FERREN, Associate Judge:

This case presents two principal questions: (1) When all parties stipulate to a settling defendant's liability in a common law tort action (here a claim for medical malpractice) — without formal adjudication of that liability — is the settling defendant a "joint tortfeasor" for purposes of determining the allowable credit to a nonsettling defendant against whom the jury has awarded full damages for the plaintiff's injury? (2) If so, will Martello v. Hawley, 112 U.S.App. D.C. 129, 300 F.2d 721 (1962), control, limiting the nonsettling defendant to a pro rata credit, even when the amount of the settlement is more than half the jury's verdict? Or will the "one satisfaction rule" applied in Snowden v. D.C. Transit Sys., Inc., 147 U.S.App.D.C. 204, 454 F.2d 1047 (1971), entitle the nonsettling defendant to a dollar-for-dollar ( pro tanto) credit when the settlement exceeds half the verdict, in order to limit the plaintiff's recovery to no more than the jury's valuation of the claim?

We conclude that such uncontested stipulation of a settling defendant's liability will be as effective as an adjudication of liability to make the defendant a joint tortfeasor for purposes of applying the appropriate credit. We further conclude that, whenever the plaintiff settles with a joint tortfeasor, the nonsettling defendant shall receive a pro rata credit under Martello, reflecting the defendant's equitable right to contribution and no more, even when the plaintiff's recovery from all defendants will exceed the amount of the verdict and thus violate the one satisfaction rule.

The trial court correctly awarded the nonsettling defendant a pro tanto credit for the amount paid by a group of settling defendants that was not a joint tortfeasor, but the court erred in also applying a pro tanto, rather than pro rata, credit for the amount paid by the settling joint tortfeasor. We therefore affirm in part, reverse in part, and remand for recalculation of damages payable by appellee Footer, the nonsettling defendant.

I.

Patricia Berg died of cervical cancer in June 1990. R. Christian Berg, as Patricia Berg's personal representative, brought a negligence action against Ms. Berg's gynecologist, Marvin Footer, M.D., the George Washington University Medical Center (GWU Medical Center), and a group of defendants connected with the Cancer Cytology Medical Laboratory (the laboratory defendants). Berg alleged that the defendants' negligent medical treatment delayed the diagnosis of cervical cancer and caused Patricia Berg's premature death.

Dr. Footer performed a surgical procedure called a fractional dilatation and curettage on Patricia Berg at GWU Medical Center in June 1987, in order to locate the cause of vaginal bleeding Ms. Berg was experiencing between menstrual periods. When performing the fractional dilatation and curettage, Footer was supposed to obtain separate tissue specimens from the cervix and from inside the uterus. It is undisputed on appeal that Footer negligently submitted the two tissue specimens to the pathologists at GWU Medical Center as a single uterine specimen. The pathologists read the specimen as normal, although it was later discovered that the cervical tissue mixed into the specimen showed an abnormal precancerous condition called dysplasia. It is further undisputed that, if the dysplasia had been detected in 1987, the possibility of eradicating the precancerous lesion — and curing Patricia Berg — would have been extremely high. The two-year delay in diagnosing Ms. Berg's condition, however, caused her condition to go from curable to fatal.

In the Joint Pretrial Statement, signed by counsel for all the parties, GWU Medical Center admitted that, despite Dr. Footer's negligence in submitting the specimens, its own "pathologist should have detected that Mrs. Berg's biopsy specimen on June 4, 1987 showed that she had a pre-invasive condition known as dysplasia" and that "the failure to detect this condition, combined with the conduct of [the other defendants], was a substantial factor in the delay of two years in the detection of Mrs. Berg's cervical cancer."

As to the laboratory defendants, Berg alleged they negligently had read and reported the Pap smears Dr. Footer had taken from Patricia Berg on five separate occasions during 1987 and 1988. Berg contended that these defendants had interpreted the Pap smears as normal when they showed marked inflammation or dysplasia, or at least "should have been reported as lacking sufficient cells for evaluation with a recommendation for a repeat smear."

Berg reached settlement agreements with GWU Medical Center and the laboratory defendants but not with Dr. Footer. The case proceeded to trial against Footer alone. The doctor did not implead either settling defendant to seek contribution. Because this is not a comparative negligence jurisdiction, neither the jury nor the judge was asked to make findings ascribing proportionate fault to Dr. Footer, to GWU Medical Center, or to the laboratory defendants. After trial, the jury found that Dr. Footer's negligent failure to detect Patricia Berg's precancerous condition had caused her premature death and awarded the plaintiff $1,406,071 in damages.

In Berg's settlement agreement with the laboratory defendants, the defendants agreed to pay Berg $150,000 but denied every allegation of negligence and wrongdoing. The agreement also provided that Berg would indemnify the laboratory defendants for any contribution or indemnification actions brought by Footer.

GWU Medical Center's offer to settle included an immediate payment to Berg of $800,000 and an additional $200,000 "if after good faith efforts to collect full damages from the remaining defendants by way of settlement or verdict, plaintiff's total recovery is less than $1.5 million." GWU Medical Center and Berg further agreed that Berg would indemnify the Medical Center for all sums the Medical Center was ordered to pay in contribution or indemnification actions.

In response to the jury's verdict, Dr. Footer filed a Motion to Alter or Amend Judgment pursuant to Super.Ct.Civ.R. 59(e), requesting dollar-for-dollar ( pro tanto) credits totalling $1,150,000 for Berg's settlements with GWU Medical Center and the laboratory defendants. The trial court agreed. The court ruled that the jury's award of $1,406,071 represented " all damages to be awarded Plaintiffs" and stressed that "the Court knows of no authority in this jurisdiction that would allow" a plaintiff to recover more than that amount. The trial court accordingly ordered that Footer receive credits for the $150,000 paid by the laboratory defendants and the $800,000 paid by GWU Medical Center. The court also credited the additional $200,000 specified under the settlement agreement with GWU Medical Center because "Plaintiffs' total recovery determined by the jury was less than $1.5 million." The trial court then concluded that Dr. Footer owed $256,071 after receiving the $1,150,000 in credits due.

GWU Medical Center was granted leave to intervene, and Berg and the Medical Center noted their appeals, contending that Dr. Footer, as a nonsettling defendant, was only entitled to credits totaling $778,035.50, represented by the $150,000 settlement with the laboratory defendants plus $628,035.50 for the GWU Medical Center settlement, reflecting GWU's pro rata (50%) share of the $1,406,071 jury verdict after subtraction of the $150,000. According to Berg and the Medical Center, Berg is entitled to $1,503,035.50 comprised of $150,000 from the laboratory defendants, $800,000 (not $1,000,000) from GWU Medical Center, and $628,035.50 (not $256,071) from Dr. Footer.

How appellants arrive at these results, and why we conclude that they are correct, comprise the analysis we now present.

II.

How to credit the judgment entered upon a jury verdict against a nonsettling defendant with the proceeds a settling defendant paid to the plaintiff is purely a question of law, subject to this court's review de novo. See Bertram v. Freeport McMoran, Inc., 35 F.3d 1008, 1019 (5th Cir. 1994) ( de novo appellate review of contribution and credit issues). We therefore proceed without deference to the trial court's analysis.

A.

Until almost mid-century in this jurisdiction, a plaintiff's settlement with one of several joint tortfeasors extinguished the claim against all others on the ground that there was a single injury for which the injured party was entitled to but one satisfaction. See Kaplowitz v. Kay, 63 App.D.C. 178, 179, 70 F.2d 782, 783 (1934). In 1943, however, the federal court of appeals — which at this time had the last word — abandoned this extinguishment rule in McKenna v. Austin, 77 U.S.App.D.C. 228, 134 F.2d 659 (1943). Judge Rutledge, writing for a divided court, declared that "[p]artial satisfaction taken in compromise and release of liability of one or some of the wrongdoers does not discharge the others." Id. at 234, 134 F.2d at 665. Once a plaintiff who had settled with one tortfeasor could sue a nonsettling tortfeasor, however, the court had to resolve how much credit, if any, the nonsettling defendant should receive for the settlement when the jury returned a plaintiff's verdict. That question has resulted in endless debate here and elsewhere.

The United States Court of Appeals for the District of Columbia Circuit, in cases this court has applied many times, has announced two rules for awarding credits to a nonsettling defendant found liable for a plaintiff's injuries. The first rule applies when the plaintiff settles with one defendant, later found to be a joint tortfeasor, and successfully proceeds to trial against a second defendant. The court awards the nonsettling defendant a pro rata (50%) credit against the verdict, representing the right to contribution. For example, in Martello the plaintiff automobile passenger injured in a collision with the Martellos settled with his driver's insurer for $700. The jury then awarded him $2,000 damages against the Martellos and expressly found that the plaintiff's driver had been negligent as well. The court of appeals held that the Martellos were entitled to a $1,000 credit representing the driver's pro rata (50%) share of the legal responsibility for the injury.

[I]n the factual circumstances of this case . . . when settlement is made with one joint tort-feasor and later a verdict is obtained against the other, and the jury finds that the settling tort-feasor should contribute, then the verdict should be credited with one-half its total amount and the defendant tort-feasor should be required to pay only the remaining balance, namely, one-half the total original verdict. It is true that in a case like the present one, where the verdict figure is in excess of twice the settlement figure, the application of this formula will necessarily reduce the amount of the injured plaintiff's recovery. However, the answer to objection to the formula on this account is that, by his [or her] settlement, the plaintiff has sold one-half of his [or her] claim for damages.

Martello, 112 U.S.App.D.C. at 132, 300 F.2d at 724. A pro rata credit is appropriate, therefore, as "a substitute for the non-settling defendant's actual claim for contribution that persists after the dismissal of the principal claim against a settling defendant." Washington v. Washington Hosp. Ctr., 579 A.2d 177, 187 (D.C. 1990).

Martello, however, provides no basis for awarding a pro rata credit when the settling defendant's liability has not been determined, because the nonsettling defendant may be the sole legal cause of the plaintiff's injury and, if so, will not be entitled to contribution. But this does not mean the nonsettling defendant will have to pay 100% of the judgment without credit for payments to the plaintiff by the settling defendant whose liability has not been adjudicated. A separate rule — premised on an altogether different rationale — applies in that case.

When the plaintiff has settled with a party whose culpability has not been determined, or with a party whom the finder of fact has found not liable, the court awards the nonsettling defendant a credit against the verdict in the amount of the settlement, dollar-for-dollar ( pro tanto). See Snowden, 147 U.S.App. D.C. at 205-06, 454 F.2d at 1048-49. The reasons for this are less related to the equities affecting defendants than to doctrinal concerns dictating the amount properly recoverable by a plaintiff.

In Snowden, a passenger was injured on a bus that collided with an automobile. During trial against both the bus company and the automobile driver, the plaintiff settled with the driver for $5,000. The jury then returned a plaintiff's verdict against the bus company for $12,500 and exonerated the automobile driver from liability. The court of appeals ruled that the bus company was entitled to a $5,000 credit, despite the fact that the automobile driver was found not to have been a tortfeasor. The court applied the "cardinal principle of law" that traditionally has limited the plaintiff to one recovery for a single injury: "in the absence of punitive damages a plaintiff can recover no more than the loss actually suffered." Id. at 205, 454 F.2d at 1048; see McKenna, 77 U.S.App.D.C. at 233, 134 F.2d at 664; Kaplowitz, 63 App. D.C. at 179, 70 F.2d at 783. In applying this "one satisfaction rule," Kaplowitz, 63 App. D.C. at 179, 70 F.2d at 783, the court rejected plaintiff's argument that the settlement with the driver reflected a gratuity from a collateral source that should not be credited. Settlement was "made under Damoclean pressure, not gratuitously," and it therefore justified a credit. Snowden, 147 U.S.App. D.C. at 206, 454 F.2d at 1049.

The Snowden court adopted the pro tanto credit rule not to lessen the liability of the nonsettling defendant § who had been adjudicated solely liable for the plaintiff's injury § but to mitigate against "potential chicanery on the part of plaintiffs." Id. The court apparently was concerned that a plaintiff might collude with a potentially liable settling defendant to impose the greater liability on another defendant who had greater resources or against whom the plaintiff had more spiteful feelings. See McDermott, Inc., v. AmClyde and River Don Castings, Ltd., ___ U.S. ___, ___ __ ___, 114 S.Ct. 1461, 1467-68, 128 L.Ed.2d 148 (1994); REVISED UNIF. CONTRIBUTION AMONG TORTFEASORS ACT § 4 (1955 Revision), 12 U.L.A. 63, 99 (1975) (Commissioners' Cmt.). If, for example, the plaintiff in Snowden had succeeded in proving that the settling defendant was a joint tortfeasor, the plaintiff's recovery against the non-settling defendant would have been diminished by the pro rata credit for the plaintiff's sale of half the claim to the settling defendant. See Martello, 112 U.S.App.D.C. at 132, 300 F.2d at 724. But because the plaintiff failed to prove the settling defendant's culpability, the plaintiff — absent the Snowden rule — would have recovered the entire amount of the verdict from the non-settling defendant in addition to the proceeds from settlement with the other defendant. The Snowden rule, therefore, reduces the potential anomaly of a plaintiff's "benefiting by losing at trial to tortfeasors with whom he [or she] has settled." Snowden, 147 U.S.App.D.C. at 206, 454 F.2d at 1049. It precludes plaintiffs from gaining a windfall by collecting both (1) settlement monies from defendants whose liability has not been adjudicated, and (2) full damages from nonsettling defendants whose liability has been confirmed. See id.; see also Kane v. Ryan, 596 A.2d 562, 569 (D.C. 1991); Washington, 579 A.2d at 186.

A nonsettling defendant is not necessarily helpless to improve the situation when the other defendant settles. Under the Snowden rule, a nonsettling defendant can attempt to prevent such collusion between the plaintiff and a settling defendant — and thus obtain a pro rata, not a smaller pro tanto, credit — by filing a cross-claim against the settling defendant and proving that this other defendant is a joint tortfeasor. See Washington, 579 A.2d at 186-87; Washington Healthcare Corp. v. Barrow, 531 A.2d 226, 230 (D.C. 1987); Otis Elevator Co. v. Henderson, 514 A.2d 784, 786 (D.C. 1986) (per curiam); see also Hall v. General Motors Corp., 207 U.S.App.D.C. 350, 359, 647 F.2d 175, 184 (1980). A situation can arise, on the other hand, when the non-settling defendant, relying on Snowden (and the one-satisfaction rule), might elect not to pursue contribution because of its estimate that a pro tanto credit would be larger than a pro rata credit.

In both Martello and Snowden, the settlements were for less than half the amounts of the eventual verdicts. Therefore, whether the credit was pro rata or pro tanto, the plaintiff could receive, in toto, no more than the jury awarded; the one satisfaction rule could not have been violated. But what happens when a settlement is for more than half the verdict and the settling defendant is a joint tortfeasor? Will the plaintiff be entitled to Martello's pro rata credit, resulting in recovery from both settlement and judgment of more than the jury awarded? Or, when Martello would give plaintiff a windfall, will the one satisfaction rule interdict the result, converting Martello's pro rata rule normally applicable to settlements with joint tortfeasors into Snowden's pro tanto rule once the settlement gets too large? Neither Martello nor Snowden — nor any other case in this jurisdiction — resolves this collision between the "contribution" and the "one satisfaction" doctrines. We address this issue afresh.

B.

Theoretically, the problem — violation of the one-satisfaction rule — can arise whether the settling defendant is, or is not, a joint tortfeasor if the nonsettling defendant were entitled to receive only a pro rata, not a pro tanto, credit when the settlement exceeds half the eventual verdict. See Rose, 163 U.S.App.D.C. at 248, 501 F.2d at 808-09 (declining, on grounds of unjust enrichment, to award Snowden credit when settlements by two defendants found not to have been joint tortfeasors exceeded their proportionate shares of the verdict). Thus, presumably we could decide, broadly, whether the one-satisfaction issue should be resolved in the same way whether the settling defendant is a joint tortfeasor or not.

We choose not to do so. We have noted the policy differences that generated the Martello and Snowden rules, respectively, in cases where the one-satisfaction rule was not violated. In this particularly complex area of the law, we see no reason to explore whether such distinctions would suggest different treatments in the two situations when that rule is violated. We confine our analysis to the facts here.

Because the Medical Center has stipulated its liability as a joint tortfeasor, we first must resolve whether the Medical Center's stipulation of liability achieves its purpose.

All parties, including Dr. Footer, agree that the Medical Center's stipulation of liability is enough to make it a joint tortfeasor. See supra note 10. We concur. Martello and cases that interpret it speak in terms of adjudicating the settling defendant's liability as a joint tortfeasor, although at least one decision, in addition to noting the litigating defendant's failure to cross-claim for contribution, intimated possible significance in the fact that the settling defendant "expressly did not admit any liability for the accident." Otis, 514 A.2d at 786. In any event, none of our decisions, see supra note 11, in focusing on whether the settling defendant's liability had been adjudicated, ever had to determine whether an all-parties stipulation of the settling defendant's liability would be enough, in lieu of adjudication, to confirm that a Martello, not a Snowden, credit should be applied. The case law focus, therefore, has merely been on whether the settling defendant's liability had been conclusively determined. We see nothing in this court's decisions to preclude our holding that, for purposes of applying the proper credit, a stipulation of the settling defendant's liability by all parties is as effective as an adjudication.

C.

We therefore turn to the dispositive question: when a joint tortfeasor settles for more than half the amount of the eventual verdict so that a Martello ( pro rata) credit for the nonsettling defendant would allow the plaintiff to recover more than the jury found the claim was worth, shall the plaintiff be allowed to recover that full amount? Or does the one satisfaction rule require a pro tanto credit, limiting recovery to the amount of the verdict and thus allowing the nonsettling defendant to pay less than its pro rata share?

We begin by reemphasizing the fact that Martello is the law used to determine the credit for a settlement with a tortfeasor, at least when, as in Martello, that settlement turns out to be less than half the amount of the verdict against the other, nonsettling tortfeasor. We also note that, according to the usual operating assumption based on litigation experience over many years, see McDermott, ___ U.S. at ___, 114 S.Ct. at 1471, the settling tortfeasor typically does settle for less than half the eventual verdict, leaving a plaintiff's total recovery at less than the amount of the jury's verdict after the litigating defendant's pro rata credit has been applied. Therefore, unlike the pro tanto credit regime where the plaintiff is always made whole in the amount of the jury's verdict, the pro rata regime (when, for example, two defendants are involved) keeps the non-settling defendant's contribution to the recovery constant — 50% of the verdict — even when the plaintiff thereby will recover less than 100% of that amount.

The somewhat narrower question, then, is whether the one satisfaction rule poses any sound reason to stop applying Martello, which typically works against plaintiffs and favors nonsettling defendants, on those occasions when the plaintiff — atypically — happens to negotiate a more favorable settlement for half the claim than the settling defendant's pro rata (50%) share of the eventual verdict.

Dr. Footer says "yes." He argues that the rationale underlying the pro tanto credit rule should limit application of the pro rata credit rule when, as in this case, a pro rata credit would allow the plaintiff to recover more than the amount of the jury verdict. The trial judge agreed that Snowden's rule trumped Martello here: "I know of no authority in this jurisdiction to say that a plaintiff can recover more than the amount of the jury verdict." According to the trial court and Dr. Footer, the plaintiff would be unjustly enriched unless a pro tanto credit were awarded.

Appellants, on the other hand, contend that the trial court's allowance of a pro tanto credit in this case, substantially larger than a pro rata credit would be, violates Martello, unjustly enriches the nonsettling defendant (Dr. Footer), and discourages settlement. They point out that if GWU Medical Center had not settled and had been found jointly liable, Dr. Footer would have been entitled to seek contribution for the Medical Center's pro rata share of the damages — and no more. Furthermore, they say, applying the pro tanto credit rewards Dr. Footer for not settling with the plaintiff when all other defendants were willing to do so. Because Dr. Footer knew the terms of GWU Medical Center's settlement with plaintiff Berg, appellants insist that Dr. Footer had reason to know there was a likelihood, perhaps a substantial likelihood, that the settlement would exceed half a reasonably anticipatable verdict. If, under these circumstances, a pro tanto credit were applicable, Dr. Footer would have no incentive to settle, let alone to implead GWU Medical Center in order to seek contribution, because the doctor (more likely his insurer) stood a good chance of paying less than half his pro rata share of the verdict.

Dr. Footer replies that appellant's argument, premised on the nonsettling defendant's windfall from a pro tanto credit, is the very argument the court rejected in Snowden in favor of the one satisfaction rule. If there had been no settlement in Snowden, the nonsettling defendant, judicially determined to be the sole tortfeasor, would have been required to pay the full amount of the verdict. However, because of the one satisfaction rule, the only party adjudicated a wrongdoer, the non-settling defendant, was allowed to pay substantially less than the full damages award. According to Dr. Footer, the fact that GWU Medical Center is a joint tortfeasor, whereas the settling defendant in Snowden was not, affords no principled basis for rejecting the one satisfaction rule here; the injustice appellants' perceive is the same one the court considered and accepted in Snowden.

These arguments show that in any given case the parties are likely to emphasize the equitable principles underlying the Martello rule (Berg and GWU), or the doctrinal principles underpinning Snowden (Footer), depending on how a particular credit rule plays out. If Berg's settlement with GWU Medical Center had been small — less than half the verdict — then the doctor presumably would be advocating a pro rata cap on his liability while plaintiff would be championing pro tanto. We therefore must look beyond this case to ascertain the better approach overall.

A number of courts and scholars have endeavored to develop a universally optimal set of credit rules. See supra note 6; RESTATEMENT (SECOND) OF TORTS § 886A cmt. m (1977). Not long ago, Justice Stevens, writing for a unanimous Supreme Court, analyzed "how a settlement with less than all of the defendants in an admiralty case should affect the liability of nonsettling defendants." McDermott, ___ U.S. at ___, 114 S.Ct. at 1464. The Court appraised the following three principal approaches to determining the amount of a credit the court should award a nonsettling defendant (applied here, for illustrative purposes, to two defendants, one of whom settles):

1. Pro Tanto with Contribution

The nonsettling defendant receives a dollar-for-dollar ( pro tanto) credit in the amount of the settlement, and the nonsettling defendant is permitted to seek contribution from the settling defendant in the same or a separate action.

2. Pro Tanto without Contribution

As in the first alternative, the nonsettling defendant receives a pro tanto credit, but in this case the nonsettling defendant is precluded from pursuing any claim for contribution against the settling defendant. This second alternative often incorporates an additional requirement: a showing that the settlement was made in "good faith." See, e.g., Velsicol Chem. Corp. v. Davidson, 107 Nev. 356, 811 P.2d 561 (1991) (per curiam); Tech-Bilt, Inc. v. Woodward-Clyde Assoc., 38 Cal.3d 488, 213 Cal.Rptr. 256, 698 P.2d 159 (1985).

3. Pro Rata

The nonsettling defendant receives a credit reflecting the settling defendant's equitable share of the legal responsibility for the plaintiff's injury. The question of contribution by the settling defendant becomes moot, because the nonsettling defendant never pays more than its pro rata (50%) share of the judgment.

Each of these alternatives, reflecting a series of approaches by the National Conference of Commissioners on Uniform State Laws, reflects different problems and benefits. The Supreme Court in McDermott evaluated each alternative for (1) promotion of efficient settlements, (2) promotion of judicial economy, and (3) consistency with judicial precedent.

The Court noted that the first alternative is "clearly inferior to the other two." Id. at ___, 114 S.Ct. at 1467. It is ideal in the twofold sense that the plaintiff is made whole (in the amount of the jury verdict) and that, after contribution, each defendant pays an equal ( pro rata) share. But the first alternative tends to discourage settlements because the settling defendant remains exposed to contribution and the settlement thus lacks finality. Most jurisdictions bar the settling defendant from seeking contribution if it settles for more than its proportionate share of the damages. Under this first alternative, therefore, the settling defendant always will end up paying as much or more than it would have paid if it had not settled and gone to trial. See McDermott, ___ U.S. at ___ n. 13, 114 S.Ct. at 1467 n. 13. The plaintiff, of course, could agree to indemnify the settling defendant for any contribution eventually required, but that would not forestall additional litigation or necessarily cover all the additional expenses involved. The failure of this approach to foster settlements led the Commissioners on Uniform State Laws to propose elimination of the nonsettling defendant's right to contribution from the settling defendant. See REVISED UNIF. CONTRIBUTION AMONG TORTFEASORS ACT § 4(b), supra note 15, at 99-100 (Commissioners' Cmt.).

The Supreme Court noted that the choice between the second and third alternatives "is less clear." McDermott, ___ U.S. at ___, 114 S.Ct. at 1467. The second pro tanto alternative, by eliminating the right to contribution, provides all defendants with a greater incentive to settle than the first. Again, the plaintiff is made whole, but this time the defendants — while equally culpable (in this jurisdiction) — almost inevitably will pay different shares of the judgment. This alternative, therefore, can be overly harsh on the nonsettling defendant when the settlement is less — often considerably less — than half the eventual verdict, and the nonsettling defendant is left to pay more — often much more — than its pro rata share.

Some jurisdictions, recognizing that a plaintiff may deliberately settle cheaply with one defendant to obtain a "war chest" of litigation expenses to pursue another (who may have a deeper pocket), have required a "good faith hearing" to assure that the settlement is not so low that the nonsettling defendant will bear an unjustifiable share of the eventual judgment. Such good-faith hearings, while facilitating fair settlements, may add substantially to the expenses of litigation and to the burden on judicial resources that settlements are supposed to avoid — unless, of course, the hearings are too perfunctory to achieve their intended purpose. This second approach, structured to make the injured party whole while assuming the finality of the settlement agreement — but also exposing the nonsettling defendant to the risk of paying much more than its pro rata share — tends to be favored by plaintiffs, as well as by those defendants (and their insurers) inclined to settle. See RESTATEMENT (SECOND) OF TORTS § 886A cmt. m.

The third alternative assures that the nonsettling defendant will pay no more — and no less — than its pro rata share of the verdict, and thus that the settling defendant will not be exposed to a claim for contribution. But when, as in virtually every case, the settlement differs from the settling defendant's pro rata share of the eventual verdict, the plaintiff will receive less — or more — than the jury has decided the claim is worth; the plaintiff will rarely if ever receive exactly what the jury awards. Probably more often than not, however, the plaintiff settles for an amount less than the settling defendant's pro rata share of the eventual verdict. See McDermott, ___ U.S. at ___, 114 S.Ct. at 1471. It does so "because of the uncertainty of recovery at the time of settlement negotiations and because the first settlement normally improves the plaintiff's litigating posture against the nonsettlors." Id. Accordingly, "[t]his solution works against the interest of the injured party and may have the effect of discouraging him [or her] from entering into a settlement." RESTATEMENT (SECOND) OF TORTS § 886A cmt. m. Furthermore, by limiting the litigating defendant's payment to its pro rata share of the verdict, this third alternative tends to be favored by defendants (and their insurers) who are disinclined to settle. Cf McDermott, ___ U.S. at ___, 114 S.Ct. at 1468.

Having concluded that neither the pro tanto (without contribution) rule nor the proportionate share (here pro rata) rule presents a clear advantage in promoting settlements or judicial economy, the Supreme Court next compared the two rules for their consistency with existing judicial precedent. The Court decided that the pro tanto (without contribution) rule would "likely lead to inequitable apportionments of liability" contrary to admiralty case law apportioning damages based on relative fault. Id. at ___, 114 S.Ct. at 1468. The Court, therefore, concluded that "although the arguments for the two approaches are closely matched, we are persuaded that the proportionate share approach is superior," especially given its consistency with the comparative fault rule. Id. at ___, 114 S.Ct. at 1470. The Supreme Court's analysis in McDermott informs our decision today.

If Martello were not the law, recognizing the nonsettling tortfeasor's right (at least when the settlement is less than half the verdict) to a pro rata credit, we might look at this case differently. There is much to be said for an across the board pro tanto (without contribution) rule, coupled with a required showing of a good faith settlement. Under such a rule, whether the settling defendant is a joint tortfeasor or not, (1) settlement pressure on all defendants is great, and (2) a plaintiff, by settling with one defendant and successfully litigating against another, will always be made whole in the amount of the jury's verdict.

Bound by precedent, however, we are not free to consider overruling our bifurcated system of credit rules under Martello and Snowden even if one uniform credit rule regime were demonstrably superior. But we do have some freedom. Because both Martello and Snowden concerned credits where the settlements did not reach the settlers' pro rata shares, neither case presented the question whether the plaintiff's recovery against joint tortfeasors always must be limited by the one satisfaction rule.

Both practical considerations — encouragement of settlements with accompanying judicial economy — and concerns about fairness lead us to conclude that a straight Martello ( pro rata credit) rule should be applied in all cases of joint tortfeasors, even when the one satisfaction rule must give way.

Given Martello for cases where the one satisfaction rule would not be violated, we believe a switch to Snowden in cases such as this one would chill prospects for settlement. The plaintiff would know its recovery could never exceed the amount of the jury's verdict, so even the prospects for an especially favorable settlement would be likely to lose force. On the other hand, a defendant disinclined to settle would know it could pay no more, and might even be able to pay less, than its pro rata share of an eventual verdict if it refused to settle. If under Martello, therefore, a nonsettling defendant will never have to pay more than its pro rata share — a rule defendants generally favor over the pro tanto rule — then settlement with all defendants is most likely to occur if each defendant also knows that it can never pay less than its pro rata share by continuing with the litigation — i.e., that there is no prospect for a pro tanto credit greater than its pro rata allocation.

We have noted that the Martello regime, because of its pro rata credit, often cuts a plaintiff's total recovery down well below the amount of the jury's verdict. Under such circumstances, we see no principled basis for changing the rule simply because a plaintiff negotiates an unusually favorable settlement. See McDermott, ___ U.S. at ___, 114 S.Ct. at 1471. Under the pro tanto regime, the plaintiff's equitable recovery (jury verdict) is assured in the amount of the jury verdict; in the pro rata regime, the nonsettling defendant's equitable contribution ( pro rata share) is assured without regard to the plaintiff's total recovery. The equities involved, therefore, do not persuade us to shift from pro rata to pro tanto at the very first moment a credit rule that ordinarily favors the litigating defendant shifts in favor of a hard bargaining plaintiff.

The principles underlying the one satisfaction rule suggest no persuasive reason to refrain from applying the pro rata rule to the exceptional facts of this case. The Supreme Court acknowledged in McDermott that its decision potentially conflicted with the one satisfaction rule. The Court concluded nonetheless:

Even if the Court of Appeals were correct in finding that the proportionate share approach would overcompensate McDermott, we would not apply the one satisfaction rule. The law contains no rigid rule against overcompensation.

___ U.S. at ___, 114 S.Ct. at 1470. The Court was unwilling to say that a defendant's generous settlement should benefit a nonsettling defendant through a large pro tanto credit amounting to a windfall, rather than benefitting the plaintiff for successful negotiation of half the claim. As the Court observed, a plaintiff often settles for too low an amount because of the uncertainty of recovery and of the need for a settlement to help strengthen its litigation prospects against a nonsettler. See id. at ___, 114 S.Ct. at 1471. On the other hand, the Court also recognized that, in some cases, settlements exceed what a defendant would have to pay after litigation because savings in litigation expenses and other factors occasionally cause a defendant to offer a higher amount than its anticipated share of the judgment alone would indicate. See id. The Court concluded that the two situations did not justify different rules:

It seems to us that a plaintiff's good fortune in striking a favorable bargain with one defendant gives other defendants no claim to pay less than their proportionate share of the total loss.

Id.

Whatever credit rule is chosen inevitably will play into all parties' settlement calculations. Making particular credit rules contingent upon the relative amounts of the settlement and the jury's verdict, as Dr. Footer asks us to do here, would only increase uncertainty and make it even more difficult for litigants to negotiate settlements. We are persuaded, as the Supreme Court was in McDermott, that consistency with established precedent applicable to settlements with joint tortfeasors weighs in favor of applying Martello, not the one satisfaction rule in Snowden, to the unusual facts of this case.

We conclude, accordingly, that when the settling defendant is a joint tortfeasor, the Martello rule applies even though the nonsettling defendant's pro rata contribution to the amount of the jury verdict, when added to the amounts recovered in settlement, will result in a plaintiff's recovery that exceeds the verdict.

D.

Dr. Footer argues that if he is only entitled to a pro rata credit, that credit should be two-thirds of the $1,405,071 verdict because, as in Rose v. Associated Anesthesiologists, supra, there were two classes of settling defendants and one nonsettler. There is a fundamental problem with this argument. In Rose, after two of the defendants had settled, the trial court found that the two had not been joint tortfeasors and awarded a pro tanto, Snowden credit covering all but $25,000 of the jury's verdict. The court of appeals reversed, declining to sustain a " Snowden credit in an amount exceeding the pro rata share of the verdict ascribable to the settling defendants," because the nonsettling defendant — the only defendant found negligent — would have been unjustly enriched. Rose, 163 U.S.App.D.C. at 248, 501 F.2d at 808. The court, therefore, held only that the nonsettling defendant must pay at least its pro rata share of the verdict when pro tanto credits, taken together, would result in its paying less. Rose did not say, contrary to Snowden, that a nonsettling defendant should receive a pro rata credit for settlement with a non-tortfeasor when that settlement is for less than the settler's pro rata share — as the laboratory defendants' $150,000 settlement was here. In the settlement agreement with plaintiff Berg, the laboratory defendants expressly denied any responsibility for Patricia Berg's delayed diagnosis; they have never admitted liability and have never been adjudicated joint tortfeasors; and there is no indication that the parties at trial considered them joint tortfeasors.

We hold, accordingly, that the trial court correctly applied Snowden in awarding Dr. Footer a pro tanto credit of the $150,000 plaintiff received under the settlement agreement with the laboratory defendants, leaving only two defendants, Dr. Footer and GWU Medical Center, for purposes of calculating Dr. Footer's pro rata share.

E.

A mechanical question remains: whether the laboratory defendants' pro tanto credit should be applied before the GWU Medical Center's pro rata credit is determined. Berg and GWU Medical Center contend, and Footer does not contest, that the laboratory defendants' $150,000 pro tanto credit should be subtracted from the $1,406,-071 jury verdict first, with the remaining $1,256,071 divided by two to determine Footer's pro rata share of the damages. This is correct. Subtraction of the pro tanto credit after determining Dr. Footer's pro rata share of the damages would have the effect of granting Footer the entire pro tanto credit of $150,000, rather than allowing him the benefit of half that sum as one of two joint tortfeasors. See Johnson v. Strouse, 697 F. Supp. 535, 540 (D.D.C. 1988) (subtracting $15,000 pro tanto credit from $100,000 verdict and then dividing remaining $85,000 to calculate pro rata credit of $42,500).

III.

We consider, finally, whether the trial court erred in awarding plaintiff an additional $200,000 from GWU Medical Center under the settlement agreement and crediting that amount to Dr. Footer. GWU Medical Center's offer to settle the case, incorporated by reference in the final settlement agreement, provided for paying Berg "$800,000 at this time and, in addition . . . another $200,000 if after good faith efforts to collect full damages from the remaining defendants by way of settlement or verdict, plaintiff's total recovery is less than $1.5 million." See supra note 2. The trial court, applying the one compensation rule, concluded that the total amount the plaintiff could recover was the amount of the jury verdict, $1,406,071, and thus determined that the settlement agreement obligated GWU Medical Center to pay the additional $200,000, credited pro tanto toward Dr. Footer's liability.

The results of our earlier analysis show that the plaintiff in this case will recover: $150,000 from the laboratory defendants, $800,000 from the GWU Medical Center, and $628,035.50 from Dr. Footer, see supra note 21, for a total of $1,578,035.50. Because the plaintiff's total recovery is greater than $1.5 million, GWU Medical Center, under the plain language of the settlement agreement, is not obligated to pay the additional $200,000.

IV.

In sum, we conclude that (1) GWU Medical Center is a joint tortfeasor, based on all parties' pretrial stipulation, not adjudication, of liability, and that the laboratory defendants are not joint tortfeasors; (2) a Snowden, pro tanto credit representing the settlement between Berg and the laboratory defendants ($150,000) should be deducted from the jury verdict ($1,406,071) for the benefit of the nonsettling defendant, Dr. Footer, leaving a balance of $1,256,071; (3) a Martello, pro rata credit of one-half that balance should be deducted for the benefit of Dr. Footer, resulting in a judgment against Dr. Footer of $628,035.50 ($1,256,071/2 = $628,035.50); and (4) GWU Medical Center is not obliged to make the contingent payment of $200,000 contemplated in the settlement agreement because the condition precedent did not occur.

We therefore affirm the trial court's award of a pro tanto credit for the laboratory defendants' settlement, reverse the court's award of a pro tanto credit for GWU Medical Center's settlement, reverse the court's ruling that the Medical Center is liable for an additional $200,000 under its settlement agreement with appellant Berg, and remand for recalculation of credits and reentry of judgment in a manner consistent with this opinion.

So ordered.