The FINAL VERSION of this paper will be published in the November, 1994 Santa Clara University Law Review. .
[ Intro Page | Table of Contents | Sec. 1 | Sec. 2 | Sec. 3 | Sec. 4 | Sec. 5 | Sec. 6 | Sec. 7 | Law & Econ Page ]
[ Go to Previous Section, Section V | Go to Next Section, Section VII ]
For most conspiracy situations, the allocation of damages per capita[234] is preferable to a comparative fault[235] or a comparative benefit allocation.[236] The per capita liability allocation rule will allow for administrative cost savings, reduction of influence costs (legal expenditures and information hiding and collecting costs by the parties), increased certainty for the plaintiff in settling, increased detection of conspiracies, and ex ante conspiracy fairness. In certain contexts, however, such as when one conspirator is able to control another conspirator, liability allocated by comparative fault will lead to better results than per capita liability allocation.
Conserving judicial resources and efficiently resolving issues are current concerns of many courts. One benefit of a per capita allocation of liability is the ease with which it can be administered.[237] Both the comparative fault and comparative benefit liability allocation rules require not just a determination of who the defendants are, but also a determination of the fault or benefit of each defendant in the conspiracy.[238] Certainly it will be easier for a court to determine only who the conspirators are as opposed to determining both who the conspirators are and the proportion of damages caused or benefit received by each conspirator.
The comparative benefit liability allocation rule may be simpler to administer than the comparative fault liability allocation rule in situations where the benefit of each defendant can be measured. Sometimes a proxy for benefit received, such as market share, may be an appropriate measure.[239]
The per capita liability allocation rule is an objective rule because it would be difficult to distort the number of defendants involved in the conspiracy. The clarity of the per capita allocation rule will lead litigating parties to spend less trying to influence the apportionment of liability. Conspirators may also exert less wasteful effort to hide and gather information and evidence during the conspiracy under the per capita liability allocation rule.[240]
Both the comparative fault and the comparative benefit liability allocation rules may lead courts to make imprecise allocations regarding liability. The imprecision in allocating liability under these rules will lead defendants to exert effort in order to obtain the "best allocation" for themselves, an allocation that leaves them with low liability for the damages.
Under the contribution with claim reduction rule, the plaintiff will also be involved in exerting effort to show that a defendant it has settled with caused little damage, thus allowing the plaintiff to seek more from the other defendants.[241]
Under the comparative fault liability allocation rule, conspirators may hide evidence of the amount of damage that they have caused, while simultaneously trying to obtain evidence that can be used to show that other conspirators have caused more damage. Similarly, under the comparative benefit liability allocation rule, conspirators may exert effort to receive their share of the conspiratorial benefit in a form that cannot easily be measured.
Of course, even the per capita liability allocation rule will lead to some wasteful hiding costs, although less than under the other two liability allocation rules.[242] Under the per capita allocation liability rule, the parties will attempt to hide the conspiracy, but not to distort the allocation of liability. The comparative benefit liability allocation rule would seem to result in less wasteful hiding and collecting of allocation evidence than the comparative fault liability allocation rule, as the comparative benefit liability allocation rule appears to be a more objective measure.
Under the contribution with claim reduction rule, the plaintiff will want to determine the liability of a defendant that it settles with, as it is the liability amount, not the settlement amount, that is subtracted from the amount of damages the plaintiff can seek from the other defendants. The per capita liability allocation rule will make it easier for the plaintiff to determine the allocation of damages attributable to each defendant. The plaintiff will thus expend fewer resources in trying to determine the allocation of liability among the defendants. The per capita liability allocation rule will also lead to less uncertainty for the plaintiff in the settlement process, and may increase the likelihood of settlement under the contribution with claim reduction rule. As former Assistant Attorney General William Baxter noted:
If a plaintiff could calculate [the defendant's] share [of the damages] accurately, it could then settle with a defendant if the amount appeared adequate. But if a plaintiff cannot calculate the share, it may shun settlements because of the risk of giving up an unknown proportion of its total claim.Certainly, the plaintiff's certainty in this respect depends critically on the method chosen to calculate contribution shares. The most straightforward method is the "pro rata" or equal shares method. . . . Comparative fault is a slightly more sophisticated method of computing contribution shares. . . . Comparative Benefit is still another method of computing contribution shares.[243]
In situations where the benefit of each party is easy to determine, the comparative benefit liability allocation rule may be preferred over the comparative fault liability allocation rule.
Professors Kornhauser and Revesz argued that the per capita liability allocation rule could lead to over-deterrence in joint and several liability cases for some defendants by exposing them to too much liability relative to the benefit they would receive in a worthwhile activity.[244] But for most conspiracy situations, parties can bargain over the division of benefits and, thus, the per capita liability allocation rule will not overly deter conspirators. In cases where the allocation of benefits can be negotiated by the conspirators, those conspiracies that produce more total benefit than total liability for the conspirators will take place.[245]
The transaction costs in reaching an additional agreement on the division of the benefit will have some deterrent effect. Yet parties will already incur some transaction costs in reaching the agreement on the conspiratorial act itself. These costs will include finding other members with whom the agreement can be made as well as the actual agreement on the act. Because parties must already reach an agreement on the act, the marginal costs of bargaining over the division of the benefit may be low in many situations.[246] In addition, if the parties know that they are to be held liable for the damage per capita, then the negotiations regarding the splitting of the benefit should be even easier for the parties than in situations where they must negotiate over the division of the benefit but do not know what their precise proportions of liability will be.
There could be some situations where the additional transaction costs are high or asset-constrained parties are unable to make needed side payments such that a profitable conspiracy will not take place. In these situations, we might wish to use the comparative fault or comparative benefit liability allocation rules.[247]
A per capita contribution rule may also lead to more detectable conspiracies and lessen the likelihood that hidden conspiracies will be able to operate indefinitely. Conspiracies may be more detectable under a per capita liability allocation rule because side payments to account for the per capita division of potential liability may make the conspiracy more visible.
In addition, under the comparative fault liability allocation rule, parties are more likely to have asymmetric information as to who will be found at fault ex ante the conspiracy. This could affect their decisions to enter into a conspiracy.[248] If parties hide information about their own fault while collecting information about others' fault, then all parties might believe that they will be liable for little of the damages, and thus, they may be more likely to enter into a "non-beneficial" conspiracy than if the damage allocation is known more precisely ex ante the conspiracy.[249] The per capita rule makes it less likely that the parties will have asymmetric information regarding their liability.[250]
Finally, I recommend the comparative fault liability allocation rule for situations where deterring the conspirator that causes much of the damage is necessary to deter the conspiracy as a whole.[251]
Another argument against per capita contribution is that it may operate "inequitably" and thus produce a form of "rough justice."[252] However, this is not true if the conspirators know that the per capita liability allocation rule will be applied before they enter into a conspiracy. If each conspirator knows that all conspirators will be held equally liable for the damages, then each conspirator will demand a share of the benefits that covers its expected liability, or else it will refuse to take part in the conspiracy.
Information regarding which conspirators are taking risks or receiving a benefit, or how the benefit is being received may be difficult for the court or plaintiff to determine after the conspiracy has been detected. A per capita liability allocation rule may be more fair than the comparative fault or comparative benefit liability allocation rules because it places the conspirators, the parties with the most knowledge of the risks and benefits of the conspiracy, in the position of determining the fairness of their conspiracy.
Further, the per capita liability allocation rule will yield the same results as the comparative fault rule in some situations. In circumstances where each defendant is essential to the carrying out of the conspiracy, it can be argued that each defendant should be equally liable for the damage. Of course, this is simply a case where the results of the per capita and the comparative fault liability allocation rules happen to reach the same result. But explicitly stating that the per capita rule will be used in some situations may lead to quicker resolution and cost-savings in determining the liability of each defendant.
Allocating the damages by either the comparative fault or comparative benefit liability allocation rules may allow the conspirators to lower their transaction costs in forming the conspiracy, as the benefit each conspirator receives without side payments may correspond to their expected damages. In such cases, the comparative fault or comparative benefit liability allocation rules are not increasing fairness among the conspirators but simply lowering the conspirators transaction costs in forming the conspiracy (while increasing the administrative, influence and uncertainty costs at trial). There is no reason to believe that as long as the conspirators know that liability will be allocated among them per capita that the conspiracy will be unfair, although the conspirators may have higher transaction costs in allocating the benefit and making any necessary side payments. The fairness argument against the per capita liability allocation rule, as applied to conspiracies, really comes down to favoring the court's allocation of liability after a trial has begun over the conspirators allocating the benefit when forming the conspiracy. The reasons we should believe that courts will make fairer decisions at trial than the conspirators make in entering into the conspiracy do not seem at all clear.
Suppose two conspirators, firm A and firm B, are contemplating entering into a conspiracy. Suppose that the expected benefit of the conspiracy to both firms is $120, but the benefit will flow to the firms such that firm A receives $80 and firm B receives $40. Suppose that the total expected damages of the conspiracy is $100,[253] and that the expected liability is allocated per capita such that each firm has an expected liability of $50.
In this case, firm A will want to enter into the conspiracy, as its expected benefit of $80 is greater than its expected liability of $50, while firm B will not want to enter into the conspiracy, as its expected benefit of $40 is less than its expected liability of $50.
But firm A and firm B can reach an agreement, in addition to an agreement on the act of the conspiracy itself, on a side payment from firm A to firm B. In this case, if firm A agrees to make a side payment to firm B anywhere in the range of $10 to $30, then the conspiracy will occur, as each party will have an expected benefit that is greater than its expected liability.[254]
The example shows that if the conspirators are able to reach an agreement on the division of the conspiratorial benefit, then those conspiracies that produce a net benefit to the conspirators as a whole will take place.
In contrast to most situations where I recommend the per capita liability allocation rule, I strongly recommend the comparative fault liability rule for unfair conspiracies. Unfair conspiracies would include situations where one conspirator has effective control over another conspirator.[255] For example, an original supplier could have control over an agent. If the original supplier wants the agent to boycott a new supplier, then the original supplier may threaten the agent with cancellation of its agent's status, effectively putting the agent out of business if it fails to go along with the boycott. I do not recommend, however, the comparative fault liability allocation rule for situations where the conspirators merely have disparities in culpability.
When examining the deterrence of liability a controlled party faces, we must compare the penalty threatened by the controlling party if the controlled party does not join the conspiracy with the penalty of potential damage liability from a plaintiff if the controlled party does join the conspiracy. If the controlling party can threaten the controlled party with a large penalty, then it is unlikely that the controlled party will face much deterrence from being held accountable for some of the damages in the conspiracy.
In such a case where there is control by one conspirator over another, we will want to place more of the damage liability on the conspirator who has the control, as it is this party that can be deterred.[256] It is the additional fault that the controlling conspirator has by threatening the controlled conspirator into the conspiracy that can be used as the basis for increased liability.
Control does not necessarily have to result from threats by the controlling party. Control can also manifest itself in the form of misrepresentation of fact. For example, a conspirator may have information that the conspiratorial act violates the law, but misrepresents its lawfulness to another conspirator, preempting the information-receiving conspirator from obtaining legal advice regarding the act. In this case, we might want the conspirator who made the misrepresentation to be liable for more of the damages than the other conspirator because of its fault or control through misrepresentation.
We need to separate good threats from bad threats. For good threats, such as price competition, the comparative fault liability allocation rule should not apply. We do not want a conspirator to be able to claim that it joined a price-fixing conspiracy because others threatened to compete with the conspirator on price. We only want to use the comparative fault liability allocation rule in the control situations where we feel that one conspirator is unfairly controlling another conspirator through a bad threat.
Of course, we do not want every conspirator to claim that it has been controlled by another conspirator. Thus, the standard at which the comparative fault liability allocation rule is used should be quite high. For example, we might want to use the requirement of economic ruin in antitrust cases. In addition, a set rule could be used, such as allocating 50% of the controlled conspirators liability to the controlling conspirator.[257] While this would lead to some failures in the internalization of damage, the set rule would reduce the administrative and influence costs in determining the allocation of fault.
The following table is a summary of my recommendations for liability allocation under the contribution with claim reduction rule.
| Per Capita Liability Allocation Rule | Comparative Fault Liability Allocation Rule | Comparative Benefit Liability Allocation Rule |
|---|---|---|
| Most conspiracies. Default rule for conspiracies. |
Unfair conspiracies where one conspirator has control over another through use of a bad threat. Joint harm caused by non-conspirators. |
Costly for parties to bargain ex ante over the division benefits and easy to measure the benefits ex post. Joint harm caused by non-conspirators. |
[ Go to Previous Section, Section V | Go to Next Section, Section VII ]
In order to apportion liability, it would be necessary to determine a relevant market and calculate the market share of each firm. This is no mean feat; the National Commission for the Reform of Antitrust Laws and Procedures concluded in 1979 (Report Vol. II, at 95-96) that in complex cases market determination takes some 55% of trial time, averaging several months of trial per case. Market definition is complex because there is no "right" market. A market definition is just a proxy for power over price, and in any given case there may be several equally plausible market definitions.Antitrust Equal Enforcement Act: Hearings Before the Comm. on the Judiciary of the United States Senate, 97th Cong., 1st & 2d Sess. 182, 201-02 (1981-82) (statement of Frank H. Easterbrook, Professor of Law, University of Chicago Law School). Return to text
[Note 239] See Antitrust Equal Enforcement Act: Hearings on S995 Before the Comm. on the Judiciary of the United States Senate, 97th Cong., 1st & 2d Sess. 11, 25-26 (1981-82) (statement of William Baxter, Assistant Attorney General). But see supra note 238. Return to text
[Note 240] The effort spent on the illegal activity, such as setting up the conspiracy or hiding and collecting information of the conspiracy, is a social waste. That effort could be spent doing productive, socially beneficial activities. See Gordon Tullock, The Welfare Costs of Tariffs, Monopolies and Theft, 5 W. Econ. J. 224 (1967). See generally C. Charles K. Rowley et al., The Political Economy of Rent-Seeking (1988). Return to text
[Note 241] The plaintiff will have the same incentives under the contribution with settlement reduction rule where a Mary Carter agreement has been signed between the plaintiff and a settling defendant. See supra Return to text
[Note 242] See Tullock, supra note 240. Return to text
[Note 243] Antitrust Equal Enforcement Act: Hearings on S995 Before the Comm. on the Judiciary of the United States Senate, 97th Cong., 1st & 2d Sess. 11, 24 (1981-82) (statement of William Baxter, Assistant Attorney General). Return to text
[Note 244] See Kornhauser & Revesz, supra note 28. Kornhauser and Revesz do not focus on the conspirator, non-conspirator dichotomy, and their analysis ignores bargaining among the conspirators over the division of benefits of a conspiracy. If the parties can bargain about the division of the benefit ex ante the conspiracy, and the additional transaction costs of reaching agreement on the division of damages are low, then their criticism of the per capita liability allocation rule for joint and several liability does not apply. Their analysis is applicable to situations where harm is caused by non-conspirators or where the ex ante bargaining over division of the benefits is costly. Return to text
[Note 245] This is simply a Coasian analysis applied to the bargaining on the division of the benefit of the conspiracy. See Ronald H. Coase, The Problem of Social Cost, 3 J.L. & Econ. 1 (1960). Return to text
[Note 246] For a non-conspiracy case, the division of the benefit among the parties would require the parties to find each other ex ante the conspiracy and to possibly agree on the extent of each party's participation in the act or acts as well as the division of the benefits. This is unlikely in many joint and several liability cases. For example, in automobile accidents it would be extremely unlikely that the participants would be able to find each other and agree on their actions in an upcoming accident. Thus, using the per capita allocation rule in non-conspiracy situations would seem ill-advised because of the much higher transaction costs of the parties reaching an agreement. Of course, if the parties reached an agreement, they would have formed a conspiracy by definition. Return to text
[Note 247] For example, besides the cost of the benefit division agreement itself, parties may have monitoring costs to make sure that others are not cheating on the agreement. Return to text
[Note 248] The court in Professional Beauty Supply asserted that deterrence would be increased by allocating damages using the per capita liability allocation rule. The court explained: "[W]e believe a rule of pro rata contribution will serve as a more effective deterrent to antitrust violations (i.e., under a comparative fault rule some parties may feel they have little to lose in joining in an antitrust violation in a minor capacity)." Professional Beauty Supply, Inc. v. National Beauty Supply, Inc., 594 F.2d 1179, 1182 n.4 (8th Cir. 1979). Return to text
[Note 249] See supra Return to text
[Note 250] However, the per capita rule will not necessarily eliminate asymmetric information as a conspirator may not know how many other conspirators are in the conspiracy or the full amount of damage caused by the conspiracy. An actor may become a co-conspirator without knowing every other conspirator, the full extent of the conspiracy, or all activities undertaken pursuant to the conspiracy. Hoffman-La Roche, Inc. v. Greenberg, 447 F.2d 872, 875 (7th Cir. 1971). Return to text
[Note 251] See supra note 125 and accompanying text. Of course, in situations where deterrence is a priority, raising the total amount of damages the conspirators are liable for may also achieve the desired deterrent effect. Return to text
[Note 252] See The Antitrust Equal Enforcement Act: Hearings on S995 Before the Comm. on the Judiciary, 97th Cong., 1st and 2nd Sess. 11, 25 (1981-82) (statement of William Baxter, Assistant Attorney General). Return to text
[Note 253] There is a difference between the conspirators expected damages and the real damages caused. The conspirators expected damages could be $100 because the real damages are $100 and the conspirators expect to have to pay these damages with certainty. However, the conspirators expected damages could also be $100 because the real damages are $300 and the conspirators expect to have to pay those damages 1/3 of the time.Suppose the real damages were $300 and the conspirators expected to pay the damages 1/3 of the time. In order for the conspirators to have an expected damage payment of $300, the damages the conspirators would need to face would be $900. Return to text
[Note 254] If the firms find that per capita liability leads to a focal point of splitting the benefits, then the parties might reach an agreement of $20 as the side payment. However, it is not necessary for the benefits to be divided per capita for the conspiracy to take place under the per capita liability allocation rule. All that is necessary is that each party receive more in expected benefits than expected liability.Of course, the additional transaction costs for reaching this further agreement on the division of the benefit of the conspiracy must be less than the surplus benefit the conspiracy produces for the conspirators. In this case, the additional transaction costs must be less than the total conspiratorial benefit of $20 for the conspiracy to take place. Return to text
[Note 255] See supra note 123. Return to text
[Note 256] One might suggest that we increase the liability on the controlled conspirator because it needs a greater incentive not to enter into the conspiracy. But if we did increase the penalty on controlled conspirators, then controlled conspirators would simply claim they were not being controlled. Thus, it is doubtful that such a step would increase deterrence.By increasing the penalty on the controlling conspirator, we can get the conspiracy as a whole to internalize the damages to the plaintiffs. Return to text
[Note 257] Thus, suppose there are two conspirators and damages are ordinarily allocated per capita. If the conspiracy involves one conspirator controlling another, then the controlling conspirator will be liable for 75% of the damages (its per capita share plus 50% of the controlled conspirators per capita share) while the controlled conspirator will be liable for only 25% of the damages (50% of its per capita share). Return to text
[ Go to Previous Section, Section V | Go to Next Section, Section VII ]