Coats v. Penrod Drilling Corp., (5th Cir. 1995)

Docket Number:92-07378


No. 92-7378 EARL WAYNE COATS, Plaintiff-Appellee, Cross-Appellant, v. PENROD DRILLING CORPORATION, ET AL., Defendants, PENROD DRILLING CORPORATION, and HYTORC, M.E., Defendants-Appellants, Cross-Appellees. Appeal from the United States District Court for the Southern District of Mississippi (August 8, 1995) Before POLITZ, Chief Judge, KING, GARWOOD, JOLLY, HIGGINBOTHAM, DAVIS, JONES, SMITH, DUHÉ, WIENER, BARKSDALE, EMILIO M. GARZA, DeMOSS, BENAVIDES, STEWART and PARKER, Circuit Judges.

KING and HIGGINBOTHAM, Circuit Judges: For more than a century, general maritime law has held joint tortfeasors jointly and severally liable for all of the plaintiff's damages suffered at their hand. Under that rule, the risk of noncollection is borne by the defendants. The plaintiff can collect his entire judgment from a single defendant, leaving to the defendants allocation of fault among themselves. We reheard this case en banc to consider the contention that we should adopt a new rule of "modified joint liability." This proposal would limit each joint tortfeasor's maximum liability to the amount for which that tortfeasor would have been liable to the plaintiff if only the negligence of that tortfeasor and the negligence of the plaintiff were compared. The new rule would, for the first time in maritime history, shift the risk of noncollection to the plaintiff. It would allocate the risk of noncollection of an admiralty judgment among the contributorily-negligent plaintiff and the defendants in proportion to their respective faults. Because replacing joint and several liability in the general maritime law with modified joint liabi lity would be neither authorized nor prudent, we affirm the judgment of the district court.

I. FACTS AND PROCEDURAL BACKGROUND The facts and procedural history of this case were set forth in the panel opinion, Coats v. Penrod Drilling Corp. , 5 F.3d 877 (5th Cir. 1993), cert. denied , 114 S. Ct. 1303, r eh'g en banc granted , 20 F.3d 614 (5th Cir. 1994), and only those portions necessary to the issues discussed herein are restated.

Maritime Industrial Services is a corporation organized under the laws of Ras Al-Khaimah, United Arab Emirates with branch offices in Dubai and Abu Dhabi. It performs repair and maintenance services for oilfield and marine vessels, and its employees are all expatriates, primarily from India, Pakistan, and the United States.

MIS uses Lee's Materials Services , Inc. in Houston, Texas to perform various services in the United States. Through Lee's, MIS advertised its job openings in the Houston Chronicle (Texas), Lafayette Advertiser (Louisiana), and Mobile Register (Alabama).

In 1987, David Shelton, manager of the Hytorc Division of MIS, travelled from the United Arab Emirates to Mississippi on vacation and to interview prospective employees for MIS. During his trip, Shelton held a meeting in Laurel, Mississippi that was attended by several young men, including the plaintiff, Earl Wayne Coats.

Shelton explained that he was soliciting employees to operate MIS equipment on certain offshore vessels. At the meeting, Shelton offered a job to Coats, and Coats accepted. Their agreement included thirty days per year of paid vacation with airfare back to Mississippi. MIS also promised to pay for Coats' return to Mississippi at the termination of his employment. The term of Coats' employ ment was indefinite. Coats obtained an updated passport as instructed by Shelton, and MIS, through Lee's, sent him a plane ticket to Dubai. Coats arrived in the United Arab Emirates and started work on December 1, 1987.

While working for MIS, Coats lived on shore and worked on various jack-up rigs owned by different customers of MIS. The majority of Coats' work consisted of operating a hydr aulically powered torque wrench used to loosen and tighten large nuts and bolts. During Coats' employment with MIS, Penrod Drilling Corporation, a Delaware corporation with its principal place of business in Dallas, Texas, contracted for MIS to perform pressure testing on Penrod's Rig 69. The pressure testing was necessary to prepare the rig for its next drilling operation. At the time, Rig 69, a jack-up drilling rig, was located in the Port of Mina Saqr in the territorial waters of the United Arab Emirates. Although it was twenty feet from shore in forty feet of water and connected to land by a gangway, it was prepared to sail and did so three days after the accident. Rig 69 flies the United States flag, and its home port is New Orleans, Louisiana. Penrod maintained a local office in the Uni ted Arab Emirates to assist in the operation of Rig 69.

MIS assigned Coats to perform the pressure testing for Penrod.

Coats was inexperienced at this task and had to ask for assistance from Penrod personnel. All safety procedures were prepared to meet stand ards of the United States. As Coats was working aboard Rig 69, Penrod's bullplug failed at a pressure less than it was rated to withstand, causing the fluid under pressure to erupt. The eruption knocked Coats down, resulting in a sever e and disabling injury to his knee. After the accident, MIS flew Coats to Hattiesburg, Mississippi for treatmen t and started paying his medical expenses. Most of these payments were made through Lee's.

Meanwhile, MIS filled Coats' job with Chris Stennett, another Mississippi resident who attended Shelton's meeting in Laurel.

On April 10, 1989, Coats sued Penrod, MIS, and Lee's 1 in the Southern District of Mississippi. The complaint asserted federal jurisdiction base d on diversity of citizenship and admiralty and alleges, inter alia , negligence on the part of Penrod and MIS, the unseaworthiness of Rig 69, and entitlement to maintenance and cure from MIS under the Jones Act. Soon thereafter, MIS terminated its payment of benefits to Coats. Coats then amended his complaint against MIS to seek compensatory and punitive damag es under the general m aritime law for wrongful termination of maintenance and cure and to allege wrongful termination of health insurance benefits under ERISA. Penrod cross-claimed against MIS for indemnity and contribution under the general maritime law. Before trial, the district court issued a number of orders in response to motions filed by the parties. The court ruled that MIS had sufficient contacts with Mississippi to justify the assertion of personal jurisdiction and that it would apply United States law, rather than the law of the United Arab Emirates, to Coats' personal injury clai ms. MIS was estimated to be doing over one million dollars a year of business in Texas at the time o f the accident.

Under American law, the court determined that Coats was not a Jones Act seaman and was not entitled to maintenance and cure (and associated damages), but the court found that Coats qualified as a Sieracki seaman with the attending right to sue under the warranty of seaworthiness. See Seas Shipping Co. v. Sieracki , 328 U.S. 85 (1946). 2 The court also declined to dismiss the case under the doctrine of forum non conveniens .

The case proceeded to trial on Coats' claim s against Penrod for negligence and unseaworthiness and against MIS for negligence, wrongful termination of maintenance and cure, and wrongful termination of benefits under ERISA. After the court directed a verdict against Coats on his claim for punitive damages based on MIS' termination of maintenance and cure, the jury returned a verdict for Coats, assessing damages of $925,000 and assigning 20% fault to Coats, 20% to Penrod, and 60% to MIS. The court reduced the award by Coats' comparativ e fault to $740,000 and entered judgment against Penrod and MIS jointly and severally. The court also awarded costs to Coats in the amoun t of $7,889.04. In addition, the court awarded Coats $26,524.82 in penalties against MIS alone for its wrongful nonpayment of benefits as required by ERISA. MIS did not contest on appeal its liability under ERISA for benefits payable to Coats under his contract of employment. All parties appealed.

In this opinion we address only the choice of law issue and Penrod's proposal for modifie d joint liability. The portions of the panel opinion addressing personal jurisdiction over MIS (Part II), see Coats , 5 F.3d at 881-85; forum non conveniens (Part IV), see id. at 889; and Coats' cross-appeal (Part VI), see id. at 890- 92, are reinstated.

II. CHOICE OF LAW A. Subject Matter Jurisdiction in Admiralty Turning to the district court's application of United States law, MIS first argues that the choice of law is between the law of the United Arab Emirate s and M ississippi law , rather than the general maritime law. This conclusion rests on the contention that the district court lacked subject matter jurisdiction in admiralty, and therefore, the only basis for federal jurisdiction is diversi ty. If so, the district court should have applied Mississippi's choice of law rules in deciding between foreign and state law. S ee Klaxon Co. v. Stentor Elec. Mfg. Co., Inc. , 313 U.S. 487, 496 (1941); E rie R.R. Co. v. Tompkins , 304 U.S. 64 (19 3 8). 3 MIS asserts that Mississippi would apply the law of the United Arab Emirates to this case. 4 MIS argues that the activity giving rise to Coats' accident does not have a sufficient connection to traditional maritime activi ty to support admiralty tort jurisdiction. S ee Jerome B. Grubart, Inc. v. Great Lakes Dredge & Dock Co. , 115 S. Ct. 1043, 1048 (1995); Sisson v. Ruby , 497 U.S. 358, 365 (1990); Foremost Ins. Co. v. Richardson , 457 U.S. 668, 674 (1982); Executive Jet Aviation, Inc. v. Cleveland , 409 U.S. 249, 268 (1972). While this circuit formerl y applied a multi-factor approach to determine whether there was a substantial relationship to traditional maritime activity, see, e.g. , Kelly v. Smith , 485 F.2d 520, 525 (5th Cir. 1973), cert. denied , 416 U.S. 969 (1974), that approach was rejected by the Supreme Court in Grubart . According to Grubart , the "connection" inquiry for admiralty tort jurisdiction involves two inquiries. A court must first "assess the general features of the type of incident involved to determine whether the incident has a potentially disruptive impact on maritime commerce." Grubart , 115 S. Ct. at 1048 (citations omitted) (internal quotations omitted). Second, a court must determine "whether the general character of the activity giving rise to the incident shows a substantial relationship to traditional maritime activity." I d. (citations omitted) (internal quotations omitted). For this second inquiry, we ask "whether a tortf easor's activity, commercial or nonc ommercial, on navigable waters is so closely related to activity trad itionally subject to admiralty law that the reasons for applying special admiralty rules would apply in the case at hand." Id. at 1051.

MIS performs repair and maintenance services for oilfield and marine vessels. Penrod is engaged in offshore oil drilling.

Penrod contracted with MIS because Rig 69 needed pressure testing before its next drilling operation. As to the first "connection" inquiry, the incident can be described in general terms as an injury to a worker while repairing and maintaining a jack-up rig in navigable waters. Without a doubt, worker injuries, particularly to those involved in repair and maintenance, can have a disruptive impact on maritime commerce by stalling or delaying the primary activity of the vessel. As to the second inquiry, the repair and maintenance of a jack-up drilling rig on navigable waters is certainly a traditional maritime activity. Moreover, we note that this tort occurred aboard a vessel on navigable waters. Providing compensation for shipboard injuries i s a traditional function of the admir alty laws. S ee Sisson , 497 U.S. at 368-75 (Scalia, J., concurring) (arguing that all vessel-related torts fall within the admiralty jurisdiction). Thus, the activity giving rise to Coats' accident has a sufficient connection to traditional maritime activity to support exercise of our admiralty tort jurisdiction.

MIS' reliance on Sohyde Drilling & Marine Co. v. Coastal Gas Producing Co. , 644 F.2d 1132 (5th Cir. 1981), is misplaced. There, we applied the Kelly factors and concluded that admiralty jurisdiction was lacking in a suit for property damage arising from the blowout of a high-pressure gas well located in a dead-end canal slip in Louisiana. Coastal, the operator of the well, had hired Sohyde to perform workover operations to corr ect a loss of production. While denying jurisdiction over the property damage at issue, the court remarked that claims for personal injury suffered on navigable waters would certainly fall within admiralty. I d. at 1136-37. Therefore, S ohyde actually supports the exercise of admiralty jurisdiction in this case, one involving only personal injury. MIS' arguments are without merit.

B. The Lauritzen - Rhoditis Factors The Lauritzen - Rhoditis factors govern the choice of law: (1) the place of the wrongful act; (2) the law of the flag; (3) the allegiance or domicile of the injured worker; (4) the allegiance of the defendant shipowner; (5) the place of the contract; (6) the inaccessibility of the foreign forum; (7) the law of the forum; and (8) the shipowner's base of operations. H ellenic Lines, Ltd. v. Rhoditis , 398 U.S. 306, 308-09 (1970); Lauritzen v. Larsen , 345 U.S. 571, 583-91 (1953). "The test is not a mechanical one in which the court simply counts the relevant contacts; instead, the significance of each factor must be considered within the particular context of the claim and the national interest that might be served by the application of United States law." F ogleman v. Aramco , 920 F.2d 278, 282 (5th Cir. 1991). "The significance of each factor in a nontraditional maritime context like offshore oil production may vary from that in the traditional shipping context in which the Lauritzen - Rhoditis test arose." Id. ; see also Bailey v. Dolphin Int'l, Inc. , 697 F.2d 1268, 1275 (5th Cir. 1983) (involving a jack-up drilling rig); C uevas v. Reading & Bates Corp. , 770 F.2d 1371 (5th Cir. 1985) (same); Jack L. Albritton, Choice of Law in a Maritime Personal Injury Setting: The Domestic Jurisprudence , 43 La. L. Rev. 879 (1983) (discussing the difference between "bluewater" and "brownwater" cases). The place of the wrongful act, the allegiance or domicile of the injured, and the place of the contract, which are less important in the shipping context, are more significant in nontraditional cases such as this one. C hiazor v. Transworld Drilling Co. , 648 F.2d 1015, 1019 (5th Cir. 1981). Our review of the district court's decision to apply United States law is de novo . S ee, e.g. , Fogleman , 920 F.2d at 282.

T he first factor is the place of the wrongful act. Coats' accident occurred in the territorial waters of the United Arab Emirates, and because this is a nontraditional maritime case, this factor is entitled to considerable weight.

The second factor is the law of the flag. "The law of the flag has traditionally been of cardinal importance in determining the law applicable to maritime cases." I d. (citing Lauritzen , 345 U.S. at 583-84). MIS is not a shipowner and therefore this factor has no specific application to it. Penrod's Rig 69 flew the United States flag. Penrod argues that the flag of the vessel in this case is fortuitous, because Coats was assigned to six different drill ing rigs with different owners and allegiances. The record indicates that in addition to the PENROD 69, Coats worked aboard the MARESK VICTORY, the TRIDENT III, the TRANSOCEAN V, the W.T. ADAMS, and the SEDCO 91. Penrod, however, does not say what flag each of these vessels fl ew and we are unable to find this information in the record. We cannot conclude that Coats' injury aboard a United States flag vessel, as opposed to a vessel registered in another country, was fortuitous without knowing what flags these other rigs flew.

The third factor is the allegiance o r domicile of the plaintiff. Coats is a United States citizen, and despite his move overseas, he maintained his residence in Mississippi, where MIS agreed to fly him for his vacations, and where he ret urned after the accident. Indeed, MIS purchased insurance to pay costs of " repatriation" in the event of an accident. Nevertheless, defendants contend that Coats' domicile was in the United Arab Emirates. They argue that he moved to that country with the intent to remain because his job with MIS was for an indefinite term and one is generally domiciled where he works. In Fogleman , however, the plaintiff was a Louisiana resident who had worked in Saudi Arabia for eight years, and we determined his domicile to be in the United States. Coats is, a fortiori , domiciled in the United States.

Fourth is the allegiance of the defendant shipowner. Penrod's allegiance is without question to the United States. Rig 69 flies the United States flag and Penrod's principal place of business is Dallas, Texas. MIS is not a shipowner, but we still take into account its organization under the laws of the United Arab Emirates. The place of the contract is the fifth factor, and another that is here entitled to weight. As the district court stated, Coats apparently executed an Arabic contract in the United Arab Emirates for the purpose of obtaining a work visa; however, the parties agreed to all of the contract terms in Mississippi. Thus, as the district court clearly found, Coats' employment contract was formed in Mississippi, and this factor favors United States law. Cf. Fogleman , 920 F.2d at 283 (noting that plaintiff signed all eight of his contracts in Saudi Arabia). The sixth factor, inaccessibility of the forum, is only relevant to forum non conveniens . Lauritzen , 345 U.S. at 589-90.

The seventh factor is the law of the forum; here, general maritime law. Fogleman , 920 F.2d at 283.

The final factor is the base of operations. In the nontraditional context, we have held that "'it is the base from which the rig is operated on a day-to-day basis rather than the base of operations of the corporate or ultima te owner of the rig which is important for choice of law purposes.'" I d. at 284 (quoting Bailey , 697 F.2d at 1275 n.22). Penrod has a local office in the United Arab Emirates to assist in the operation of Rig 69.

The record shows that this office is occupied by the rig superintendent who frequently communicates with Penrod's office in Dallas, Texas by facsimile. We addressed a similar situation in Bailey . There, the local office in Singapore "was in daily contact with the Houston office by telex or telephone, usually providing it with drilling reports." 697 F.2d at 1271 n.6. In addition, "the day-to-day decisions respecting the activities and operations of the [rig] were made by [the area manager] or [the rig manager and drilling superintendent] or by personne l on the rig." I d. We nevertheless agreed that the base of operations was not in the United States. I d. at 1274. Therefore, we are constrained to find that Penr od's base of operations for purposes of this case is in the United Arab Emirates. MIS' base of operations is also in the United Arab Emirates; it has no offices anywhere else. Despite the business it conducts through Lee's in the United States and the fact that it has a substantial number of American employees, its day-to-day operations are conducted in the United Arab Emirates.

Considering these factors and weighing them in this offshore oil drilling context, we agree with the district court's decision to apply general maritime law. Of the factors deemed more significant in this context, only the place of the wrongful act favors foreign law; the allegiance of the plaintiff and the place of contract refer us to United States law. The law of the flag and the allegiance of the defendant shipowner also point to United States law. In short, the United States has a greater interest in applying its law to this case than the United Arab Emirates. Coats was recruited in the United States, accepted the job while in this country, was supervis ed by American employees, suffered injury aboar d an American vessel, and was flown home to recover. After his return, MIS willfully terminated benefits due Coats under ERISA, resulting in liability that it never questioned on appeal.

See Albritton, Ch oice of Law , s upra (noting the unlikelihood of courts denying the benefit of American maritime law to an American citizen who is recruited to work overseas and does not give up his permanent United States residence).

Prior cases are less instructive in such a fact-specific inquiry as here. Regardless, our decision today is consistent with precedent. With one exception, our decisions involving nontraditional, "brownwater" vessels have involved a foreign plaintiff injured off the coast of a foreign country seeking the protections of American law. We have uniformly rebuffed these attempts. See, e.g. , C uevas v. Reading & Bates Corp. , 770 F.2d 1371 (5th Cir. 1985); Koke v. Phillips Petroleum Co. , 730 F.2d 211 (5th Cir. 1984); Bailey v. Dolphin Int'l, Inc. , 697 F.2d 1268 (5th Cir. 1983); Vaz Borralho v. Keydril Co. , 696 F.2d 379 (5th Cir. 1983); Chiazor v. Transworld Drilling Co. , 648 F.2d 1015 (5th Cir. 1981).

The one exception is Fogleman , where we refused to allow an American plaintiff to sue under United States law for an injury that occurred in Saudi Arabia. Fogleman, a Louisiana resident, went to work for Fluor Arabia in Saudi Arabia. He applied for the job by completing a "Foreign Employment Application" and mailing it to Saudi Arabia. Fluor Arabia is a subsidiary of Fluor Corporation, a Delaware corporation with its principal pla ce of business in California, but is only authorized to do business in Saudi Arabia. Fogleman worked under a series of eight one-year contracts, all signed in Saudi Arabia, and lived aboard a boat flying the Saudi Arabian flag. Fluor Arabia had a contract with ARAMCO, and pursuant to that cont ract, Fluor Arabia assigned Fogleman to work with ARAMCO. Fogleman sustained a sharp pain in his chest while transferring from an oil platform to a workboat that flew the Panamanian flag and later suffered a heart attack, allegedly cau sed by excessive work hours aboard ARAMCO's oil platform. Fogleman sued ARAMCO and Fluor Arabia, and we affirmed the district court's application of Saudi Arabian law to ARAMCO and Fluor Corporation. 920 F.2d at 281.

The contacts with the United States in Fogleman were not as strong as in this case. The vessels involved did not fly the United States flag, an d all of the plaintiffs' contracts were signed in the foreign country. Moreover, the allegiance of both defendants was foreign. I d. at 282-83. "[T]he only significant factor pointing to the application of United States law [was] the domicile of the plaintiff." I d. at 284. We are persuaded that the connections with the United States in this case are substantial and require a different result than Fogleman .

III. JOINT AND SEVERAL LIABILITY Penrod argues that traditional joint and several liability, under w hich even a contributorily-negligent plaintiff may recover his entire damages award from any defendant held to be partially responsible, has no place in a world where comparative negligence is the norm. Penrod points out that under the present scenario, Coats, who was found by the jury to be 20% responsible for his in juries, will be able to satisfy 100% of his judgment from the equally-negligent Penrod. 5 In Penrod's view, Coats should have to bear part of the risk that the judgment against MIS, found 60% responsib le for Coats' injuries, may be wholly or partially uncollectible. 6 Consequently, Penrod seeks to modify the di strict court's judgment by limiting Coats' ability to recover the entire judgment from either Penrod or MIS in propor tion to Coats' own contributory negligence. To accomplish this change in the judgment, Penrod advoca tes the adoption of modified joint liability in the general maritime law. According to Penrod, the modified joint liability proposal is a fairer allocation of the responsibility of each party, and is consistent with developments in the state law that have abolished or modified traditional joint and several liability. To understand why we reject the invitation to adopt modified joint liability, we must begin by understanding the chan ges that the proposal would work in the general maritime law. A. Understanding the Proposal Penrod's modified joint liability proposal adopts an approach advocated sixty years ago by Charles O. Gregory, a professor of law at the University of Chicago. S ee Charles O. Gregory, Legislative Loss Distribution in Negligence Act ions , 77-79, 142-48 (1936).

Judge Garwood in turn advocated Professor Gregory's approach, using the example of a three-car accident in which all three parties plaintiff A, defendant B, and defendant C are equally at fault: "the risk that C will not compensate plaintiff A . . . is borne by A and B in the respective ratios that the fault of each of them bears to the total fault of both." S imeon v. T. Smith & Son, Inc. , 852 F.2d 1421, 1436-48 (5th Cir. 1988) (Garwood, J., concurring in part and dissenting in part), cert. denied , 490 U.S. 1106 (1989).

The court would divide B's negligence by A's and B's combined negligence (A + B) to calculate the extent of B's maximum liability, which in this hypothesis would be 50%, or 33% / 66%.

Thus, at a maximum, A can collect from B half of the damages awarded rather than the two-thirds A would have been able to collect from B under traditional joint and several liability. B would then have a contribution claim against C for that amount of the judgment it actually pays over its 1/3 share of fault. The jury here awarded total damages of $925,000 and found the plainti ff, Coats, to be 20% responsible, Penrod to be 20% responsible, and MIS to be 60% responsible. The adoption of Penrod's modified joint liability proposal would provide Coats with a judgment that includes a joint liability component and a several liability component against each defendant. 7 Penrod's proposal would work as follows: Penrod's maximum liability would be $462,500: 20 X 925,000 20 + 20 (Penrod's negligence divided by the sum of Penrod's and Coats' negligence, multiplied by the total damages award) Similarly, MIS' maximum liability would be $693,750: 60 X 925,000 20 + 60 (MIS' negligence divided by the sum of MIS' and Coats' negligence, multiplied by the total damages award) The trial court would then subtract MIS' maximu m liability ($693,750) from the amount Coats can collect ($740,000) 8 to arrive at $46,250 for which Penrod is solely liable. This $46,250 figure is Penrod's several liability component. Similarly, when Penrod's maximum liability ($462,500) is subtracted from the total amount that Coats can collect ($740,000), MIS is solely liable for $277,500 of the judgment. This $277,500 figure is MIS' several component . Finally, Penrod's and MIS' joint liability component ($416,250) is calculated by subtracting the s um of Penrod's sole liability ($46,250) and MIS' sole liability ($277,500) from Coats' maximum overall recovery ($740,000). 9 Coats could pursue Penrod for the amount of Penrod's maximum liabilit y ($462,500), and then seek recovery from MIS for the remaining $277,500 ($740,000 - $4 62,500) that Coats can collect.

Because Penrod would have paid more ($462,500) than its 20% share of fault ($185,000), 10 Penrod would have a contribution claim against MIS for the extra $277,500 ($462,500 - $185,000) in damages that it paid over to Coats. Similarly, Coats could pursue MIS for the amount of MIS' maximum liability ($693,750), and then seek recovery from Penrod for the remaining $46,250 ($740,000 - $693,750) that Coats can collect. Because MIS would have paid more than its 60% share of fault ($555,000 ), 11 MIS would have a contribution claim against Penrod for the extra $138,750 ($693,750 - $555,000) in damages that it paid over to Coats.

The modified joint liability proposal benefits defendants.

Penrod would be liable for $740,000 under traditional joint and several liability, but only for $462,500 under modified joint liability. On the other hand, the proposal hurts plaintiffs, because full recovery of damages is harder to get under modified joint liability than under the traditional scheme. Mathematically, "[s]ince a defendant's joint liab ility would become defined by a sum which is less than the total amount of the defendants' combined liabilities, a plaintiff could recover the total amount he is owed only by enforcing the judgment against each and every defendant." Simeon , 852 F.2d at 1449 (King, J., specially concurring). It also goes without saying that the plaintiff will h ave to expend additional effort and money to collect the award from two different defendants, a circumstance which becomes more expensive with each additional co-defendant. Furthermore, in the event that one defendant is statutorily immune, insolvent, or otherwise judgmentproof, the plaintiff will receive less than his total recoverable damages as found by the trier-of-fact, even if he recovers against all remaining defendants. For example, at best, Coats would receive only 63% of his maximum recovery ($462,500 / $740,000) if MIS is insolvent or otherwise judgment-proof. Although there is no evidence of insolvency or uncollectibility in the case before us, Penrod and the proponents of modified joint liability justify this result by arguing that a partially-negligent plaint iff, such as Coats, should bear part of the risk of noncollection, rather than placing the entire burden upon the defendants. 12 B. The Case for Modified Joint Liability Penrod asserts that modified joint liability should be adopted for two basic reasons. First, Penrod argues, the traditional rule of joint and sever al liability was not intended to apply to a contributorily-negligent plaintiff. As Penrod sees it, the removal of the requirement that the plaintiff be wholly innoc ent has unfairly allowed a contributorily negligent plaintiff to recover the entire judgment from a defendant whose fault is minuscule.

Second, Penrod notes that the general maritime law has been responsive to changes in the common law and to legislative enactments. In light of the modifications to joint and several liability enacted by many states, Penrod argues that admira lty courts should change the general maritime law to respond to these developments. We disagree with both of these contentions. 1. The traditional rule of joint and several liability The traditional rule of joint and several lia bility can be traced back to eighteenth century England and the case of Hill v. Goodchild , 98 Eng. Rep. 465, 5 Buff. 2790 (K.B. 1771). The rule was derived from the principle that a cause of action was "unitary," and therefore, apportionment of damages by the jury was n ot permitted. S ee W. Page Keeton et al., Prosser and Keeton on The Law of Torts § 46, at 323 & n.5 (5th ed. 1984) [hereinafter "Prosser & Keeton"] (collecting cases). Consequently, it was impossible to impose upon the individual defendants anything less than entire liability. S ee Larry Pressler & Kevin V. Schieffer, Joint and Several Liability: A Case for Reform , 64 Denv.

U. L.

Rev. 651, 655 (1988).

Originally, joint and several liability was conf ined to situations where the joint tortfeasors acted "in concert." See Pressler & Schieffer, supra , at 660; see also Prosser & Keeton, supra , § 46, at 322-23. The rule was combined with the common-law rules of procedural joinder, which were limited in application to tortfeasors acting "in concert." Consequently, under the restricted joinder rules, defendants could not be joined, and joint liability could not be imposed, unless the defendants had in fact acted together to cause the harm. This circumstance apparently led the American courts to equate "joinder" and "joint liability." S ee Pressler & Schieffer, supra , at 660; see also id. ("At common law, the concepts of procedural joinder and joint and several liability were indistinguishable because there could be no joinder of parties unless it was alleged that they were jointly responsible for acts done in concert.").

A separate rationale of imposing "entire liability" developed alongside the concept of joint liability for those acting "in concert." Under this corollary reasoning, "a defendant might be liable for the entire loss sustained by the plaintiff, even though the defendant's act concurred or combined with that of another wrongdoer to produce the result or, as the courts have put it, that the defendant is liable for all consequences proximately caused by the defendant's wrongful act." Prosser & Keeton, supra , § 47, at 328. This notion reflected the belief that a tortfeasor should be responsible for all consequences st emming from his actions, regardless of the fortuitous circumstance that others may also have contributed to the injury.

By 1876, the common-law rule of joint an d several liability was being discussed in the admiralty setting: Nothing is more clear than the right of a plaintiff, having suffered such a loss, to sue in a com mon-law action all the wrong-doers, or any one of them, at his election; and it is equally clear, that, if he did not contribute to the disaster, he is entitled to judgment in either case for the full amount of his loss. He may proceed against all the wrong-doers jointly, or he may sue them all or any one of them separately . . . . Acts wrongfully done by the co-operation and joint agency of several persons constitute all the p arties wrongdoers, and they may be sued jointly or severally; and any one of them, said Spencer, C.J., is liable for the injury done by all . . . .

The Atlas , 93 U.S. 302, 315 (1876). I n this context , as in others, the concern that the innocent plaintiff receive full recove ry of his damages was offered as one of the primary justifications for joint and several liability. In fact, as we will explain, this consideration has apparently taken on an elevated significance in maritime law because of special concerns unique to the admiralty, especially its role as "protector" of seamen. J oint and several lia bility has the benefit of allowing a seaman to pursue and to collect his entire damages award from on e co-defendant when the generally international character of his profession might make it diffi cult or impossible to locate or to collect from other tortfeasors. Penrod argues that joint and seve ral liability was "historically one of two counterbalancing principles arising out of the legal theory of the 19th Century that all parties are responsible for all of the consequences of their negligence." According to Penrod, the "second half" of this couplet is the concept of contributory negligence, which at common law would cut off a ll recovery for the partially-responsible plaintiff. Thus, only a wholly-innocent plaintiff had the advantage of collecting his entire damages award from any of the jointly liable defendants.

As Penrod argues, when the tide of comparative negligence swept the nation, and a plaintiff's own negligence was no longer an absolute bar to recovery, the balance of the "couplet" was destroyed. Thus, because joint and several liability was born in the context of the wholly-innocent plaintiff, Penrod argues that it should be limited to that context. 13 In support of its position, Penrod points to T he Atlas , contending that it was the fi rst case to adopt joint and several liability in admiralty. S ee The Atlas , 93 U.S. at 314 ("[P]roof of entire innocence or freedom from fault . . . entitles the promoter of a suit for such a claim to full compensation for his loss from the guilty party."). While T he Atlas may have been such a foundational case, see Edmonds v. Compagnie Generale Transatlantique , 443 U.S. 256, 260 n.7 (1979) ("We s tated the common-law rule in The At las and adopted it as part of admiralty jurisprudence . . . . "), Penrod's position is simply not compelling because its statement of history is incomplete. The move to comparative negligence in maritime personal injury law occurred over a century ago with the deci sion in T he Max Morris , 137 U.S. 1, 14-15 (1890). 1 4 S ee also Socony-Vacuum Oil Co. v. Smith , 305 U.S. 424, 429 (1939); The Arizona v. Anelich , 298 U.S. 110, 122 (1936); Prosser & Keeton, supra , § 67, at 471 ("Outside of admiralty, comparative negligence did not appear in A merican jurisprudence until the early twentieth century."). Indeed, as early as 1920, the rule of comparative negligence was incorporated into the Jones Act and into the Death on the High Seas Act. S ee 46 U.S.C. § 688 (Jones Act); 46 U.S.C. § 766 (DOHSA); see also United States v. Reliable Transfer Co. , 421 U.S. 397, 408 n.13 (1975) (noting that comparative negligence is applicable under the Jones Act and under DOHSA). The move to comparative negligence, however, did not abrogate admiralty's application of joint and several liabi lity, and the two doctrines have continued to work side-byside in the maritime law. Three Supreme Court decisions strongly support this proposition, and Penrod points to no decision, Supreme Court or otherwise, to the contrary. 15 In Pope & Talbot, Inc. v. Hawn , Hawn, a carpenter, was working for Haenn, an independent contractor, aboard Pope & Talbot's ship.

See 346 U.S. 406, 407 (1953 ). After suffering injuries on the ship, Hawn brought a negligence and unseaworthiness action against Pope & Talbot under general maritime law. I d. Pope & Talbot impleaded Haenn and sought contribution or indemnity. The district court rendered judgment against Pope & Talbot for 100% of the damages less 17.5% for the proportion due to plaintiff's fault. It also awarded contribution to Pope & Talbot for Haenn's share of fault. S ee id. at 408. The court of appeals affirmed Hawn's judgment agai nst Pope & Talbot, but reversed the judgment of contribution because there was n o right of contribution in such cases. S ee id. The Supreme Court affirmed the application of joint and several liability even though Pope & Talbot had no right of contribution against Haenn and even though Hawn was at fault.

In other words, the Court treated comparative fault as consistent with traditional joint and several liability. The plaintiff was responsible for his own share of comparative fault, but the risk of noncollection fell on defendants, rather than on an injured victim.

In Edmonds v. Compagnie Generale Transatlantique , 443 U.S. 256 (1979), Edmonds, the plaintiff longshoreman, was injured on a ship not owned by his employer stevedoring company. S ee id. at 258. He received benefits from his employer under the Longshore and Harbor Workers' Compensation Act (LHWCA), and he additionally brought a negligence action against the shipowner under 33 U.S.C. § 905(b) a provision of the LHWCA added by Congress in 1972 to specifically authorize such suits. See id. The jury apportioned the $100,000 of damages as follows: 10% to the plaintiff, 20% to the shipowner, and 70% to the stevedoring company. The trial court reduced the award by the amount of the plaintiff's negligence, but the court could not permit Edmonds to recover any of the judgment from the employer stevedoring company because the employer was not a party to the suit, as the LHWCA had specifically limited its liability.

See id. at 261. Consequently, at the end of the day, the shipowner, whose conduct was determined to have caused only 20% of the harm, was held to be liable for 90% of the entire judgment even though it had no contribution rights against the statutorilyimmune stevedore.

The shipowner made two arguments to the Supreme Court: (1) that in t he process of specifically authorizing negligence suits against shipowners through the LHWCA's 1972 a mendments, Congress limited a shipowner's liability to only that proportion of the plaintiff's damages which the shipowner actually caused in this case, 20%; and (2) that even if Congress did not decree proportionate liability, the Supreme Court should, using its authority to fash ion the general maritime law, limit the shipowner's liabil ity to its proportionate share of the longshoreman's damages.

The Supreme Court began its examination of the a rguments by remarking tha t "[a]dmiralty law is judge-made law to a great extent," and by commenting that prior to 1972, a longshoreman's negligence action against a shipowne r was recognized by general maritime l aw, not by statute . S ee id. at 259-60. As the Court explained, prior to 1972, the general maritime law fashioned an injured plaintiff's recovery pursuant to the principles of joint and several liability . S ee id. at 260 & n.7. Importantly , the Court made it clear that the contributory negligence of a plaintiff had never changed the traditional rule : As [admiralty law] had evolved by 1972, a longshoreman's award in a suit against a n egligent shipowner would be reduced by that portion of the damages assignable to the longshoreman's own negligence; bu t, as a matter of maritime tort law, the shipowner would be responsible to the longshoreman in full for the remainder, even if the stevedore's negligence contributed to the injuries. This latter rule is in accord with the co mmon law, which allows an injured party to sue a tortfeasor for the full amount of damages for an indivisible injury that the tortfeasor's negligence was a substantial factor in causing, even if the concurrent n egligence of others contributed to the incident.

Id. at 259-60.

After establishing the state of the law prior to the LHWCA's 1972 amendments, the Court turned its attention to the amendments themselves and to their effect on the judicially-created doctrine of joint and several liability. An analysis of the amendments led the Court to conclude that Congress had not upset the "long- established and familiar principl[e] of maritime law by imposing a proportionate-fault rule." I d. at 263 (internal quotation omitted). The Court moved, therefore, to the remaining issue of whether it should make the vessel lia ble only for the damages in proportion to its own negligence when a longshoreman sues the vessel owner for negligence under the LHWCA. See id. at 271. To this question, the Court answered "no": [W]e are mindful that here we deal with an interface of statutory and judge-made law. . . . By now changing what we have already established that Congress understood to be the law, and did not itself wish to modify, we might knock out of kilter this delicate balance. As our cases advise, we should stay our hand in these circumstances.

Once Congress has relied upon conditions that the courts have created, we are not as free as we would otherwise be to change them. A change in the conditions would effective ly alter the statute by causing it to reach different results than Congress envisioned.

Id. at 271-73 (citations omitted) (footnote omitted). 16 Furthermore, the Court's more recent decisi on in M cDermott, Inc. v. AmClyde , 114 S. Ct. 1461 (1994), again recognized the continued application of joint and several liability in the general maritime law. At issue in M cDermott was the proper method of accounting for a settlement with certain defend ants in the calculation of the amount of a plaintiff's injuries for which nonsettling defendants could be held liable at trial. S ee id. at 1463. The Court a dopted a "proportionate share" sett lement rule that would diminish the claim of the injured party against the remaining defendants in proportion to the set tling defendant's share of fault, as found by the trier-of-fact. S ee id. at 1470-72.

The Court was persuaded that the proportionate share approach was superior to other options, in part because it was consistent with the Court's previous decision in United States v. Reliable Transfer Co. , 421 U.S. 397 (1975). In R eliable Transfer , the Supreme Court adopted a rule requiring the assessment of damages on the basis of proportionate fault, and the Court abandoned a century-old "divided damages" rule, whereby property damages were divided equally among co-defendants, primarily in collision cases, without regard to their relative degrees of fault. See id. at 410-11 .

The respondents in McDermott argued that the proporti onate share approach was inconsistent with the Court's earlier decision in Edmonds , as joint and several liability was applied in Edmonds , and no reduction in the shipowner's judgment was made for the proportio nate fault attributed to the stevedore. S ee McDermott , 114 S. Ct. at 1471. In rejecting the argument, the M cDermott Court noted that Edmonds was "primarily" a LHWCA statutory construction case, but the Court also observed that: one can read [ Edmonds ] as mer ely reaffirming the wellestablished principle of joint and several liability . As the Court pointed out, that principle was in no way abrogated by Reliable Transfer 's proportionate fault approach. . . . [T]here is no tension between joint and several liability and a proportionate share approach to settlements. J oint and several liability . . . can r esult in one defendant's paying more than its apportioned share of liability when the plaintiff's recov ery from other defendants is limited by factors beyond the plain tiff's control, such as a defendant's insolvency. . . . Unlike the rule in E dmonds , the proportionate share rule announced in this opinion only applies when there has been a settlement. In such cases, the plaintiff's recovery against the settling defendants has been limited not by outside forces, but by its own agreement to settle. There is no reason to allocate the shortfall to the other defendants, who were not parties to the settlement. 114 S. Ct. at 1471-72 (footnotes omitted) (emphasis added). It is worth repeati ng that the Court specifically recited that the principle of joint and several liability "was in no way abrogated by Reliable Transfer 's proportionate fault approach," thus treating joint and several liability and proportionate fault as compatible in admiralty. Id. at 1471.

Numerous lower court decisions also recognize that joint and several liability is the maritime rule, even when the case involves a contributorily-negligent plaintiff. S ee, e.g. , Drake Towing Co. v. Meisner Marine Constr. Co. , 765 F.2d 1060, 1063, 1067 (11th Cir. 1985) (stating, in a maritime tort case involving the Suits in Admiralty Act, that a contributorily-negligent plaintiff "may recover i ts entire damages, less that proportion attributable to its own fault, from the United States," even though the United States was only 20% at fault while a nother defendant was 60% at fault); Ge le v. Chevron Oil Co. , 574 F.2d 243, 245, 250-51 (5th Cir. 1978) (stating, in a general maritime lawsuit, t hat the plaintiff's own negligence "would not bar recovery of damages" and that the plaintiff has a "right to collect all his damages from one party in the event he is unable to obtain the relative portion of damages from each party at fault"); see also Maritime Comparative Responsibility Act as Referred to House Com mittee on Judiciary, § 2, H.R. 3318, 102d Cong., 1st Sess. (1992) comments, reprinted in 2 Benedict on Admiralty § 7, at 1-35 (7th ed. 1994) ("The existing maritime rule of joint-and-several liability of joint tortfeasors c ontinues to apply under this Act. This is true whether the claimant was contributorily negli gent or not."); 1 Thomas J. Schoenbaum, Admiralty and Maritime Law § 5-5, at 167 (2d ed. 1994) ("The adop tion of comparative fault has not affected the well established rule that there is joint and several liability in admiralty tort actions."); cf. Empire Seafoods, Inc. v. Anderson , 398 F.2d 204, 217 & n.21 (5th Cir.) (observing that with regard to the divided damages rule for mutual fault, "[t]he authorities state the rule in t erms of `innocent third parties.'. . . [W]e are convinced that the reduction of [the contributorily-negligent plaintiffs'] respective recoveries under the comparative negligence doctrine is to be considered full penalty for their fault and that they must, thereafter, be treated in the same manner as `innocent third parties.'"), cert. denied , 393 U.S. 983 (1968) . 17 In addition, Penrod's characterization of the contributory negligence bar and joint and several liability as a couplet ignores the important fact th at the principles of contribution and indemnity were also developed as a procedural means to counteract the danger that one defendant would be unduly burdened. S ee Marie R. Yeates et al., C ontribution and Indemnity in Maritime Litigation , 30 S. Tex. L. Rev. 215, 217 (1989). Contribution principles distribute a loss "by requiring each tortfeasor to pay that proportion of the damages attributable to his actions." Id. Indemnity "permits one tortfeasor to shift all of the loss onto another tortfeasor if it is determined that the latter should rightfully answer for all of the plaintiff's damages." Id. Both contribution and indemnity provide a mechanism for apportioning the plaintiff's damages among the tortfeasors themselves. S ee id. This apportionment is designed to alleviate any unfairness resulting from joint and several liability because the tortfeasor paying the entire judgment can recoup some or all of the payments from the other tortfeasors. I n short, Penrod's argument that joint and several liability is reserved only for innocent plaintiffs is plainly inconsistent with the accepted p ractice of the maritime law. 2. Signals from the states Admiralty courts have historically been responsive to commonlaw developments and to legislative enactments. S ee, e.g. , Moragne v. States Marine Lines , 398 U.S. 375, 392 (1970). In M oragne , for example, the Supreme Court crea ted a general maritime wrongful death cause of action after observing that federal and state law had changed to allow recovery for wrongful death. S ee id. at 390- 93, 401. It is also true that the Supreme Court has been wil ling to forge a general maritime position when no wholesale consensus has developed. In M cDermott , the Court noted that no uniform consensus had developed for an approach to the issue of settlement credits, but the Court went on to evaluate and to choose from the three "principal" alternatives identified by the American Law Institute.

See McDermott , 114 S. Ct. at 1465-67. Nevertheless, on more than one occasion, the Supreme Court has counseled against the adoption of a distinct ly minority view. S ee id. at 1466 n.8 ("We are unwilling to consider a rule that has yet to be applied in any jurisdiction."); Miles v. Apex Marine Corp. , 498 U.S. 19, 35 (1990) (noting the Court's discomfort with "adopting a distinctly minority view," and implying that the Court prefers a more "wholesale" and "uniform" policy judgment).

With these principles in mind, an examination of the state law and model law changes to traditional joint and several liability is striking , as the wide variety of alternatives reveals a fragm entation in approaches far greater than the Court was presented with in McDermott . To begin with, approximately twenty years ago, joint and several liability was the rule in every state.

See Pressler & Schieffer, supra , at 656 . 18 Since that time, a majority of the states have modified the concept, either by substantial limitation or by outright elimination. S ee id. at 656- 5 7. Thirteen states, however, still adhere to traditional joint and several liability.

The 1977 Uniform Comparative Fault Act adopted an a pproach that begins with a joint and several judgment, but permits a defendant to return to the court that entered the judgment to request the court to reallocate a defendant's equ itable share of the judgment after it is established to be uncollectible among the remaining tortfeasors and the contributorily-negligent plaintiff: Upon motion made not later than [one year] after judgment is entered, the court shall determine whether all or part of a pa rty's equitable share of the obligation is uncollectible from that party, and shall reallocate any uncollectible amount among the other parties, including a claimant at fault, according to their respective percentages of fault.

Uniform Comparative Fault Act § 2(d), 12 U.L.A. 50 (Supp. 1993).

The Maritime Law Association also adopted this courtordered reallocation approach in the legi slation that it proposed to Congress in 1992. 19 S ee Maritime Comparative Responsibility Act as Referred to House Committee on Judiciary, § 3(d), H.R. 3318, 102d Cong., 1st Se ss. (1992) comments, r eprinted in 2 Benedict on Admiralty § 8, at 1-46, 1-47 (7th ed. 1994). 20 Four states employ this post-judgment reallocation approach. S ee Mich .


Laws Ann. § 600.6304; Minn. Stat. Ann. § 604.02; Mo. Ann. Stat. § 537.067; see also Conn. Gen. Stat. Ann. § 52-572h(g) (allowing the reallocation of uncollectible non-economic damages among all parties and uncollectible econo mic damages among the remaining defendants).

The operation of these reallocation schemes, however, differs in crucial respects f rom the modified joint liability proposal urged upon us by Penrod. Under the Uniform Comparative Fault Act, for example, traditional joint and several liability is maintained : The common law rule of jo int-and-several liability of j oint tortfeasors continues to apply under this Act. This is true whether the claimant was contributorily negligen[t] or not. The plaintiff can recover the total amount of his judgment against any defendant who is liable.

Uniform Comparative Fault Act § 2, 12 U.L.A. 50 (Supp. 1993) (comment). The Maritime Law Association's proposed legislation uses similar language: The existing maritime rule of joint-and-several liability of joint tortfeasors continues to apply under this Act. This is true whether the claimant was contributorily negligent or not. The plaintiff can recover the total amount of his judgment against any defendant who is liable.

Mar itime Comparative Responsibility Act as Referred to House Committee on Judiciary, supra , § 7, at 1-35; see also Minn.


Ann. § 604.02(1) ("When two or more persons are jointly liable, contributions to awards shall be in proportion to the percentage of fault attributable to each, except that each is jointly and severally liable f or the whole award."); Mo. Ann.

Stat. § 537.067(3) ("This section shall not be construed to expand or restrict the doctrine of joint and several liability except for reallocation as provided in subsection 2.").

More importantly, even though these schemes reallocate an insolvent defendant's share of liability, reallocation applies only after a party's share is determined to be uncollectible. S ee Uniform Comparative Fault Act § 2(d), 12 U.L.A. 50 (Supp. 1993); Maritime Comparative Responsibility Act as Referred to House Committee on Judiciary, supra , § 8, at 1-46, 1-47; Conn. Gen. Stat.

Ann. § 52-572h(g); Mich. Comp. Laws Ann. § 600.6304; Minn. Stat.

Ann. § 604.02; Mo. Ann. Stat. § 537.067. In contrast, Penrod's modified joint liability proposal would "reallocate" as part of the initial judgment before it is determined that a defendant's share is uncollectible. Thus, Penrod's proposal builds the risk of noncollection (and the expense and delay of collection) into the judgment, disadvantaging the pl aintiff r egardless of whether a defendant's share actually proves to be uncollectible. 21 This is particularly noteworthy in the instant case, where no evidence of insolvency or uncollectibility is present in the record. The rule proposed by Penrod severs the principle of joint and several liability from its collectibility moorings in a manner that no state, uniform law, or even the Maritime Law Association, has embraced. 22 As mentioned, aside from the reallocation schemes, there are further approaches to the modification of joint and several l iability. Colorado, Idaho, and North Dakota, for example, have suspended the joint and several liabi lity principle except where the co-defendants were "a cting in concert" (or were vicariously liable). S ee Colo. Rev. Stat. Ann. § 13-21-111.5; Idaho Code § 6- 803(5); N.D. Cent.

Code § 32-03.2-02.

As a third approach, several states have preserved joint and several liability only when th e plaintiff is determined to be wholly without fault. S ee, e.g. , Ga. Code Ann. §§ 51-12-31 to -33; Wash. Rev. Code Ann. § 4.22.070; see also Boyles v. Oklahoma Natural Gas Co. , 619 P.2d 613, 616-17 (Okla. 1980). A fourth approach limits the application of the traditional rule to situations where the defendant from whom sa tisfaction is sought bears at least a minimum percentage of the responsibility. S ee, e.g. , Fla. Stat. Ann. § 768.81 (permitting the plaintiff to recover economic damages jointly and severally only from those defendants whose negligence is equal to or exceeds that of the plaintiff); Iowa Code Ann. § 668.4 (allowing joint and several liability only whe re a defendant's negligence exceeds 50% of the total responsibility); Mont. Code Ann. § 27-1-703 (same); Tex. Civ. Prac. & Rem. Code Ann. § 33.013 (measuring a defendant's negligence against total liability for some actions and agai nst the plaintiff's percentage responsibility for others, and permitting joint and several liabili ty only where a defendant's negligence exceeds the enumerated percentages). A fifth approach eliminates joint and several liability with respect to non-economic damages, but maintains joint and several liability for economic damages. S ee, e.g . , Cal. Civ.

Code § 1431.2; Fla. Stat. Ann. § 768.81; Ohio Rev. Code Ann. § 2315.19; Or. Rev.

Stat. § 18.485; cf. Ill. Rev.

Stat., ch. 735, para. 5/2- 1117 (permitting joint and several liability for medi cal and medically-related services). Due to public policy concerns, a sixth approach preserves joint and several liability with respect to certain enumerated causes of action. S ee, e.g. , Ariz. Rev. Stat. Ann. § 12-2506(D) (permitting joint and several liability when the cause of action involves hazardous wastes); Nev. Rev. Stat. § 41.141(5) (retaining joint and several liability for strict liability, intentional tort, hazardous substances, and products liability cases); N.M. Stat. Ann. § 41-3A-1 (stating that joint and several liability is available for st rict liability claims, intentional torts, and situations "having a sound basis in public policy"). A seventh approach eliminates joint and several liability altogether, instead imposing pure several liability.

See, e.g. , Alaska Stat. § 09.17.080(d) ("The court shall enter j udgment against each party liable on the basis of several liability in accordance with that party's percentage of fault."); Utah Code Ann. § 78-27-38(3) ("[N]o defendant is liable to any person seeking recovery for any amount in excess of the proportion of fault attributa ble to that defendant."); Wyo. Stat. § 1-1- 109(e).

Significantly, however, most of the states that modify joint and several liabi lity have adopted a hybrid approach by enacting statutory schemes that incorporate more than one of the abovementioned trends. S ee, e.g. , Haw. Rev. Stat. Ann. § 663-10.9 (combining limitations relating to causes of action, types of damages, and a defendant's percentage of responsibility); Ill. Rev. Stat., ch. 735, para. 5/2-1117 (same); Minn. Stat. Ann. § 604.02 (combining claims-related, percentage liability, and reallocation schemes); Wash. Rev. Code Ann. § 4.22.070 ( preserving joint and several liability for tortfeasors "acting in concert" and for wholly-innocent plaintiffs). Each scheme represents a unique blend of policy considerations weighed by the respectiv e state legislatures. 23 From this examination, we make one important observation: n o state has adopted a modified joint liability scheme that functions in the manner proposed by P enrod. It bears repeating that the Supreme Court has been willing to consider positions where no uniform consensus has developed, see McDermott , 114 S. Ct. at 1465- 67, but the Supreme Court has explicitly stated that it is "unwilling to consider a rule that has yet to be applied in any jurisdiction." S ee id. at 1466 n.8. We too are unwilling, and we may be unauthorized, to adopt the modified joint liability proposal urged upon us by Penrod. 24 C. Other Factors Informing Our Ability to Change General Maritime Law 1. Uniformity and "harmonization" As noted, the general marit ime law applies the century-old doc trine of joint and several liability. Similarly, the two principal federal maritime statutes, the LHWCA and the Jones Act, apply joint and several liability principles as well. A fter recognizing the origins of joint and several liability in the common law and in the general maritime law, the Edmonds Court approved the application of joint and several liability in the context of § 905(b) actions under the LHWCA. S ee Edmonds , 443 U.S. at 260 & n.7, 271-73.

The doctrine of joint and several liability is crystallized in the Jones Act context as well, although it stems from a different source. By incorporating the remedies afforded to railway employees under the Federal Emplo yers' Liability Act with attendant judicial glosses 2 5 Congress evidenced its intention that joint and several liability apply in Jones Act cases. S e e Simeon , 852 F.2d at 1450-51 (King, J., specially concurring); see also Cox v. Roth , 348 U.S. 207, 209 (1955) (explaining that in drafting the Jones Act to refer to the FELA, Congress effectively declared that "those contingencies against which Congress has provided to ensure recovery to ra ilroad employees should also be met in the admiralty setting"). Specifically, section 53 of the FELA provides that a plaintiff may recover the total amount of his judgment less that part representing his own contrib utory negligence. S ee 45 U.S.C. § 53. There is no exception in the Act for cases in which one or more of the defendants fails to pay its share. Congress, therefore, provided seamen the remedy of joint and several liability that was prevalent at the time that the Jones Act was adopted. S ee Simeon , 852 F.2d at 1450 (King, J., specially concurring). Indeed, Penrod concedes that Congress has statutorily declared that an injured Jones Act seaman is entitled to be made whole with the benefit of joint and several liability. As Penrod notes: We do not suggest that the proposed "modified joint liability" apply in any case where Congress has declared that a particular c lass of litigant, such as the Jones Ac t seaman or the Longshoremen's Act employee, is entitled to special consideration . . . . Statutorily, Congress has declared that injured Jones Act seam[e]n and longshoremen with claims under 33 U.S.C. § 905(b) are to be made whole. . . . [T]he policy established by Congress for dealing with injured Jones Act seamen was given controlling weight in the majority opinion in Simeon , as well it should have been . . . . (emphasis added).

Penrod's concession is realistic. Numerous cases have recognized joint and seve ral liability for Jones Act violations.

See, e.g. , Joia v. Jo-Ja Serv.

Corp. , 817 F.2d 908, 917 (1st Cir. 1987) ("The joint and several loss allocating mechanism which serves to provide an injured seaman his full judgment is consonant with the policy behind the Jones Act . . . ."), cert. denied , 484 U.S. 1008 (1988); D icola v. American Steamship Owners Mut. Protection and Indem.

Ass'n, Inc. (In re Prudential Lines, Inc.) , 170 B.R. 222, 235 (S.D.N.Y. 1994) ("[U]nder the Jones Act, . . . a tortfeasor can be held jointly and severally liable for the entirety of the damages a seaman sustains, even if [the tortfeasor's] negligence was minimal."); Johnson v. National Steel & Shipbuilding Co. , 742 F. Supp. 1062, 1065 (S.D. Cal. 1990) (noting that a defendant "can be adjudged jointly and severally liable with the cross-claimants in the cases where plaintiffs or their decedents w ere seamen pursuant to the Jones Act"); T exaco, Inc. v. Addison , 613 So.2d 1193, 1202 (Miss. 1993) (stating that plaintiff "was a Jones Act seaman at the time of his injury and [is] thus entitled to collect damages . . . from [defendants], who are jointly and severally liable .") (emphasis added). 26 Penrod's proposal would sanction a form of recovery for general maritime law that is different from the form of recovery under the LHWCA and the Jones Act. The present uniformity would be replaced by a lack of uniformity among the legislative and the judicial schemes. Such dissonance has concerned the Supreme Court in many cases. S ee, e.g. , Miles , 498 U.S. at 27, 33 (observing that applicable statutes "both direct and delimit our actions" in shaping the general maritime law, and taking action to "restore a uniform rule applicable to all actions for the wrongful death of a seaman, whether under DOHSA, the Jones Act, or general maritime law " (emphasis added)); Mobil Oil Corp. v. Higginbotham , 436 U.S. 618, 624 (197 8) (noting that "[a]s M oragne itself implied, DOHSA should be the courts' primary guide as they refine the nonstatutory death remedy, both because of the interest in uniformity and because Congress' considered judgment has great force in its own right," and allowing a coastal waters / high seas distinction in remedies to remain in the general maritime law because "a desire for uniformity cannot override th e statute" (emphasis added) (footnote omitted)); Moragne , 398 U.S. at 395, 401 (creating a maritime wrongful death action to remedy "the present nonuniformity in the effectuation of the duty to provide a seaworthy ship" that existed between federal statutory schemes and the general maritime law); see also American Dredging Co. v. Miller , 114 S. Ct. 981, 989 (1994) ("While there is an established and continuing tradition of federal common lawmaking in admiralty, that law is to be developed, insofar as possible, to harmonize with the enactments of Congress in the field ." (emphasis added)). We need not decide precisely how far the uniformity principle extends becaus e, given the Supreme Court's concern with differences between the legislat ive and judicial maritime law, it would be problematic, to say the least, to accept Penrod's proposal and thereby create a lack of uniformity among the legislative and judicial schemes in an area where uniformity currently exists.

A different concern, namely, the concern for uniformity within the general maritime law, produces a further complication.

Penrod's change in the general maritime law would directly affect not only Coats-like brown water seamen who are not covered by any federal maritime statute, but also blue water seamen the general maritime law's most common plaintiff. We are therefore compelled t o address the "special solicitude" afforded to seamen and th eir families. 27 S ee Miles , 498 U.S. at 36; Sea-Land Serv. v. Gaudet , 414 U.S. 573, 583 (1974). The law of the sea has developed principles unk nown to the common law specifically, a special solicitude for seamen, as they are considered to be the "wards of admiralty." S ee, e.g. , O'Donnell v. Great Lakes Dredge and Dock Co. , 318 U.S. 36, 40 (1943); Garrett v. Moore-McCormack Co. , 317 U.S. 239, 246-47 (1942) (quoting Harden v. Gordon , 11 F. Cas. 480, 485 (C.C.D. Me. 1823) (No. 6047) (Story, J.)). Justice Jackson eloquently described the rationale for affording special concern to seamen: From ancient times admiralty has given to seamen rights w hich the common law did not give to landsmen, beca use the conditions of sea service were different from conditions of any other service, even harbor service. . . . While his lot has been ameliorated, even under modern conditions, the seagoing laborer suffers an ent irely different discipline and risk than does the harbor worker . His fate is still tied to that of the ship. His freedom is restricted.

Pope & Talbot , 346 U.S. at 423-24 (Jackson, J., dissenting).

Traditional joint and several liability offers protection for seamen. One of the realities of a complex and international commercial maritime system is that se amen come into contact with multi-national entities who may be difficult to find, pursue, and collect from in the event of an injury. The ability to recover the judgment from any one defendant, however, as provided by joint and several liability, helps to alleviate this concern. The abolition of joint and several liability for seamen plaintiffs travels a gainst the powerful current of a special protection for seam en. Cf. Edmonds , 443 U.S. at 270 (stating that although "[s]ome inequity appears inevitable in the present statutory scheme, [] we fin d nothing to indicate and should not presume that Congress intended to place the burden of inequity on the longshoreman whom the [LHWCA] seeks to protect."). 28 Preserving joint and several liability for seamen plaintiffs, while modifying joint liability for other gen eral maritime plaintiffs, would introduce a new disuniformity within the general maritime law. Finally, it would be difficult to cabin Penrod's proposal for the general maritime law in the manner suggested by Penrod that its proposal would not apply to a LHWCA employee suing under § 905(b). When the E dmonds Court considered changing marit ime law applying joint and several liability in § 905(b) actions, the Court noted that such a change "would effectively alter the [LHWCA] by causing it to reach different results than Congress envisioned." Edmonds , 443 U.S. at 271-73. In short, Penrod's proposal would affect an "interface of statutory and judge-made law" that is, a statute (the LHWCA) whose provisions are defined in part by Congress and in part by the common law of admiralty. Id. at 272.

Given Edmonds , it is no answer that this case does not involve the i nterface between the LHWCA and the general maritime law beca use accepting Penrod's proposal will have some effect on this interface. If the developing case law under the LHWCA does not incorporate a change to modified joint liability, the LHWCA and the general maritime law will be out of step. If the LHWCA does evolve with our change, the LHWCA and the Jones Act will differ in their recovery schemes, as the Jones Act is built upon the FELA not upon the general maritime law. In either case, the proposed change would engender uncertainty and frustrate the principles of Edmonds .

See Edmonds , 443 U.S. at 273 ("By now changing what we have already established that Congress understood to be the law, and did not wish itself to modify, we might knock out of kilter this delicate balance. As our cases advise, we should stay our hand in these circumstances." (footnote omitted)). I n short, given our uniformity and harmonization concerns, we are not inclined to adopt Penrod's modified joint liability proposal for the general maritime law. 29 2. Stare decisis Uniformity and predictability are important in admiralty, and Moragne counsels that "[v]ery weighty considerations underlie the principle that courts should not lightly overrule past decisions." 398 U.S. at 403. In Moragne , the Court enun ciated three factors in the s tare decisis analysis which must be weighed prior to rejection of a longstanding rule: [1] the desirability that the law furnish a clear guide for the conduct of individuals, to enable them to plan their affairs with assurance against untoward surprise; [2] the importance of furthering fair and expeditious adjudication by eliminating the need to relitigate every relevant proposition in every case; and [3] the necessity of maintaining public faith in the judiciary as a source of impersonal and reasoned judgments. 398 U.S. at 403. With respect to the first factor, considered to be "the mainstay of stare decisis ," id. , we recognize that the need for predictability in the commercial m aritime arena is arguably greater than in other areas o f law and commerce. This is true because there are already numerous and inherently unpredictable fac tors stemming from the perils of the sea and the continual and frequently fortuitous interaction with enterprises of other nations. It is axiomatic that when the rules of law are clear, parties may contract within or around their boundaries, and the commercial system i s facilitated in many ways, including reduced litigation, more favorable insurance coverage, and overall ease of application. This factor therefore counsels against the proposed change. S ee Lewis v. Timco , 716 F.2d 1425, 1428 (5th Cir. 1983 ) ("The [maritime law's] values of uniformity, with their companion quality of predictability, a prized v alue in the extensive underwriting of marine risks, are best pres erved by declining to recog nize a new and distinct doctrine without assuring the completeness of its fit."). The second factor similarly points away from the proposed change. Our adoption of modified joint liability in this case would promote forum shopping and would add another level of complication to maritime litiga tion. Modified joint liability would apply onl y to general maritime law claims, whereas traditional joint and several liability would apply to certain statutory claims.

The final strand of the Moragne inquiry affords an opportunity for changing "a rule unjustified in r eason, which produces different results for breaches of duty in situations that cannot be dif ferentiated in policy." M oragne , 398 U.S. at 405. No such situation exists here. The traditional doctrine of joint and several liab ility, which preserves the right of the injured maritime worker to recover for his injuries, represents a conscious policy choice to shift the burden of uncollec tibility to defendants, and it has su bstantial justification in history. In addition , the traditional rule currently applies uniformly to statutory and to general maritime law claims producing the same results for the same breaches of duty often entwined in maritime litiga tion. In contrast, adopting modified joint liability for general maritime law claims, or only for g eneral maritime law claims of non-seamen, would produce different res ults than the statutory schemes for the same "tortious" conduct. Thus, the third Moragne factor also counsels against change. 3. Deferral to legislative action The wide spectrum of legislative enactments across the country demonstrates the various policy objectives attainable by altering joint and several liability. Most notable are the distinctions based upon causes of action and types of damages. The Congress is in a better position than a court to evaluate various policy objectives. We are persuaded that deferring to congressional action here is the wiser course. Even here we stand on maritime tradition, for even the earliest jurists appear to have counseled deference to the legislature: If, within its proper scope, any chang e is desired in [ the] rules [of admiralty], other than those o f procedure, it must be made by the legislative department.

It cannot be supposed . . . that the law should forever remain unalterable. Congress undoubtedly has authority under the comm ercial power, if no other, to introduce such changes as are likely to be needed.

The Lottawanna , 88 U.S. (21 Wall.) 558, 577 (1874). Congress could evaluate the desirability of modifying joint and several liability n ot only for the general maritime law, but also for t he many maritime statutes that it superin tends. Congress could, for example, limit the application of joint and several liability to situations where a defendant bears either a statutory percentage of the total fault or at least more than that of the plaintiff. These precise remedies would be more problematic for a federal court to enact because our instruments of revision are gen erally blunt.

Indeed, as mentioned, the vast majority of the states that have altered tradi tional joint and several liability have done so legislatively , while only four states have modified the traditional rule through the judicial process. S ee Joia , 817 F.2d at 917 ("[T]he decision whether to continue this trend [away from joint and several recovery] is more properly before a legislature.").

We are keenly aware of the fast-moving political forces now calling to heel excesses of the tort law. This turn of the political light upon tort law only stiffens our resolve not to attempt to run in front of the Congress. Leading a political charge is not an appropriate role for a federal court, not even for a federal court sitting in admiralty. 4. Private ordering Furthermore, contractual allocation of risk by private parties better accomplishes the goal of allocating risk. Private parties, rather than courts, are better able to assess the risks of noncollection and to decide who is in the best position to collect a judgment. Sophisticated maritime parties, intertwined in contractual relationships, can usually foresee the risk of insolvency and can allocate or insure against it. For example, Penrod and MIS were in a position to address the risk of insolvency when they wrote their contract. Coats had no such "bargaining" position. Indeed, "where potential co-defendants can contract in advance regarding their apportionment obligations among themselves, rules that leave both traditional joint and several liability and traditional apportionment rights in place might create optimal incentives and be consistent with equitable concerns." 2 American Law Institute, Enterprise Res ponsibility for Personal Injury 156 (1991). The rules now in place are clear and easily administered.

They leave the allocation of the risks of noncollection to the parties best equipped to evaluate these r isks. Absent congressional intervention, private ordering of the risks is far superior to the proposed effort. Our ability by definition is infer ior to the market's ability to tailor and allocate these risks.

IV. CONCLUSION We decline Penrod's proposal to adopt modified joint liability for the general maritime law. The judgment of the district court is AFFIRMED.

GARWOOD, Circuit Judge, dissenting, joined by Judges JOLLY, JONES, SM ITH, EMILIO M. GARZA, and DeMOSS (except that Judges JONES, SMITH, and DeMOSS do not join in the second paragraph of footnote two): This cas e involves an accident on an American-owned jack-up rig undergoing repairs in a United Arab Emirates (UAE) port in which the American plaintiff is an employee of the UAE company hired by the rig's owner to perform the repairs. The plaintiff, Coats, and the rig owner, Penrod, were each found twenty percent at fault, and the plaintiff's employer, MIS, sixty percent. The rig owner, cast in judgment for eighty percent of plaintiff's damages, contends that its liability to the plaintiff should not exceed fifty percent of his total damages, as thei r fault was equal. I agree, largely for the reasons stated in my dissent in Simeon v. T. Smith & Son, Inc ., 852 F.2d 1421, 1436-38 (5th Cir. 1988), c ert. denied , 490 U.S. 1106 (1989). The subject matter of this suit plainly is not within the scope of the Longshoremen's and Harbor Workers' Compensation Act (LHWCA), 33 U.S.C. § 901 et seq ., the Jones Act, 46 U.S.C. App. § 688, or the Death on the High Seas Act (DOHSA), 46 U.S.C. App . § 761 e t seq . 30 Accordingly, S Q assuming 31 For the rea sons stated by Judge DeMoss, in my view the choice of law issue should be resolved in favor of UAE law. While I find much to agree with in Judge DeMoss's comments regarding Seas Shipping Co. v. Sieracki , 66 S.Ct. 872 (1946), I do not believe that issue is before us. Penrod was found guilty of negligence, as well as of unseaworthiness, and while those faults were combined in an "and/or" form in the percentage of fault determination, no complaint of the form of that submission appears to have been made on this appeal. Moreover, both the negligence and the unseaworthiness related to the same condition. Finally, it is undisputed that a defendant liable for unseaworthiness is as much entitled to a reductio n of plaintiff's recovery for plaintiff's negligenc e as is a defendant liable only for neglige nce. S ee, e.g., Fontenot v. Teledyne Movible Offshore, Inc ., 714 F.2d 17, 19-20 (5th Cir. 1983); Sc ot v. Fluor Ocean Services, Inc ., 501 F.2d 983, 984 (5th Cir. 1974). 32 This opinion addresses only instances of indivisible damages, where the fault of each of the parties is a proximate cause of all the damages.

United States law applies at all 31 SQ the governing law is the general maritime law of the United States.

Overview of the Issue The a rchetypal general maritime law tort litigation is the collision case. Assume that a shrimper, under the command of its master-owner, and an Exxon crew boat collide while each is trying to avoid a suddenly appe aring small pleasure craft. Only the crew boat suffers significant damage. 32 Each vessel is under separate ownership and acting independently of the others, and none is so related to either of the others as to be vicariously liable for the other's fault. If only the shrimper and crew boat are causatively at fault, and their fault is equal, the shrimper is liable for half the crew boat's damages. Should the amount for which the shrimper is ultimately liable to the crew boat change if the small pleasure craft is also found to have been causatively at fault ( e.g., lacking adequate lights) equally with the other two vessels? S imple logic tells us it should not. The crew boat's damages remain the very same, the fault of the shrimper and of the crew boat are each still a proximate cause of all such damage, and the causative fault of the crew boat and shrimper remain equal to each other. As the pleasure craft has acted independently of the crew boat and shrimper, neither of which is vicariously liable for the pleasure craft's wrongs, there is no basis on which to charge the pleasure craft's fault to either the shrimper or the crew boat, and hence, for purposes of the shrimper's ultimate responsibility to the damaged crew boat, the relevant comparative fault is that as between those two. However, this straightforward approach is opposed at two opposite extremes. At one extreme is that approach generally known as pure several liability, in which all of the pleasure craft's fault is charged to the crew boat, so that SQ despite the crew boat's damages not having decreased, its share of the total fault having decreased, and the shrimper remaining equally at fault with it SQ the shrimper is nevertheless liable for only a third SQ not a half SQ of the crew boat's damages. At the other extreme is the majority's approach here SQ pure joint liability r eflexively, and rather oxymoronically, app lied in a case where recovery depends on comparative fault SQ which charges all of the pleasure craft's fault to the shrimper, so that the shrimper's total exposure to the crew boat goes up from one-half to two-thirds of the crew boat's damages, even though the damages remain identi cal, the relative fault as between the crew boat and shrimper is unchanged , and the shrimper's percentage of the total negligence has gone down .

The majority's approach leads to the absurdity that in certain situations a slightly negligent defendant could nev ertheless be liable for ninety percent of the damages of a plaintiff whose negligence proximately causing all those damages was ten times as great as the negligence of that particular defendant. 33 That is the very same kin d of absurdity that caused most common law jurisdictions to abandon the doctrine that the plaintiff's contribu t ory negligence barred all recovery, even though such negligence was minimal and far less than that of the defendant.

Of course, until the relatively recent arrival of pure several liability in many jurisdictions, a negligent defendant was traditionally liable to a non-negligent plaintiff for all the latter's damages proximately caused by that defendant's negligence, notwithstanding that an independent third party's fault may also have proximately caused all those same damages. See P rosser and Keeton on Torts (West 5th ed. 1984) § 47. 3 4 This rule, however, does not afford a principled just ification for rejecting the approach espoused in this dissent. The result in the justmentioned instance comes about only because none of the plaintiff's damages are proximately caused by the plaintiff's fault and all are proximately caused by the defendant's. The logic of such a regime, however, dictates that the plaintiff whose contributory negligence proximately causes all his damages may recover nothing, even though a defendant's much greater fault may also have been a proximate cause of all plaintiff's damages. And that indeed was the almost universal common law rule. Relatively recently, dissatisfaction with this result led most common law jurisdictions to abandon the contr ibutory negligence bar in favor of some form of comparative negligence. Under such an approach, there arises the question of what the plaintiff' s negligence is compared to and how the comparison is to be made. The question, of course, does not arise if the plaintiff is not negligent (nor, obviously, does it arise if no defendant is at fault). If the plaintiff and one defendant are the only ones guilty of causative fault, the obvious and universally accepted answer is that the fault of each is compared to that of the other. A question arises if, but o nly if, both the plaintiff and a defendant are guilty of causative fault and so also is at least one other independent actor (whether or not that actor is likewise a defendant). In that relatively rare setting, apportioning all of the other (or third) independent actor's fault to the defendant may not be logically justified by the principle that every party is responsible for all the proximate results of his own fault, even though such results are also contributed to by the fault of another, because that principle equally well justifies apportioning all the other actor's fault to the contr ibutory negligent plaintiff, and also becau se, in any event, where comparative fault is applied the above-refe renced principle has been abandoned both by allowing the contributorily ne gligent plaintiff to recover at all and by limiting his recovery to less than the full amount of the loss he suffered. 35 As University of Chicago law professor Charles Gregory explained nearly sixty years ago: "At common law joint tortfeasors are virtually guarantors of each other's solvency so far as concerns the injured plaintiff's joint judgment for damages; and the introduction of contribution between joint tortfeasors does not affect that situation in the slightest degree.

The plaintiff receives his damages at all costs, leaving the defendants to even up the loss between themselves if and as they may and can. But under a comparative negligence statute, where the plaintiff, although negligent, may still recover, the situation is fundamentally different. Here absolutely no reason exists why the defend ants, even if they are treated as joint tortfeasors and thus subjected to joint judgment liability for certain purposes, should be made to assume the entire risk of each other's insolvency with respect to plaintiff's rec overable damages. For when the plaintiff and the solvent tortfeasor are both negligent, they share the stigma which at common law seems to have furnished the justification for the somewhat arbitrary allocation of this risk on joint judgment debtors.

Furthermore, it is quite possible to have a plaintiff who is as negligent as, or more negligent than, either of his defendants, but is still entitled to recover. Under such c ircumstances, it seems idle to suppose that a joint liability to the plaint iff should carry absolutely the same incidents as the common-law joint judgment; and distribution of the risk of insolvency of one of the joint defendants in accordance with the apportionment of fault would seem to be the only method of administration consistent with the terms of the comparative negligence statute." C. Gregory, Legislative Loss Distribution in Negligence Actions 142 (1936) (footnote omitted).

It is true that, by virtue of the modern availability of contribution, which was generally not available at common law, see Prosser and Keeton on Torts (West 5th ed. 1984) § 50, a particular defendant to whom all an independent third actor's fault is charged, rather than shared proportionate ly with the negligent plaintiff, suffers an ultimate economic injustice o nly if he i s unable to realize adequate recovery of contribution from the third actor. Such a recovery would be unavailable if that third actor were insolvent, enjoyed some legal immunity, or could not be found.

Thus, in one sense, as observed in the above quotation from Gregory, the issue is how the risk of the third actor's insolvency, immunity, or lack of amenability to process should be borne: should it be borne entirely by the defendant, as the majority would have it; or should it be borne entirely by the negligent plaintiff, as under pure severa l liability; or should it be borne by the negligent plaintiff and defendant in the ratio that their respective degrees of fault bear to each other, as espoused here and by Professor Gregory. But in another sense, the question is why should the defendant e ver be liable to the plaintiff for a greater proportion of plaintiff's damages than the de fendant's negligence (including any for which he is vicariously liable) is of the total negligence of the plaintiff and tha t defendant (again, including any for which he is vicariously liable).

Development in Common Law Jurisdictions As previously observed, at common law what we now call "joint liability" was predicated either on concerted action (or vicarious liability) or on the principle that the defendant is liable for all consequences proximately caused by his own wrongful acts, even though the wrongful conduct of one other than the plaintiff was also a proximate cause of the harm (see notes 5 & 6, supra ). There was no occasion to consider the application of these rules to instances where the plaintiff's negligence was a proximate cause of his damages, because such a plaintiff was barred from any recovery. This was the virtually uniform rule in the United States until 1908, when the Federal Employees Liability Act (FELA), 45 U.S.C. § 51 et seq ., was enacted. Act of April 22, 1908, c. 149, 35 Stat. 65-66. The FELA provided interstate railroad employees a cause of action against their employer for injuries in the course of employment caused by the railroad's negli gence and provided that "the fact that the employee may have been guilty of contributory negligence should not bar a recovery, but the damages shall be diminished by the jury in proportion to the amount of negligence attributable to such employee." 45 U.S.C. § 53. Thereafter, in 1910 Mississippi enacted a "pure" comparative negligence statute.

V. Schwartz, Comparative Negligence (Mi chie 3rd ed. 1994), § 1- 4(b)(2). In 1913 Georgia, through a combination of judicial decision and much earlier legislation applicable to those injured in railroad operations, adopted "a rule that the plaintiff in all cases may recover an apportioned part of his damages if the defendant's negligence is greater than the plaintiff's." I d ., § 1- 5(a)(2) at 19 (citing Elk Cotton Mills v. Grant , 79 S.E. 836 (Ga. 1913)). Also in 1913, Nebraska by legislation allowed diminished recovery where the plaintiff's negligence was slight in comparison to the defendant's. I d ., § 1-4(b)(4) at 15. So matters stood until 1920, wh en Congress enacted the Jones Act, providing "any seaman" injured "in the course of his employment" an action against his employer in which the FELA would apply. 36 The same year, DOHSA was enacted, providing a cause of a ction for death wrongfully caused "on the high seas," 46 U.S.C. § 761, in which the decedent's negligence did "not bar recovery" but "the court shall take into consideration the degree of negligence attributable to the decedent and reduce the recovery accordingly." 46 U.S.C. § 766. No other jurisdiction adopted comparative negligence until in 1931 Wisconsin passed legislation allowing a plaintiff recovery if his negligence was "not as great at that of the defendant." Schwartz, s upra , § 1- 4(b)(3).

Thus, in 1909 all states generally applied the complete bar of contributory negligence; by 1930 only three states had lifted the bar to any extent; by 1940 only four had; by 1954 only five ; 37 and, as late as 1968, only seven states and Puerto Rico had any form of comparative negligence. 38 Then the rush to some form of comparative fault began, so that now in only Alabama, Maryland, North Carolina, Virginia, and the District of Columbia does plaintiff's negligence, no matter how slight, bar any recovery whatever. Schwartz , supra , § 1-1 at 2.4, § 1-5(e)(3); McIntyre v. Balentine , 833 S.W.2d 52, 55 (Tenn. 1992). This was accomplished by judicial decision in twelve states, and by legislation in thirty-four states. M cIntyre at 55, 56 & ns. 3 & 4. The results are summarized in Kionka, Recent Developments in the Law of Joint and Several Liability and the Impact of Plaintiff's Employers' Fault , 54 La. L.

Rev. 1619 (1994): "Four states still do not have comparative fault . . ., and they retain joint and several liability. Of the forty-six states that have some form of compa r ative fault, ten states still have the pure form of joint and several liability, and twelve sta tes now have pure several liability. The remaining twenty-four states, . . . have some mixture of joint and several and several liability. These statutory schemes can be quite complex.

The common thread, however, is that they all represent a compromise position between the two extremes SQ pure joint and sev eral liability on the one hand and pure several liability on the other." I d . at 1621 (footnotes omitted).

In other words, eighty percent of the states reject the rule espoused by the majority, "pure" joint and several liability in a system of comparative fault. The ten states that follow that approach are outnumbered by the twelve states at the other extreme, which follow "pure" several liability. The remaining twenty-four state s have, indeed, adopted a variety of approaches, but, as Kionka observes, a "com mon thread" runs through them, as "all represent a compromise position between the two extremes." Id. One such position between the two extremes is that of the Uniform Comparative Fault Act (UCFA) approved by the National Conference of Commissioners on Uniform State Laws in 1977. 12 U.L.A. at 42-60 (West Supp. 1994). 39 Section 1 of the UCFA provides that "fault chargeable to the claimant diminishes proportionately the amount awarded as compensatory damages . . . but does not bar recovery," and section 6 provides that the claimant's recovery is also reduced by the percentage of fault of any party with whom the claimant has settled. Sections 3, 4, and 5 deal with set-off and contribution. Section 2 is the operative section. 40 Under it the respective percentages of fault of all concerned are dete rmined, the judgment sets forth the corresponding "equitabl e share" of each, and the plaintiff is awarded the amount of his total damages, reduced by his and any settling party's percentages of fault, "against each party liable on the basis of rules of joint-andseveral liability." H owever , under section 2(d) if any party' s "equitable share" of th e judgment is (wholly or partly) "uncollectible from that party," the court "shall rea llocate any uncollectible amount among the other parties, including a claimant at fault , according to their respective percentages of fault" (emphasis added). The reason for this provision is set forth in the official comments to section 2 as follows: "Reallocation of the equitable share of the obligation of a party takes place when his share is uncollectible.

Reallocation takes place among all parties at fault.

This includes a claimant who is contributorily at fault.

It avoids the unfairness both of the common law rule of joint-and-several liability, which would cast the total risk of uncollectibility upon the solvent defendants, and of a rule abolishing joint-and-several liability, which wou ld cast the total risk of uncollectibility upon the claimant. " (Emphasis added). 41 That is precisely the rationale and effect of the position here espouse d. The official comments likewise illustrate the application of the reallocation rule by an example in which the pla intiff, whose total damages are $10,000, is forty percent negligent and two defendants are each thirty percent negligent. If one defendant is insolvent, the plaintiff's recovery from the other is ultimately $4,286, which is 3/7ths of the $10,000, the exact result here advocated. 42 This dissent, however, would, in the example given, initially limit each of the defendants' liability to $4,286. For the reasons stated below, this essentially procedural modification to the UCFA approach is practically fair and analogous to traditional admiralty practice. 43 The UCFA was judicially adop ted by the Supreme Court of Missouri when it eliminated the common law contributory negligence bar. G ustafson v. Benda , 661 S.W.2d 11, 15-16 (Mo. 1983) (en banc). 44 Later, this was legislatively ratified, and similar legislation has also been adopted in Arizona, Minnesota, Montana, Connecticut, and New Hampshire. Schwartz, s upra , § 3-5(c)(5), citing Mo. Rev. Stat. § 537.067(2); Ariz. Rev. Stat. § 12-2508; Minn.

Stat. § 604.02(2); Mont.

Code Ann. § 27-1-703(3); N.H. Rev.


Ann. § 507:7-e (III). 45 Similarly, the Restatement (Second) of Torts § 886A, comment i (1977), states in pertinent part: "In determining equitable shares of the obligation, it seems wise, particularly in comparative-negligence sta tes, to confine the determination to parties to the action rather than to attempt to calculate the equitable shares for alleged tortfeasors who are not parties and not bound by the decisions. I f o ne tortfeasor's equitable share turns out to be uncollectible it should be spread proportionately among t he other parties at fault ." (Emphasis added).

Essentially the present approach was adopted in Haney Electric Co. v. Hurst , 624 S.W.2d 602 (Tex. Civ. App. SQ D allas 1981, writ dismissed as moot). That case involved a three-car collision. In separate actions, two of the drivers SQ each later found to be thirty percent negligent SQ sued the third driver SQ later found to be forty percent negligent. The cases were consolidated, and one question was whether the defendant (the third driver, found forty percent at fault) should be liable to a particular plaintiff for forty percent of the harm (that is, only the defendant's share) or seventy percent (that is, the defendant's share added to the entire thirty pe rcent share of the other plaintiff/tortfeasor, neither of the plaintiffs being also a defendant). The court adopted neither approach and chose inst ead to hold the defendant third driver liable to each plaintiff for 40/70ths of total damages. That fraction represented the ratio of the defendant's negligence (forty percent) to the total of his negligence and the negligence of the party seeking recovery (thirty percent). The court thus placed on the defendant a portion of the unsued tortfeasor's share of fault, but only that portion represented by the ratio of the defendant's fault (forty percent) to the combined fault of the defendant and plainti ff (seventy percent). The court considered this result mandated not only by the Texas comparative negligence scheme, Tex. Rev. Civ. Stat. art. 2212a (codified as amended at Tex. Civ. Proc. & Rem .

Code § 33.001), but also by "[e]lementary fairness." 624 S.W. 2d at 612.

An analogous approach has been taken under the Louisiana statute, LSA-C.C. art. 2324, in respect to employer fault in an employee's suit against a third party. Thus, in P restenbach v. Rains , 4 F.3d 358 (5th Cir. 1993), the Louisiana employer, immune by virtue of the worker's compensation law, was found seventy-five percent at fault, the plaintiff-employee fifteen percent, and the defendant-third party ten percent. The plaintiff appealed the judgment which awarded him only ten percent of his damages against the third party. We applied the "'ratio approach'" of G uidry v. Frank Guidry Oil Co ., 579 So.2d 947 (La. 1991), as carried forward by Gauthier v. O'Brien , 618 So.2d 825, 832-33 (La. 1993), and held that plaintiff was entitled to recover forty percent of his damages from the defendant because the defendant's percentage of negligence (ten percent) was forty percent of the combined negligence (twentyfive percent) of the plaintiff (fifteen percent) and the defendant (ten percent). P restenbach at 360-61. Similarly, in D avis v. Commercial Union Ins. Co. , 892 F.2d 378 (5th Cir. 1990), the plaintiff-employee was fo und sixty percent at fault, his immune Louisiana employer thirty percent, and the de fendant third party ten percent. The plaintiff appealed the judgment, which awarded him only ten percent of his damages against the third party. We held that the employer's fault should be allocated between the plaintiff and the defendant "in proportion to their previously determined degrees of fault," with the result that plaintiff was to be granted judgment against the defendant for 1/7th (14.29%) of his total damages. Id . at 384-385. 46 A thorough review and analysis of the relevant decis ions, legislation, and scholarly writing is contained in the American Law Institute's Restatement of the Law (Third) Torts: Apportionment of Liability, Pr eliminary Draft No. 1 (May 31, 1995) (Reporter, Pr ofessor William C. Powers, Jr., University of Texas School of Law; Associate Reporter, Professor Michael D. Green, University of Iowa College of Law) (hereafter " Apportionment of Liability "). The recommendations made there include provisions in substance the same as those of the UCFA. 47 The reporters' notes to section 25A (see note 18, supra ) explain the rationale for these recommendations as follows: "The critical question is who should bear the risk of insolvent parties. The advent of comparative fault, at least when some fault is attributed to the plaintiff, removes the traditional justification for imposing that risk on defendants. See Pearson, A pportionment of Losses Under Comparative Fault LawsAn Analysis of the Alternatives , 40 La. L.

Rev. 343, 362 (1980) ('When the plaintiff himself has been negligent, the logical support for joint and several liability evaporates.').

Nevertheless, even with the plaintiff sharing some fault, each defendant is still the legal cause of all of plaintiff's da mages. Shifting the entire risk of insolvency to plaintiff 'merely transform[s] the inequity of imposing that risk entirely on solvent defendants into the equal and oppos ite inequity of imposing the risk entirely on the plaintiff.' II A M ERICAN L AW I NSTITUTE R EPORTERS ' S TUDY , E NTERPRISE L IABILITY FOR P ERSONAL I NJURY 147 (1991); see also Wade, Should Joint and Several Liability of Multiple Tortfeasors be Abolished? , 10 A M .J. T RIAL A DV . 193, 197 (1986).

Professor Charles O. Gregory made this point quite eloquently many years ago: . . .[W]hen the plaintiff and the solvent tortfeasor are both negligent, they share the stigma which at common law seems to have furnished the justification for the somewhat arbitrary allocation of this risk on joint judg ment debtors. . . . [D]istribution of the risk of insolve ncy of one of the joint defendants in accordance with the apportionment of fault would seem to be the only method of administration consistent with the terms of the comparative negligence statute.


Numerous comment ators have advocated reallocating the share of an insolvent or immune party to the remaining responsible parties in proportion to their responsibility for plaintiff's injuries. See . . .

Wade, S hould Joint and Several Liability of Multiple Tortfeasors be Ab olished? , 10 A M . J. T RIAL A DV . 193, 198 (1986); U NIFORM C OMPARATIVE F AULT A CT § 2(d) (1977); II A MERICAN L AW I NSTITUTE R EPORTERS ' S TUDY , E NTERPRISE L IABILITY FOR P ERSONAL I NJURY 127-57 (1991) (advocating reallocation of insolvent party's share when defendants are independent tortfeasors without a prior relationship); Zavos, Comparative Fault a n d the Insolvent Defendant: A Critique and Amplification of American Motorcycle Ass'n v. Superior Court, 14 L O Y . L.A. L. R EV . 775 (1980-81); W ILLIAMS , J O INT T ORTS AND C ONTRIBUTORY N EGLIGENCE § 110, at 414 -20 (1951); Sobelsohn, Comparing Fault , 60 I ND . L.J. 413, 456 (1985); Miller, Extending the Fairness Pr inciple of L i and American Motorcycle: A doption of the Uniform Comparative Fault Act , 14 P AC . L.J. 835, 86 1-63 (1983); Boyette, Note, Reconciling Comparative Negligence, Contribution, and Joint and Several Liability , 34 W ASH . & L EE L. R EV . 1159, 1174-76 (1977); see also Steenson, Recent Legislative Responses to the Rule of Joint and Several Liability , 23 T ORT & I NS . L.J. 482 (1988) (describing the variety of reallocation schemes that exist in a number of states).

The reallocation provision in § 25A also co mports with provisions in a number of states that have abolished joint and several liability for independent tortfeasors, except where the plaintiff is attributed no responsibility for the injury. . . . Of course, this exception reflects the common law rule before the adoption of comparative fault, which made independent tortfeasors jointly and severally liable for a plaintiff's indivisible injury. Section 25A results in the same outcome in tho se instances in which the plaintiff is found free of responsibility." I d . at 248- 251.

In sum, pure joint and several liability was an incidental and logical application of a regime in which the plaintiff's causative fault, no matter how slight in comparison to that of a defendant, barred any recovery whatever. Until the late 1960s, that was the almost universal rule in common law jurisdictions. Since then, the vast majority of jurisdictions that have abandoned the common law ban on any recovery f or a plaintiff whose negligence is to any extent a cause of the accident in question have likewise abandoned across-the-board pure joint and several liability. Where the plaintiff and a defendant are both guilty of causative fault, and so also is a third actor, there is no justification for allocating, as between that plain tiff and defendant, ultimate responsibility for the fault of the third a ctor on any basis other than on the ratios which the fault of the plaintiff and the defendant respectively bear to the total fault of them both.

Under the UCFA and Apportionment of Liability , this allocation will frequently SQ though by no means always SQ not be made until after judgment. For that reason, it has been subject to the justifiable criticism tha t it may be somewhat unwieldy, administratively burdensome, and may tend to undermine the finality of judgments. 48 However, these criticisms are not applicable to the position taken by this opinion, which is that the allocation will always be made in the judgment, and that there is no reason to make any defendant the plaintiff's collection agent for any portion of the damages for which plaintiff bears the ultimate responsibility. We turn now briefly to the mechanics of allocation in the judgment.

Mechanics of Judgment Damages Allocation Let us revert to our collision invo lving Exxon's crew boat, the shrimper, and the small pleasure craft. Exxon sues one or both of the other two vessels for the damages to its crew boat. There is no problem if Exxon is not at fault, for then the fault of any defendant (no matter how many are at fault) will necessarily be one hundred percent of the combined fault of Exxon and that defendant, so that defendant is liable for one hundred percent of Exxon's damages. Likewise, there is no problem if Exxon is at fault, but of the other two vessels only the shrimper is found at fault, the pleasure craft either being found not at fault or no finding being made as to its fault (as might often be the case if it were not a party). In that situation, all agree that Exxon recovers from the shrimper the same fraction of its total damages as its fault is of the total fault of itself and the shrimper. A problem arises o nly if Exxon, the shrimper, and the pleasure craft are all three found to be guilty of causative fault. Assume each is assessed one-third of the fault and that Exxon's total damag es are $100,000. The judgment should award Exxon a total recovery of $66,666.67 (2/3rds of $100,000), 49 with provision that no more than $50,000 (1/3 ÷ 2/3 x $100,000) thereof may be collected from the shrimper and no more than $50,000 (1/3 ÷ 2/3 x $100,000) thereof may be collected from the pleasure craft. 50 If the causative fault percentages are changed somewhat, to correspond to those in this case, so that Exxon's percentage of fault is 20%, the shrimper's is 20%, and the pleasure craft's is 60%, then the judgment should award Exxon a total recovery of $80,000 (80% of $100,000), with provision that no more than $50,000 (20/40 x $100,000) thereof may be collected from the shrimper and no more than $75,000 (60/80 x $100,000) thereof from the pleasure craft. 51 In other words, in such a situation the total judgment is for the amount which equals the same fraction of plaintiff's total damages as the total fault of all except the plaintiff is of the total fault of all including the plaintiff; but the judgment will provide that the plaintiff may not recover more of said sum from any particular defendant than the amount which equals the same fraction of plainti ff's total damages as that particular defe ndant's fault is of the total fault of both plaintiff and that particular defendant. For example, if plaintiff suffers total damages of $100,000 and is 10% at fault, defendant A is 40%, defendant B is 30%, and defendant C is 20% at fault, then plaintiff's total judgment is for $90,000, but provides that no more than $80,000 (40/50 x $100,000) thereof may be collected from defendant A, no more than $75,000 (30/40 x $100,000) thereof may be collected from defendant B, and no more than $66,666.67 (20/30 x $100,000) thereof may be collected from defendant C. 52 This simple system will work in all cases and serve to authorize appropriate recovery, while at the same time limiting any particular defendant's ultimate potential liability to an amount no greater than the fraction of plaintiff's damages which is that defendant's percentage of fault divided by the total of the percentages of fault of the plaintiff and that defendant.

If it is desired that the expression of this re sult in the judgment be in terms of some several liability and some joint and several liability, then that, too, can be accomplished, although in some cases an algebraic formula must be employed. A case such as this, with only the plaintiff and two defendants at fault, will be by far the most frequent instance in which any allocation question arises, and in such an instance a fairly simple set of steps may also be utilized to arrive at the appropriate several and joint and several liability figures to be set forth in the judgment. Assume plaintiff's total damages are $100,000, and , as here, causative fault is distributed 20% to the plaintiff, 60% to defendant A, and 20% to defendant B. First plaintiff's maximum recovery is calculated at $80,000 (80% of $100,000); then the maximum liability of defendant A is calculated at $75,000 (60/80 x $100,000) and the maximum liability of defendant B is calculated at $50,000 (20/40 x $100,000), all as above explained. Next, the amount of A's maximum liability ($75,000) is subtracted from plaintiff's maximum recovery ($80,000), the result being $5,000 ($80,000 - $75,000 = $5,000), which is the several liability of B. Next, the amount of B's maximum liability ($50,000) is likewise subtracted from plaintiff's maximum r ecovery ($80,000), the result being $30,000 ($80,000 - $50,000 = $30,000), which is the several liability of A. Then, the several liability of B ($5,000) and the several liability of A ($30,000) are added together, and the total of $35,000 ($30,000 + $5,000 = $35,000) is subtracted from plaintiff's maximum recovery ($80,000), the result being $45,000, which is the joint and several liability of A and B. Cast in this form, plaintiff would have judgment against A alone for $30,000, and also against B alone for $5,000, and furth er against A and B jointly and severally for an additional $45,000. These figures total $80,000 ($45,000 + $30,000 + $5,000 = $80,000). B's exposure is limited to $50,000 ($45,000 + $5,000 = $50,000); and A's exposure is limited to $75,000 ($45,000 + $30,000 = $75,000). Contribution would also be provided for as between B and C (see note 22, supra ).

In c ertain circumstances where three or more defendants and the plaintiff are each found guilty of causative fault SQ surely an extremely rare occurrence SQ an algebraic formula must be employed to arrive at the appropriate amounts of the several liability of each defendant and of the joint liability. Appropriate formulas are set out in the appendix to this d issent. It is important to recall, however, that it will always suffice to simply provide in the judgment a maximum amount which may be collected from each particular defendant, w hich is easily arrived at merely by multiplying the plaintif f's total damages by the fraction whose numerator is that particular defendant's percentage of the total fault of all parties and whose deno minator is the total of that particular defendant's and the plaintiff's respective percentages of the total fault of all parties. As previously noted, a judgment in that form, with appropriate provisions for contribution (see notes 22 and 23, supra ) will be wholly adequate. 53 General Maritime Law The m ajority concludes that it has always been a clearly established rule of United States general maritime law that in our hypothetical collision involving the three vessels, if all three were equally at fault the crew boat could recover two-thirds of its damages from the shrimper even though the shrimper was no more at fault than the crew boat. Not surprisingly, however, the majority cites no Supreme Court opinion so holding or stating, and only some general language in a few scattered lower court decisions, the earliest being in 1968, which do not directly address the question.

First, some background.

In The Catherine , 58 U.S. [17 How.] 170, 15 L.Ed. 233 (1855), two vessels collided, each being at fault. The Supreme Court held that the total loss should be divided equally, thus allowing a party to recover despite its own negligence, albeit only half of its loss. The Court stated: ". . . [I]t becomes necessary to settle the rule of damages in a case where both vessels are in fault.

The question, we believe, has never until now come distinctly before this court for decision. The rule that p revails in the District and Circuit Courts, w e understand, has been to divide the loss. . . .

This seems to be the well-settled rule in the English admiralty. . . .

Under the circumstances usually attending these disasters, we think the rule deviding [sic] the loss the most just and equitable, and as best tending to induce care and vigilance on both sides in the navigation." Id ., 58 U.S. at 177-178.

Thereafter, in The Washington , 76 U.S. [9 Wall] 513, 19 L.Ed. 787 (1869), a passenger on a ferry, who sustained serious personal inj ury when the ferry and the steamboat Washington collided, libeled both vessels, each of which was fou nd at fault. The libelant-passenger, of course, was not at fault. The Supreme Court stated: "Both vessels being in fa ult, both were liable to the libelant, and both could be proceeded against in the same libel. The damages were properly apportioned equally between the two vessels, the right being reserved to the libelant to collect the entire amount of either of them in case of the inability of the other to respond for her portion." Id ., 76 U.S. at 516.

The same result obtained in The Alabama , 92 U.S. 695, 23 L.Ed. 763 (1876), where the bark Ninfa, in tow of the tug Game-Cock, collided with the Alabama. The Ninfa libeled both The Game-Cock and The Alabama. "[B]oth The Alabama and The Game-Cock were in fault, and . . .

The Ninfa, which was in tow of The Game-Cock, and suffered the loss, was not in f ault." 92 U.S. at 695-96. "The district court rendered a decree against both [The Alabama and The GameCock] for the whole [of The Ninfa's loss], regarding them as liable in solido. The circuit court, on appeal reversed this decree, and divided the loss between them, rendering a decree against each for one half the amount." I d . at 696. The Ninfa appealed to the Supreme Court, which held: "Conceding, therefore, th at a vessel in tow, and without fault, is to be regarded as sustaining the same relation to the collision which is sustained by cargo (and it seems fair thus to consider it), we think that the decree of the circuit court was erroneous, and that a decree ought to be made against The Alabama and The Game-Cock, and the respective stipulators, severally, each for one moiety of the entire damage , interest, and costs, so far as the stipulated value of said vessel shall extend; and a ny balance of such moiety , over and above suc h stipulated value of either vessel, or w hich the libelant shall be unable to collect or enforce, shall be paid by the other vessel or her stipulators to the extent of the stipulated value thereof, beyond the moiety due from said vessel.

This is substantially the form of decree sanctioned by this court in The Washington and The Gregory , 9 Wall. 516, 19 L. ed. 788, a case involving similar principles, although the particular point was not fully discussed in that case." Id . at 697-98 (emphasis added).

Next came The Atlas , 93 U.S. 302, 23 L.Ed. 863 (1876), so heavily relied on, and evidently misunderstood, by the ma jority.

There, a canal boat laden with cargo was under tow by The Kate when t he canal boat and The Atlas collided, and as a result the canal boat sank and its cargo was lost. The subrogated insurers of the cargo lib eled The Atlas alone, and it was the only vessel before the court, as The Kate was not brought in. Id . at 308-309. The district court found that the cargo loss was "caused by the mutual fault of the steam-tug Kate and the steamboat Atlas, and that the libelants do recover against the steamboat Atlas one half of the damages by them sustained . . . . " Id . at 309. The libelants appealed, and the Supreme Court held that, as The Kate wa s not a party to the suit and The Atlas had not attempted to bring it in, the libelants, innocent of any wrongdoing, were entitled to recover their full damages against The Atlas, not simply on e half. The Court cites The Washington with full approval and states that "[m]uch care was taken in framing the decree in that case." T he Atlas at 318. The Court explains its holding as follows: "Contributory negligence on the part of the libelant cannot defeat a recovery in collision cases . . . .

Proof of the kind will defeat a recovery at common law; but the rule in the admiralty is, that the loss in such a case must be apportioned between the offending vessels, as having been occasioned by the fault of both; but the rule of the common law and of the admiralty is the same where the suit is promoted by an innocent party , except that the moiety rule may be applied in the admiralty, if all the parties are before the court , and each of the wrong-doers is liable to respond for his share of the damage. Subject to that qualification, the remedy of the innocent party is substantially the same in the admiralty as in an action at law, the rule being, that in both he is entitled to an entire compensation from the wrong-doer for the injury suffered by the collision. . . .

Goods shipped as cargo , and their owners, as in the case before the court, are innocent of all wrong ; and the owners of the cargo may sue the owners of one of the ships, or both, and they may sue at law or go into the admiralty, at their election, and having p roved their case, they are as much entitled to full compensation in the admiralty as they would have been if they had elected to pursue their common law remedy, saved to them by the proviso contained in the 9th s ection of the Judiciary Act. 1 Stat. at L., 77.

Co-wrong-doers, not parties to the suit, cannot be decreed to pay any portion of the damage adjudged to the libelant , nor is it a question in this case whether the party served may have process to compel the other wrongdoers to appear and respond to the alleged wrongful act. . . . .

Parties without fault, such as shippers and consignees, bear no part of the loss in collision suits, and are entitled to full compensation for the damage w hich they suffer from the wrong-doers, and they may pursue their remedy in personam, either at common law or in the admiralty, against the wrong-doers or any one or more of them, whether they elect to proceed at law or in the admiralty courts." I d . at 316-319 (emphasis added).

Plainly, The Atlas intended no departure from the moiety rule of The Washington and The Ala bama , but did not apply it s olely because the other vessel at fault, The Kate, was not before the Court, and no one had tried to bring her in. Just as plainly, the majority errs in suggesting that The Atlas ' numerous references to plaintiffs who are "without fault" or "innocent of all wrong" are explainable as having been made fourteen years before the bar of contributory negligence was lifted by The Max Morris , 11 S.Ct. 29 (1890). However, at least since the 1855 decision in T he Catherine , contributory negligence had been no bar. Moreover, T he Atlas was a collision case, its remarks were directed to such cases, and it openly recognized that "[c]ontributory negligence on the part of the libelant cannot defeat a recovery in collision cases" although "[p]roof of the kind will defeat a recovery at common law; but the rule in the admiralty is, that the loss in such a case must be apportioned between the offending vessels." I d . at 316-317. The majority has clearly misread The Atlas .

The Juniata , 93 U.S. 337, 23 L.Ed. 930 (1876), follows the same principles as The Atlas . 54 However, where both vessels at fault are before the court, the proper decree in favor of an innocent third party (such as a passenger, a tow, or cargo) continued to be the "moiety" rule, granting judgment for half the innocent party's dam ages against each of the two offending vessels, with provision that if the libelant should be unable to collect from one vessel its moiety, the other vessel would then be responsible for the deficiency. In such a case, it was reversible error to enter a judgm ent for the innocent plaintiff's damages against both vessels at fault jointly.

See, e.g., The Sterling , 1 S.Ct. 89 (1882), where the Court stated: "This was a suit in admiralty against the ship Sterling and tow-boat Equator, for damages sustained by the bark Sif in a collision. Both the ship and tow-boat were found to be in fault, and they were condemned in solido for the whole amount of the loss. From a decree to that effect this appeal was taken.

It is conceded that upon the facts found the owners of the Sif are entitled to a decree against the ship and the tow-boat, as both were in fault. The wellestablished rule in such cases is to apportion the damages equ ally between the two offending vessels, the right being reserved to the libelant to collect the entire amount from either o f them in case of the inability of the other to respond for her portion. . . . [citations] As in this case the decree was against both vessels for the full amount of the loss, it should be modified so as to be against the Sterling and the Equator, and their respective stipulators, severally, each for one-half of the entire damage and costs; any balance of such half which the libelant shall not be able to enforce against either vessel to be paid by the other vessel or her stipulators." Id . at 89-90. 55 And this rule continued to be enforced. Thus in C rain Br others, Inc. v. Wirman and Ward Company, 223 F.2d 256 (3d Cir. 1955), a suit by innocent cargo against the barge charterer, Union, and barge owner, Crain, who were both at fault, the Court stated: "We disagree, however, with the manner in which damages were awarded. Judgment was entered against both Union and Crane in the full amount. In admiralty, we have the rule of divided damages. . . . Where two parties are jointly responsible f or injury to a third, each is primarily liable for only one-half the damages. the charterer and owner of the barge should each be assessed with one-hal f the cargo loss with a provision that if the libellant cannot collect any part from one, that amount should be assessed against the other in addition to the one-half for whic h it is primarily liable." Id . at 258.

See also Gilmore & Black, Admiralty (2d ed. 1975) at 528 ("Where a third party is damaged, and sues two ships that are at fault, he is not prejudiced by the half-damages rule, but may collect his full damages from one if the other is unable to respond in damages, or may collect any deficiency if one cannot pay its full half" [footnote omitted]). 56 The Max Morris was a personal injury suit by a longshoreman against the vessel he was loading, and both parties having been found at fault the question certified to the Supreme Court was whether "the libelant . . . is entitled to a decree for divided damages," which the Supreme Court "answered in the affirmative." Id . at 31, 33. The Court noted that under T he Catherine fault did not bar all recovery in collision ca ses, and, after discussing, among other decisions, The Washington , The Alabama, T he Juniata , and Atlee v. Packett Co ., 21 Wall. 389, 22 L.Ed. 619 (1875) (where a vessel struck a pier), observed that "this court has extended the rule of the division of damages to claims other than those for damages to the vessels which were in fault in a collision." I d . at 32. The court then reviewed several lower court decisions concerning damage to cargo or tows, caused by mutual fault but not involving any collision, where the divided damages rule was applied to allow some recovery despite the plaintiff's fault. 57 The Court concluded that these cases had properly held that the lib elant's fault should only diminish recovery, not completely bar it, and that such a rule was appropriate "as in harmony with the rule for the division of damages in cases of collision." Id . at 33.

Accordingly, it held that the libelant "is entitled to a decree for divided damages." Id . 58 There are basically two things that one can say about all these cases. First, n one of them involved a situation in which the instant question could have ever been presented; that is, none involved a negligent plaintiff and at least two negligent other parties or actors. Second, the general maritime law did not slavishly follow the common law. Nor was the only difference that admiralty allowed the negligent plaintiff some (albeit diminished) recovery, for the i nnocent plaintiff's rights were also somewhat different, as Judge Addison Brown explained in The Hudson , 15 F. 162, 164 (S.D.N.Y. 1883): "And where the innocent owner of the cargo, or of a tow in charge of one vessel, sues and recovers against both vessels, the libelant cannot recover a judgment in solido against both for his whole damage, with a right to levy his execution in full against either alone, as at common law , but only a judgment for a moiety of the damages again st each vessel, with an alternative right of recour se against either for so much of the moiety adjudged t o be paid by the other as he is unable to collect from the latter." (Emphasis added).

Of course, U nited States v. Reliable Transfer Co. , 95 S.Ct. 1708 (1975), abandoned the rule that loss was always to be divided equally SQ or per vessel SQ among vessels at fault, and held that instead the allocation was to be based on the actual comparative fault of each. 59 However, there is nothing to indicate that the divided damages rule or its operation was changed otherwise than by replacing automatic equal per vessel at fault allocation with allocation by actual comparative degree of fault. The allocation was merely made more precise, so as to be fairer. 60 Suppose in The Juniata (see note 25, supra ) the libelant United States had also been at fault equally with The Juniata.

Would it have recovered two-thirds of its loss from The Juniata because the proof showed that The Neafie also was guilty of equal fault, even though the United States libeled only The Juniata (and The Neafie was not brought in)? No authority suggests such a perverse result. Suppose that the United States libeled both The Juniata and The Neafie, and all three were found equally at fault.

The United States would then presumably have recovery for one-third of its damages ag ainst The Juniata and The Neafie each, but what would the United States' alternative right of recovery be if, for example, the full third could not be collected from The Neafie? Could the United States then collect all of that shortfall from The Juniata, or only half of it? The only decision we have found addre ssing this question is P etition of Kinsman Transit Company , 338 F.2d 708 (2d Cir. 1964), cert. denied , 85 S.Ct. 1026 (1965), decided by a distinguished panel of the Second Circuit. There three parties, the City of Buffalo, Continental Grain Company, and Kinsman Transit Company, were each at fault and each suffered damages. Kinsman, however, was held entitled to limit its liability under the Limitations of Vessel Owner's Liability Act, 46 U.S.C. §§ 181-188. Judge Friendly, writing for himself and Judges Waterman and Moore, 61 held as follows: "A separate problem is how to deal, among the negligent parties, with that part of Kinsman's responsibility of which its limitation frees it. We think the fair soluti on is to divide that deficiency equally between Buffalo and Continental, rather than to h old Continental liable to Buffalo for the e ntire unsatisfied portion of Kinsman's share and vice versa . . . .

The decree is modified so that the City of Buffalo may recover two-thirds of the damages to its property from Continental and Kinsman subject to limitation by the latter but with Continental bearing only half of Kinsman's deficiency, that Continental may recover twothirds of the damages to its property from the City and Kinsman subject to limitation by the latter but with the City bearing only half of Kinsman's deficiency, and that Kinsman, which ma de no claim against Continental, may recover half of the damages suffe red by [Kinsman's vessel] the Shiras at the bridge from the City of Buffalo, which may then obtain contribution of half that amount from Continental." Id . at 726.

Kinsman SQ which is squarely contrary to the majority's approach SQ is di rectly on point and should control. We are aware of no contrary authority.

The few decisions cited by the majority are not persuasive of a contrary result. The only relevant issue in E mpire Seafoods, Inc. v. Anderson , 398 F.2d 204 (5th Cir.), cert. denied , 89 S.Ct. 449 (1968), w as whether Anderson and Gates, two employees of Cleary, a contractor working on a bridge, should have been awarded recovery directly against Cleary, as well as against Empire, whose vessel struck the bridge and who was awarded recovery over against Cleary for half of what the j udgment required it to pay Anderson and Gates. All parties were at fault. There was n o i ssue on appeal as to how much Anderson and Gates should recover from either Cleary or Empire, but only whether their recovery could be directly against Cleary at all or, if not, whether whatever they were awarded against Empire could be included as Empire's damages in Empire's action against Cleary. In our initial opinion, we held Anderson and Gates could recover directly from Cleary, as well as Empire, and supported this by quoting with approval the following passage from Benedict on Admiralty § 416 (6th ed. 1940), viz : "'The decree, therefore, should provide that each vessel . . . pay one-half of the entire damages, interest and costs, . . . and it should further provide that any part of the one-half damages assessed against either vessel, which libe lant may not be able to collect from that vessel, be assessed against the other vessel, in addition to the one-half which she is in t he first instance compelled to pay.'" Empire at 217.

Recognizing that this text was addressing liability for the damages of an innocent thi rd party 6 2 SQ as is obvious from the reference to the two vessels at fault each being primarily liable for "one-half of the entire damages" SQ our original opinion appended a footnote at the end of the above quotation, as follows: " 21. The authorities state the rule in terms of 'innocent third parties.' While it might be argued that these authorities can have no applicati on to the instant situation since Anderson and Gates were negligent, we are convinced that the reduction of their respective recoveries under the comparative negligence doctrine is to be considered fu ll penalty for their fault and that they must, thereafter, be treated in the same manner as 'innocent third parties.'" Empire at 217 n.21.

But, this footnote does not address SQ and there was no issue before our Empire panel concerning SQ whether the plaintif f, whose negligence is equal to that of each of the two defendants so that each defendant is initially liable for a third of damages, if unable to recover his thi rd from one of the defendants may then recover it all fr om the other, or may recover only half of the deficiency, as in Kinsman . Moreover, on rehearing in E mpire we withdrew our holding that Anderson and Gates could recover directly from Cleary SQ the holding made in that portion o f the opinion to which footnote 21 was appended SQ and stated: "Upon reconsideration, we are convinced that what we said about the District Court decree in our original opinion was apropos only to those instances where , aside from a statutory prohibition, an innocent third party is injured by the mutual fault of vessels in a collision." E mpire at 217.

The majority also relies on Gele v. Chevron Oil Co ., 574 F.2d 243 (5th Cir. 1978), involving a collision between a pleasure craft and a Chevron structure in the Gulf of Mexico in which Gele, a guest on the pleasure craft, was injured. The district court held Chevron solely at fault. On appeal, both Gele and Chevron contended that the pleasure craft, operated by Herr, was also at fault, and we agreed. We remanded to determine whether or not Gele also played such a role in the pleasure craft's operation so as to be chargeable with its fault, and the degrees of comparative fault as between the pleasure craft and Chevron. We next held that neither Chevron nor Herr were liable to the other in indemnity.

Imm ediately following this latter holding appears the following passage relied on by the majority here, viz : "This decision, of course, does not affect Gele's right to collect all his damages from one party in the event he is unable to obtain the relative portion of damages from each party at fault. E mpire Seafoods, Inc. v. Anderson , 5 Cir., 1968, 398 F.2d 204, 217, 1968 A.M.C. 2664, cert. denied , 393 U.S. 983, 89 S.Ct. 449, 21 L.Ed.2d 444." Gele at 251 (emphasis added).

This passage appears to address only a situation in which Gele was not chargeable with any fault, as else he would not be entitled to "collect all his damages" from anyone (and reference to his collection "from each party at fault" would like wise be inappropriate). 63 In any event, there is nothing to indicate that there was any issue before the Gele court concerning how much Gele, if negligent , could recover from Chevron or Herr in the event that collection could not be affected from one of them. As to the issue now before us, the quoted Gele language is no more than a passing and inapposite remark.

The majority's reliance on Drake Towing Co., Inc. v. Meisner Marine Const. Co ., 765 F.2d 1060 (11th Cir. 1985), is plainly misplaced. In that case, Drake's vessel was damaged when it struck a piece of concrete left in the channel by Meisner. Drake sued Meisner and the United States, the latter for its misplacement of a marking buoy. However, Drake settled with Meisner prior to trial. On trial, fault was allocated twenty percent to Drake, twenty percent to the United States and sixty percent to Meisner, and the district court awarded Drake judgment against the United States for twenty percent of Drake's total damages. Drake appealed, contending "that the district court erred in decreasing its recovery against the United States by the percentage of liability attributed to Meisner, a nonparty to the trial of the case. " 64 The Eleventh Circuit agreed. Its holding, however, is plainly contrary to McDermott, Inc. v. Am Clyde , 114 S.Ct. 1461 (1994). Just as the majority does here, the Drake Towing panel relie d on, and misread, E dmonds v. Compagnie General Transatlantique , 99 S.Ct. 2753 (1979). S ee Drake Towing , 765 F.2d at 1067. All that aside, however, the majority here clearly misreads Drake Towing itself and wholly ignores its actual holding.

The majority relies on the opinion's statement that "Drake may recover its entire damages, less that portion attributable to its own fault, from the United States." Id . The majority apparently believes that this means that the United States was charged with all Meisner's fault and thus Drake was held entitled to recover eighty percent of its damages from the United States. But, if that were so, then the Eleventh Circuit would simply have reformed the judgment (or ordered the district court to do so) to so reflect (there being no issue as to the amount of Drake's total damages).

However, that is not what the Eleventh Circuit did. Rather, it held that "[t]he issue of Meisner's liability is irrelevant to the determination of that of the United States," id . (emphasis added), and the court "therefore remand[ed] the case to the district court to reallocate liability between Drake and the United States without considering the negligence of Meisner ." I d . at 1068 (emphasis added). 65 In other words, D rake Towing held that the relevant comparison was not , as the majority here would have it, that between the negligence of Drake, on the one hand, and the combined negligence of Meisner and the United States, on the other hand, but rather was simply that between the negligence of Drake and the negligence of the United States, without considering whether or to what extent Meisner was negligent. Drake Towing does not support the majority here; rather, it rejects the very position which the majority contends for. These are essentially the general maritime law cases cited by the majority. They simply do not sustain its assertion of a wellestablished general maritime law rule allowing a plaintiff, in an accident or collision caused by his fault and that of two others acting independently of each other, with each of the three equally guilty, to hold either one of the other two liable for mo re than half his damages. The decision that comes closest to really addressing this issue is Kinsman , and it plainly supports the approach advocated in this dissent. It is not contended that that approach is well established either. The point simply is that the issue is essentially open. We should choose the fairest and most logical approach.

LHWCA Cases The majority also relies on cases involving injuries covered by the LHWCA, principally Pope & Talbot, Inc. v. Hawn , 74 S.Ct. 202 (1953), and Edmonds . These cases are plainly inapposite, as the subject matter of the instant case is not within the scope of the LHW CA, which does not reach injuries or activities within the territorial waters of foreign nations.

Examination of these decisions likewise reveals not only that they were driven by their LHWCA s etting, as is made plain by McDermott , but also that they did not purport to address or consider the issue here presented.

In Pope & Talbot , Hawn, a ship repairman employed by Haenn, was injured while on board Pope & Talbot's vessel. Hawn began receiving LHWCA compensation payments from Haenn and then sued Pope & Talbot for negligence, agreeing with Haenn to refund to it the LHWCA payments it had made out of any sums Hawn recovered from Pope & Talbot. Pope & Talbot brought in Haenn, seeking contribution or indemnity from it. A jury found Pope & Talbot, Haenn, and Hawn each negligent; seventeen-and-a-half percent negligence was assigned to Hawn, but no percentage was assigned to either Haenn or to Pope & Talbot, and it does not appear that any complaint was ever made of this manner of submission. The district court rendered judgment for Hawn against Po pe & Talbot for 87œ% of his total damages and awarded Pope & Talbot contribution against Haenn in the amount of half of Pope & Talbot's liability to Hawn (but not more than Haenn's maximum potential LHWCA liability to Hawn). H awn v. Pope & Talbot , 99 F.Supp. 226 (E.D. Pa. 1951). 66 On appeal, the Third Circuit affirmed the award against Pope & Talbot, but reversed the award against Haenn, holding that contribution was not available. H awn v. Pope & Talbot , 198 F.2d 800 (3d Cir. 1952).

The Supreme Court granted Pope & Talbot's application for certiorari, but affirmed the Third Circuit. It held that under Halcyon Lines v. Haenn Ship Ceiling & Refitting Corp ., 72 S.Ct. 277 (1952), contribution against Haenn was barred. Pope & Talbot , 74 S.Ct. at 204. 67 It likewise rejected Pope & Talbot's alternative contention that because Hawn had agreed to refund his LHWCA payments to Haenn out of his recovery from Pope & Talbot, therefore "the judgment against it [Pope & Talbot] should be reduced by this amount." I d . at 206. The Court rejected this contention as being inconsistent with section 33 of the LHWCA and as in effect allowing contribution from the employer contrary to Halcyon . 68 The Court lik ewise rejected Pope & Talbot's contention that "contributory negligence should have been accepted as a complete bar to Hawn's recovery," stating: "The harsh rule of the common law under which contributory negligence wholly barred an injured person from recovery is completely incompatible with modern admiralty policy and practice. Exercising its traditional discretion, admiralty has developed and now follows its own fairer and more flexible rule which allows such consideration of contributory negligence in mitigation of damages as just ice requires . Petitioner presents no persuasive arguments that admiralty should now adopt a discredited doctrine which automatically destroys all claims of injured p ersons who have contributed to their injuries in any degree, however slight . P ope & Talbot, 74 S.Ct. at 204- 205 (emphasis added; footnote omitted). 69 That is just what this dissent asks for, a "fairer and more flexible" rule allowing "consideration of contributory negligence in mitigation of damages as justice requires." Other than Pope & Talbot's "ingeniou s argument" that Hawn's recovery from it should be reduced by what he received under the LHWCA, which the Court rejected as contrary to LHWCA section 33 (see note 39, supra ), Pope & Talbot's position vis-a-vis Hawn was simply an all or nothing one SQ Hawn should not recover at all from it, not that his recovery was not properly calculated. The point here in issue was simply not before the Court in Pope & Talbot , nor did the Court there in any way address it. 70 Pope & Talbot was an LHWCA-driven case, and simply does not speak to the present question.

We turn now to Edmonds , the majority's lead case. There, Edmonds, a longshoreman, was injured in 1974 on a vessel in the course of his employment. He received LHWCA compensation from his employer, the stevedo re, and brought suit against the vessel's owner for negligence. The jury found Edmonds suffered a total of $100,000 damages, that he was 10% at fault, that the vessel was 20% at fault, a nd that the stevedore, which was not a party to the suit, was 70% at fault. The district court granted judgment for Edmonds against the vessel owner for $90,000. The Court of Appeals held that Edmonds could recover no more than $20,000 from the vessel owner, its percentage of the total fault of all three actors tim es the total damages. E dmonds v. Compagnie General Transatlantique , 577 F.2d 1153 (4th Cir. 1978). The Supreme Court reversed, holding that Edmonds was entitled to recover $90,000 from the vessel owner. E dmonds v. Compagnie General Transatlantique , 99 S.Ct. 2753 (1979).

Two things may be said about Edmonds . First, it was driven by the LHWCA. Edmonds extensively reviews how the pure several liability approach of the Court of Appeals would affect the stevedore's and longs horeman's rights under the LHWCA and particularly the 1972 amendments thereto. I d . at 2761-62. The Court concludes by observing "we are mindful that here we deal with an interface of statutory and judge-made law," id . at 2762, and expressing re luctance to "knock out of kilter" the "delicate balance" struck by Congress between the rights of longshoremen, stevedores, and shipo wners in the 1972 amendments to the LHWCA.

Id . at 27 63. Any doubt on this score is surely removed by McDermott where the Court states that " Edmonds was primarily a statut ory construction case and related to special interpretive questions pos ed by the 1972 amendments to the Longshoremen's and Harbor Workers' Compensation Act." McDermott , 114 S.Ct. at 1471.

This was not idle dicta, for in McDermott a principle argument of respondents was that "the proportionate share rule," which McDermott ultimately approved, "is inconsistent with Edmonds ." McDermott , 114 S.Ct. at 1471. Moreover, several courts, including the Eleventh Circuit in Self v. Great Lakes Dredge & Dock Co ., 832 F.2d 1540, 1548 (11th Cir. 1987) ("bound by the Supreme Court's guidance and the rule in Edmonds "), and this Court in Hernandez v. M/V Rajaan , 841 F.2d 582, 591 (5th Cir.), cert. denied , 109 S.Ct. 530 (1988) (following reasoning of Self ), previously had rejected the proportionate fault settlement credit rule adopted in McDermott on the theory that it was inconsistent with Ed monds . We should indeed be wary of again reading Edmonds too broadly. Finally, we cannot ignore McDermott 's express and apparently approving reference to section 2 of t he UCFA (quoted in note 11, s upra ), particularly to th a t section's provision for "reallocation of insolvent defendant's equ itable share." M cDermott at 1471 n.32 ( see also id . n.31). Seemingly, M cDermott considers such an approach at least an unforeclosed option in the non-LHWCA context.

Second, all the parties and courts in volved in E dmonds conside red only two alternatives, namely whether to apply pure several liability, with the vessel being liable only for its 20% share and bearing no part of the stevedore's 70%, or whether, on the other hand, to apply joint and several liability, as would be the case if the plaintiff had not been negligent, so that the vessel would be liable for 90 % and would bear all of the stevedore's fault. No consideration was given to, and there was even no recognition of, the possibility that the stevedore's fault should simply be ignored or, what is essentially the same thing, that the stev edore's fault should be allocated between the longshoreman and the vessel in the same ratio that the negligence of each bore to that of the other. Apart from its concern with the LHWCA, the thrust of Edmonds amounts to a questioning of the proposition that a third party's fault should reduce the liability which the defendant would otherwise have. E dmonds cites no authority or general principles addressing how the negligence of a plaintiff is to be compared where there are two or more other independent a ctors also guilty of causative fault. 7 1 E dmonds ' approach in this respect is well illustrated by its posing of the question: "'one is still left to wonder why the longshoreman injured by the negligence of a third party should recover less when his employer has also been negligent than when the employer has been without fault.'" Id . at 2761 n.24 (emphasis added) (quoting Zapico v. Bucyrus-Erie Co ., 579 F.2d 714, 725 (2d Cir. 1978)).

There is no satisfactory answer to that question. This case poses the flip side of the same question, namely why should negligent A, injured in a three-person accident also involving B, likewise negligent, and C, recover more from B if C is negligent than if C is without fault. There is similarly no satisfactory answer to this question. The reason in each instance is that the independent third party's fault is irrelevant to what the plaintiff should ultimately recover from the other party, just as Drake Towing held.

Certainly, the result in Edmonds is binding on us in suits on LHWCA-covered injuries. But outside of that class of case, E dmonds is not a proper basis on which to evaluate an approach it (and the parties before it) wholly failed to address or consider. S ee United States v. Mitche ll , 46 S.Ct. 418, 419-20 (1926). 7 2 As p reviously observed (see note 38, s upra ), the Court in Cooper Stevedoring Co., Inc. v. Fritz Kopke, Inc ., 94 S.Ct. 2174 (1924), refused to extend the Halcyon ban on contribution beyond its context of a contribution claim against the LHWCA employer of the injured plaintiff, and McDermott recognized that Edmonds was LHWCAdriven and refused to extract from it a general principle to govern the effect of settlement in general maritime law multiple party cases. In this general maritime law case, we, too, should not expand Edmonds beyond its LHWCA context to speak to something it never addressed even in that special context.

FELA, Jones Act, and Miles v. Apex Marine The majority argues that the Jones Act incorporates the FELA, that the re sult it reaches would be reached under the FELA and hence under the Jones Act, and that therefore under Miles v. Apex Marine Corp ., 111 S.Ct. 317 (1990), should be reached in this general maritime law case.

There are several answers to this. Most obviously, the s ubject matter of this case is not governed by the Jones Act. Coats was not a Jones Act seaman, nor was he any sort of employee of Penrod. 73 M iles considered "whether the parent of a seaman who died from injuries on . . . [the defendant's] vessel may recover under general maritime law for loss of society, and whether a claim for the seaman's lost future earnings survives his death." I d . at 319-20. It answered both questions in the negative, because neither such recovery was available under the Jones Act. I d . at 325-26, 328. The Court stated "we restore a uniform rule applicable to all actions for the wrongful death of a seaman," id . at 326, and " [b]cause this case involves the death of a seaman, we must l ook to the Jones Act." I d . at 328 (emphasis added). Our recent en banc opinion in Guevara v. Maritime Overseas Corporation , F.3d (No. 92-4711, 5th Cir., , 1995), states: "In order to decide whether (and how) Miles applies to a case, a court must first evaluate the factual setting of the case and determine what statutory remedial measures, if any, apply in that context. I f the situation is covered by a statute like the Jones Act or DOHSA , and the statute informs and limits the available d amages, the statute directs and delimits the recovery available under the general maritime law as well." (Emphasis in original).

Clearly the factual setting o f this case is n ot covered by the Jones Act. 74 Accordingly, the above methodology stated in G uevara wou ld appear not to support application of the M iles uniformity principle here.

But even were the Miles uniformity principle applicable, the majority has not demonstrated any established or consistent body of law sustaining the result in this case under either the FELA or the Jones Act. Turning first to the FELA, when it was adopted in 1908 none of the states authorized any recovery whatever by a plai ntiff whose negligence proximately contributed, in even the slightest degree, to the accident in question. Thus, when the FELA was adopted it could not possibly have inferentially incorporated any common law rule on how the recove ry of a negligent plaintiff was to be computed in an instance in which two or more other independent actors, at least one of whom was a defendant, were also guilty of causative fault. The common law simply did not address such a situation. Nor does the wording of the FELA. It provides that "[e]very comm on carrier by railroad . . . shall be liable in damages to any person suffering injury while he is employed by such carrier . . . resulting in whole or in part from the negligence . . . of such carrier," 45 U.S.C, § 51, and that "the fact that the employee may have been guilty of contributory negligence shall not bar a recovery, but the damages shall be diminished by the jury in p roportion to the amount of negligence attributable to such employee." 45 U.S.C. § 53. The FELA neither creates nor even speaks to any cause of action or suit against anyone other than the emp loyer-railroad. 7 5 The facial inference from the wording and structure of the statute is that the plaintiff's negligence is to be compared to, and only to, that of the defendant employerrailroad. In N orfolk & Western Railway Company v. Earnest , 33 S.Ct. 654 (1913), the Court, speaking of what is now section 53, stated: ". . . [T]he statutory direction that the diminution shall be 'in proportion to the amount of negligence attributable to such employee' means, and can only mean, that, where the causal negligence is partly attributable to him and partly to the carrier, he shall not recover full damages, but only a proportional amount, bearing the same relation to the full amount as the negligence attributable to the carrier bears to the entire negligence attr ibutable to b oth . . . ." I d . at 657 (emphasis added).

Nothing in section 53 suggests that the negligence comparison thereby called for involves the consideration of the negligence of anyone other than the plaintiff-employee and the defendant-employer railroad.

The only FELA case which the majority cites as being to the contrary is Gaulden v. Burlington Northern, Inc ., 654 P.2d 383 (Kan. 1982). There the plaintiff railroad employee, injured in a crossing collision involving a truck driven by James, a third party, sued the employer railroad under the FELA and James under state law, but settled with James prior to trial. The court held that the proportionate fault rule to account for the settlement was app licable, so that the plaintiff's recovery from the railroad would be reduced by the proportion which the total of plaintiff's negligence and that of the settling James bore to the total negligence of all three parties. The court went on in dicta, however, and without citation of any authority, to state that had James not settled then the railroad would be liable to the plaintiff for the same fraction of his total damages as the total of the fault of the railroad and James was of the total fault of all three. Id . at 392.

If this dicta in a 1982 Kansas decision is the best the majority can do, it can hardly be said that there is or was any well-established and settled FELA rule in this respect.

As previo usly observed, the Jones Act, passed in 1920, gave seamen injured in the course of employment an action against their employer to be governed by the FELA. 76 The situation then was essentially the same as in 1908 when the FELA was adopted, namely that in all b ut three states S Q as opposed to in all states in 1908 SQ any causative negligence on the part of the plaintiff, no matter how slight, barred any recovery whatev er. 77 Accordingly, what has been said about the FELA is applicable to the Jones Act. When the Jones Act was passed, it could not have impliedly adopted any general or established common law rule or practice for computing the recovery of a negligent plaintiff when two (or more) independently acting defendants SQ or a defendant and one or more others SQ were also neglige nt. That is true because there was no such rule or practice, as the plaintiff's negligence barred any recovery. Nor is there any showing that by 1920 there had grown up under the FELA any such established rule or practice which the Jones Act could be said to have impliedly adopted. 78 The majority cites four cases under the Jones Act which it claims would support the result here if this were a Jones Act case.

All of these cases were decided within the last decade. In three of the cases, the court was not presented with, and did not purport to speak to, a situation involving a negligent plaintiff, so the cited general language concerning joint and sever al liability of the defendants is plainly consistent with the position of this dissent. 79 These three cases tell us absolutely nothing relevant here.

The fourth and final Jones Act case cited by the majority in this respect, Joia v. Jo-Ja Service Corp., 817 F.2d 908 (1st Cir. 1987), me rits more detailed consideration. There the plaintiff- seaman, Joia, sued his employer, Niagra, under the Jones Act, and in the same action also sued Jo-Ja, the owner of another vessel contributing to the injury, under the general maritime law. Joia's total damages were found to be $360,000, and fault was allocated 5% to plaint iff Joia, 30% to his employer Niagra, and 65% to Jo-Ja.

Niagra alone appealed. The First Circuit held that Joia was entitled to jud gment against Niagra and Jo-Ja, jointly and severally, for $342,000 (95% of $360,000). Several things are significant about Joia . First, the only contentions of the parties in this respect, and all that the First Circuit addressed or considered, was whether the limit of Niagra's liability should be $108,000, 30% of the total damages, or $342,000, 95% of the total.

Id . at 914, 917. The court found pure several liability, in which all of Jo-Ja's fault is charged to Joia, too harsh considering "the remedial nature of the Jones Act." Id . at 917. In its rejection of the pure several liability contended for by Niagra, the First Circuit relied on Edmonds , which plainly likewise rejected such an approach (as does this dissent). J oia at 916-917. 8 0 The J oia panel SQ like Edmonds SQ simply never adverted to the possibility that Niagra's maximum liability should instead be fixed by the ratio of its perce ntage of fault (30%) to the total of the percentages of fault of Joia (5%) and it (30%) SQ in other words, on the basis of a comparison of Niagra's negl igence to Joia's. That would have limited Niagra's exposure to $308,571.42 ( 30/35 S Q or 6/7 SQ of $360,000). Further, the J oia panel noted Niagra's reliance on Leger v. Drilling Well Control, Inc ., 592 F.2d 1246 (5th Cir. 1979), but instead chose to follow Ebanks v. Great Lakes Dredge & Dock Co ., 688 F.2d 716 (11th Cir. 198 2), cert. denied , 103 S.Ct. 1774 (1983), which rejected the proportionate share credit approach of Leger as being inconsistent with Edmonds . 81 J oia at 915-17. We now know from M cDermott SQ which the Joia panel did not h ave the benefit of SQ that Leger was right and that Ebanks and its progeny erred in concluding otherwise and in reading Edmonds overbroadly.

Finally, it is significant that Joia treats the issue before it essentially as res nova . J oia does not purport to find any settled or recognized body of Jones Act (or FELA) law, or general maritime law, addressing how a given defendant's maximum exposure is to be fixed in a multi-party case involving a negligent plaintiff and two (or more) independently act ing negligent defendants (or one negligent defendant and one or more negligent, independently acting others).

In sum, the subject matter of this case is not one covered by the Jones Act. Moreover, there is no settled body of Jones Act law addressing the issue now before us, and there certainly was no settled body of law, either under the FELA or otherwise, addressing the issue at the time the Jones Act was adopted (or when the FELA was). 82 Hence, the Jones Act does not dictate the result here.

Other Considerations Other than its appeal to authority, which is largely nonexiste nt, the majority levels essentially three objections to the rule espoused by this dissent.

First, the majority seems to suggest that this is a matter which shou ld be taken care of by contract, or by avoiding doing business with potential co-defendants who might be or become insolvent. This is obviously a make-weight, at best. We do not normally justify adoption of rules that are illogical and unfair on the basis that parties might often be able to contract around the illogic or unfairness we are thus c reating. Moreover, the majority's rationale in this respect is inconsistent with the settled rule that one is not ordinarily liable for the independent fault of an independent contractor. The majority would have it that such liability should always be imposed because the owner can recover in indemnity or contr ibution from the contractor, and if that is precluded by the contractor's insolvency or unavailability, then the owner has no one but himself to blame, as he sho uld not have done business with the contractor. Presumably under this approach, no o ne would contract with an immune entity such as a county. Finally, how, in our hypothetical collision involving the shrimper, the Exxon crew boat, and the pleasure craft, can any of these parties SQ having no prior contact one with the other SQ be expected to have contracted with each other in advance of the accident? Next, the majority invokes the notion tha t the plaintiff's recovery should be maximized. However, if that were the guiding principle, we could simply disregard, or perhaps give only half weight to, the plaintiff's contr ibutory negligence. And, again reverting to our hypothetical collision, why do we want to strain to make the s hrimper's master-owner pay Exxon more than half Exxon's damages, even though Exxon's crew boat is every bit as much at fault as the shrimper, just because the little pleasure craft, acting wholly independently of the shrimper with which it has never had any contact, was also negligent? Finally, the majority objects because the dissent's approach, in contrast to that of the UCFA, does not require the defendant to first, and after judgment, es tablish that another defendant or actor i s insolvent or unreachable (or immune) so that full contribution is unavailable from that other defendant, before that other defendant's "equitable share" is partially reallocated to the negligent plaintiff. Of course, under the UCFA and Apportionment of Liability , if the other defendant is determined to be insolvent before judgment, then the "reallocation" will be in the o riginal judgment, and the origina l judgment will give the negligent plaintiff the same maximum recovery from the solvent defendant as would be the case under the rule advocated in this dissent. More importantly, however, the substance of the UCFA and Apportionment of Liability approach is clearly that the negligent plaintiff should not be able to cause any one defendant to ultimately bear a greater fraction of the plaintiff's damages than the fault of that defendant is of the total of the fault of the plaintiff and that defendant. In other words, beyond that limit, that defendant is simply not ultimately liable to the negligent plaintiff. 83 Why should that defendant be made the collection agent fo r the negligent plaintiff as to sums for which that defendant is not ultimately liable? 84 Why should one party be the collection agent for another? Moreover, to say that the approach of this dissent puts an unfair collection burden on the negligent plaintiff is certainly to ignore the longstanding general mari time law rule that even the innocent plaintiff who is personally injured in an accident as to which two defendants equally at fault are before the court (either by being sued directly or brought in under Rule 14(c) or its precursors) recovers judgment initially from each defendan t for only half his damages, and can go beyond that as to each o nly by first showing his (plaintiff's) inability to collect from the other defendant the latter's half. The majority has simply ignored this long-standing rule of the general maritime law. 85 The Judicial Role in the General Maritime Law As the Supreme Court said in Edmonds , "[a]dmiralty law is judge-made law to a great e xtent." I d . at 2756. Indeed, "the Judiciary has traditionally taken the lead in formulating flexible and fair remedies in the law maritime." R eliable Transfer at 1715.

And, as the Supreme Court observed on yet another occasion: "Absent a relevant statute, the general maritime law, as developed by the judiciary , applies . . . the general maritime law is an amalgam of traditional common-law rules, modifi cations o f those rules, and newly created rules." E ast River Steamship Corp. v. Transamerica Delaval, Inc ., 106 S.Ct. 2295, 2299 (1986) (emphasis added).

In East River , the Court "join[ed] the Courts of Appeals" in adopting stri ct products liability, thus doing away with the traditional requirement of negligence in such instances. The Court in both Reliable Transfer and Moragne v. States Marine Lines, Inc ., 90 S.Ct. 1772 (1970), overruled its own long-standing and consistently and frequently enforced pr ecedents. But, no such departure from settled and clearly established precedent is required to adopt the approach taken in this dissent.

Nevertheless, it is appropriate to note that in Reliable Transfer the Supreme Court declined to continue with a rule it considered "unnecessarily crude and inequitable," despite that rule's lesser "problems of proof" and "facile application." I d . at 1714.

Reliable Transfer departed from a settled rule that "has continued to prevail in this country by sheer inertia rather than by reason of any intrinsic merit." Id . at 1715.

There is no relevant statute here, nor any clearly established rule, and we should reject an approach based on abstract doctrinal reflex a nd inertia rather than intrinsic merit. And, the majority's approach is one rejected by four-fifths of the states.

To borrow from Pope & Talbot , admiralty should rather employ a "fairer and more flexible rule which allows such consideration of cont ributory negligence in mitigation of damages as j ustice requires ." I d . at 204 (emphasis added). Justice, it seems to us, requires that the negligent defendant bear no greater S Q or lesser SQ part of the negligent plaintiff's total damages than that fraction which such defendant 's negligence is of the total negligence of the two of them; and that there is no reason to charge a negligent defendant with all , while charging the negligent plaintiff with none , of the fault of an independent third party, just as there is no reason to charge all such third party fault to the negligent plaintiff. The majority repeats "joint and several liability" as if it were some kind of magical mantra or totem which both banishes all dangers of rational analysis and dispenses with the need for authority dealing with the issue here posed: how does admiralty compare fault when both the plaintiff and a defendant are at fault and so also is at least one other independent actor (whether or not likewise a defendant). The majority abjures any meaningful normative defense of its position, and the best authority it can come up with is Joia , a 1987 Jones Act case that treats the matter as res nova and wrongly assumes that the only alternative to its result is pure several lia bility that assigns all the third party's fault to the plaintiff. An approach such as that taken by Judge Friendly in Kinsman SQ whic h is flatly inconsistent with that of the majority SQ simply accomplishes a just result, as does the position of this dissent.

Conclusion For the foregoing reasons, I respectfully dissent. i Appendix Formula for Calculation of Several Liability of Each Defendant and Joint Liability of Defendants 1. Three defendants and plaintiff each guilty of causative fault (a) VARIABLES Q = Plaintiff's total damages X = Plaintiff's % of total fault D1, D2, D3 = each defendant's % of total fault M1, M2, M3 = each defendant's maximum liability Y1, Y2, Y3 = each defendant's several liability Z = Joint liability of defendants (b) FORMULAS M1 = D1 Q D1 + X M2 = D2 Q D2 + X M3 = D3 Q D3 + X Y1 = M1 - Z Y2 = M2 - Z Y3 = M3 - Z Z = [(1.00 - X) * Q] - Y1 - Y2 - Y3 ii (c) EXAMPLE 1 Q = $100,000 X = 10 D1 = 20 D2 = 30 D3 = 40 M1 = D1 Q = 20 $100,000 = (2/3) * $100,000 = $66,667 D1 + X 20 + 10 M2 = D2 Q = 30 $100,000 = (3/4) * $100,000 = $75,000 D2 + X 30 + 10 M3 = D3 Q = 40 $100,000 = (4/5) * $100,000 = $80,000 D3 + X 40 + 10 Y1 = M1 - Z = $66,667 - Z Y2 = M2 - Z = $75,000 - Z Y3 = M3 - Z = $80,000 - Z Z = [(1.00 - X) * Q] - Y1 - Y2 - Y3 Z = [(1.00 - .10) * $100,000] - Y1 - Y2 - Y3 Z = (.90 * $100,000) - Y1 - Y2 - Y3 Z = $90,000 - ($66,667 - Z) - ($75,000 - Z) - ($80,000 - Z) Z = 3Z - $131,667 2Z = $131,667 Z = $65,833 Y1 = $66,667 - $65,833 = $834 Y2 = $75,000 - $65,833 = $9,167 Y3 = $80,000 - $65,833 = $14,167 The judgment thus provides plaintiff a total recovery of $90,000 composed of the following: Total D1 D2 D3 Recovery Several Liability $ 834 $ 9,167 $14,167 $24,168 Joint Liability 65,833 65,833 65,833 65,833 Total $66,667 $75,000 $80,000 $90,001 iii (d) EXAMPLE 2 Q = $100,000 X = 25 D1 = 25 D2 = 25 D3 = 25 M1 = D1 Q = 25 $100,000 = (1/2) * $100,000 = $50,000 D1 + X 25 + 25 M2 and M3 are calculated the same as M1 Y1 = M1 - Z = $50,000 - Z Y2 and Y3 are calculated the same as Y1 Z = [(1.00 - X) * Q] - Y1 - Y2 - Y3 Z = [(1.00 - .25) * $100,000] - Y1 - Y2 - Y3 Z = (.75 * $100,000) - Y1 - Y2 - Y3 Z = $75,000 - ($50,000 - Z) - ($50,000 - Z) - ($50,000 - Z) Z = 3Z - $75,000 2Z = $75,000 Z = $37,500 Y1 = $50,000 - $37,500 = $12,500 Y2 = $50,000 - $37,500 = $12,500 Y3 = $50,000 - $37,500 = $12,500 The judgment thus provides plaintiff a total recovery of $75,000 composed of the following: Total D1 D2 D3 Recovery Several Liability $12,500 $12,500 $12,500 $37,500 Joint Liability 37,500 37,500 37,500 37,500 Total $50,000 $50,000 $50,000 $75,000 iv (e) EXAMPLE 3, FOUR DEFENDANTS Q = $100,000 X = 20 D1 = 20 D2 = 20 D3 = 20 D4 = 20 M1 = D1 Q = 20 $100,000 = (1/2) * $100,000 = $50,000 D1 + X 20 + 20 M2, M3 and M4 are calculated the same as M1 Y1 = M1 - Z = $50,000 - Z Y2, Y3, and Y4 are calculated the same as Y1 Z = [(1.00 - X) * Q] - Y1 - Y2 - Y3 - Y4 Z = [(1.00 - .20) * $100,000] - Y1 - Y2 - Y3 - Y4 Z = (.80 * $100,000) - Y1 - Y2 - Y3 - Y4 Z = $80,000-($50,000- Z)-($50,000-Z)-($50,000-Z)-($50,000-Z) Z = 4Z - $120,000 3Z = $120,000 Z = $40,000 Y1 = $50,000 - $40,000 = $10,000 Y2 = $50,000 - $40,000 = $10,000 Y3 = $50,000 - $40,000 = $10,000 Y4 = $50,000 - $40,000 = $10,000 The judgment thus provides plaintiff a total recovery of $80,000 composed of the following: Total D1 D2 D3 D4 Recovery Several Liability $10,000 $10,000 $10,000 $10,000 $40,000 Joint Liability 40,000 40,000 40,000 40,000 40,000 Total $50,000 $50,000 $50,000 $50,000 $80,000 v (f) EXAMPLE 4, TWO DEFENDANTS Q = $100,000 X = 20 D1 = 20 D2 = 60 M1 = D1 Q = 20 $100,000 = (1/2) * $100,000 = $50,000 D1 + X 20 + 20 M2 = D2 Q = 60 $100,000 = (3/4) * $100,000 = $75,000 D2 + X 60 + 20 Y1 = M1 - Z = $50,000 - Z Y2 = M2 - Z = $75,000 - Z Z = [(1.00 - X) * Q] - Y1 - Y2 Z = [(1.00 - .20) * $100,000] - Y1 - Y2 Z = (.80 * $100,000) - Y1 - Y2 Z = $80,000 - ($50,000 - Z) - ($75,000 - Z) Z = 2Z - $45,000 Z = $45,000 Y1 = $50,000 - $45,000 = $5,000 Y2 = $75,000 - $45,000 = $30,000 The judgment thus provides plaintiff a total recovery of $80,000, composed of the following: Total D1 D2 Recovery Several Liability $ 5,000 $30,000 $35,000 Joint Liability 45,000 45,000 45,000 Total $50,000 $75,000 $80,000 vi 2. SPECIAL INSTANCES (a) With as many as four parties (including the plaintiff) guilty of causative fault, the formula will, in certain instances of unusual fault distributions, produce a negative several liability number for a particular defendant (where as many as five parties, including plaintiff, are guilty of causati ve fault, as many as two such negative several liability numbers are possible in certain instances of unusual f ault distributions). In such instances, additional steps must be added to the formula, resulting in a lowered joint liability for that particular defendant and an additional component of joint liability for the remaining defendants (if there are two defendants with initial negative several liability numbers, there will be two additional components of joint liability for other defendants). The additional necessary steps are set out below. (b) Formula for three defendants showing additional steps where under initial steps one defendant's several liability is a negative number Q = $100,000 X = 30 D1 = 30 D2 = 30 D3 = 10 M1 = D1 Q = 30 $100,000 = (1/2) * $100,000 = $50,000 D1 + X 30 + 30 M2 = D2 Q = 30 $100,000 = (1/2) * $100,000 = $50,000 D2 + X 30 + 30 M3 = D3 Q = 10 $100,000 = (1/4) * $100,000 = $25,000 D3 + X 10 + 30 Y1 = M1 - Z = $50,000 - Z Y2 = M2 - Z = $50,000 - Z Y3 = M3 - Z = $25,000 - Z Z = [(1.00 - X) * Q] - Y1 - Y2 - Y3 Z = [(1.00 - .30) * $100,000] - Y1 - Y2 - Y3 Z = (.70 * $100,000) - Y1 - Y2 - Y3 Z = $70,000 - ($50,000 - Z) - ($50,000 - Z) - ($25,000 - Z) Z = 3Z - $55,000 2Z = $55,000 Z = $27,500 vii Y1 = $50,000 - $27,500 = $22,500 Y2 = $50,000 - $27,500 = $22,500 Y3 = $25,000 - $27,500 = -$2,500 CAVEAT: If Y3 is

Z = Joint liability of D1, D2, and D3 J = Joint liability of D1 and D2 STEP 1: Z = M3 Z = $25,000 STEP 2: Y1 = Y1 + Y3 Y1 = $20,000 STEP 3: Y2 = Y2 + Y3 Y2 = $20,000 STEP 4: J = Y3 * -2 J = $5,000 STEP 5: Y3 = 0 Y3 = 0 The judgment thus provides the plaintiff a total recovery of $70,000 composed of the following: Total D1 D2 D3 Recovery Several Liability $20,000 $20,000 $ -0- $40,000 Joint D1 & D2 5,000 5,000 -0- 5,000 Joint D1, D2 & D3 25,000 25,000 25,000 25,000 Total $50,000 $50,000 $25,000 $70,000 (c) Formula for four defendants showing additional steps when under initial steps one defendant's sev eral liability is a negative figure Q = $100,000 X = 25 D1 = 25 D2 = 25 D3 = 20 D4 = 5 M1 = $50,000 M2 = $50,000 M3 = $44,444.44 M4 = $16,666.67 Z = $28,703.70 viii Y1 = $21,296.30 Y2 = $21,296.30 Y3 = $15,740.74 Y4 = - $12,037.03 CAVEAT: If Y4 is

Z = Joint liability of D1, D2, D3, and D4 J = Joint liability of D1, D2, and D3 STEP 1: Z = M4 Z = $16,666.67 STEP 2: Y1 = 1/2Y4 + Y1 Y1 = $15,277.78 STEP 3: Y2 = 1/2Y4 + Y2 Y2 = $15,277.78 STEP 4: Y3 = 1/2Y4 + Y3 Y3 = $9,722.22 STEP 5: J = 1/2Y4 + Y4 J = -$18,055.55 STEP 6: J = J * -1 J = $18,055.55 STEP 7: Y4 = 0 Y4 = 0 The judgment thus provides the plaintiff a total recovery of $75,000 composed of the following: Total D1 D2 D3 D4 Recovery Several Liab'y $15,277.78 $15,277.78 $ 9,722.22 $ -0- $40,277.78 Joint All D's 16,666.67 16,666.67 16,666.67 16,666.67 16,666.67 Jnt D1, D2, D3 18,055.55 18,055.55 18,055.55 -0- 18,055.55 Total $50,000.00 $50,000.00 $44,444.44 $16,666.67 $75,000.00 (d) Formula for four defendants showing additional steps when under initial steps the several liabili ty of each of two defendants is a negative number Q = $100,000 X = 30 D1 = 20 D2 = 40 D3 = 6 D4 = 4 ix M1 = $40,000 M2 = $57,142.86 M3 = $16,666.67 M4 = $11,764.70 Z = $18,524.75 Y1 = $21,475.25 Y2 = $38,618.11 Y3 = - $1,858.08 Y4 = - $6,760.05 CAVEAT: If Y3 and Y4 are

Secondly, if United States law is to be applied, we s hould apply United States law as it existed at the time of the casualty in this case not as it existed prior to 1972.

I. Whether United States Law My differences with the panel on the Lauritzen - Rhoditis choice of law factors involve the first factor (place of the wrongful act); the fourth factor (allegiance of the defendant ship owner); and the fifth factor (place of the contract). Looking first at the place of the wrongful act, the majority opinion devotes one sentence to analysis of this subject. It recognizes that "the accident occurred in the territorial waters of the United Arab Emirates" and that since this is a "nontraditional maritime case," that factor is entitled to "considerable weight." Coats was injured while on board the Penrod 69, a jackup drilling rig owned and operated by Penrod Drilling Corporation ("Penrod"). At the time of the acciden t, there is no doubt that the Penrod 69 was "located in the Port of Mina Saqr in the territorial waters of the United Arab Emirates." In my judgment, the fact that the rig was "in port" has crucial significance in this case, because it makes xi clear that the vessel was within the boundary recognized for i nternational law purposes as the boundary of the United Arab Emirates and within what would be referred to under United States nomenclature as the "inland waters" of Ra's Al Khaymah, the particula r emirate in which that port is located. The Penrod 69 was within the inland waters of Ra's Al Khaymah just like a jackup rig in the Port of Galveston is considered to be within the inland waters of the State of Texas, and like a rig in the Port of Biloxi is within the inland waters of the State of Mississippi.

Furthermore, the Penr od 69 had been "in port" for some eight or nine months prior to the date of Coats' injury. The records are clear that on August 12, 1987, the Penrod 69 was surveyed for its annual condition certificate, and at that time, the survey report indicates, the "vessel lay jacked-up" in this port. The Penrod 69 was out of service, deactivated, not operated, and not occupied by any personnel other than a watchman, up until January 1988, when as a result of a new contract for the rig's use in a Persian Gulf drilling activity, Penrod commenced the task of preparing the Penrod 69 to g o back into service. During this interval of deactivation, the Penrod 69 functioned solely as an a rtificial wharf or dock for the purpose of storing the equipment and facilities thereon, with its legs standing on the bottom of the port and its hull up out of the water. Substantial repairs, replacements and refurbishing activities were required to prepare the Penrod 69 to resume its offshore drilling function. This work took some four months to accomplish and included the installation xii of a new derrick. In performing the refurbishing work, Penrod used its own personnel (assumptively the crew of the Penrod 69) and other categories of "contract labor, catering, and service personnel." Penrod hired MIS to assist in the refurbishing work, and MIS designated Coats to operate the MIS pump that was brought on board to provide pressure to test certain pressurized systems of the rig. The daily reports as to the personnel working on board the rig, which are in the record, reflect that the total number of contract labor, catering, and service personnel always exceeded the number of Penrod personnel. The record does not clearly indicate whether on the date of the injury , April 12, 1988, the Penrod 69 was still in a "jacked-up" position, or whether its hull had been lowered into the water. Obviously, if it was still in a jacked-up position, its categorization as a "vessel" in navigation is in serious doubt. Even if it had been lowered into the water, however, the nature and extent of the work going on, and the number of outside personnel deploye d in such work, clearly demonstrate that the repair and refurbishing activities were beyond the capacity of the "crew" of the Penrod 69 to accomplish, and that such work could be accomplished only with the ready availability and access of shore-based personnel and facilities. In my view, under these facts, the "place of the wrongful act" element of the Lauritzen - Rhoditis factors should be given more than just the "considerable weight" that the majority gave it. It should be the controlling factor in the choice of law decision. I have looked for and have been unable to find any Supreme Court decision or xiii Fifth Circuit decision applying United States law to reso lve the claim of a shore-side worker injured while assisting in the refurbishing of a jacked-up drilling rig while it was located within the inland waters of anoth er nation. In my view, the majority opinion constitutes an unjustifiable extension of United States law into areas where simple comity amo ng nations requires that the law of the place of the casualty apply.

My second area of disagreement with the panel regarding the Lauritzen - Rhoditis factor s concerns the factor of "allegiance of the defendant shipowner." I do not quarrel with the majority's determination that the allegiance of Penrod, as owner of the Penrod 6 9, is to the United States. But, in my view, the factor of "allegiance of the defendant shipowner" has materiality only in the circumstance where the flag of the vessel and the allegiance of the defendant shipowner are different (i.e. the vessel's flag is a flag of convenience), and the law of the nation of allegian ce of the defendant shipowner can appropria tely be applied to the determination of rights between that shipowner and his seaman employee when that vessel is engaged in international commerce. In this case, however, the allegiance of the defendant shipowner is an inconsequential factor for two reasons: First, the Penrod 69 is documented under the United States flag; Penrod's allegiance is to the United States and there is no flag of convenience involved.

Secondly, and more importantly, bot h the district court and the majority opinion recognize that there was no employment relationship as seaman or otherwise between Penrod and Coats. The xiv majority's use of the allegiance of the defendant shipowner as a factor in tipping the scales in favor of application of United States law would, in my judgment, be improvident even if the only defendant in this case were Penrod, because that factor should be applied only where there is an employment relationship between the injured plaintiff and the defendant shipowner. But Penrod is not the only defendant in this case, and the other defendant, MIS, is not a shipowner; it is an entity which was created by and whose allegiance is owed to the laws of the United Arab Emirates, and it is in fact the employer of Coats. T he majority gives no seri ous consideration to the key distinctions in this case (1) that Coats was not an employee of the defendant shipowner, Penrod, but was an employee of MIS; and (2) that the trial court found that Coats was not a Jones A ct seaman of Penrod. I suggest that the L auritzen - Rhoditis factors assume an employment relationshi p between the injured seaman-plaintiff and the defendant shipowner, and that when that relationship does not exist, the allegiance of the shipowner should be considered less sig nificant than that of the defendant e mployer. I have looked and have not found any Supreme Court or any Fifth Circuit decision applying United States law to determine the rights and obligations between a United States citizen injured in a foreign country during the course and scope of his employment with a corporate entity organized under that foreign country's law. In my judgment, the panel opinion improvidently extends United States la w to the set of circumstances involved in this case by xv giving greater weight to the allegiance of the defendant shipowner than to the allegiance of the defendant employer.

Finally, I question the correctness of the panel decision in evaluating the "place of contract" factor in the Lauritzen - Rhoditis analysis. Here again, the majority misconstrues the significance of this factor. I start out with the language used by the Supreme Court in concluding its discussion of this factor in Lauritzen itself: "We do not thi nk the place of contract is a substantial influence in the choice between competing laws to govern a maritime tort. " Lauritzen , 345 U.S. at 589 (emphasis added). Furthermore, the contract referred to in both Lauritzen and Rhoditis is the contract of employment between the injured s eaman plaintiff and the defendant ship owner. There is no such contract between Coats and Penrod in this case; whatever contract of employment there was in this case existed between Coats and MIS, who did not own any vessel and was essentially a shore-based supplier of services to companies engaged in exploration and development of oil and gas. I recognize that Coats was recruited by representatives of MIS at his home in Mississippi and that the basic terms of his employment agreement were verbally negotiated and orally a greed upon during this recruitment visit. However, it is clear beyond doubt that he was recruited and "em ployed" to work for MIS, not for Penrod, and to work in the United Arab Emirates, not aboard any particular vessel.

Furthermore, it is clear that in order for Coats to get the necessary visa to enter the United Arab Emirates, Coats and MIS xvi "executed an Arabic contract," and that Coats then applied for and received the necessary work permit from the UAE which permitted him to reside ashore there in the UAE during his employment. The record is clear that Coats performed his duties for MIS at locations of oil and gas wells on shore in the UAE as wel l as offshore in the Persian Gulf, and at warehouses and d ockside facilities in the UAE. This existence of a work permit is a special factor present in this case which has not been present in any of the other choice of law cases cited in the majority opinion.

Presence in the UAE and acceptance of a UAE work permit would unquestionably subject Coats to the criminal laws and civil laws of the UAE had his injury occurred on land. In my view, Coats' acceptance of a work permit necessitates a determination that the law of the UAE should apply to an injury occurring on the waters of a UAE port during employment under that UAE work permit.

In his original appellee's brief, Coats argued: "U.S. Maritime Law applies whenever a U.S. citizen i s injured on a U.S. flag drilling vessel anywhere in the world." (p. 52). The cases cited b y Coats for that proposition do not support his assertion. But the majority opinion in effect arrives at the same conclusion by misinterpretation and misevaluation of the Lauritzen - Rhoditis factors. Because I think such a conclusion is bad law under the facts of this case, and that it will produce undesirable effects when applied as a precedent, I would reverse the district court's choice of law determination and remand the case to the district xvii court for retrial in accordance with the laws of the United Arab Emirates.

In arri ving at this result, I rely on the following line of Fifth Circuit cases: C hiazor v. Transworld Drilling Co., Ltd. , 648 F.2d 1015 (5th Cir. 1981), cert. denied , 455 U.S. 1019 (1982); Zekic v. Reading & Bates Drilling Co. , 680 F.2d 1107 (5th Cir. 1982); Bailey v. Dolphin Intern., Inc. , 697 F.2d 1268 (5th Cir. 1983); Koke v. Phillips Petroleum Co. , 730 F.2d 211 (5th Cir. 1984); Schexnider v. McDe rmott Intern., Inc. , 817 F.2d 1159 (5th Cir. 1987), cert. denied , 484 U.S. 977 (1987); and Fogleman v. Aramco , 920 F.2d 278 (5th Cir. 1991). All of these cases involve "nontraditional" vessels similar in nature and function to the Penrod 69, and all of these determi ned that the law of another nation, other than the United States, applied.

II. What United States Law Having decided that United States law shall apply, the district court expressly reached the following conclusions (which the majority opinion inferentially affirms): 1. That the United States Longshoremen and Harborworkers Compensation Act ("LHWCA") could not apply because it applies only to injuries or death occurring "on navigable waters of the United States." 2. That the Jones Act was in applicable (i) because there was no employment relationship between the 86 Penrod's appeal from these choice of law decisions by the district court clearly raises the propriety of Coats' status as a " Sieracki seaman" for our determination, even though, as Judge Garwood notes in his dissent, there was no specific objection raised by Penrod to the submission of an issue on unseaworthiness and to the joint submission of negligence and unseaworthiness in the jury issues as to percentages of fault.

When a party contests and objects to a trial court's choice of law determination, I can see no need for repetitious and futile objections to the implementation by the trial court of its choice of law determinations during the trial. xviii plaintiff Coats and Penrod, the owner and operator of the Penrod 69; (ii) because Coats was aboard the Penrod 69 only for one day, the day he was injured, and therefore had no permanent connection to that vessel; and (iii) because there was no common owner ship or control by his employer, MIS, of any of the six offshore drilling rigs on which Coats worked during his employment. 3. That the general maritime law of the United States would be applied, including specifically the concept of Sieracki seaman status originating in the case of Seas Shipping Co. v. Sieracki , 328 U.S. 85 (1946), with its concomitant availability of the warranty of unseaworthiness for the benefit of Coats. 86 The judicially manufactured concept of Sieracki seaman 's status remained a vibrant part of United States law from its creation in 1946 until Congress passed the 1972 amendments to the LHWCA, which expressly removed the right of recovery under the warr anty of seaworthiness for individuals covered by that Act. The district xix court recognized some disagreement in the law as to whether the 1972 amendments to the LHWCA also abolis hed Sieracki relief for individuals no t covered by the LHWCA. But relying on two Fifth Circuit cases, Aparicio v. Swan Lake , 643 F.2d 1109 (5th Cir. 1981), and Cormier v. Oceanic Contractors, Inc. , 696 F.2d 1112 (5th Cir.) cert. denied , 464 U.S. 821 (1983), the district court concluded that, "Coats comes squarely within one of the so-called `pockets of Sieracki seamen remaining after the 1972 amendments.' Aparicio , 643 F.2d at 1118 n.17." Accordingly, the district court allowed Coats to proc eed under the general maritime law against Penrod on both negligence and unseaworthiness theories. I disagree with the maj ority's affirmance of the district court, (1) because I believe the Fifth Ci rcuit precedents relied upon by the district court can no longer be supported in light of the policies stated by a unanimous Supreme Court in Miles v. Apex Marine Corp. , 498 U.S. 19 (1990); and (2) because the majority opinion wholly ignores the impact of M iles v. Apex Marine on the su bstantive content of United States g eneral maritime law, even though we were sitting en banc and were in a position to consider such impact. For these two reasons, I respectfully register my dissent. In my view, M iles v .

Apex Marine constitutes a major restatement by a unanimous Supreme Court as to th e role to be played by the federal judiciary in defining the substantive content of general maritime law. I recognize that the holdings in M iles v. Apex Marine relate only to the specific issues of whether the cause of action for wrongful death of a seaman exists under general 87 In another en banc case, this court relied upon the broad policy implications of Miles v. Apex Marine to overturn prior precedents regarding recovery of punitive damages in cases involving claims of willful nonpayment of maintenance and cure.

Guevara v. Maritime Overseas Corporation , No. 92-4711, F.3d, 1995 WL 437211 (5th Cir. July 26, 1995) (en banc). xx maritime law and whether recovery for loss of society in a general maritime wrongful death actio n would be permitted. However, the statements of philosophy and approach as to the role of courts in eliminating "anomalies" and achieving "uniformity in the exercise of admiralty jurisdiction" constitute a major redefinition of the interplay between the role of Congress and the role of the courts in defining general maritime law. 87 Because of their extreme relevance to the issue before us in this case, I cite four passages from Miles v. Apex Marine that clearly set forth the new approach and philosophy: We no longer live in an era when seamen and their loved ones must look primarily to the courts as a source of substantive legal protection from injury and death; Congress and the States have legislated extensively in these areas. In this era, an admiralty court should look primarily to these legislative enactments for policy guidance. We may supplement these statutory remedies where doing so would achieve the uniform vindication of such policies consistent with our constitutional mandate, but we must also keep strictly within the limits imposed by Congress. Congress retains superior authority in these matters, and an admiralty court must be vigilant not to overstep the wellconsidered bo undaries imposed by federal legislation. These statutes both direct and delimit our actions.... The general maritime claim here alleged that Torregano had been killed as a resu lt of the unseaworthiness of the vessel. It would be inconsistent with our place in the xxi constitutional scheme were we to sanction more expansive remedie s in a judicially-created cause of action in which liability is without fault th an Congress has allowed in cases of death resulting from negligence. ...

We sail in occupied waters. Maritime tort law is now dominated by federal statute, and we are not free to expand remedies at will simply because it might work to the benefit of seamen and those dependent upon them. ...

Congress has limited the survival right for s eamen's injuries resulting from negligence .

As with loss of society in wrongfu l death actions, this forecloses more expansive remedies in a general maritime action founded on strict liability. We will not create, under our admiralty powers, a remedy disfavored by a clear majority of t he States and that goes well bey ond the limits of Congress' ordered system of recovery for seamen's injury and death....

Apex , 498 U.S. at 27-36.

The differences in jurisprudential outlook between Aparicio and Apex are the differences between night and day. A paricio considers judge-made mari time law to be paramount and requires statutory changes to expressly cover all possible circumstances to be effective; A pex recognizes constitutional limitations to the s cope of judge-made law and requires accommodation of judge-made law to statutory policy from similar, though not identical, areas of the law. Aparicio looks for and encourages the recognition of "pockets" where judge-made maritime law can survi v e statutory changes; A pex abhors anomalies and encourages the tailoring a n d adjus ting of maritime law to promote uniformity of rights and remedies. A paricio puts the burdens on Congress to speak to the intended scope of its 1972 amendments to the LHWCA Act; Apex puts xxii the burden on the federal courts to construe the continued vitality of the Sieracki doctrine in a manner consistent with the 1972 amendments.

I am disappointed that my colleagues in the majority of this en banc consideration failed to follow the counsel of Apex and to seize the opportunity to make United States maritime law applicable to the casualty i n this case the same as the law that would have been applicable had this casualty occurred in United States waters.

What reason in logic or good public policy is t here for federal judges to extend the benefits of the warranty of unseaworthiness to a United States citizen working as a longshoreman or harborworker in a foreign port when that same warranty of unseaworthiness was statutorily removed as a protection for United States citizens working as longshoremen and harborworkers in United States waters? The precedential effect of the majority's decisio n will be truly dramatic, for once this remedy is established for the benefit of United States citizens working as longshoremen and harborworkers in fo reign ports, there will not be any logical reason to deny the extension of the warranty of seaworthin ess to citizens of other nations worki ng in those same foreign ports on United States vessels. And as a result, the United States courts will become the preferred forum for every worker who is injured on board a United States vessel in foreign ports