OPINION OF THE COURT

A week before trial of a negligence action against two defendants, plaintiff executed a document in favor of one defendant entitled “Covenant Not To Sue”, accepted $9,750 from him, and on the eve of trial withdrew his complaint as against that defendant. The other defendant, appellant in this case, moved for production of the document and amended his answer to allege that plaintiff had been fully compensated.

At trial the document was not shown to the jury or alluded to. The court refused to allow defendant’s counsel to ask plaintiff if he had received payment. The jury returned a verdict for plaintiff against both defendants in the amount of $50,000.^[1] The court then reduced the verdict by $9,750, the amount set forth in the so-called “Covenant Not To Sue”.

The first issue posed by this appeal is whether, as a matter of law, the covenant not to sue ran only to appellant’s co-defendant. Appellant contends that the document clearly ran to him and, if not, a question of fact as to the intent of the parties should have been presented to the jury. On the record before us it does not appear that appellant successfully preserved his right to press the latter position here, but that is of no moment in the light of our conclusion.

We conclude that the trial judge properly construed the document as a covenant running only to the named defendant. The relevant portions of the document provided:

“For * * * $9750.00, * * * the undersigned does hereby covenant * * * to forever refrain from * * * prosecuting or in any way aiding any claim or suit against Anderson Stanley * * * and all agents, successors and assigns of his, and all other persons, firms and corporations for whose acts or to whom they or any of them might be liable, (hereinafter referred to as said parties), [on account of the accident in question] * * *.

“For said consideration it is hereby further agreed: * * * that all rights which the undersigned may have to proceed against parties other than said parties hereto are expressly reserved; * * *.” (Emphasis in the original.)

The document is not an example of adept draftsmanship. The reference to “successors” is one example of inappropriate language. But the language that causes the difficulty is that which appellant seizes on to bring himself within the scope of plaintiff’s undertaking. He argues that, since plaintiff agreed to refrain from suing a person “to whom * * * [Stanley] might be liable”, and since Stanley as a joint tort-feasor would be liable to appellant for contribution (in[*467] deed, would be liable to no other person), appellant is included in the class sought to be protected. Appellant also argues that, since the reservation clause uses the words “said parties”, he is also included in the exception to the reservation.

Were this the correct interpretation of the document, the most prominent part— the italicized clause reserving rights to proceed against “parties other than said parties” — would have no meaning whatsoever. We think the trial court correctly interpreted the document as a covenant running only to the defendant named therein. Were extrinsic evidence relevant, it would be supplied by plaintiff’s action in attempting to withdraw his action against Stanley only.

The second issue in the case is whether it was error for the judge not to have allowed counsel to inquire on the issue whether plaintiff considered his settlement with Stanley as full satisfaction. Appellant places justifiable stress on the opinion of Associate Justice Rutledge in McKenna v. Austin, D.C.Cir., 1943, 77 U.S.App.D.C. 228, 134 F.2d 659. He fairly quoted the following language;

“Whether the settlement is made and accepted as full satisfaction or merely as the best obtainable compromise for the settler’s liability is the crucial issue, and ordinarily one of fact. If however the agreement’s terms leave no room for doubt, the decision should be made as a matter of law.” Id. at 664.

In this case, as we have indicated in discussing the first issue, we say that the document itself leaves no room for doubt. The judge properly ruled in effect that under the facts of this case, plaintiff, as a matter of law, did not accept the $9,750 (or the promise to pay this amount) in full satisfaction of his injuries.

Affirmed.

On Petition for Rehearing

Following our decision on the appeal, a petition for rehearing was filed urging us to consider the further question: to what extent, if at all, should the judgment against appellant be reduced to reflect any right of contribution he might have. To help us answer this question we requested and received additional briefs from counsel.

We approach the issue with full recognition that the result we reach should be consistent with the long-standing policy considerations underlying any system of contribution: the encouragement of voluntary settlements, the equitable distribution of liability among joint tort-feasors, and the elimination or minimization of collusion.

The first question is whether a right of contribution generally exists among joint tortfeasors. That no such right is recognized at common law has long been clear. See, e. g., Merryweather v. Nixan, 101 Eng.Rep. 1337 (1799). Nor do we find reference to such a right in the statutory or reported decisional law of the Virgin Islands. After extended consideration, however, we have decided that there is no compelling reason to defer or deny applying what we deem a rule of reasoned fairness.^[1] There is no longer a legitimate place in our system, if, indeed, there ever was, for a rule of law which places the full burden of restitution upon one who is only in part responsible for a plaintiff’s loss. Such a philosophy stems from a more ancient day which more easily intermingled[*468] penal and compensatory considerations. See generally Prosser, Torts § 47 (3d ed. 1964).

Our next inquiry is whether a covenant not to sue obtained by one tortfeasor should operate as a bar to an action for contribution against him by a joint tortfeasor. This question, too, is easily answered. Voluntary settlements are to be encouraged and a rule permitting contribution under such circumstances would not work to that end. Not only would a joint tortfeasor defendant be stripped of any incentive to settle but he would have a positive incentive to stand trial and actively participate in his defense in order to minimize his liability.

We state our views on these two questions to make clear our interest in a system which provides for the equitable distribution of liability for joint fault and yet encourages out-of-court settlements. We recognize that a rule which maximizes one of these goals must to some extent work against the other. Our task is to harmonize them as best we can in deciding upon a formula to determine the amount owed a successful plaintiff by a non-settling defendant joint tortfeasor.

There are several options open to us. We might, for one, consider him liable for the full amount of the verdict irrespective of any amount received by the plaintiff as consideration for a covenant not to sue. Or we might subtract from the verdict an amount representing the settlement figure and allow plaintiff the difference, as did the district court in this case. Should we adopt the pro rata reduction rule^[2] we would view settlement as a partial sale of plaintiff’s cause of action, albeit for perhaps less than what it is worth, in an amount representing the fraction.

No. of settling tortfeasors

-. Finally,

No. of tortfeasors

we could order that special questions be put to the jury to determine comparative negligence among joint tortfeasors and assess judgment against non-settling defendants on the basis of a percentage of the verdict.

The first two approaches we reject principally because they leave the field of settlement very much open to collusive arrangement between a plaintiff and a favored joint tortfeasor. This is especially so where, as we have held here, no right of contribution exists against settling tortfeasors.^[3] Of the two remaining choices we are persuaded that a system which taxes non-settling tortfeasors to the extent of their negligence, and no further, is the more equitable.

The fairness of such a system is evident. The troublesome aspect in our view is whether its implementation will serve to encourage or discourage settlement prior to trial. The argument made is that a plaintiff, unable to know with any certainty before verdict the extent of his[*469] recovery, will be reluctant to discharge one or more of several tortfeasors from liability. We find this circumstance of no greater weight under a comparative negligence system than under a pro rata reduction rule. In each case a plaintiff negotiates settlement largely on the basis of an estimated gross recovery and the negotiating tortfeasor’s legal liability. Thus, in a pro rata reduction jurisdiction, negotiating parties will bargain high or low depending upon what they consider to be the likelihood of a jury finding of negligence on the part of the alleged tortfeasor, and the effect such a finding will have in reducing the amount plaintiff may realize on the verdict from non-settling tortfeasors. In a comparative negligence jurisdiction negotiating parties will bargain high or low depending upon what they consider to be the extent of the alleged tortfeasor’s negligence. Each system presupposes an element of chance, and, depending on a given set of facts, will speed settlement or make it more difficult to achieve. See Bielski v. Schulze, 16 Wis.2d 1, 114 N.W.2d 105, 111 (1962). On balance the equities inherent in a comparative negligence system convince us of its ultimate merit.

We might add that the introduction of the doctrine of comparative negligence to a settlement and contribution context is not novel. The Uniform Act of 1939 recognized it to a limited extent in an optional paragraph of section 2 which reads: “When there is such a disproportion of fault among joint tortfeasors as to render inequitable an equal distribution among them of the common liability by contribution, the relative degrees of fault of the joint tortfeasors shall be considered in determining their pro rata shares.” And some states have enacted into law the theory of comparative negligence as among joint tortfeasors. See, e. g., Ark.Stat.Ann. § 32-1002(4) (1962); Hawaii Rev.Laws § 246-11 (1955); S.D. Code § 33.04A03(4) (Supp.1960).

Nor by announcing such a rule judicially do we think we display a particular boldness.^[4] The Supreme Court of Wisconsin took the important step before us in Bielski v. Schulze, supra. Concededly, the Wisconsin court emphasized how natural such a step was, given its long experience with statutory comparative negligence as between plaintiffs and defendants and its possession of procedural and practical devices allowing for meaningful implementation. See also Note, 63 Colum.L.Rev. 380 (1963). But we do not view the functioning of juries in Wisconsin and the Virgin Islands as so different that judicial pronouncement of the comparative negligence doctrine, successful in one, must be unworkable in the other. Indeed, the Wisconsin comparative negligence statute, Wis.Stat.Ann. § 331.045 (1958), limited as it is to the effect on defendant’s liability of plaintiff’s contributory negligence, does no more than legislate the simple idea of partitioning fault. The relevant background, then, is not statutory but decisional, as evidenced by the some seventy pages of annotations following the statement of the act.^[5]

The Virgin Islands is certainly as capable of careful judicial developement in the future. And as for procedural devices necessary to the implementation of such a rule, we note that the all-important special verdict is a long-standing and familiar tool in the federal workchest. Fed.R.Civ.P. 49(a).

In applying the rule which we have announced, imposing liability upon joint tortfeasors in proportion to their[*470] comparative negligence, it is necessary that special questions be stated and submitted to the jury, in a case tried to a jury, devised to determine the degrees of fault of the several alleged tortfeasors. This not having been done here, there are several alternative dispositions. We could affirm the action of the court in reducing the judgment by the amount of the settlement, but there is no special sanctity in or reliance by the parties on such a resolution. We could apply to this case the pro rata rule, ordering that the judgment against appellant be reduced to one half the amount of the verdict. But there is no special merit in adopting for such cases as have gone to verdict and are presently in the process of appeal a rule which we have rejected as bearing no necessary relationship to the degrees of fault of joint tortfeasors.

Finally, there exist several possibilities in remanding for a new trial. A complete new trial on all issues would not seem to us to be fair to plaintiff-appellee, who has already prevailed in establishing his freedom from contributory negligence, his damages, and defendant-appellant’s negligence. Nor would it seem necessary to reopen the issue of damages, there being no contention that these are excessive. The only issue which our present opinion has affected is the allocation of damages. While trial on this issue alone may present some problems, we think that fairness compels such a restriction.

We accordingly grant the petition, reverse the judgment of the district court, and remand the cause for a new trial limited to determining, under proper instructions, the proportion of the damages for which appellant should be held responsible. If his negligence is found to be less than the percentage which the verdict diminished by the settlement bears to the total verdict, the judgment against him shall be correspondingly reduced. If more, it shall not be increased, since to do so would in effect give appellee more than the total damages to which he is entitled.