The FINAL VERSION of this paper will be published in the November, 1994 Santa Clara University Law Review. .

Copyright 1994-95 by the Santa Clara 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 IV | Go to Next Section, Section VI ]

V. Extensions

A. Lifting of Model Assumptions

We have made many assumptions in order to more easily show the differences among the rules. Most of these assumptions have not been necessary to the conclusions reached, and will be lifted in this section.

1. Uncertain Detection

We have assumed thus far that the plaintiff has been able to detect the conspirators with certainty. Thus, the probability that a plaintiff would be able to obtain damages from the defendants was the same as the plaintiff's probability of success at trial. We now lift the assumption that the plaintiff's detection of the conspiracy occurs with certainty. In determining the probability that a plaintiff will be able to collect its damages at trial from a conspirator, the probability of detection can be separated from the probability of the plaintiff winning at trial after detection has occurred.[172]

In the diagram above, q is the probability of detection of the defendant by the plaintiff, and p is the probability of the plaintiff winning at trial given the plaintiff has detected the defendant. The probability of the plaintiff obtaining damages from the defendant is thus (p x q).

If q is equal to 1, detection is assumed to occur with certainty, and the differences between the no contribution rule and the contribution rules are extenuated. This is what we have assumed above in comparing the rules. However, in many, if not most, situations the probability of detection will not be 1. If the probability of detection is uncertain, the different rules are closer in their results.

We can reexamine the example in subsection [173] In that example, we had two defendants who were equally liable for the total damages of $100. Detection of the conspiracy occurred with certainty. Both defendants had beliefs that each defendant would lose at trial to the plaintiff with probability 0.5. The expected damage a defendant faced under the no contribution rule was $33.33, while under the contribution rules the expected damage was $25. The difference in expected damages of the rules, and thus deterrence, being $8.33.

Now suppose that each defendant believes that the probability of detection is 0.7071 and the probability of losing to the plaintiff at trial after detection is 0.7071. This also will lead each defendant to believe that the probability of paying damages is 0.5.[174] Under the no contribution rule, if detection occurs, the defendants will have expected damages of $41.42,[175] while if detection does not occur, the defendants will have zero in expected damages. Under the contribution rules, if detection occurs the defendants will have expected damages of $35.36,[176] while if detection does not occur, the defendants will again have $0 in expected damages.

When detection does not occur, 29.29% of the time, the defendants will pay zero regardless of which rule is used. The other 70.71% of the time, they will have expected damages of $41.42 under the no contribution rule and $35.36 under the contribution rules. Thus, the expected damages will be $29.28 under the no contribution rule,[177] and $25 under the contribution rules.[178] The difference between the expected damages under the rules is only $4.28, which is less than the difference in expected damages when detection occurs with certainty of $8.33. In both examples, the defendants have a probability of paying damages to the plaintiff of 0.5. But because the strategic bargaining under the no contribution rule only occurs after detection of the conspiracy, the differences between the no contribution rule and the contribution rules are greater when a defendant's uncertainty of paying damages is due solely to the its uncertainty of loss at trial after detection.

We can look at one more example and observe that if all of the uncertainty of the defendants' having to pay damages to the plaintiff is due to the uncertainty of detection, then the no contribution rule and the contribution rules will lead to the same results. Suppose that the probability that the conspiracy is detected is 0.5, and that if the conspiracy is detected the defendants will lose to the plaintiff. In this case, both the no contribution rule and the contribution rules will give the defendants expected damages of $25.

Under the no contribution rule, if the conspiracy is not detected, then each defendant will have damages of $0. If the conspiracy is detected, then each defendant will have expected damages of $50 (as each defendant will be just as likely to have to pay more than its share of the damages). As the probability of detection is 0.5, each defendant has expected damages of $25 under the no contribution rule.

Under the contribution rules, if the conspiracy is not detected, then each defendant will have damages of $0. If the conspiracy is detected, then each defendant will have damages of $50 (as each defendant is equally liable for the total damages of one-hundred). As the probability of detection is 0.5, each defendant has expected damages of $25 under the contribution rules. Thus, where all of the uncertainty of a defendant's paying damages is due to the uncertainty of the conspiracy being detected, the contribution and the no contribution rules lead to the same outcomes.

The results of the three cases we have seen separating the probability of detection from the probability of loss at trial are summarized below.

An additional question that should be addressed here is whether the rule that is used affects the probability of detection. One could argue that, because the plaintiff's total damages under the no contribution rule will be greater than its total damages under the contribution rules, the plaintiff is more likely to extend resources in detecting conspiracies under the no contribution rule.[179] The defendants will know that the plaintiff will have higher total damages under the no contribution rule than under the contribution rules, and that the plaintiff may expend more resources in detecting conspiracies; thus, the defendants may believe that the conspiracy is more likely to be detected under the no contribution rule, and may be more deterred.[180]

2. Independent Trial Outcomes and Sequential Settlement

The plaintiff's probabilities of winning at trial against different defendants are more likely to be correlated in conspiracy cases than in non-conspiracy cases. For example, in a two firm price-fixing conspiracy case, if the plaintiff is to win at trial against the first defendant, it will be likely to win at trial against the second defendant (since for the plaintiff to win at trial against the first defendant for fixing prices with the second defendant, the plaintiff will also be showing that the second defendant is fixing prices with the first defendant).

However, in some situations the defendants' probabilities of loss at trial may not be perfectly correlated and possibly even independent. We will briefly examine the situation where the probabilities of the defendants' losing at trial are independent.

Under the assumption of independent probabilities of loss at trial, the expected damages under the no contribution rule and contribution rules are the same. Under the no contribution and contribution rules, if the plaintiff negotiates simultaneously with defendants, then the plaintiff and the defendants will be unable to agree on individual settlements.[181] Under the no contribution rule, however, if the plaintiff negotiates sequentially with the defendants, then settlements may be reached.[182] Where the defendants' probabilities of loss at trial are independent, the plaintiff will have credibility in sequentially bargaining with the defendants, although the defendants will only be willing to settle under the no contribution rule.[183] The following example where probabilities of loss at trial are independent will demonstrate the settlement and deterrence effects of the rules.

Suppose there are two identical defendants, each defendant equally contributing to the damage of the plaintiff, the probability of each defendant losing at trial is 0.5 and the total damage caused is $100. Assume that the probabilities of each defendant losing at trial to the plaintiff are independent.

First, note that if the plaintiff sues both defendants in a joint trial, then under both the contribution rules and the no contribution rule the plaintiff has an expected damage award of $75, and each defendant has expected damages of 37.50.[184]

Under the no contribution rule, the plaintiff can achieve a total settlement of $75 through sequential negotiations and settlements with the defendants. Under the contribution rules, the defendants will not want to settle, as they would be better off going to trial.

The settlement pattern under the no contribution rule is as follows. Suppose the plaintiff decides to first negotiate with defendant A and then with defendant B. The plaintiff will first approach defendant A and threaten to sue it for the entire $100 in damages. Defendant A will settle with the plaintiff for the expected damage payment of $50.[185] The plaintiff will then approach defendant B and threaten suit for the remaining $50 in damages. Defendant B will settle with the plaintiff for the expected damage payment of $25. The plaintiff's total expected damage award will be $75. If the defendants do not know ex ante the conspiracy what the plaintiff's sequential settlement order will be, then each defendant will have expected damages of $37.50.

Suppose that instead of negotiating with the defendants sequentially, the plaintiff makes offers to settle with each defendant for $37.50. In this case, there will be no settlement unless both defendants simultaneously or conditionally settle, because each defendant will prefer to let the other defendant settle first and then be sued by the plaintiff--while the plaintiff will not be willing to settle with the first defendant unless it settles for its full share of the damages. Suppose the plaintiff first settles with defendant A for $37.50, then the plaintiff can sue defendant B for $62.50. But then the plaintiff will only have an expected award of $31.25 at trial against defendant B. Thus, the plaintiff will have an expected damage award of only $68.75, which is less than the $75 the plaintiff can expect by suing both jointly at trial.[186] Both defendants will prefer to settle second in such a situation, and the plaintiff will prefer to take them both to trial rather than settle with either individually.

Professors Kornhauser and Revesz show that where the plaintiff's probabilities of winning at trial against the defendants are independent, and the plaintiff makes a pair of settlement offers to both defendants concurrently, there will be no pair of settlements that are agreeable to both defendants and the plaintiff.[187] Thus, where negotiations are simultaneous, there will be a trial regardless of the rule applied to contribution.[188] Only where the plaintiff settles with the defendants in a group, or, under the no contribution rule, sequentially, will litigation be avoided.

The following table summarizes the outcomes of the rules depending on the circumstances. Under the no contribution rule, the plaintiff sequentially negotiates with the defendants and settles first with defendant A and then with defendant B. Under the contribution rules, the plaintiff sues both defendants at trial.

3. Uncertain Contribution

In the analysis above we assumed that, if allowed to do so under the rule, a defendant could obtain contribution from another defendant with certainty. One can think of this as contribution in its strongest form. The no contribution rule can be seen as contribution with probability zero. Moving from contribution with probability one to probability zero is thus moving from this strong contribution to no contribution. Uncertain contribution falls between these two extremes.[189]

Both deterrence and settlement will be affected by the defendants' ability to obtain contribution. Under the contribution with settlement reduction rule, expected damages will be higher when contribution cannot be obtained with certainty than when contribution can be obtained with certainty.

Settlement patterns will also be different where contribution cannot be obtained with certainty. Unlike the case where contribution is certain, if contribution is uncertain then under the contribution with settlement reduction rule, a defendant may be willing to settle with the plaintiff.[190] The defendant will settle even though it may have to pay additional damages if a later losing defendant is able to obtain contribution from it. The settlement amount will be less than the defendant's expected damages but the settlement will put an upper bound on the defendant's potential liability at its attributable share of the damages. An example will show the affects of uncertain contribution on deterrence and settlement.

Suppose there are two identical defendants, each defendant equally contributing to the damage of the plaintiff, the probability of each defendant losing at trial is 0.5, and the total damage caused is $100. Suppose that the probability of one defendant obtaining contribution from another defendant is 0.5.

First, we can examine the defendants' expected damage payments under the contribution with claim reduction rule.[191] Suppose defendant A settles for some amount s. The plaintiff will then sue defendant B for $50 in damages, an amount equal to the total damages of $100 less the $50 in damages attributable to defendant A. Thus, defendant B will have an expected damage payment of $25. Defendant A will settle for $25, as it cannot do better by settling second, and the plaintiff will not accept less than $25. Thus, under the contribution with claim reduction rule, we have the same result as when contribution occurred with certainty: both defendants will have expected damages of $25 and the plaintiff will have an expected award of $50.[192]

Under the contribution with settlement reduction rule, we reach a different result when contribution cannot be obtained with certainty than when contribution can be obtained with certainty.[193] First, it should be noted that there is now an advantage to settling. A defendant that settles removes the chance that it might pay more than its share of the damages. After settlement, the most the defendant will ever have to pay is its attributable share of the damages. A trial-losing defendant may be liable for more than its share of damages if it fails to obtain contribution from the other defendants.

Suppose defendant A is considering settling with the plaintiff. If defendant A settles for some amount s, it will have a total expected liability of its settlement s plus the expected liability from possibly having to pay contribution to defendant B. Defendant A will only have to pay defendant B contribution if defendant B loses at trial to the plaintiff and defendant B successfully sues defendant A for contribution. These are uncertain events, as the probability of each occurring is 0.5. The probability that defendant A will pay contribution is 0.25. The amount of contribution that defendant A would have to pay defendant B would be $50 - s. Therefore, defendant A's total expected liability is s + [(0.25) x ($50 - s)].

We can now calculate defendant B's expected liability if defendant A first settles with the plaintiff for s. The plaintiff will sue defendant B for the total damages less defendant A's settlement, $100 - s. However, if defendant B loses at trial against the plaintiff, defendant B may be able to obtain contribution from defendant A for defendant A's attributable share of the damages. In this case, defendant A will be liable for $50 in damages, but has already paid s in damages through its settlement with the plaintiff. Thus, defendant B can seek contribution from defendant A of $50 - s. Now this contribution is uncertain, as defendant B will only be able to obtain it 50% of the time. Thus, defendant B's expected damages from losing at trial to the plaintiff are $100 - s - [(0.5) x ($50 - s)]. That is the damages the plaintiff sues defendant B for less the expected contribution defendant B obtains from defendant A. Of course, defendant B will only lose at trial to the plaintiff 50% of the time. Thus, defendant B's total expected damages are (0.5) x {$100 - s - [(0.5) x ($50 - s)]}.

Now we need to determine if defendant A will settle first. Defendant A will only settle first if its expected damages from settling are less than or equal to its expected damages from not settling. In this case, if we set defendant A's expected damages equal to defendant B's expected damages and solve for s, we will find that s = $25.[194] Defendant A will have total expected damages of $31.25, equal to its settlement with the plaintiff for $25, and the expected contribution it must pay to defendant B. Given that defendant B will both lose at trial to the plaintiff and obtain contribution of $25 from defendant A only 25% of the time, defendant A's expected contribution payment will be $6.25.

Defendant B will also have expected damages of $31.25. The plaintiff will sue defendant B for $75 and win 50% of the time. But if defendant B loses at trial to the plaintiff, then 50% of the time defendant B will be able to obtain contribution of $25 from defendant A. Thus, defendant B will have an expected damage payment of (0.5) x {$75 - [(0.5) x ($25)]} = $31.25. It should be noted that defendant B will not settle with the plaintiff, even if they agree on the outcome of the trial, because a settlement will keep defendant B from being able to sue defendant A for contribution.[195] Thus, the contribution with settlement reduction rule will lead to at least one trial.

We have shown that, while settlement may occur for the first defendant under the contribution with settlement reduction rule, the last defendant facing trial will not settle. The first defendant settles because settlement protects it from the possibility of paying more than its attributable share of the damages. The second defendant will not settle because if it settles it will be unable to seek contribution from previous settling defendants. The expected damages of both defendants are higher under the contribution with settlement reduction rule where contribution cannot be obtained with certainty than where contribution can be obtained with certainty.[196] As in the case of the no contribution rule, under the contribution with settlement reduction rule with uncertain contribution, the plaintiff can use the defendants' preferences for settlement to obtain a damage award that is higher than if it sues both defendants at trial.

In this example, the expected damages of the defendant are higher under the contribution with settlement reduction rule than under the contribution with claim reduction rule.[197] But the example above did not analyze the incentives to reveal information. Even where contribution cannot be obtained with certainty, the contribution with settlement reduction rule will result in a disincentive for the defendants to provide the plaintiff with information to be used against other defendants, as those other defendants may later seek contribution from the settling information-giving defendant.[198] But information exchange may greatly affect the plaintiff's probabilities of prevailing at trial against the defendants, and thus increase the expected damage award and deterrence.[199] Thus, the expected damages in the example should not be seen as necessarily a complete representation of increased deterrence under the contribution with settlement reduction rule over the contribution with claim reduction rule when contribution is not obtained with certainty.

The chart below summarizes the results of this section, where contribution cannot be obtained with certainty under the contribution rules, and compares these outcomes with the results where contribution can be obtained with certainty under the contribution rules, and the results under the no contribution rule.[200]

B. Mary Carter Agreements

The term "Mary Carter Agreement" is used rather generally to apply to any agreement between a plaintiff and a settling defendant whereby the parties place limitations on the financial responsibility of the settling defendant.[201] The amount a settling defendant actually pays is sometimes on a variable scale that decreases as the plaintiff's recovery against the other defendants increases.

Under the contribution with settlement reduction rule, a settling defendant and the plaintiff may enter into a Mary Carter agreement to protect the settling defendant from having to pay contribution to later trial losing defendants. Should the settling defendant have to pay contribution to a later trial losing defendant, the plaintiff "reimburses" the settling defendant for any amount of contribution that the settling defendant is forced to pay, limiting the settling defendant's total liability to the settlement amount. A Mary Carter agreement can thus increase a defendant's incentive to settle under the contribution with settlement reduction rule.

Proponents of Mary Carter agreements argue that they encourage settlement,[202] allow defendants to freely provide information,[203] and allow for risk sharing among the settling defendants and the plaintiff.[204]

Some have argued that Mary Carter Agreements are bad in that they hurt non-settling defendants, undermine the equitable apportionment of damages and contravene legal ethics.[205] If information by a settling defendant is given to the plaintiff, then the non-settling defendant may be more likely to lose at trial. However, this is not necessarily bad, so long as the information is truthful. In fact, this is one of the advantages shared by the contribution with claim reduction rule and the no contribution rule. The Mary Carter Agreement, by limiting a settling defendant's liability, encourages information exchange under the contribution with settlement reduction rule.

Damages will be equitably apportioned where Mary Carter Agreements are allowed. Each defendant will be liable for its share of the damages, and no defendant will have to pay more than its share without being able to seek contribution from others who pay less than their shares. Of course, the equitable apportionment of liability is not the same as each defendant paying the same amount (or percentage of its share) in damages. Any rule that allows for information exchange between the defendants and the plaintiff may lead to earlier settling defendants settling for less than later settling defendants.[206] But each defendant will not face potential damages caused by others without the right to seek contribution (as is the case under the no contribution rule). The Mary Carter Agreement will simply keep the plaintiff from collecting the full share of damages from an agreeing defendant; it does not increase the damages that a non-agreeing defendant faces.[207]

The legal ethics question is somewhat more murky. Certainly if a settling defendant will receive a lower settlement amount if the non-settling defendant pays more at trial, the settling defendant will have an incentive to exaggerate or even lie. However, a non-settling defendant at trial will also have an incentive to exaggerate (or forget) or lie in order to pay less in damages. The point is that as long as a party has an interest in the outcome of the trial, the party will have an incentive to be somewhat less than truthful. Countervailing this incentive not to be truthful are penalties, such as perjury, and ethical rules. It seems odd that to think that a Mary Carter Agreement will cause a settling defendant to be more untruthful in situations where it helps the plaintiff, and thus itself, than in situations where defendants are simply representing their own interests at trial against the plaintiff.

If we want situations where defendants will have no incentive to lie at trial, we could allow only those Mary Carter Agreements that fix the settling defendants' liability at a set amount, and not allow agreements where the payment is on a sliding scale that depends on the plaintiff's recovery against other defendants. Thus, by receiving a payment from the settling defendant, the plaintiff will be reducing the amount it can obtain from a non-settling defendant by the settling defendant's share of the damages. This results because if the plaintiff wins at trial against a non-settling defendant, to the extent that the non-settling defendant is able to obtain contribution from the settling defendant, the plaintiff will have agreed not to collect the damages. This in effect allows the parties to change the contribution with settlement reduction rule into the contribution with claim reduction rule.[208] By limiting its liability to a set amount regardless of the plaintiff's suit against other defendants, the settling defendant will not have an incentive to be untruthful at trial, and can freely give the plaintiff and the court information that will allow the tribunal to make a correct decision.

C. Fairness and Equity Among Defendants

The no contribution rule has been critiqued as unjust for two reasons. First, defendants who cause little damage often must pay more than their share of the damages.[209] A defendant may face more damages than its attributable share because under the no contribution rule a defendant may be liable for all of the damages. Because damages are not apportioned under the no contribution rule, those defendants who cause little damage face the same potential liability as those who cause substantial damage. Second, defendants who fail to settle early often face excessive damages at trial.[210] A defendant that fails to settle with the plaintiff may face high damages at trial if the plaintiff settles with other defendants cheaply because the plaintiff can seek damages from non-settled defendants that are reduced only by a settling defendant's settlement, not the settling defendant's attributable share of the damages.[211] Both complaints with respect to the equity of the no contribution rule follow from the plaintiff's ability to strategically apply the rule.

There are two questions regarding inequities under the no contribution rule that need to be answered. First, when should we look for the inequities, before or after the conspiracy? Second, even if inequities exists, who should remedy the situation, the courts or the conspirators themselves?

The issue of when to determine the equity was addressed by Easterbrook, Landes and Posner.[212] They noted that we should look for inequities at the time the conspiracy is formed, not after the conclusion of the settlements or trial.[213] In focusing on ex ante equity, they stated "a no contribution rule would be unfair only if one group of defendants were, for inappropriate reasons, more likely than the other [group of defendants] to be selected as the defendant[s] called on to pay the full damages."[214] Of course, even this might not lead to ex ante unfairness if the conspirators know ex ante which "group" of defendants are more likely to be sued.[215] Prior to the conspiracy each defendant should know that it is potentially liable for all of the damages, and will take into account its beliefs about being sued by the plaintiff when entering into the conspiracy. If a conspirator believes it will be held liable for much of the damage, then it should ask for a greater share of the conspiratorial benefit. Thus, if a conspirator feels that it will be unable to settle quickly with the plaintiff and will face higher damages than other conspirators, then it should ask for a larger share of the conspiratorial benefit to cover this increased expected liability.

Easterbrook, Landes and Posner explain: "A fairness argument from the mouth of the intentional wrongdoer is unappealing because the wrongdoer can avoid his Ôpredicament' by conforming his conduct to the law's demands."[216] This seems to be a reasonable response to a conspirator complaining about inequities of the no contribution rule, since before entering into the conspiracy, each conspirator must have expected a net benefit, otherwise it would not have joined the conspiracy.

Of course, there might be situations where a conspirator joins the conspiracy because of coercion on the part of another conspirator.[217] In such cases, it may be unfair to impose additional liability on the coerced conspirator. In addition, a conspirator's belief as to damages will depend upon the information it has when it enters into the conspiracy. It could be that a conspirator enters into a conspiracy without realizing the full extent of the activities or damage that the conspiracy will cause.[218] In this case, the conspirator's limited information may have led it to a bad decision. We might want to investigate the reason for the conspirator's inadequate information. If the reason involved a misrepresentation by another conspirator, we might punish the conspirator that made the misrepresentation.[219] Indeed, if the conspirator realizes that it might be liable for all of the damages, it will have an enhanced incentive to keep track of the activities of the conspiracy.

Thus, while ex ante equity will be prevalent in most conspiratorial situations, there may be some situations where the conspirators are treated unfairly prior to the conspiracy. This unfairness does not result from the no contribution rule, but rather from the actions of the other conspirators. In situations where a controlling conspirator unfairly controls the actions of a controlled conspirator, we may wish to increase the liability of the controlling conspirator, possibly by allowing contribution with liability allocated by comparative fault.[220]

This leads us to the question of whether the courts or the conspirators should remedy the inequities of the no contribution rule. The costs for the courts to remedy the inequities might not be higher than having the conspirators determine the equity of the conspiracy. In many situations, we should be able to expect the conspirators to reach agreements regarding the division of the conspiratorial benefits among themselves.[221]

Even if we believe there is too much inequity among conspirators under the no contribution rule, we will need to determine where we place the desire for equity for conspirators among our other goals, such as deterrence, the plaintiff's recovery and the reduction of litigation costs. It would not be difficult to argue that equity among the defendants should probably come somewhere near the end of our list of desirable attributes for the rule.[22] When conspirators can make rational decisions to enter into a conspiracy, the inequity argument against the no contribution rule holds little weight. However, in situations where a controlling conspirator controls the actions of a controlled conspirator, we may wish to allow contribution in order to guard against the unfairness among conspirators.

In addition, the actual application of the no contribution rule may not be as unfair as it theoretically could be. There may already be a thumb on the scale for fairness for defendants that cause a small proportion of the damage. A defendant's damage payment under the no contribution rule theoretically does not depend upon the proportion of damages attributable to the defendant,[223] as the damage payment should only depend on the probabilities that the defendant and other defendants will lose at trial to the plaintiff. However, there are situations where plaintiffs treat defendants differently, depending on their size. The following example illustrates some of these situations.

The settlement pattern in the Corrugated Container litigation has been used to show the unfairness of the no contribution rule to later settling defendants.[224] We will use the Corrugated Container settlement data to show that the actual application of the no contribution rule may not be unfair to smaller defendants.[225] The settlement pattern is shown in the following table.[226]

The above data shows that for defendants that are similarly situated (at the same level of federal indictment) the settlement per market share point of earlier settling defendants is less than for later settling defendants. The court noted that the plaintiffs had engaged in a settlement pattern that was meant to obtain "maximum aggregate recovery" for the plaintiffs.[227] However, it is not clear that this pattern of settlement actually maximized the plaintiffs' total damage award.

We need to ask why under the no contribution rule the plaintiffs would base their settlement demands on a defendant's market share when damages are not allocated per market share and each defendant is potentially liable for all the damages. The important numbers are the actual settlement amounts (the bold numbers in the above table), as under the no contribution rule, the settlement of each defendant should be independent of the share of damages caused by the defendant. That the plaintiffs' settlement offers are related to market share is indicative of plaintiffs bringing fairness into the settlement negotiations, possibly unintentionally, for the defendants causing smaller amounts of damage (or receiving fewer benefits).

There seem to be four plausible explanations for the plaintiffs' decisions to settle with smaller defendants for less than larger defendants that were similarly situated in terms of federal indictment status. First, some of the smaller defendants were judgment proof.[228] The damages estimated in the Corrugated Container litigation ranged from $200 million to $1 billion and were subject to trebling.[229] For example, Olinkraft, a smaller defendant, did not have enough assets to cover the total damage the plaintiffs suffered. Because Olinkraft was judgment proof, it is not surprising that Olinkraft settled for less in absolute terms than the larger firms.[230]

Second, a firm's probability of loss at trial or the determination of damages at trial could be correlated with the defendant's market share.[231] Juries may be more likely to empathize with smaller defendants. Smaller defendants thus face lower expected damage payments and settlement costs. However, neither judgment proofness nor jury empathy for small defendants fully accounts for some of the settlements observed. Not every firm with a small market share was small in size. It seems unlikely that the plaintiffs would have won less in damages at trial against a large federally indicted firm such as Boise-Cascade than a smaller non-indicted firm such as Williamette Industries.[232] But, because the plaintiffs decided to base their settlement strategy on a particular market share, and though both firms settled on the same day, Williamette Industries ended up settling for over $1,000,000 more than Boise-Cascade. Thus, firm size and culpability do not seem to fully explain the plaintiffs' settlement demands.

Third, the plaintiffs may have been trying to lower their transaction costs in the settlement negotiations. By asking for a dollar amount per market share point, the plaintiffs would not have to communicate to each defendant what the other defendants were also being offered. But with the amount of damages involved in this case, it would appear that the plaintiffs could have come up with a cheaper way to save transaction costs.

Fourth, the plaintiffs simply may have thought that basing their settlement demands on market share was the correct way to maximize settlement.[233] If this were the case, then the plaintiffs did not truly understand how the no contribution rule works--as they could have gained millions in potential settlement dollars by demanding more from smaller market share defendants. Certainly, the plaintiffs could have demanded more from Boise-Cascade than they did from Williamette Industries.

Giving the plaintiffs and their attorneys the benefit of the doubt, we can assume that it was probably a combination of all these reasons that led to the market percentage-based settlement pattern. In this case, the plaintiffs were effectively being "fair" to the smaller defendants as compared with the larger defendants. If any firm seems to be in a position to complain about the unfairness of the no contribution rule, at least as it was applied in this instance, it would be those large unindicted defendants, such as Williamette Industries, who ended up paying more in damages simply because they were larger. Of course, one could argue that the plaintiffs were fair to Williamette Industries and too easy on the other defendants. The point, however, is that there may be some forces of fairness that are outside the model of the no contribution rule.

[ Go to Previous Section, Section IV | Go to Next Section, Section VI ]

Notes

This contrasts to the case where the probabilities of the defendants losing at trial to the plaintiff are not independent where the plaintiff may not have credibility in sequentially negotiating with the defendants because if the negotiations are not going in the plaintiff's favor, then it will be better off entering into simultaneous negotiations with all defendants. See supra note 57. Return to text

For this case of identical defendants, the defendants have expected damages of $37.50, even if the defendants are unable to obtain contribution from each other with certainty, as each defendant will be equally likely to pay more than its share of the damages. Return to text

Under the no contribution rule, because the plaintiff can do better through sequential settlement or group trials, the plaintiff will no longer have an incentive to settle with the defendants for the simultaneous negotiating equilibrium amount of $33.33. However, if the probabilities are somewhat but not perfectly correlated, then $33.33 will be the minimum that the plaintiff could obtain from each of the defendants under the no contribution rule.

Under the contribution rules, defendant A will not want to settle for $50 since it will not be able to seek contribution from defendant B. If defendant A loses at trial and pays the plaintiff damages of $100, then defendant A can seek contribution of $50 from defendant B. This will make defendant A's net loss at trial only $50. Even if defendant A wins at trial, there will still be a chance of defendant B later losing at trial and seeking contribution from defendant A, but defendant A's total expected damages will still only be $37.50, which is less than the proposed settlement amount of $50. Return to text

Defendant A will lose at trial to the plaintiff 50% of the time and have to pay the damages of $100. However, 50% of the time defendant A pays damages, it will be able to obtain contribution of $50 from defendant B. Thus, defendant A will have an expected damage payment from a loss at trial of $75. $75 = $100 - [(0.5) x ($50)]. Since defendant A only will lose at trial 50% of the time, its expected damages will be $37.50.

The plaintiff will be better off settling than going to trial for any settlement above $25. Thus, the settlement reached between the plaintiff and defendant A will be between $25 and $37.50. The actual settlement will depend on characteristics of the plaintiff and defendant such as risk preferences, time value of money and other possible factors.

After settling with defendant A, the plaintiff will sue defendant B for $50, equal to the total damages of $100 less defendant A's attributable share of the damages of $50. The plaintiff and defendant B will settle for the expected damages of their trial of $25. Thus, under the contribution with claim reduction rule, when the plaintiff sequentially negotiates with the defendants, the plaintiff's total damage award will be between $50 and $62.50. It is the uncertain contribution that raises defendant A's expected damages (from the $25 in the certain contribution case to $37.50 in this uncertain contribution case) in a sequential trial that drives the result. Return to text

See supra Return to text

We analyze the [California] Supreme Court decisions as creating a hierarchy of interests. First in the hierarchy is maximization of recovery to the injured party for the amount of his injury to the extent fault of others has contributed to it . . . . Second is encouragement of settlement of the injured party's claim . . . . Third is the equitable apportionment of liability among the tortfeasors.

Sears, Roebuck & Co. v. International Harvester Co., 147 Cal. Rptr. 262, 264 (Ct. App. 1978). In addition, one may want to add deterrence to this list of interests before "equitable apportionment of liability among the tortfeasors." Id. Return to text

This data has also been analyzed by others. See Contribution and Claim Reduction, supra note 2, at 15-19; Cavanagh, supra note 11, at 1288-90. Return to text

Willamette Industries negotiated a settlement for $11,286,000, or $2.7 million per percentage point of its 4.18 percent share of the defendants' market.

Willamette was investigated by the Corrugated grand jury but neither it nor any of its employees were indicted.

. . . Willamette's 1977 net working capital was $67 million and its 1977 after-tax income was $48 million.

. . . The class's chances of prevailing against Willamette [(had they not settled) would not have been] very substantial.

. . . .

Boise-Cascade negotiated a settlement for

$9,891,000, or $3.5 million per point of its 2.8 percent share of defendants' market.

Boise-Cascade was indicted for pre-1975 violations of the Sherman Act. The Government identified nineteen employees as co-conspirators, and two of them, both general managers, were also indicted at the misdemeanor level. Boise and its two employees pled nolo contendere and were sentenced . . . .

. . . .

Boise is a large company. [Boise-Cascade's] 1977 net working capital was $275 million and its 1977 after-tax income $115.6 million. It is, therefore, capable of responding in judgment for the whole range of damages.

The class's chances of prevailing against Boise [(had they not settled) would have been] relatively good.

In re Corrugated Container Antitrust Litig., 1981-1 Trade Cas. (CCH) ¦ 64,114, at 76-710 to 76-711, (S.D. Tex. 1981). Return to text