Dilution of Liability and Multiple Tortfeasors in the Context of Liability for Unrequested Precautions

Assaf Jacob

A Response to Ariel Porat, Private Production of Public Goods: Liability for Unrequested Benefits, 108 Mich. L. Rev. (2009).


One of the more intriguing questions in tort law is the case of joint and several tortfeasors and the dilution-of-liability puzzle. When harm materializes and there are multiple potential tortfeasors, the law tends to limit the number of joint tortfeasors, focusing the final burden on a small number of actors. This limitation is achieved by several legal mechanisms, such as a no duty rule, a narrow interpretation of negligence, a restrictive implementation of the causal link (be it the but for test, the proximate cause test or the rule of intervening cause test), and a doctrine of remoteness of damage. Thus, in the typical accident example, if A, B, and C inflicted risk upon D, often times the tort system will filter out A and B and leave only C to carry the final burden. Ariel Porat's outline of the Expanded Duty of Restitution in his article, Private Production of Public Goods: Liability for Unrequested Benefits, provides an interesting and provocative solution to the dilution of liability puzzle.

I. The Current Problem with Dilution of Liability

Why is the law so reluctant to impose liability on every tortfeasor? After all, a prerequisite for liability is the defendant acting or inacting in a faulty manner (putting aside for a moment the legal regime of strict liability). Only if a defendant deviates from a given standard of conduct will he be held liable. So, back to the example, if A, B, and C all deviate from the reasonable person standard and harm D, the law will often hold only one of them responsible; yet shouldn't they all be liable for the materialized harm?

Limiting the number of defendants, so the argument goes, runs against the interest of the plaintiff, who has fewer tortfeasors from whom to recover his loss. This practice also seems to run against the intuition of corrective justice. Adding tortfeasors improves the plaintiff's position by providing him with more defendants to sue and increases his chances of receiving full compensation, thus shifting the risk of limited resources to the tortfeasors. Yet this is not the case in current tort law.

I have explained this puzzle in a prior article with Professor Alon Harel. Increasing the number of tortfeasors actually aggravates the risk that liability will be diluted. Indeed, ex post-once the damage has materialized-the injured party is best served by having as many tortfeasors as possible, since this maximizes the chances for full recovery. But this does not hold true to the plaintiff's ex ante interests. Having more potential tortfeasors ex ante actually enhances the risk that the damage will materialize because of the phenomenon of dilution of liability. To illustrate this point it may be useful to draw an example similar to one used by Porat:

Assume that A and B pollute a river. The pollution creates harm of $1000 with a probability of ten percent. Installing a filter can prevent the accidental harm. The cost of installing a filter is $60 and any one of the polluters can single-handedly prevent the damage.

In this example, installing the filter is the socially optimal decision. However, under a regime of joint and several liability, each polluter would only carry half the damage (50), which is less than the cost of the filter (60). As Porat points out, if A installs the filter ex ante, he will not be entitled to remuneration from B. Thus, instead of investing 60, reducing the expected loss from 100 to 0, A will do nothing. By doing nothing he is taking the chance of paying half of the loss, which is less than the investment to prevent the loss in the first place. Therefore, instead of reaching the socially optimal outcome, the polluters will do nothing and the harm will materialize.

Notice that in this example concentrating liability on just one tortfeasor provides the parties with the right incentives. Picking only one of the polluters and making this depiction a salient one, causes the private and public incentives to align. Picked individually, both A and B have an incentive to invest in precautions. But it is the addition of an extra tortfeasor that creates the dilution problem. The more tortfeasors there are, the more acute the problem will be, since the addition of extra tortfeasors further reduces the final burden each of them will have to carry.

In a world with low transaction costs, the dilution problem is not very serious and often does not even exist. The parties will contract between themselves to achieve the optimal outcome. Thus, in the example above, when no transaction costs are involved, both A and B would have been better off had they contracted between themselves, splitting the costs of precautions. High transaction costs, however, often impede cooperation. In the typical scenario, not only do the parties have to identify each other but they also have to negotiate how much each of them will pay and who will take the actual precautions. This often proves impossible due to lack of familiarity between parties and the crippling problem of free-riders.

Thus, until Porat's seminal article, the legal system had only two choices: ex ante deterrence or ex post recovery. Should society compromise on ex ante incentives and enhance the probability the plaintiff will receive full recovery, or should it compromise the chances for full recovery in order to reduce the probability of the occurrence of accidents? Now, there is a third choice.

II. The Expanded Duty of Restitution and Joint and Several Liability

In his article, Porat argues that the current law should be modified to include an Expanded Duty of Restitution ("EDR"), which would compel a beneficiary to compensate his benefactors for unrequested benefits. This thesis provides a solution to the problem of dilution of liability. In the multiple would-be-tortfeasor scenario, each tortfeasor is a recipient. By analogy to Porat's argument, the benefactor should be entitled to sue all the recipients for their share in the costs of precaution. Porat's model applies particularly convincingly to multiple tortfeasors situations because not only are the benefits derived from the activity directly enjoyed by the recipients, but they also have positive externalities-preventing damage to unspecified parties who cannot be involved in the process of taking precautions and whose identity cannot be verified prior to the materialization of the harm.

If EDR is applied to the multiple tortfeasor situation, private production of the public good would be achieved through the ability of each tortfeasor to receive indemnification from his fellow tortfeasors. The public good in this case is the reduction of accident costs. In this particular area, it is important to produce the public good by private means because normally the government cannot control nor regulate all private activities that impose risks on society. The costs of information, intervention, and enforcement are prohibitive. Furthermore, such intervention by the government runs against the foundation of a democratic society. The EDR solution minimizes these social costs without the need to single out a limited number of tortfeasors or the enforcement of precautions. Rather than the government, the participants in the activity themselves choose to take precautions and later divide the costs among themselves.

The legal framework Porat suggests would promote an efficient outcome by providing each tortfeasor with the right to sue the others for the costs of precautions. Furthermore, the injured party may be considered one of the tortfeasors if he would have been placed with some of the fault (contributory fault) had the damage actually occurred. If the legal system adopted Porat's EDR, it would no longer have to single out a very limited number of tortfeasors in order to manage the ex ante incentives. The would-be tortfeasors would lose their ability and incentive to refuse to chip in and to veto their participation in taking precautions, and therefore, the number of joint and several tortfeasors can be expanded to the benefit of the injured party without sacrificing economic efficiency.

III. Measures of Recovery

In his article, Porat raises the issue of the measures of recovery-how much should the benefactor be allowed to charge? This is easily dealt with in the case of joint and several liability. There are several alternatives: a share of the costs of precautions (B); a share of the reduction of the expected harm (PL); or a share of the benefits derived f rom continuing/stopping the activity. Sometimes the first and second options converge, but there are many situations where a small i nvestment in precautions reduces the expected loss dramatically.

Porat suggests that the measures of recovery should be either the indisputable benefit or the relative share of the reasonable costs of producing the benefit-whichever is lower. This standard can be easily implemented in the case of joint and several liability. Each tortfeasor should pay his part of the costs of precautions according to his share of the loss were it to materialize. This could be limited by the reasonableness standard in the Learned Hand formula, or by other mechanisms. Basically, each injurer would have had to carry a part of the total harm, had the damage materialized. This expected loss should constitute the maximum amount a would-be tortfeasor has to pay. In this context, EDR is wonderfully simple. For courts, both B and PL are native benchmarks that they apply on a regular basis under the Learned Hand formula. Additionally, such a damage measure would maximize social welfare. Society does not want to encourage over-investment in precautions. Moreover, from a moral perspective, had the other tortfeasors known that there would be an over-investment in precautions, they would have preferred to do nothing and pay their share in the materialized loss.

IV. The Counter Arguments

Acknowledging the controversial nature of EDR reform, Porat discusses several objections that could be raised against it and deflates them with counter arguments. The specific scenario of joint and several liability helps to make the case for EDR reform and renders the general objections to it significantly weaker.

The special case of joint and several tortfeasors lowers autonomy-based objections to EDR. Porat explains that EDR could infringe on recipients' autonomy by obliging them to pay for benefits to which they never consented, but he provides several convincing arguments in rebuttal. In this respect, the paradigm of joint and several tortfeasors is an easy case. The new remedy does not compromise the benefactor's autonomy above and beyond the standard negligence law. Negligence law allows people to participate in activities while taking due care-a term based on objective standards. Once a tortfeasor chooses to participate in a potentially harmful activity, he has a duty of care and should therefore either invest in precautions or pay for the loss. Thus, an injurer cannot argue that EDR obligates him to pay for benefits to which he never would have consented. Since the costs of the required precautions are, by definition, smaller than the expected loss, he cannot convincingly argue that his autonomy is compromised. Moreover, in joint and several liability cases, compromising the injurer's autonomy, if at all, should be balanced against the infringement of the injured party's autonomy. The impact on the injured party's autonomy, especially if we are dealing with bodily harm, makes EDR, in this context, an easy case.

The autonomy argument is closely related to another objection that is based on the notion that different people differ in their attitudes toward risk. My analysis has assumed risk neutrality on the part of the potential tortfeasors. One may, however, argue that the reluctance of a potential tortfeasor to invest in precautions is not necessarily founded on a hope to be a free-rider, but instead on a different risk-aversion. If, for example, A is a risk-taker and B is risk-averse, arguably, granting B the right to force A to invest in precautions may be inefficient. Recall, however, that current legal doctrine requires courts, ex post, to implement the risk neutrality criterion. That is, ex post decisions rely on the assumption that agents are risk neutral. Hence in circumstances in which multiple potential tortfeasors have different levels of tolerance to risk, a risk neutral approach should be adopted-namely, regulations that are based on the risk neutrality assumption along the lines of the rules of precautions proposed above.

Another possible obstacle raised by Porat is the over-valuation problem. By adopting the EDR mechanism, society faces the risk of over-valuing the benefits, which will produce too many non-cost-justified benefits. This problem, however, does not pose a real concern in negligence law, where courts generally measure and evaluate the objective costs of precautions vis-à-vis the expected loss. Once, as suggested, EDR is limited to the lowest of either one's share in the costs of precautions or the expected loss, the risk of over-valuation is rather small.

In rebutting the counter arguments, Porat distinguishes between harm and benefit cases using two issues relevant to this analysis: liquidity (the inability of the recipient to pay for the benefits as he receives them); and the volume of litigation. His article suggests several mechanisms to deal with the liquidity matter, the most relevant to this analysis being delaying payment. In the context of potential tortfeasors, however, liquidity issues should receive a different treatment. Liquidity is one of tort law's main concerns. If tortfeasors believe they will not be able to pay for the materialized loss, they will have fewer incentives to invest in precautions. Why pay now when you will not be able to pay later? This leads to under-deterrence. Tortfeasors may be involved in dangerous activities and not take any precautions, knowing they will not be able to pay for the loss. This argument makes the case for EDR in the context of joint and several tortfeasors, because an early remuneration suit by another would-be tortfeasor is exactly what society would encourage. If the injurer has no funds to invest in precautions he will most probably have no funds to pay the damage. Therefore, he should stop his risky activity at once and switch to a safer activity. An early EDR suit might drive him away from the risky activity before the risk materializes.

The volume of litigation presents a similar argument for EDR in the context of joint and several tortfeasors. One might argue that the proposed amendment of legal norms may increase the overall amount of litigation-thereby producing additional social costs. This, however, is not necessarily the case. According to the proposed mechanism, some of the cases litigated today would never reach the courts because the parties would simply invest in an optimal amount of care. In addition, as Porat points out, potential tortfeasors will be reluctant to turn to court unless they have a strong case, knowing that courts provide reimbursement for reasonable investment in precautions. Moreover, even in cases where damage materializes, the suggested procedure simplifies the litigation and reduces its cost. This is particularly true if the parties receive a "pre-ruling" on the level of care they should adopt.

Pre-rulings themselves have important ramifications. It is well accepted that courts exhibit a hindsight bias in the assessment of the ex ante expected damage. Ex post, the focal point of the courts is the damage itself and objectively estimating the expected damage ex ante is a rather complicated task. But even this problem is resolved under Porat's proposed rule, since, in their pre-ruling perspective, courts are not exposed to the actual cost of the damage and may, therefore, experience no bias in assessing the necessary precautions vis-à-vis the expected loss. Pre-rulings can also play an important role when the parties disagree about the normative need to take precautions-whether because they have different information about the magnitude or probability of the risk, or because they differ in their attitude towards risks.

V. Preventative Measures with Varying Costs

Porat's argument leaves some open questions in the situation where parties can take different precautions at different costs. If the costs of precautions of A and B are different, the legal system should generally aim to place responsibility on the cheapest cost avoider. But how can the tort system expand the number of potential tortfeasors yet provide the right incentives to the cheapest cost avoider? To understand this, consider the following variation on the earlier example: A can prevent the damage at a cost of 60 and B can prevent it at a cost of 80. Society would be better off by giving A the incentives to prevent the harm. However, if A takes no precautions, society would want B to prevent the damage rather than let the damage materialize.

The solution to this example is unclear. One option would be to impose all liability on A. Such a rule would provide A with the right incentives. However, this requires the regulator to estimate ex ante (or the courts, ex post) who is in a better position to take precautions. Oftentimes, this is not an easy task. Moreover, such a rule would provide no incentives for B to take precautions in case A decides not to act.

Another option is to place liability on both A and B and let them decide who is going to invest in precautions. Thus, if A takes precautions he will be entitled to get 30 from B. If B takes precautions, he will be able to charge A 40. The question is how to prevent B from taking precautions or even worse, both parties taking precautions at the same time. One might think of different scenarios-for example, whether each party knows about the other's costs of precautions, or whether there is an emergency that requires immediate response. This is not the right place to elaborate and explore all the possibilities, but there are general principles, similar to those mentioned by Porat, that should guide the courts. In case of an emergency or when there is no way to communicate with the other would-be-tortfeasors, the first mover should take precautions and receive remuneration from the others for his reasonable costs. In other cases, it is better to adopt other mechanisms such as notice or even voting. Thus, if B decides to take precautions he will give notice to A stating his intentions. Upon this notice, A will communicate to B that he can do it in a cheaper manner and therefore B should take no precautions. When timeliness is not an issue, society would want B to communicate with A about his intentions, because such a notice can harness A's private information. Applying EDR to this situation, it is within both A's and B's interests to minimize their costs since both will carry the final burden. They will probably even agree on the outcome. In cases of disagreements, the parties will be able to bring the case to the court to decide whether A or B should take precautions, assuming time is not of the essence.


If we accept Porat's argument and this analysis of the advantages EDR will yield in the realm of multiple tortfeasors, we are left with some interesting questions: How would it affect the time honored doctrines of causation? Remoteness of damage? Intervening causes? If, due to Porat's reform, society cares less about dilution of liability, we might want to expand the circle of potential tortfeasors for a given activity. This has direct implications on the interpretation and application of many legal doctrines that until now have narrowed the circle of liability. Perhaps they will be reexamined as EDR begins to reshape the doctrines of tort law.


Assaf Jacob is a professor at the Radzyner School of Law, Interdisciplinary Center Herzliya.

Suggested citation: Assaf Jacob, Response, Dilution of Liability and Multiple Tortfeasors in the Context of Liability for Unrequested Precautions, 108 Mich. L. Rev. First Impressions 12 (2009),

   //  VIEW PDF
Volume Archive
Monthly Archive
Tagged Posts

& Other Current Events

Crawford v. Washington: A Ten Year Retrospective

No one disputes the significance of Crawford v. Washington, 541 U.S. 36 (2004), which fundamentally transformed Confrontation...

Come Back to the Boat, Justice Breyer!

I want to get Justice Breyer back on the right side of Confrontation Clause issues. In 1999, in Lilly...

Crawford v. Washington: The Next Ten Years

Imagine a world . . . in which the Supreme Court got it right the first time. That is,...

The Crawford Debacle

First a toast-to my colleague Jeff Fisher and his Crawford compatriot, Richard Friedman, on the...

Confrontation and the Re-Privatization of Domestic Violence

When the Supreme Court transformed the right of confrontation in Crawford v. Washington, the prosecution...