­turers through changes in products--liability laws. First, products tend to move in national markets, while people tend to live in one place. Some goods and many services are made for local use, but most products with major liability exposure are distributed nationally or even internationally. A state’s decision makers will therefore observe that nearly all the consumers injured in the state are local resi­dents and constituents, while most of those who can be sued for making the prod­ucts are residents of other states. By contrast, in related fields of tort law, such as medical malpractice and municipal liability, producers as well as consumers are likely to be local residents.

Second, product manufacturers cannot generally price their products differently to reflect the costs imposed by differing state standards of liability.  Calculating the dollar cost of different liability rules, where hard data come only years after sale, is a daunting if not impossible task. To compound the difficulty, differential pricing would give consumers and middlemen an incentive to buy products in states with less onerous liability laws (and hence lower prices) and transport them to states with more stringent laws (and hence higher prices). This arbitrage would leave the manufacturer doubly defeated: it would be exposed to just as much lia­bility as before but would collect less revenue for its product.

In addition, the Robinson-Patman Act of 1936 broadly outlaws differential pricing. Although a seller can defend such pricing on the basis of actual cost differ­ences, the burden of proof is stringent and the outcome uncertain. While the Robinson-Patman Act goes virtually unenforced today, its vestigial presence may discourage state-by-state price differentials. Note that the providers of local ser­vice (e.g., hospitals, taverns, municipalities) and of the minority of products that are more or less immobile can generally set prices to reflect estimated tort liability with little fear of either transhipment or Robinson-Patman challenge.

It follows from these factors that the cost of a given state’s liability laws, as they apply to mass-market products, is borne by consumers nationwide. In effect, consumers in states with less generous products-liability laws pay a portion of the more generous recoveries won by plaintiffs in other states. This imbalance introduces an incentive for strategic behavior that would not be present if states made rules for themselves alone. Each state can profit at the expense of the others by expanding its scope of liability, at least until the others catch up. A rich litera­ture has arisen documenting the states’ propensity to succumb to just such temptations. In fields ranging from taxation to antitrust to utility rate regulation, they have avidly embraced policies whose costs are borne by other states.

Resolving suits under the law of the plaintiff's own state would, in itself, create a structural bias toward proplaintiff state laws. But developments in recent years have added a different sort of bias, by making it easier for plaintiffs to shop around for the most favorable law--whether that of their own state, the defendant's, or a third state.

It was once the general rule that tort cases were governed by the law of the state where the injury occurred. Thus, if a consumer living in Ohio purchased a product in Michigan from a New York manufacturer and was injured by it while traveling in North Dakota, North Dakota law would apply. Most often, this meant in practice that the law of the plaintiff's home state would apply, since consumers ordinarily use products, and thus are injured by them, at or near home. The distinctive feature of this choice-of-law rule was its so-called mechanical operation. The applicable substantive law was fixed at the time of the injury without regard to its relative attractiveness to plaintiff or defendant or to subsequent litigation tactics.

This rule came under intense attack by progressive legal commentators and courts. Essentially, their argument was that the rule was arbitrary and failed to take into account the degree of "interest" of the states involved in the conflict. For example, in the hypothetical situation above, it was argued that both Ohio and New York have a stronger interest in the outcome of the case than does North Dakota. Ohio has an interest in ensuring that its residents are adequately compensated for injuries and do not become public charges. New York has an interest in protecting the profitability of its manufacturers and also in seeing to it that safe and wholesome products are produced within its borders. North Dakota’s interest in the outcome was said to be relatively slight. In this instance, the reformers concluded, to apply North Dakota law was to sacrifice real sovereign interests to imaginary ones.

The traditional rule of applying the law of the place of injury has therefore given way to "interest analysis," under which the law of the state with the most substantial interest in the conflict will apply. Most often, courts take this to mean either the plaintiff's or the manufacturer's home state. At this level of abstraction, interest analysis seems a major improvement over the prior rule. The prior rule was arbitrary and led to senseless results in some cases. The difficulty, however, is that interest analysis is not neutral; it systematically loads the dice in favor of plaintiffs and against defendants.  It tends to select whichever states’ law—the plaintiff’s or the defendant’s—seems likely to allow a higher recovery. (If their laws are equivalent, there is no choice-of-law problem.)

Interest analysis begins with the assumption that states are interested only in obtaining benefits for their own residents. The interests reflected in a proplaintiff law are said to be greater compensation for injured consumers within the state, or, alternatively, greater deterrence of unsafe manufacturing within the state. The interest reflected in a promanufacturer law is said to be the promotion of manufac­turing within the state.

In roughly half the cases, where the defendant’s state law is more favorable to the plaintiff, interest analysis almost automatically favors the plaintiff. The defendant’s state is said to be expressing: its interest in deterring unsafe manufacturing. The plaintiff’s state has no countervailing interest in applying its own promanufacturer law, since the purpose of such a law is not served by extending its protection to an out-of-state manufacturer. And so the proplaintiff law applies. David Cavers has approvingly summarized the situation: “A forum where the victim resided would have an obvious interest in his compensation if its law were pro-claimant. If its law were pro-producer, the court might well conclude that, having a resident victim but no local producer to protect, it should apply the law of the state of acquisition or of production if either were pro-claimant."[4] The exceptions seem to be relatively rare.

In the other half of the cases, where the plaintiff's own state law would allow higher recovery, the analysis gets more complicated. The plaintiff's state has an interest in gaining greater compensation for its injured consumers, but the defen­dant's state also has an interest in promoting manufacturing within its borders. This situation is called a "true conflict," because both states appear to have an interest in applying their own law. There is no uniformly agreed-upon solution to a "true conflict"; but two leading solutions turn out to be proplaintiff. One is to apply the law of the forumthat is, the state in which the lawsuit is brought. Since the plaintiff chooses where to file suit, he will get to select the most favorable forum, in this case his own home state.

The other solution is more straightforward—to choose the law more favorable to the plaintiff.  Legal commentators generally agree that this is the usual result of modern conflicts analysis and that it is a satisfactory resolution. “Among the factors that should be borne in mind in framing choice-of-law rules,” as Willis Reese has put it,  "are the policies underlying the relevant substantive field." In he field of torts this points to "a policy favoring compensation for injury and spreading the risk."[5] This is not the place for a full discussion of the merits of importing substantive "policy" into choice of law. But if one sees the aim of liability policy as striking a balance between values, rather than simply maximizing compensation, one will wish to guard against errors in either direction, toward either too much or too little compensation for injury. And one will not so easily dismiss state laws that are less favorable to plaintiffs than others as obviously mistaken or illegitimate. At any rate, introducing a systematic bias in choice of law is certainly apt to affect the states' underlying substantive decisions about products liability.

The involvement of third, fourth, or fifth states in the transaction gives the plaintiff even more chances. If neither the plaintiff's nor the manufacture’s state allows liberal recovery, there is always the state where the injury took place or the state where the product was sold. The involvement of added plaintiffs and defendants complicates the issue enormously but without changing the bias. In short, although courts still occasionally apply the rules of a prodefendant state in conflicts, the current system of choice of law is heavily weighted in the other direction.

At the same time that a proplaintiff style of interest analysis was sweeping the field of choice of law, new modes of constitutional interpretation were eroding old constraints on plaintiffs' ability to shop for a favorable forum and on forum states' ability to apply a law favorable to plaintiffs.

The Supreme Court has acknowledged that forum-shopping has-become less difficult: "The limits imposed on state jurisdiction by the Due Process Clause, in its role as a guarantor against inconvenient litigation, have been substantially relaxed over the years."[6] A dramatic illustration is Keeton v. Hustler Magazine, Inc., in which a resident of New York alleged that she had been libeled by Hustler maga­zine, an Ohio corporation with its principal place of business in California.[7]  By the time she filed her lawsuit, the statute of limitations had expired in New York, Ohio, and every other state except New Hampshire. And so she sued in New Hamp­shire, a state with which she had had no previous contact whatsoever and in which only a tiny percentage of Hustler magazines were sold. The Supreme Court up­held her suit, holding that a corporation can be sued in any state in which it dis­tributes its products. And under New Hampshire law, she was entitled to collect damages for injury to her reputation in all fifty states, including the forty-nine whose statutes of limitations had already expired. (The court did not resolve the question whether the state could constitutionally apply its substantive law.)

For forum-shopping to work, not only must the plaintiff be entitled to sue in a state with a favorable law, but that state must also be entitled to apply its own law to the suit. This second step raises independent constitutional concerns of its own. Before 1981, the Supreme Court struck down as unconstitutional lower-court decisions that applied the law of a jurisdiction that did not have a substantial in­volvement with the events giving rise to the lawsuit, whether or not it was the forum state. This still allowed plaintiff-oriented courts to choose between the states where the plaintiff resided, the manufacturer was located, and the injury occurred to find the law most favorable for the plaintiff. Nonetheless, it did provide at least some constraint. But in 1981, with Allstate Insurance Co. v. Hague, even this con­straint was effectively eviscerated.[8]

Allstate grew out of an accident in Wisconsin near the Minnesota border. A Wisconsin resident was killed while a passenger on a motorcycle driven by a second Wisconsin resident that collided with an automobile driven by a third Wisconsin resident. The crash victim had purchased an insurance policy in Wisconsin from Allstate, a nationwide insurer. Under Wisconsin law, the deceased was entitled to $15,000 in benefits under the uninsured-motorist provisions of the insurance. But the victim’s estate sued Allstate in Minnesota for additional compensation under Minnesota law. In Minnesota, as in many other states, someone killed in similar circumstances would be entitled to $45,000 under the same policy.

The Supreme Court upheld the Minnesota state court's application of its own law for three reasons. First, after the accident, the victim’s widow, who was exec­utor of his estate, moved to Minnesota. Second, the victim had been employed in Minnesota and commuted to work from his home in Wisconsin. Third, Allstate does business in Minnesota.

If those three contacts between Minnesota and the litigation are sufficient, it is difficult to imagine what state’s law a plaintiff would not be able to invoke be­fore a sympathetic court, at least in a liability case involving a nationally dis­tributed product. The first Allstate factor, the current residence of the plaintiff’s executor in Minnesota, allows the litigation to be governed by the strategic behavior of the plaintiff. As Justice Lewis Powell commented in dissent: "If a plaintiff could choose the substantive rules to be applied to an action by moving to a hospitable forum, the invitation to forum shopping would be irresistible."[9]  The second factor, the victim’s employment in Minnesota, allows the case to be governed by an irrelevancy. The victim's employment in Minnesota surely increased the odds that he would come to some harm in Minnesota, but in fact the accident had nothing to do with his employment and occurred in Wisconsin. Had the case revolved around minimum wages, workers' compensation, or the terms of the employer's benefits package, the employment relation might have been significant. Under the circumstances, his employment in Minnesota was no more relevant to the lawsuit than his state of birth or the site of his alma mater.

The final Allstate factor, that the insurance company does business in Minnesota, is the most troubling. Under this theory, a manufacturer that distributes products nationwide can be sued under the country’s most stringent state law –whether or not the plaintiff or the product at issue ever entered the state. To introduce his factor is to eliminate all constitutional constraint regarding choice of law for nationally distributed products.

The Allstate decision was not, to be sure, an abrupt departure from precedent. In actuality it had been decades since the Supreme Court last reversed a state court for applying its own law to an interstate conflict. In the intervening years the constitutional standard had seemingly relaxed. Nonetheless, Allstate was a signal that the relaxed standards had become formal doctrine. The extent of the change is not yet clear. Four years after Allstate, in Phillips Petroleum Co. v. Shutts[10] the Court held that the forum state in a class action could not apply its substantive law to transactions with no relation to the state. It is too early to tell whether Shutts represents a retreat from the extremes of Allstate or merely shows that some applications of forum law are beyond the pale even under the relaxed standards of Allstate.

Even under the old rule in conflicts – place of the injury governs – any one state legislature would have little incentive to enact reforms that reduce the scope of ability, because the reforms would mostly benefit out-of-state manufacturers but would inflict most of their cost on plaintiffs within the state (and, not incidentally, on the plaintiffs' bar). If plaintiffs are forced to sue in other states, they will suffer some cost and inconvenience and will sacrifice the advantage of a jury predisposed to the local party. But the new approach to choice of law accelerates this end. Even if a state is willing to say no to some of its own plaintiffs in order to help defendants nationwide, plaintiffs can sidestep the prodefendant law through combination of forum-shopping and choice-of-law doctrine. Either they will sue elsewhere or they will obtain application of a more favorable law. Thus the most proplaintiff states will set the law for most of the country.

Under these conditions, states no longer serve as laboratories for social ex­perimentation. The resultant law of products liability does not and cannot reflect actual judgments and experiences, regarding the effect of liability on safety and compensation, on the one hand, or on innovation and wealth creation, on the other. Instead, states pursue a persistent and one-directional race toward ever­-higher plaintiff recoveries, a race whose outcome does not necessarily represent the considered judgment of decision makers in the several states. 

Full Faith and Credit Legislation

Article IV, section 1, of the United States Constitution provides: "Full Faith and Credit shall be given in each State to the public Acts, Records, and judicial Proceedings of every other State. And the Congress may by general Laws prescribe the Manner in which such Acts, Records and Proceedings shall be proved, and the Effect thereof." This explicit delegation of authority to Congress to prescribe the effect of the public acts or statutes of each state in the other states has never been exercised. The sole relevant federal statute, 28 U.S.C. 1738, provides only that state statutes "shall have the same full faith and credit ... as they have by law or usage in the courts of [the] ... State ... from which they are taken." This statute is either meaningless or internally self-contradictory. If conflicting laws of two states appear to apply to a legal controversy, it is impossible for each state to give the other's law the "same full faith and credit." The statute has accordingly played no role in the development of choice-of-law principles. Nonetheless, the federal power to adopt new choice-of-law rules is evident. Such rules would interfere far less with state sovereignty than would a federal products liability code passed under the commerce power.  What choice-of-law principle would best eliminate the bias in the system?

The sources of bias in the current system, as have been discussed, are two: (1) the proplaintiff orientation of interest analysis coupled with latitude for forum shopping, and (2) the interest of the plaintiff's own state in allowing generous recovery, to be paid by consumers nationwide through higher prices. The solution to the first source of bias is to return to a choice-of-law principle under which the law governing a dispute is fixed at or before the time of injury.  Plaintiff’s should not be allowed to select the most favorable rule, and judges should not be allowed to weigh the “"interests" of the various states retrospectively.

Among the various fixed-rules for products-liability cases, some will create greater incentives for reform legislation and some less. Applying the law of the plaintiff's home state will create incentives for proplaintiff products-liability rules, since the price will be paid nationwide. Applying the law of the manufacturer's home state would have the opposite effect: most of the benefits of promanufacturer legislation under this choice-of-law regime would be enjoyed within the state, while much of the cost would be exported to injured consumers in other states, who would receive lower recoveries. Thus, while either of these fixed rules--the plaintiff’s state governs or the manufacture’s state governs—would be preferable to the current system, both depart sharply from the autarkic result.

The traditional choice-of-law rule, the place of injury, is not much better. In most cases, the place of injury will also be the plaintiff’s home state and will thus generate the same incentives for legislative strategic behavior. In the other instances, when the plaintiff is injured in a state other than his own, the place of injury will likely have no interest in favoring either party. This is an improvement, but it still does not create any incentive for reform.

Although further exploration of the incentive effects of choice-of-law rules on legislatures may lead to superior solutions, the best rule on the basis of current thinking is the law of the place of sale.  "Place of sale," in this context, means the place of delivery of a product to the ultimate purchaser through ordinary commercial channels. If a Pennsylvania consumer drives to New York to buy a product, the place of sale is New York. If he calls a New York retailer, which mails the product to him in Pennsylvania, the place of sale is Pennsylvania. If a New Yorker buys a product in New York and resells the product years later in Pennsylvania to a Pennsylvania consumer, the place of sale remains in New York. Although there will be some difficult problems of proof, it will be feasible to determine the place of sale of most products involved in products-liability cases. The place of sale of drugs can be determined through prescription records; the place of sale machine tools and automobiles is recorded on documents of sale; mail orders have a paper trail. There will be relatively little opportunity for manipulation by the parties.

The incentive effects of this choice-of-law rule are likely to be modest, but positive.  It will tend to discourage extreme liability rules. 

Manufacturers can refuse distribute products in a state that adopts an unreasonable rule of damages. Consumers can, at least theoretically, avoid purchases in states with insufficient protection. State legislatures have an interest in increasing sales activity within their states, which is likely to give them a constructive interest in avoiding rules that repel either side of the transaction. One can imagine the consumer outrage if, for example, a New Jersey decision causes the producer of an important birth-control device to withdraw from sales in that state, when the device remains available across the border in New York. In theory, the effect of this choice-of-law rule would to create an incentive for each legislature to strive for the optimal products liability law: one that will attract the largest aggregate of producers willing to sell and consumers willing to buy.

Moreover, the proposal might make it possible for manufacturers to respond to different state products-liability laws by differential pricing—a response that is virtually impossible under the current regime. To the extent that differential icing is now precluded by the ability of purchasers to buy a product in one state and use or resell it to users in another, the proposal would alleviate the problem. If it became possible for manufacturers to price the product in each state according the liability exposure created by sales in that state, then the proposal would virtually replicate the autarkic result. The costs and benefits of products-liability law would be felt largely within the state: consumers would pay more for more generous products-liability protection.

Although the proposal would furnish a major impetus in theory for legislatures to design optimal products-liability laws, its effect in practice is likely to be less dramatic for three reasons.  First, unless a state's products-liability law is far outside the mainstream,  it is likely to prove a relatively minor factor in manufacturers' decisions to distribute products and consumers' decisions to purchase them.  Manufacturers are unlikely to sacrifice sales in a state on account of small differences in liability exposure, however irksome. Consumers are unlikely to forgo a useful product, or travel to a distant state to purchase it, merely because the law of their own state provides for a somewhat lower recovery in the event of injury. Thus, while the proposal is likely to discourage the extremes, it may do little to advance the optimal solution. Second, though manufacturers would have more latitude to adjust price according to liability exposure in each state, the technical difficulty of doing so with any sort of exactness would remain, so price differentials are likely to be quite rough if they emerge at all. Third, the place of sale will commonly also be the consumer's home state, and many state legislatures are likely to continue responding to those constituent incentives rather than seek increased sales.

The proposal thus falls short of perfect neutrality in state incentives; nothing could achieve that in a system of nationwide distribution of products and state determination of products-liability laws.  Nonetheless, it would be an improvement over the current system. It would discourage litigants from taking their lawsuits to the most proplaintiff forums in the country. It would also discourage them from manipulating the choice of law to their advantage, since that choice would hinge on fixed factors outside the litigants' control. It would bring greater certainty to private planning: producers would have a fair idea that the distribution of state laws they faced as defendants would roughly match the distribution of underlying complaints. It would create a modest incentive for state legislatures to promote sales within their states by passing rules that balance the interests of producer and consumer and would accordingly curb the tendency of every state to emulate the most proplaintiff law in the country.

The recent experience of liability reform shows the importance of getting both costs and benefits into the legislative trade-off. More than thirty states have passed liability-reform legislation since 1985. Certain reforms are by far the most common--limits on medical-malpractice claims, on municipal liability, and on the liability of taverns and social hosts for accidents caused by persons to whom they serve liquor. The common feature of these reforms is that defendants as well as plaintiffs reside in the state in question. Limitation on ski-resort liability is likewise very common; it is surely no coincidence that in those cases the defendants are in-staters and the plaintiffs are mostly from other places. By contrast, reform of products liability, in which defendants tend to be from out of state, is compara­tively less common, even though the issue is more significant in dollar terms. This suggests that where both the cost and the benefit of a liability rule are felt within a state, legislatures are more disposed toward reform.

It is more difficult to compare the effects of this proposal with those of a possible federal products-liability code, since so much would depend on what the federal code would contain. There are at least three theoretical reasons to expect state legislation to be preferable to federal.

First, thinking about products liability is prone to dramatic shifts. Over the past twenty years the notion of basing liability on fault went out of fashion and was replaced by the notions of risk-spreading and enterprise liability. Then there was a swing back, as society came to see the ill effects of excessive liability and the tort system’s inefficiencies in spreading risk. Legislative activity is thus taking many experimental forms around the state capitals. Some states are bucking the trend by continuing to expand the scope of liability in particular areas.

It is no accident that state legislatures in more than thirty states have passed liability-reform laws in recent years, while Congress has only debated. The Constitution was deliberately designed to slow the pace of change and guard against fads and fashions. It takes longer for Congress both to make and to correct a mistake. When the state of knowledge and opinion on an issue is in flux, it is generally sensible for the federal government to hold back. If a federal products-liability code is enacted in 1988, Americans may well be living with it for the next half century, much as they still live with New Deal throwbacks like the Agricultural Marketing Agreements Act of 1937 and with the energy-crisis package that was passed in 1978 and became obsolete by 1982.

Second, public preferences about products liability probably vary from region to region. Relying on state legislation allows more people to live under laws that suit them. A single national solution necessarily sacrifices the views of more people whose opinions differ from the norm.

Third, if incentives are properly structured so that states cannot succeed in benefiting their own citizens at the expense of others, they will have an extra reason to adopt policies more sensible than those the federal government would adopt. People and businesses can migrate far more freely from state to state, in response to changes in law than from country to country. If New York enacted a liability law that manufacturers or consumers found harsh or threatening, they could move to New Jersey or Connecticut to escape it. Even gradual and marginal migrations of this sort influence state officials. If Congress enacted a harsh liability law, on the other hand, the dissidents would have to move abroad, which is a lot harder. Just as a competitive business has more incentive to give good service than a monopolist, state governments have a greater incentive to pass wealth-generating legislation than Congress does. The advantages of state over federal legislation should not be exaggerated. But given the poor prospects that comprehensive products liability reform legislation will clear Congress, it is worth considering ways of working within federalism. 


State law cannot be treated simply as a given. It is the result of a process of political deliberation, political power, and political negotiation. The outcome of the process is powerfully affected by the structure of decision making, just as the outcome of a contest of brawn or brain is affected by the rules of the game. The choice-of-law rules for products liability have been set without considering their effect the substantive state laws themselves. A shorthand description of the rules--much oversimplified, to be sure, but basically accurate--is that plaintiffs in an interstate case can choose which states law will apply to their lawsuit. If this is accurate description, the prospects for products-liability reform at the state level are not promising. More precisely, individual states will be unable to enforce, and disinclined to enact, products-liability rules more favorable to manufacturers than the laws of other states.

In part, this bias is the result of changes in choice-of-law and personal-jurisdiction rules over the past thirty years. One possible avenue for reform is to modify the choice-of-law rules to eliminate proplaintiff bias without replacing it with prodefendant bias. Such reform would give each state an incentive to develop wealth-­generating rules. And it would discourage the sort of strategic behavior by which states attempt to profit at the expense of their neighbors but succeed only in ensuring the disadvantage of all.

[1] Note: this article is slightly modified from the original. 

[2] New State Ice Co. v. Liebmann, 285 U.S. 262, 311 (1932). 

[3] Russell Weintraub, “A Defense of Interest Analysis in the Conflict of Laws and the Use of that Analysis in Products Liability Cases," Ohio State Law Journal 46 (1986): 493, 504. 

[4] See David Cavers, "The Proper Law of Producer's Liability," International and Comparative Law Quarterly 26 (1977): 703, 715.

[5] Willis Reese, "Products Liability and Choice of Law: The United States Proposals to the Hague Conference," Vanderbilt Law Review 25 (1972): 34.

[6] World-Wide Volkswagen Corp. v. Woodson, 444 U.S. 286, 292 (1980).

[7] 465 U.S. 770 (1984).

[8] 449 U.S. 302 (1981) (plurality opinion). 

[9] 449 U.S. at 337. Accord, Reich v. Purcell, 67 Cal. 2d 551, 63 Cal. Rptr. 31, 432 P. 2d 727 (1967) (Traynor, C.J.). 

[10] 472 U.S. 797 (1985).