The common issue in these three consolidated cases is whether, in a tort action in which there is settlement with one or more defendants before trial, and a verdict is subsequently obtained against the remaining defendants which must be reduced both by the settlement amount and a comparative negligence factor, a court should first subtract the settlement amount or the comparative negligence factor. In Rittenhouse, the appellant also contends that the prejudgment interest against it was improperly calculated.
Facts
The facts leading to this appeal are largely undisputed in each case.
Rittenhouse
Karen Rittenhouse is the daughter of plaintiff Donald Rittenhouse. On the evening of June 14, 1973, Rittenhouse, then a sixteen-year-old high school student, was at home with her boyfriend, seventeen-year-old Kerry Erhart. At approximately 10:00 p.m., they left in Erhart’s 1969 Chevrolet, saying that they would return soon.
About 7:00 a.m. the following morning, the Er-hart family automobile was found parked in a nearby subdivision with the two teenagers inside. The ignition switch was in the "on” position, but the engine was not running. Carbon monoxide poisoning caused the death of Erhart, and Rittenhouse suffered brain injuries from which she has not fully recovered.
*173On December 20, 1974, plaintiff obtained an order appointing him next friend of Karen Rittenhouse, and he filed a complaint listing Earl M. Erhart (father of the deceased) as the sole defendant. Appellant North Woodward Mufflers, who had installed the exhaust system on the Erhart auto, entered the case when a second amended complaint was filed on October 12, 1976.
Other defendants were also named, but they settled for the combined amount of $395,000 prior to or during trial. At trial the jury found that appellant was negligent and that plaintiff had suffered damages in the amount of $1,500,000. Earl Erhart was not found to have been negligent. The jury also found that plaintiff had been twenty percent comparatively negligent. The trial judge computed prejudgment interest from the date of the filing of the second amended complaint and the amount of judgment in the manner proposed by appellant:
$1,500,000 Plaintiff’s damages
- 300,000 Twenty percent comparative negligence
$1,200,000
- 395,000 Settlements
$ 805,000 Judgment against appellant
The Court of Appeals affirmed1 the circuit court’s judgment, except that it recomputed the amount of the judgment against appellant in the following manner:
$1,500,000 Plaintiff’s damages
- 395,000 Settlements
$1,105,000
- 221,000 Twenty percent comparative negligence
$ 884,000 Judgment against appellant
*174The Court of Appeals also noted that plaintiff had failed to cross-appeal the issue of when the prejudgment interest should have begun to run. Subsequently, plaintiff filed a motion for leave to appeal on a particular issue, which the Court of Appeals granted, and the Court ruled that interest should have been calculated from the date the original complaint was filed.
We granted North Woodward’s application for leave "limited to the following issues: (1) whether the Court of Appeals erred in its method of calculating the judgment amount, and (2) whether the Court of Appeals erred in its determination of the date from which the appellant is to be charged judgment interest.” 418 Mich 955 (1984).
Gagnon
Plaintiff, Michael Gagnon, suffered a broken leg and other complications as a result of an industrial accident on July 26, 1976. Plaintiff filed suit naming several defendants including appellant, Dresser Industries. Settlements totaling $100,000 were made with other defendants, and plaintiff went to trial against Dresser only.
The jury found appellant negligent and that plaintiff had suffered $200,000 worth of damages. It also found, however, that plaintiff was fifty percent comparatively negligent. The trial court entered a judgment on the verdict in the manner proposed by plaintiff:
$200,000 Plaintiff’s damages
—100,000 Settlements
$100,000
— 50,000 Fifty percent comparative negligence
$ 50,000 Judgment against appellant
*175Defendant appealed, contending that the judgment should have been calculated in the following manner:
$200,000 Plaintiffs damages
—100,000 Fifty percent comparative negligence
$100,000
—100,000 Settlements
$ 0 Judgment against appellant
In a per curiam opinion, the Court of Appeals affirmed the judgment of the trial court. 130 Mich App 452; 344 NW2d 582 (1983). We, thereafter, granted defendant’s application for leave to appeal. 419 Mich 917 (1984).
Jackson
Plaintiff, Michael Jackson, was injured while working at a construction site in 1977. He was accidentally struck in the face with a shovel wielded by a coworker. Suit was filed against the general contractor and the appellant, Barton-Ma-low. After plaintiff and the general contractor settled for $150,000, the case went to trial.
The jury, as in the companion cases, was not informed of the settlement. Plaintiff was found to have suffered $400,000 in damages. The jury found plaintiff seventy-five percent negligent and defendant twenty-five percent negligent.
Defendant successfully urged the trial judge to compute the amount of judgment accordingly:
$400,000 Plaintiffs damages
—300,000 Seventy-five percent comparative negligence
$100,000
—150,000 Settlement
$ 0 Judgment against appellant
Plaintiff appealed, and the Court of Appeals *176modified the amount of the judgment by the following method:
$400,000 Plaintiffs damages
—150,000 Settlement
$250,000
—187,500 Seventy-five percent comparative negligence
$ 62,500 Judgment against appellant
131 Mich App 719; 346 NW2d 591 (1984).
We granted defendant’s application for leave to appeal and ordered this case to be argued and submitted with Rittenhouse and Gagnon. 419 Mich 917 (1984).
We affirm the judgment of the Court of Appeals in each case.
I
In 1978, the Legislature decreed that comparative negligence be applied in products liability cases. MCL 600.2949(1); MSA 27A.2949(1) provides:
(1) In all products liability actions brought to recover damages resulting from death or injury to person or property, the fact that the plaintiff may have been guilty of contributory negligence shall not bar a recovery by the plaintiff or the plaintiff’s legal representatives, but damages sustained by the plaintiff shall be diminished in proportion to the amount of negligence attributed to the plaintiff.
One year later this Court formally2 adopted the doctrine of comparative negligence in Placek v *177Sterling Heights, 405 Mich 638; 275 NW2d 511 (1979). The pure comparative negligence form was adopted, because it is the one which "most nearly accomplishes the goal of a fair system of apportionment of damages.” Id., 660. This is the only doctrine that "truly distributes responsibility according to the fault of the respective parties.” Id. (Citation omitted.)
Citing Kirby, supra, the Placek Court also noted that one of the virtues of pure comparative negligence is that it does not " 'unjustly enrich’ anyone.” Id. Also, the following language from Kirby was approved:
The doctrine of pure comparative negligence does not allow one at fault to recover for one’s own fault, because damages are reduced in proportion to the contribution of that person’s negligence, whatever that proportion is. The wrongdoer does not recover to the extent of his fault, but only to the extent of the fault of others. To assume that in most cases the plaintiff is more negligent than the defendant is an argument not based on equity or justice or the facts. What pure comparative negligence does is hold a person fully responsible for his or her acts and to the full extent to which they cause injury. That is justice. [Id., 661.]
With these general principles in mind, we now turn to the issue in the cases at bar.
Appellants contend that the amounts of plaintiffs’ total-damages in these cases must be reduced by the percentages of comparative negligence returned by the juries before subtractions of the settlements. They assert that the percentages of comparative negligence assigned to each plaintiff are factual findings of the proportion of responsibility on each plaintiff’s part in bringing about the injury. Each plaintiff’s total damages should thus, *178according to appellants, be reduced by the degree of fault in bringing about the injury, i.e., the jury’s finding of the percentage of comparative negligence.
The fundamental flaw in this argument is that there is no indication that the juries in these cases determined the plaintiffs’ percentages of negligence by comparing their actions to all the possible tortfeasors. Rather, the juries only weighed the responsibility of the plaintiffs as to the defendants at trial. It would thus be inaccurate to reduce plaintiffs’ total damages by a percentage which only applies to the trial parties and not all tortfeasors.3
In analyzing appellants’ contention, it is essen*180tial to delineate the limited nature of the trials in these cases. In each case, the plaintiff commenced suit against multiple defendants. Plaintiffs achieved settlements with some defendants and went to trial against others. The juries were not instructed on the existence of the settlements or of the amounts received. Brewer v Payless Stations, 412 Mich 673; 316 NW2d 702 (1982). Also, importantly, the negligence or fault of the settlers, in causing plaintiffs’ damages, was not an issue litigated before the juries by virtue of the settle*181ments. Nor was the fault of the settlers in causing plaintiffs’ damages an issue as to the nonsettling tortfeasors, as the settlements guaranteed that the settlers were immune from contribution from the trial defendants. MCL 600.2925d; MSA 27A.2925(4). Moreover, it would not serve the non-settling defendants’ goal of limiting their liability to attempt to show that the settlers were more the cause of the plaintiffs’ injuries than they. This is so because if the nonsettler was found to be negligent and the cause of the plaintiff’s injuries at all, then he would be liable for all of the plaintiff’s injuries (less settlements and comparative negligence) no matter how much negligence was attributable to the settlers. We thus believe that the focus of trial, such as the ones in the cases at hand, will normally be upon only the conduct of the trial parties. Accordingly, the percentages of comparative negligence returned by these juries represent findings applicable only to the trial parties.4 It would be inaccurate to reduce plaintiffs’ total damages by a percentage not applicable to the total causal negligence. Also, once the settlements occurred, the total damages were no longer at stake for the trial parties. Rather, under MCL 600.2925d; MSA 27A.2925(4), the amount remaining in issue at trial was reduced by the settlement amounts. In relevant part, § 2925d provides:_
*182When a release or a covenant not to sue or not to enforce judgment is given in good faith to 1 of 2 or more persons liable in tort for the same injury or the same wrongful death:
(b) It reduces the claim against the other tortfeasors to the extent of any amount stipulated by the release or the covenant or to the extent of the amount of the consideration paid for it, whichever amount is the greater. [Emphasis added.]
Accordingly, by the time of trial, the "claim” of each plaintiff against the .nonsettling tortfeasors was an amount equal to the total damages minus the settlements. Mayhew v Berrien Co Road Comm, 414 Mich 399, 410; 326 NW2d 366 (1982).5
Contrary to the assertion of the dissent, our position is not that the "plaintiffs 'claim’ [against the nonsettling defendants] includes] the percentage of damages attributable to the plaintiff[’s]” own negligence. Rather, we hold that a plaintiffs percentage of fault in these trials was determined only in relation to the defendants participating at trial. Therefore, a plaintiffs comparative negligence should only be deducted from that part of the judgment to be paid by the defendant who was the only other party to a determination of fault.6 *183Prior to subtracting the plaintiffs comparative fault from this judgment, the settlement between the plaintiff and the nonparty tortfeasors should be subtracted.7 This settlement may, and presum*184ably will, be based, among other things, on an estimate of the comparative fault of the parties to the settlement.8
In Scott v Cascade Structures, 100 Wash 2d 537; 673 P2d 179 (1983), the majority, en banc, adopted the rule advocated by appellants here. We find, however, the reasoning of the dissent more convincing:
The comparative fault of [plaintiffs was] calculated not as a percentage of total fault, but as a percentage of the fault of the parties remaining in the action. Logically, this limited comparative fault figure should be applied not to total damages but to only those damages attributable to the parties remaining in the action. [Id., 546 (Utter, J., dissenting).]
Also, the dissent noted:
[A]ll allocation of fault at trial [is] among only those parties still remaining in the case. Logically, such fault apportionment figures should be applied to a correspondingly limited damages figure, i.e., only those damages attributable to the remaining parties. This limited damages figure is to be created by the deduction from total damages of the *185court approved settlement, since that settlement should approximate the damages attributable to the settling defendants. [Id., 548. Citations omitted.]
The limited nature of these trials was also succinctly analyzed by the Court of Appeals in Gagnon in the context of the facts of that case:
[P]laintifFs entire claim for injuries [is] the amount of total damages which the jury has found, in the present case, $200,000. . . . [T]his "claim” should be reduced by the amount of settlement, $100,000 here, recognizing that the settlement represented satisfaction of the third party’s liability, thus removing the issue of the third party’s fault from subsequent litigation between plaintiff and the nonsettling tortfeasor, Mayhew, supra. Applying this analysis, the trial [is] then confined to the issue of how to apportion fault as between the plaintiff and nonsettling tortfeasor for the remaining $100,000 in damages. The jury would not have to face the problem of determining the liability of non-parties; its sole duty would be to allocate fault among the parties who were represented at trial, plaintiff and the nonsettling tortfeasor. [130 Mich App 458-459.]
We concur in this reasoning, as it correctly delineates the limited scope of the trials in each of these suits and recognizes that the percentages of comparative negligence returned by the juries are applicable only to the parties , at trial and not to all possible wrongdoers.
Appellants also assert that the form jury instructions led the jurors to believe that the amount of total damages awarded would be reduced by the plaintiffs’ percentages of comparative negligence. Specifically, appellants point to SJI2d 66.01, which was given in each case here. Under *186the instructions, it was initially established that defendants were negligent and that this negligence was a proximate cause of plaintiffs’ injuries. The juries then found plaintiffs’ "total amount of damages.” Each jury also determined that plaintiffs were negligent and that their negligence was also a proximate cause of the injuries. In Jackson and Gagnon, the following instruction was given:
Using 100 percent of the total combined negligence [which proximately caused the injury to plaintiff], what percentage of such negligence is attributable to the plaintiff?
A similar instruction was given in Rittenhouse.9
In Jackson and Gagnon the forms also stated:
Please note that the Court will reduce thé total amount of plaintiff’s damages [entered] by the percentage of negligence attributable to plaintiff, if any, [found]. The remainder will be the amount which plaintiff is entitled to recover.[10]
*187Appellants urge that the juries’ findings on these instructions manifest an intent by the jury that plaintiffs receive a net damages award of an amount equal to the total damages minus the percentage of comparative negligence found by the juries.11
We agree that the instructions given could have this effect and that to that extent they are misleading. However, we do not agree that the jury’s possible intent in a limited trial between only the parties that have not settled should be determinative as to the total recovery against all tortfeasors. The juries were never told that it was their responsibility to fix the final figure to be received by the plaintiffs. To the contrary, it was made clear to them that their responsibility was to answer specific questions about damages and comparative fault. The most these instructions meant to the jurors was that the plaintiffs would be entitled to a *188certain amount from the trial defendants.12 The jurors did not know the extent of negligence or amounts of settlements with other tortfeasors. Thus, they could not accurately determine a complete net recovery for plaintiffs. Assuming arguendo, the juries were operating under the erroneous interpretation that plaintiffs would receive a set amount from these defendants, it would be equally true that they erroneously assumed that the defendants would pay a set amount. We thus decline to allow speculation on the juries’ speculation of the outcome to control our decision as long as the jury was properly informed of its responsibility. In Brewer, supra, we held that juries should not know of the settlements of joint tortfeasors and accordingly of the ultimate subtraction of the settlements from the jury’s verdict, and in May-hew, we held that the jury may not determine the comparative fault of parties not at trial. These holdings compel the conclusion that juries are *189normally in possession of enough information to find the plaintiff’s damages and the fault of the trial parties. These necessarily limited findings should not, as appellants contend, be given inordinate weight so that they control the ultimate outcome of complex litigation.
It is our conclusion that a more accurate and just verdict, consistent with the principles enunciated in Placek can be best arrived at by reducing the plaintiff’s comparative negligence vis-á-vis the nonsettling tortfeasor by deducting it from the "claim” against that same nonsettling tortfeasor rather than from the total damages. That "claim” under the statute as interpreted in Mayhew is "the total liability of the joint tortfeasors minus the amount of the settlement. . . .” 414 Mich 410.
The judgment of the Court of Appeals on this issue in each case is affirmed, with costs awarded to the appellees.
II
Karen Rittenhouse suffered her injuries on June 14-15, 1973. On December 20, 1974, the plaintiff filed his original complaint naming only Earl Er-hart (the father of the deceased) as a defendant. Appellant North Woodward Mufflers first entered the case when a second amended complaint naming it as defendant was filed on October 12, 1976. The trial court ordered that plaintiff was entitled to prejudgment interest from appellant from the date the second amended complaint was filed. The Court of Appeals held that plaintiff was allowed interest from the date of the original complaint. Appellant contends that it should only be liable for interest from the date of the second amended complaint, when it was first made a party to the action. Since the jury returned a verdict of no *190cause of action on the original complaint, appellant asserts that the interest should only be allowed on the complaint where liability was found. MCL 600.6013; MSA 27A.6013 provides in part:
(1) Interest shall be allowed on á money judgment recovered in a civil action, as provided in this section.
(2) For complaints filed before June 1, 1980, . . . the interest on the judgment shall be calculated from the date of ñling the complaint to June 1, 1980 at the rate of 6% per year and on and after June 1, 1980 to the date of satisfaction of the judgment at the rate of 12% per year compounded annually. [Emphasis added.]
When this action arose § 6013 provided in part:
Interest shall be allowed on any money judgment recovered in a civil action, such interest [is] to be calculated from the date of ñling the complaint at the rate of 6% per year .... [Emphasis added.]
Because we agree with the appellee that this language is a clear expression of the legislative intent and that no interpretation is required, we decline to join in Justice Riley’s "interpretation” of the statute. Section 6013 plainly says that "interest on the judgment shall be calculated from the date of the filing of the complaint.” Had the Legislature borne the intent ascribed by appellant, it would have simply provided, for example, that interest runs "from the date of the filing of the complaint or amended complaint first filed against the judgment debtor” or "from the date of the complaint against the party named in the judgment.” Instead, the wording of the statute is un*191ambiguous, and we will not read into it provisions that the Legislature elected not to include.13
Even an examination of the rules of statutory interpretation still compels us to conclude that the Court of Appeals reached the correct result.
Section 6013 is a remedial statute which is to be liberally construed in favor of a plaintiff, the intended beneficiary of the legislation. Denham v Bedford, 407 Mich 517, 528; 287 NW2d 168 (1980). Appellant asserts that the central purpose of § 6013 is to encourage prompt resolution of lawsuits and that it could not have settled its suit until a complaint was filed. However, we have held that " '[t]he purpose of [§ 6013] is to compensate the prevailing party for the expenses incurred in bringing an action and for the delay in receiving money damages.’ ” (Citations omitted.) Wood v DAIIE, 413 Mich 573, 589, n 17; 321 NW2d 653 (1982), affirming 99 Mich App 701; 299 NW2d 370 (1980).
The evolution of §6013 into its current form supports this conclusion. As of 1961, § 6013 provided:
Execution may be levied for interest on any money judgment recovered in a civil action, such interest to be calculated from the date of judgment at the rate of 5% per year unless the judgment is rendered on a written instrument having a higher rate of interest in which case interest shall be computed at the rate specified in the instrument if such rate was legal at the time the instrument was executed. In no case shall the rate exceed 7% per year. [1961 PA 236. Emphasis added.]
1965 PA 24014 amended §6013 in two important *192respects: 1) to provide for statutory interest "from the date of filing the complaint” rather than date of judgment, and 2) to give the trial court discretion to stop statutory interest after the date a bona fide written settlement offer is refused by a plaintiff in a tort action and the ultimate judgment is equivalent or less favorable to the settlement offer. The first modification evidences a compensatory policy of allowing interest from the date the plaintiff formally initiates an action to recover damages. The latter addition to §6013 demonstrates a policy of encouraging settlements.
1980 PA 134 rewrote § 6013 by dividing what had been one paragraph into five subsections. MCL 600.6013(1); MSA 27A.6Q13(1) provides that interest accrues from the date of the complaint. Subsections (2), (3), and (4) concern the appropriate rate of interest, which was increased by the 1980 act. Section (5) involves the trial court’s discretion to stop the accruing of interest under certain circumstances where settlement offers have been refused.
The history shows that the purpose of § 6013(1) is compensatory. The policy of encouraging settlements is embodied in subsection (5). Thus, while appellant is correct that one of the policies of §6013 is to encourage settlements, its reliance is misplaced when it focuses on subsection (1) rather *193than subsection (5). The legislative intent to encourage settlements is embodied in subsection (5) and since that subsection does not address the amended-complaint issue here we are convinced that the Legislature intended interest to accrue from the date of the filing of the complaint in these circumstances.15
We would affirm the judgment of the Court of Appeals on this issue.
Williams, C.J., and Cavanagh, J., concurred with Brickley, J. Levin, J., concurred with Brickley, J., in part I only. Riley, J.The computation issue presented by these appeals concerns a joint tortfeasor’s substantive right to a credit for amounts received by a plaintiff in settlement with other potentially liable joint tortfeasors for the same injury. Because the principles of joint and several liability have not been displaced by the adoption of comparative negligence, the computation approach approved by the majority today may result in double recovery in direct contravention of the long-established rule that a tort victim is entitled to only one satisfaction for an indivisible injury sustained, and that a defendant is entitled to a setoff for amounts received by the plaintiff in a prior settlement. Larabell v Schuknecht, 308 Mich 419; 14 NW2d 50 (1944).1
The majority’s decision in this regard is without *194any legal or factual basis and is an affront to the fundamental principles of the comparative negligence doctrine as it has been developed through the decisions of this Court. Furthermore, my colleagues’ decision requires a strained interpretation of the language of the applicable statutory setoff provision, MCL 600.2925d; MSA 27A.2925(4), codified and embodied within the contribution statutes, MCL 600.2925a-600.2925d; MSA 27A.2925(1)-27A.2925(4). The majority concludes that a plaintiff found to have been contributorily negligent should have the amount received in a prior partial settlement deducted from the factfinder’s determination of total damages which includes the percentage of the damages which the plaintiff is not entitled to recover. The effect of this computation approach is to reduce the plaintiff’s contributory negligence factor and thereby to increase the amount of the plaintiff’s recoverable damages. The plaintiff’s "claim against the other tortfeasors,” within the meaning of the contribution statute, is interpreted, by the majority, as including the percentage of the damages attributable to the plaintiff, which is, pursuant to the doctrine of comparative negligence, not recoverable.2_
*195The majority’s reasoning is premised on a number of false presumptions. The majority presumes that the relative negligence of a settling defendant is attributed to the plaintiff by the factfinder in determining what percentage of the plaintiff’s damages was proximately caused by that plaintiff’s own negligence. Theoretically, the potential relative liability of a settling defendant (percentage of fault relative to other defendants), if any, is assumed by the nonsettling defendants, and not attributable to the plaintiff. Hence, the reason for reducing the plaintiff’s claim against the nonsettling defendants by the amount of the settlement. Any potential effect upon a jury’s determination concerning the amount of a plaintiff's contributory negligence because of the absence of a released defendant is speculative at best. Indeed, none of the plaintiffs-appellees in the present cases have argued otherwise.3 Yet, the majority takes this speculative reasoning even further than its advo*196cates have argued and implies that a settling defendant’s percentage of negligence is somehow directly and equally attributed to the plaintiffs percentage of contributory negligence. Each of the hypothetical illustrations set forth within footnotes 3 and 7 of the majority opinion is based upon this false presumption.
By misconstruing this Court’s decision in Mayhew v Berrien Co Road Comm, 414 Mich 399; 326 NW2d 366 (1982), the majority assumes that in determining the amount of a plaintiffs contributory negligence the factfinder is prohibited from considering all of the factual circumstances surrounding the plaintiffs injury, and may only consider the trial defendants’ negligence in this regard. The issue decided in Mayhew, however, was not even related to this issue and the consistent reference to Mayhew in the majority opinion, as holding that the plaintiffs amount of contributory negligence must be determined by comparing the plaintiff’s negligence with the negligence of the trial defendants only, is simply unfounded.4
Although its decision is entirely premised upon this speculation and unwarranted interpretation of Mayhew, the majority further expressly states: "[W]e hold that a plaintiffs percentage of fault [is to be] determined only in relation to the defendants participating at trial.” Ante, p 182. First, it *197must be noted that this was not the accepted rule at the time of any of the trials in the present cases. Second, this proposition is inconsistent with the fundamental principles of the comparative negligence doctrine, and principles of joint tort liability generally, because it implies that a plaintiffs claim for damages may be bifurcated against separate joint or concurrent tortfeasors with regard to a single indivisible injury.5
In Placek v Sterling Heights, 405 Mich 638; 275 NW2d 511 (1979), this Court held that a plaintiff found to have been contributorily negligent in causing his own injury is still entitled to recover to the full extent that his injury was caused by the tortious conduct of others. A plaintiff’s recovery in such cases has not heretofore been limited by the percentage of fault attributable to the trial defendants; it is the plaintiff’s percentage of negligence that determines the total amount of his recoverable damages. The implications of the majority’s express, if incidental, holding in this regard may adversely affect plaintiffs in such cases, and, in any event, will result in inaccurate findings of fact concerning a plaintiff’s percentage of contributory negligence.6_
*198I. Contribution Statute
The computation issue presented by these cases concerns the appropriate application of the setoff provision incorporated within the statutory scheme embodied in the contribution statutes, MCL 600.2925a-600.2925d; MSA 27A.2925U)-27A.2925(4). The contribution-release section here in question, MCL 600.2925d; MSA 27A.2925(4), states in relevant part:
When a release ... is given in good faith to 1 of 2 or more persons liable in tort for the same injury or the same wrongful death:
(a) It does not discharge any of the other tortfeasors from liability for the injury or wrongful death unless its terms so provide.
(b) It reduces the claim against the other tortfeasors to the extent of any amount stipulated by the release ... or to the extent of the amount of the consideration paid for it, whichever amount is the greater.
(c) It discharges the tort-feasor to whom it is given from all liability for contribution to any other tort-feasor. [Emphasis added.]
This statute allows a plaintiff to settle with less than all of the alleged tortfeasors, without discharging his claim against the remaining tortfeasors. It also allows the settling tortfeasor to escape liability for contribution, whatever the settlement amount or degree of fault.7 The plaintiffs "claim against the [remaining] tort-feasors,” however, is reduced by the amount of the settlement._
*199The approach approved by the majority construes the plaintiffs "claim” against the nonsettling defendants as including the percentage of the damages attributable to the plaintiffs own negligence. This construction is inconsistent with the general comparative negligence principles adopted in Placek, supra, and by legislative act applicable in all products liability actions, MCL 600.2949(1); MSA 27A.2949(1).8
The flaw in this characterization of the plaintiffs "claim” is that it does not distinguish the plaintiffs negligence in causing his own injury from the negligence of the defendant-tortfeasors. A plaintiffs negligence in this regard relates to a failure to use due care for his own protection. Unlike the defendant’s negligence which relates to the lack of due care for the safety of others, the plaintiffs negligence is not tortious. A plaintiff found to have been contributorily negligent in causing his own injury, therefore, cannot be characterized as a "tort-feasor” with regard to his own injury. Recognizing this distinction in American Motorcycle Ass’n v Superior Court, 20 Cal 3d 578, 589-590; 146 Cal Rptr 182; 578 P2d 899 (1978), the California Supreme Court noted:
Moreover, even when a plaintiff is partially at fault for his own injury, a plaintiff’s culpability is not equivalent to that of a defendant. In this setting, a plaintiff’s negligence relates only to a failure to use due care for his own protection, while a defendant’s negligence relates to a lack of due care for the safety of others. Although we *200recognized in Li [v Yellow Cab Co of California, 13 Cal 3d 804; 119 Cal Rptr 858; 532 P2d 1226 (1975)] that a plaintiffs self-directed negligence would justify reducing his recovery in proportion to his degree of fault for the accident, the fact remains that insofar as the plaintiffs conduct creates only a risk of self-injury, such conduct, unlike that of a negligent defendant, is not tortious. (See Prosser, Law of Torts [4th ed] § 65, p 418.) [Cited in Lemos v Eichel, 83 Cal App 3d 110, 117; 147 Cal Rptr 603 (1978).]
The conclusion that the proportion of damages attributable to the plaintiffs own negligence is to be included as part of the plaintiffs "claim” against the nonsettling defendants is inapposite. The determination concerning whether and to what extent a plaintiff’s own negligence contributed to his injury is necessary, specifically, to determine whether and to what extent a claim exists. The language of the statutory setoff provision here in question does not lend itself to the majority’s characterization of the plaintiff’s "claim.” That the Legislature intended to confer a substantive right upon plaintiffs to deduct settlement amounts from the proportion of the plaintiffs damages found to have been self-inflicted is questionable. To quote Judge Marutiak’s dissenting opinion in Jackson v Barton Malow Co, 131 Mich App 719, 727; 346 NW2d 591 (1984): "The statute clearly anticipates that plaintiff’s negligence will be factored against the gross assessment of damages before the settlement amount is to be subtracted.”
Unless § 2925d is constitutionally invalid, it governs and must be applied. Notwithstanding the majority’s attempt to reconcile its decision with the language of this statutory provision, the only possible construction that is consistent with its *201decision is that "the claim against the other tortfeasors” is the total amount of damages before application of the plaintiff’s contributory fault percentage. This construction was expressly given to §2925d by the Court of Appeals in Gagnon v Dresser Industries, 130 Mich App 452, 458; 344 NW2d 582 (1983). See ante, p 185.
II. Comparative Negligence Doctrine
In Placek, supra, this Court abrogated the contributory negligence doctrine and replaced it with the doctrine of comparative negligence. "Comparative negligence” is a term of art which distinguishes contributory negligence as a partial bar to recovery from contributory negligence as a total bar. The Court explained:
As a matter of terminology we replace the doctrine of contributory negligence. In actuality, however, it is contributory negligence as a total bar to recovery which is replaced. The effect of this action is to establish contributory negligence as a partial bar to recovery by insuring that any recovery of damages by a plaintiff be reduced to the extent of his or her own negligent contribution to the injury. [Emphasis in original. Placek, supra, 650, n 1.]
In adopting the pure form of comparative negligence, the Court in Placek, 660-661, stated:
The "pure” form does not "unjustly enrich” anyone. For example, if an accident is wholly the fault of one party, then that party would not, of course, recover damages. If an injured plaintiff was 51% to blame, then there still remains 49% of the fault which was not plaintiff’s, and for which therefore the person who caused that much of the injury should be liable.
*202* * *
The doctrine of pure comparative negligence does not allow one at fault to recover for one’s own fault, because damages are reduced in proportion to the contribution of that person’s negligence, whatever that proportion is. The wrongdoer does not recover to the extent of his fault, but only to the extent of the fault of others. . . . What pure comparative negligence does is hold a person fully responsible for his or her acts and to the full extent to which they cause injury. That is justice.
Thus, in cases in which the negligence of the plaintiff is at issue, a specific determination concerning the extent to which the plaintiffs negligence contributed to his injury is necessary in order to determine the total amount of recoverable damages. It should be noted that this determination is not made in relation to each defendant. The relative degrees of fault attributable to the defendants in such cases is not an issue relevant to the plaintiff.9 It is relevant only as between the tort*203feasors themselves, in an action for contribution, when available. MCL 600.2925b; MSA 27A.2925(2), MCL 600.2925c; MSA 27A.2925(3). The present jury instructions, which are modeled after the suggested instructions set forth in the appendix to Placek, only require assigning a percentage to the plaintiffs contributory fault. Percentages are not assigned to any of the defendants in making this determination, whether they have participated at trial or settled and been released. In assigning a percentage of contributory fault to the plaintiff, the jury is instructed to consider what percentage of the total fault (one hundred percent) is attributable to the plaintiff. The balance is assumed by the defendants. In a contribution action, the factfinder is instructed to assign percentages of fault to each defendant considering the total fault of the defendants only, so that the percentages add up to one hundred percent. See SJI2d 43.01A, 43.01B, 72.01A and 72.01B.
In each of the present cases the juries were instructed and directed that if they found negligence on the part of the defendants they would then proceed to assess the total amount of the plaintiffs damages and determine whether the plaintiff was negligent. If he was, they were then to go on to express that negligence in the form of a percentage which would be used to reduce the total amount of the plaintiffs damages. In each case, the matter was submitted to the jury in special verdict form per Placek, and in each case the plaintiffs were found to have negligently contributed to their injuries. For example, in Jackson and Gagnon, the matter was submitted to the jury in the form of special questions as set forth in SJI2d 66.01. Question number three of SJI2d 66.01 is as follows:
*204Question Number 3: What is the total amount of plaintiff’s damages?
Answer: $__
Question number six of SJI2d 66.01 poses the following:
Question Number 6: Using 100 percent as the total combined negligence which proximately caused the injury or damage to the plaintiff, what percentage of such negligence is attributable to the plaintiff?
Answer: _percent.
Please note that the Court will reduce the total amount of plaintiff’s damages entered in Question No. 3 by the percentage of negligence attributable to plaintiff, if any, entered in Question No. 6. The remainder wdll be the amount which plaintiff is entitled to recover.
Thus, the juries in each case were asked to determine, using one hundred percent as the total combined negligence which caused the plaintiff’s injury, what percentage was attributable to the plaintiff. In Jackson, the jury determined that the plaintiff’s total damages were $400,000 and assessed the plaintiff’s negligence at seventy-five percent. In Gagnon, the jury determined that the total damages were $200,000 and that the plaintiff’s negligence was fifty percent responsible. In Rittenhouse, the jury assessed the plaintiff’s total damages at $1,500,000 and found the plaintiff to be twenty percent at fault.
Applying the computation method approved by the majority (i.e., deducting the settlement amount before applying the plaintiff’s percentage of negligence), reduces the plaintiff’s fault percentage below that found by the jury. Applying the method approved by the majority to the present cases has *205the following effect. In Jackson, the plaintiff’s total recovery is increased to $212,500, and, thus, the jury’s determination concerning the plaintiff’s comparative negligence is reduced from seventy-five percent to slightly less than forty-seven percent (46.875 percent). In Gagnon, the plaintiff’s total recovery is $150,000, and the jury’s finding that fifty percent of the damages were attributable to the plaintiff is reduced to twenty-five percent. In Rittenhouse, the plaintiff recovers $1,279,000, and the jury’s assessment of the plaintiff’s fault percentage is reduced from twenty percent to fifteen percent.
This approach results, in every instance, in the modification of the jury’s determination concerning the percentage of damages attributable to the plaintiff. In fact, the greater the proportion of fault attributed to the plaintiff toward one hundred percent, the greater such a modification occurs.10
*206Two other jurisdictions that have adopted pure comparative negligence, by judicial decision and by legislative act, have addressed this precise computation issue. The courts of both jurisdictions have rejected the approach approved by the majority in these cases because it is inconsistent with the fundamental principles of comparative negligence and contradictory to the statutory setoff provisions of those states, which are substantially similar to the setoff provision incorporated within Michigan’s statutory scheme.11
The California Court of Appeals addressed this precise calculation issue in Lemos v Eichel, 83 Cal App 3d 110; 147 Cal Rptr 603 (1978). The court rejected the approach approved by the majority in *207the present cases, stating that applying the percentage reduction to the total damage before deducting the settlement is necessary to comply with the principles of the comparative negligence doctrine, adopted by the California Supreme Court in Li v Yellow Cab Co of California, 13 Cal 3d 804; 119 Cal Rptr 858; 532 P2d 1226 (1975), because it "[reduces] the plaintiffs damages 'in proportion to the amount of negligence attributable to the person recovering.’ ” Lemos, supra, 118 (citing Li, supra, 829). The court said that "deducting the settlement first in effect reduces the fault percentage below that found by the jury,” and, therefore, the court concluded that "the correct procedure is to apply each plaintiffs contributory fault percentage reduction to the total damage and reduce the resulting figure by the amount of that plaintiff’s pretrial settlement.” (Citation omitted.) Id., 118-119. The California Supreme Court has approved the Lemos court’s decision in this regard. See Aceves v Regal Pale Brewing Co, 24 Cal 3d 502, 512, n 4; 156 Cal Rptr 41; 595 P2d 619 (1979).
The Washington Supreme Court rejected the approach approved by the majority in the present cases in Scott v Cascade Structures, 100 Wash 2d 537; 673 P2d 179 (1983). Applying the Washington contribution-release statute, Wash Rev Code, §4.22.060(2), which is analogous to Michigan’s in all relevant respects,12 the court held that the percentage of the plaintiff’s fault was to be deducted prior to the settlement reduction:
The purpose of the contribution statute is to ensure that a plaintiff receives that to which he or she is entitled. Accordingly, we conclude the proper method of calculation is to reduce the jury award by the percentage of fault attributable to *208the plaintiff before deducting the settlement award. [Id., 545.]
The Washington court also noted that under the gross damages approach (deducting settlement amounts from total damages), the percentage of fault attributable to the plaintiff would fall below the actual percentage determined by the jury. Id.
I agree with the analyses of the California Court of Appeals and the Washington Supreme Court and conclude that the reasoning advanced by the majority in supporting its decision affirming the Court of Appeals decisions in the instant cases is erroneous.
The Court of Appeals in Rittenhouse, 126 Mich App 674, 683-684; 337 NW2d 626 (1983), relied entirely upon DeMaris v Brown, 27 Wash App 932; 621 P2d 201 (1980), in deciding this issue. The Washington Court of Appeals decision in DeMaris, however, was directly reversed by the Washington Supreme Court in Scott, supra.
The Court of Appeals in Gagnon, supra, 458-459, based its decision upon an incorrect extension of this Court’s decision in Mayhew, supra, and a strained construction of the language of MCL 600.2925d; MSA 27A.2925(4).
In Jackson, supra, 722-723, the Court of Appeals decision was based upon the prior decisions in Gagnon, and Rittenhouse. The majority in Jackson also cited Mayhew, DeMaris, supra, and the policy of encouraging settlements in support of its decision.
The majority states:
[T]here is no indication that the juries in these cases determined the plaintiffs’ percentages of negligence by comparing their actions to all the possible tortfeasors.
*209* * *
[PjlaintifFs percentage of fault . . . was determined only in relation to the defendants participating at trial. [Ante, pp 178, 182.]
The majority asserts that in determining the extent of a plaintiffs contributory fault a jury’s " 'sole duty [is] to allocate fault among the . . . plaintiff and the nonsettling tortfeasor.’ ” Ante, p 185 (quoting Gagnon, supra, 458-459).
These statements are less than accurate and thus misleading. The juries in these cases did not determine the defendants’ percentages of relative fault in relation to the plaintiffs contributory fault at all.13
In determining the plaintiffs’ comparative negligence the juries in each of these cases were only required to assign percentages to the plaintiffs’ contributory fault. These necessary determinations were not made "only in relation to the defendants participating at trial.”
Pursuant to the fundamental principles of the comparative negligence doctrine, factual determinations concerning what percentage of a plaintiffs injury, if any, was caused by that plaintiffs contributory fault are necessary to determine the total amount of that plaintiff’s recoverable damages. Because plaintiffs in such cases are entitled to recover to the full extent of the fault of others, this necessary determination is not to be made "only in relation to the defendants participating at trial.” Rather, this determination should reflect the percentage of the plaintiffs injury that was proximately caused by the negligence of others, including those who have settled and been released from liability, and others who, for a variety *210of possible reasons, have not participated at trial. This is the only way that a plaintiffs total amount of recoverable damages can accurately be determined, and it is the only way which complies with the comparative negligence principle enunciated in Placek, supra, and with the principles of joint tort liability generally.
Nothing in the past decisions of this Court precludes a plaintiff, defending against an assertion of comparative negligence, from presenting all of the factual circumstances surrounding the injury, including the conduct of settling parties or other parties that for some reason are not present at trial.
That factual determinations concerning a plaintiffs contributory fault are to be made only in relation to nonsettling tortfeasors and that a settling tortfeasor’s relative fault is somehow attributed to plaintiffs in such cases was refuted by this Court’s analysis in Mayhew, supra. In Mayhew, this Court ruled that the statutory setoff provision incorporated within the contribution statute has survived the judicial adoption of comparative negligence, and rejected the assertion that the plaintiffs recovery against nonsettling defendants should be reduced by the percentage of fault attributable to the settling defendants rather than the amount of the settlements. The defendants in Mayhew argued that to comply with principles of pure comparative fault, the nonsettling defendants’ liability should be reduced by the settling defendants’ percentage of fault because the nonsettling defendants’ liability should be limited to their proportion of fault. See 12 ULA, Uniform Comparative Fault Act, § 6 (1985 Supp). The plaintiff, on the other hand, argued that the statutory language was clear and, therefore, should be applied. The Court, noting that the adoption of com*211parative negligence did not displace principles of joint and several liability, held that the nonsettling defendants’ liability was to be reduced by the settlement amounts only, regardless of the settling defendants’ relative fault. The Mayhew Court stated that the "defendants’ argument [was] not without logic,” id., 405, and that "allocation of liability by the tortfeasors’ relative degree of fault is an important goal.” Id., 410. Nevertheless, the Court held, applying the statutory setoff provision reflected in MCL 600.2925d; MSA 27A.2925(4), "that the liability of the defendant non-settling tortfeasors is the total liability of the joint tortfeasors minus the amount of the settlement of the settling tortfeasor.” Id., 410. The "total liability of the joint tortfeasors” from which the settlement amount is to be deducted is, as referred to in Mayhew, the total amount of damages not caused by the plaintiff’s own negligence. The rationale advanced by the Mayhew Court was that its decision was consistent with the policies of "(1) encouraging settlements and (2) assuring that a plaintiff is fully compensated for injuries sustained.” Id., 411-412. The statute encourages settlement because the nonsettling defendants, in cases in which a plaintiff receives amounts in settlement from other potentially liable parties which are less than commensurate to the proportionate fault of those parties, risk liability in excess of their proportionate fault. The statute assures that a plaintiff is fully compensated for the same reason: The non-settling defendants assume all liability in excess of amounts received by the plaintiff in settlement. In reaching its decision in Mayhew, the Court offered an example of the calculation and deduction method envisioned by the applicable statutory setoff provision. The calculation method approved by the majority in the present cases is inapposite *212to the example offered in Mayhew. See Id., 408, n 6.
The issue decided in Mayhew did not concern the method by which a jury is to determine the plaintiff’s percentage of comparative negligence. However, the arguments of the defendants and the plaintiff in Mayhew, as well as the Court’s analysis and decision, were based on the legally and factually correct premise that a settling defendant’s relative percentage of fault is assumed by the nonsettling defendants and not equally attributed to the plaintiff.
In Brewer v Payless Stations, Inc, 412 Mich 673; 316 NW2d 702 (1982), the parties were in agreement concerning the amount of settlement and the calculation method for the deduction. This Court’s decision in Brewer was limited to the following issue:
Does the judge or the jury deduct the amount of the settlement . . . under the facts of this case in which evidence of the amount has no bearing on a material issue of fact? [Id., 675.]
The Court noted, quoting MRE 402, that
[a]ll relevant evidence is admissible, except as otherwise provided by the Constitution of the United States, the Constitution of the State of Michigan, these rules, or other rules adopted by the Supreme Court. Evidence which is not relevant is not admissible. [Id., 675, n 1.]
The defendant in Brewer argued that the jury should be fully informed of settlement amounts "if it is to deliver a considered judgment.” Id., 677. The plaintiff argued that informing the jury of settlement amounts may be prejudicial to a plain*213tiff for a variety of reasons. Id. The Court agreed with the plaintiff in Brewer and held:
When there is no genuine dispute regarding either the existence of a release or a settlement between plaintiff and a codefendant or the amount to be deducted, the jury shall not be informed of the existence of a settlement or the amount paid, unless the parties stipulate otherwise. [Id., 679.]
In reaching its decision in Brewer, the Court noted that the issue in that case could cut both ways; some of the plaintiff’s arguments could be used by a defendant, and vice versa. Id., 678.
The cases considered today stand as examples of the Brewer Court’s observation concerning the double-edged nature of the arguments for and against allowing the jury to make the settlement deductions. In the instant cases the plaintiffs argue that if the juries had been informed of the settlement amounts, and had been required to make the deductions, their determinations may have been different, resulting in more favorable judgments for the plaintiffs. Likewise, the plaintiffs argue that if the juries had been required to determine the percentages of fault attributable to the settling defendants, their determinations concerning the plaintiffs’ percentage of contributory fault may have been lower.14 Therefore, the plaintiffs argue, and the majority agrees, the settlement amounts should be deducted from the total damages before the reduction for the plaintiff’s comparative negligence is applied.
It should be noted that none of the plaintiffs in *214the present cases has challenged this Court’s decisions in Brewer or Mayhew. Nor did any of the plaintiffs object to the jury instructions given in each case. They each argue that -these factors should only be considered at the calculation stage, and, thus, their arguments are limited to supporting the suggested deduction method they have advanced.
The problems with this analysis are many. First, it has no basis in fact; both arguments are premised on speculation. Second, as the Court noted in Brewer, these arguments cut both ways. The plaintiff in Brewer successfully argued that allowing the jury to make the settlement deduction would prejudice the plaintiff. Third, and most importantly, this calculation approach is contradictory to the fundamental principles of the comparative negligence doctrine and is not supported by the language of the applicable statutory setoff provision.
Since Placek, the jury instructions used by the trial courts of this state have been modeled after the suggested jury instructions set forth in the appendix to that reported decision.15 None of the *216applicable jury instructions indicate that a plaintiff’s contributory fault percentage should fluctuate depending on how many defendants are present at trial. They do, however, inform the jury that this finding of fact will be used to determine the "amount which plaintiff is entitled to recover,” and that it should be made in relation to the "total combined negligence.”16
*217As stated in Placek, by general definition, and as reflected in the jury instructions given in each of the present cases, the factual finding concerning the extent of a plaintiffs contributory fault reduces the total amount of that plaintiffs damages, and the remainder is the total amount which is recoverable. This factual determination must be applied to reduce the plaintiffs total damages before deducting settlement amounts. There is no legal or factual basis to support crediting amounts received by a plaintiff in partial settlement, in such cases, to the percentage of damages attributable to the contributory fault of that plaintiff.
For the foregoing reasons, I am in respectful dissent. I would reverse the decisions of the Court of Appeals and reinstate the judgments of the trial courts in Rittenhouse and Jackson, and I would reverse the Court of Appeals decision in Gagnon and remand that case to the trial court for entry of judgment consistent with this opinion.
III. Prejudgment Interest
With regard to the prejudgment interest issue in Rittenhouse, this Court construes "the complaint” within the meaning of the prejudgment interest statute, MCL 600.6013; MSA 27A.6013, as the formal complaint filed against the defendant upon whom the prejudgment interest is being taxed. *218Thus, we hold that prejudgment interest commences to accrue, and therefore should be calculated, from the date of the filing of the complaint upon the defendant against whom the judgment has been entered. In the instant case it was at that time that an action commenced against the defendant, North Woodward Mufflers.
We are persuaded that the purposes of the prejudgment interest statute are not furthered by the dissent’s argument with respect to this issue, and the relation back of amendments to complaints for other purposes is not analogous. That the Legislature intended plaintiffs to be compensated for periods during which no disputed claim even existed against the judgment debtor strains credulity. Likewise, the laudable purpose of encouraging settlements is not applicable to periods during which no claim existed against the defendant. A statute must be construed in the light of the purpose sought to be accomplished thereby. Lakehead Pipe Line Co v Dehn, 340 Mich 25, 34; 64 NW2d 903 (1954).
Thus, we reverse the decision of the Court of Appeals in Rittenhouse as to the calculation of prejudgment interest issue.
We assess no costs on this issue, neither party having prevailed in full.
Ryan, J., concurred with Riley, J. Levin, J., concurred with Riley, J., in part III only.126 Mich App 674; 337 NW2d 626 (1983).
In Kirby v Larson, 400 Mich 585; 256 NW2d 400 (1977), three justices of this Court voted to adopt the comparative negligence doctrine in Michigan. As expressed in the opinion by Justice Williams, that view was by and large adopted by the Court in Placek two years later.
The inaccuracies that would result by the adoption of appellants’ method are convincingly demonstrated by the following hypotheticals.
A. Plaintiff suffers $1000 worth of damages and names two tortfeasors in an action. Each of the parties is one-third responsible (negligent) in bringing about the injuries, and one tortfeasor settles for $333. The case then proceeds to trial between plaintiff and one tortfeasor. Since the trial parties are equally negligent and the juries will only assess the fault of the trial parties (Mayhew), each one will be considered fifty percent negligent. The amount of damages plaintiff is found to have suffered should not change because of the settlements.
The plaintiff ideally should receive a total recovery of $666.00, because the plaintiff is thirty-three percent comparatively negligent as to all responsible parties for $1000 worth of damages.
Also, ideally, both tortfeasors should be liable for $333 as they are each one-third responsible for the injuries.
Appellees’ proposed version:
$1000 Plaintiff’s damages
— 333 Settlement
$ 666
— 333 Fifty percent comparative negligence
$ 333 Judgment against nonsettling tortfeasor
+ 333 Settlement
$ 666 Plaintiff’s net recovery
Appellants’ proposed version:
$1000 Plaintiff’s damages
— 500 Fifty percent comparative negligence
$ 600
— 333 Settlement
*179$ 167 Judgment against nonsettling tortfeasor
+ 333 Settlement
$ 500 Plaintiff’s net recovery
As can be seen, appellants’ calculation, in effect, treats plaintiff as if he were fifty percent comparatively negligent when he was actually thirty-three percent negligent. Also, the appellants’ version has the effect of reducing the settlement amount by the comparative negligence percentage. We believe it is more accurate to assume that any possible comparative negligence on plaintiff’s part has already been taken into account in the settlement process.
B. Plaintiff suffers $1000 worth of damages and is again thirty-three percent comparatively negligent. There are two tortfeasors who are also each one-third responsible for the injuries. Ideally, each tortfeasor is liable for $333 worth of damages, but one settles for $200 while the other goes to trial. As in the first hypothetical, the jury will find each trial party fifty percent comparatively negligent, as they are equally at fault and the negligence of the settlor may not be considered. Mayhew.
Again, plaintiff’s ideal net recovery is $666.66. Since he received $200 in settlement, the optimal judgment against the trial defendant is $466.66.
Appellees’ version:
$1000 Plaintiff’s damages
— 200 Settlement
$ 800
— 400 Fifty percent comparative negligence
$ 400 Judgment against nonsettling tortfeasor
+ 200 Settlement
$ 600 Plaintiff’s net recovery
Appellants’ version:
$1000 Plaintiff’s damages
— 500 Fifty percent comparative negligence
$ 500
— 200 Settlement
$ 300 Judgment against nonsettling tortfeasor
+ 200 Settlement
$ 500 Plaintiff’s net recovery
Due to low settlement, each method results in plaintiff receiving less than the ideal amount. However, appellee’s version has the virtue of making plaintiff closer to being whole.
C. A final possible solution is where the settling tortfeasor settles for an amount greater than that for which it is actually liable. If plaintiff and two tortfeasors are each equally at fault for $1000 worth of damages, defendants are liable for $333 each. If one tortfeasor settles for $500, the ideal recovery against the nonsettlor will be $166.66, making plaintiff’s total net recovery $666.66.
*180Appellees’ version:
$1000 Plaintiff’s damages
— 500 Settlement
$ 500
— 250 Fifty percent comparative negligence
$ 250 Judgment against nonsettling tortfeasor
+ 500 Settlement
$ 750 Plaintiffs net recovery
Appellants’ version:
$1000 Plaintiffs damages
— 500 Fifty percent comparative negligence
$ 500
— 500 Settlement
$ -0- Judgment against nonsettling tortfeasor
+ 500 Settlement
$ 500 Plaintiffs net recovery
Due to the inaccurately high settlement, appellees’ version results in a recovery greater than is ideal, while appellants’ computation still achieves a recovery below the optimal result. However, appellees’ approach, as in the first two examples, is closer to the ideal. We believe that in the vast majority of cases the settlement process will insure that the settlement is not likely to exceed the tortfeasor’s liability. Even if this does not always occur, the higher accuracy of the approach adopted by the Court of Appeals is evident under the first hypothetical. A calculation that allows plaintiff to recover only to the extent of the fault of others is mandated by comparative negligence principles. In the ideal situation where a tortfeasor has settled for the exact amount of its actual liability, appellees’ version correctly distributes the liability among the remaining parties while appellants’ method assigns plaintiff an artificially high percentage of comparative negligence. When the settlement amount is below an accurate amount, and in many cases where it is above the ideal figure, the appellees’ computation will result in a net recovery closer to the optimal amount.
This conclusion is supported by the following jury instruction given here:
"If you find that each party was negligent and that the negligence of each party was a proximate cause of the plaintiffs injuries, then you must determine the degree of such negligence expressed as a percentage attributable to the plaintiff.” (Emphasis added.)
The instruction speaks only of the negligence of each party and not of that of the settlers or others whose negligence may have caused the damages. We must assume that the juries followed these instructions and determined the “degree” of negligence of each "party” expressed as a percentage attributable to the plaintiff and did not assign a percentage to plaintiff applicable to nonparties.
In Mayhew, we enunciated this same rule from the viewpoint of defendant liability as follows: "the liability of the defendant non-settling tortfeasors is the total liability of the joint tortfeasors minus the amount of the settlement of the settling tortfeasor.”
The dissent cites Lemos v Eichel, 83 Cal App 3d 110, 116; 147 Cal Rptr 603 (1978), for the proposition that the comparative negligence factor should be deducted before the settlements. The critical distinguishing aspect of that case is that there the jury was instructed:
" 'The combined negligence of the plaintiff, the deceaseds, the defendants and of other persons whose negligence proximately contributed to the injury being 100%, what proportion of such combined negligence is attributable to the plaintiff, what proportion of such combined negligence is attributable to each of the deceased, what proportion is attributable to each of the defendants, and what proportion is attributable to each of the other persons?’ ” (Emphasis added.)
*183The jury returned a verdict listing the proportion of negligence among the plaintiff, defendants, and two settling tortfeasors. Obviously there is no question in Lemos that the jury was apportioning the plaintiff’s fault vis-á-vis all of the negligent actors. Had the jury in these cases been asked to apportion mult among all those whose negligence proximately caused the damages our method of calculation should perforce be the same as in Lemos.
The appellant’s proposed computation in effect reduces a trial defendant’s liability dv the settlement amount plus a percentage of the settlement amount equal to the percentage of plaintiff’s comparative negligence.
In Rittenhouse the appellant’s version is as follows:
$1,500,000 Total damages to plaintiff
— 395,000 Settlements
$1,105,000
— 221,000 Plaintiff’s twenty percent comparative negligence
$ 884,000
— 79,000 Deduction of twenty percent of $395,000 settlement
$ 805,000 Net liability of appellant
In Gagnon:
$ 200,000 Total damages to plaintiff
— 100,000 Settlements
$ 100,000
— 50,000 Plaintiff’s fifty percent comparative negligence
$ 50,000
— 50,000 Deduction of fifty percent of $100,000 settlement
$ 0 Net liability of appellant
In Jackson:
$ 400,000 Total damages to plaintiff
— 150,000 Settlement
$ 250,000
— 187,500 Plaintiff’s seventy-five percent comparative negligence
$ 62,500
— 112,500 Deduction of seventy-five percent of $150,000 settlement
$ 0 Net liability of appellant.
This double deduction for comparative negligence is unwarranted because, as noted in footnote 7, the settlement negotiations have already factored in the possible comparative negligence defense. The plaintiffs should not receive a disincentive to settle by having the jury’s finding of comparative negligence apply to the settlement amounts already received. After the settlements were made, the *184appellants were open to liability only in the amount of plaintiffs’ total damages minus the settlement amounts. The Court of Appeals has correctly held that appellants were not entitled to further reduce their liability by using the juries’ findings on comparative negligence, applicable only to them, to in effect make an additional deduction beyond the actual settlement amounts.
One of the goals of the settlement process, including mediation, is to ascertain the likely amount of a party’s liability considered against all the factors in each individual case. Among the factors to be considered in attempting to accurately assess the value of a case are "[t]he strength or weaknesses of a party’s proofs, the relative liability or negligence of each of the parties, and normal jury verdicts in similar cases . ...” 2 Martin, Dean & Webster, Michigan Court Rules Practice (3d ed), Commentary on Rule 2.403, p 443. (Emphasis added.) These same factors, as well as others, are likely considered by parties during settlement negotiations.
"Using 100 percent as the total combined negligence . . ., what percentage of such negligence is attributable to the plaintiff [Karen Rittenhouse]?”
In Gagnon, the trial court instructed the jury as follows concerning the verdict form:
"Now, you do not worry about doing any subtractions, additions, arithmetic or anything beyond answering these questions and, please, note that the Court will reduce the total amount of Plaintiff’s damages entered in question number five by the percentage of negligence attributable to Michael Gagnon if any entered in question eight and the remainder would be the amount that the Plaintiff is entitled to recover. That is something done by the Court. Do not try to do it, okay, if you get to that point.”
The emphasized portion belies appellant Dresser Industries implicit argument that the calculations were actually carried out by the jury during deliberations. It is better to assume, as we must, that the jury followed the instructions and did nothing "beyond answering [the] questions” on the verdict form. There is also no showing in Jackson and Rittenhouse that the juries carried out the mathematics necessary to have in mind a net recovery for plaintiffs.
The dissent interprets the instruction "[u]sing 100 percent as the total combined negligence which proximately caused the injury or damage to the plaintiff” to mean that the jury considered plaintiff’s comparative fault in respect to all tortfeasors, whether present at trial or not. In Mayhew, supra, where the nonsettling defendant was seeking to apportion fault to a settling tortfeasor, we said:
"[Njumerous difficulties would be presented if we were to allow the jury to apportion damages among all tortfeasors, including a settling non-party. It would mean that the settling tortfeasor’s liability would be assessed without anyone adequately representing that interest. It would put the plaintiff in a unique trial situation. The plaintiff would not only have to advocate that he was not at fault, he would have to convince the jury that the non-party was only minimally at fault. Otherwise, there might be too great a percentage of fault attributed to the non-party, thus reducing the plaintiff’s recovery.” 414 Mich 412.
If as we said in Mayhew, it would be difficult for the plaintiff to try to defend against the relative fault of a nonparty defendant, how, indeed can we expect the plaintiff in this situation to defend against his contributory negligence vis-á-vis nonparty defendants. If the instructional reference to "total combined negligence” was interpreted ”to include the negligence of nonparty tortfeasors, it would be left to the plaintiff and defendant to argue the relative negligence of the nonparty tortfeasors, notwithstanding the "numerous difficulties” alluded to in Mayhew.
For example, in Rittenhouse, two of the defendants actively participated in the trial for several weeks until they settled before the verdict. After the settlement the court instructed the jury:
“[The] matters between those two defendants . . . and the plaintiff, have been resolved and they are no longer in this case and you’re not to consider them in anything that happens from here on with respect to this particular case.”
In Gagnon and Jackson the juries knew that plaintiffs were injured during their employment, but due to the exclusive remedies provided in the Workers’ Disability Compensation Act the fault of the employers could not be considered. Also, in Gagnon, the appellant’s theory of the cases, read by the court to the jury, provided in part:
"Defendant contends that it was the owner’s responsibility [the owner is one of the settling defendants] for maintaining the crane, that it was his improper maintenance which was the proximate cause of the Plaintiff’s accident.”
In Jackson, the person who wielded the shovel which struck plaintiff was not named as a defendant, and the jury was not able to assess his degree of responsibility, if any. It is short sighted to presume that the juries believed their findings on damages and comparative negli, gence would be the only basis upon which plaintiffs’ ultimate net recovery would be calculated. Each jury knew of facts which could have made it aware that its findings were not determinative as to plaintiffs’ net recovery.
We do not decide whether this result is also required under MCR 2.118(D) (formerly GCR 1963, 118.4).
1965 PA 240 provides:
*192"Execution may be levied for interest on any money judgment recovered in a civil action, such interest to be calculated from the date of filing the complaint at the rate of 5% per year unless the judgment is rendered on a written instrument having a higher rate of interest in which case interest shall be computed at the rate specified in the instrument if such rate was legal at the time the instrument was executed. In no case shall the rate exceed 7% per year after the date judgment is entered. In the discretion of the judge, if a bona fide written offer of settlement in a civil action based on tort is made by the party against whom the judgment is subsequently rendered and the offer of settlement is substantially identical or substantially more favorable to the prevailing party than the judgment, then no interest shall be allowed beyond the date the written offer of settlement is made.”
We also note this conclusion is the same as that reached in Michigan Mutual v Staal Buick, Inc, 41 Mich App 625; 200 NW2d 726 (1972), and Awedian v Theodore Effron Mfg Co, 66 Mich App 353; 239 NW2d 611 (1976).
See generally 4 Restatement Torts, 2d, § 885(3), p 333; Prosser, Torts (4th ed), § 49, pp 304-305; Anno: 104 ALR 931. See also Thick v Lapeer Metal Products, 419 Mich 342; 353 NW2d 464 (1984). In Thick, *194the Court held that this fundamental-common-law principle of tort liability is applicable in the workers’ compensation context, and that the settlement amount is credited to reduce a judgment against a nonsettling defendant " 'whether or not the person making the payment is liable to the injured person.’ ” Id., 348, n 1 (quoting 4 Restatement Torts, 2d, § 885[3], p 333). Writing for the Court, Justice Boyle noted:
"This construction of the act parallels the common-law rule that where a negligence action is brought against joint tortfeasors, and one alleged tortfeasor agrees to settle his potential liability by paying a lump sum in exchange for a release, and a judgment is subsequently entered against the non-settling tortfeasor, the judgment is reduced pro tanto by the settlement amount.” Thick, supra, 348, n 1.
The majority expressly states: "the 'claim’ of each plaintiff against the nonsettling tortfeasors was an amount equal to the total damages minus the settlements.” (Emphasis added.) Ante, p 182. The majority further expressly states, ante, p 189, that the "'claim’ under the *195statute as interpreted in Mayhew is 'the total liability of the joint tortfeasors minus the amount of the settlement (Quoting Mayhew v Berrien Co Road Comm, 414 Mich 399, 410; 326 NW2d 366 [1982].) The total liability of the joint tortfeasors, however, is not the total amount of the plaintiff’s damages, but only the amount that was not caused by the plaintiff’s own negligence. In order to determine the amount of total liability of the joint tortfeasors, therefore, the plaintiff’s percentage of contributory negligence must first be deducted from the plaintiff’s total damages.
The majority’s attempt to reconcile its holding that settlement amounts are to be deducted from total damages before applying the plaintiff’s contributory fault percentage with the language of the applicable statute is perplexing. Was the Court incorrect in Mayhew v Berrien Co Road Comm, 414 Mich 399, 402-403; 326 NW2d 366 (1982), when it stated: "the law does not contemplate judicial abrogation of constitutional legislative enactments because not reasonably related to a judicial decision”?
None of the plaintiff-appellees in the present cases have argued that a settling defendant’s percentage of negligence is in fact correlatively attributed to both the nonsettling defendants and the plaintiff. They each argue, in the words of plaintiff-appellee Gagnon, that "[t]he simple fact is, we don’t know” whether the jury’s finding concerning the plaintiff’s comparative negligence would have been different had the settling defendants been present at trial.
The issue decided in Mayhew was whether the settling defendants’ proportionate fault was to be used to reduce the nonsettling defendants’ liability, rather than the amount of the settlements as required by the applicable provision of the contribution statute. MCL 600.2925d; MSA 27A.2925(4). That issue did not concern the method by which a jury is to determine what percentage of a plaintiffs total damages is attributable to that plaintiffs own contributory negligence, which is necessary in order to determine what percentage is recoverable. This Court’s holding in Mayhew, therefore, does not mandate that a jury must balance the plaintiffs negligence with the negligence of the trial defendants only, in determining what percentage of the total damages the plaintiff is entitled to recover. If anything, the Court’s opinion in Mayhew implies the opposite.
This issue was not briefed in the present cases, nor has it been directly addressed in any of the past decisions of this Court. It certainly cannot be said to be the general view among other jurisdictions. See 12 ULA, Uniform Comparative Fault Act, § 2 (1985 Supp).
The majority’s holding in this regard will adversely affect plaintiffs in situations in which the amount received in settlement is less than equal to the proportionate fault of the settling defendants.
Conversely, in situations in which the amount received by a plaintiff in settlement exceeds the amount of the plaintiff’s recoverable damages attributable to the fault of the settling defendants (as subsequently determined by a jury), a plaintiff would recover amounts in excess of that to which he was entitled.
Under the majority’s holding in this regard, the only situation that will result in a plaintiff recovering that amount to which he is entitled, no more and no less, is when the amount received in settlement is exactly commensurate to the settling defendants’ per*198centage of fault. The heretofore recognized rule (adopted in Placek and reflected in the jury instructions), as noted by the Court in Mayhew, assures that plaintiffs will be fully compensated, even if amounts received in settlement are less than equal to the relative fault of the settling defendants.
A settling tortfeasor may also recover contribution if the requirements of MCL 600.2925a(3); MSA 27A.2925(1)(3), and MCL 600.2925c(4); MSA 27A.2925(3)(4) are met.
MCL 600.2949(1); MSA 27A.2949(1) provides:
"In all products liability actions brought to recover damages resulting from death or injury to person or property, the fact that the plaintiff may have been guilty of contributory negligence shall not bar a recovery by the plaintiff or the plaintiff’s legal representatives, but damages sustained by the plaintiff shall be diminished in proportion to the amount of negligence attributed to the plaintiff.”
In situations in which a plaintiff has sustained a single indivisible injury, although that injury may have been caused in varying degrees by joint tortfeasors, the plaintiff is entitled to only one satisfaction. Furthermore, a jury may not apportion damages between joint tortfeasors who have caused a single, indivisible injury. Bowerman v Detroit Free Press, 279 Mich 480; 272 NW 876 (1937). The applicable jury instruction, SJI2d 41.04, which should be given in such cases, reads as follows:
"SJI2d 41.04 Damages Not to Be Allocated among Joint Tort-Feasors
"If your verdict is for the plaintiff against [both defendants/more than two defendants], you may not divide the damages [between/ among] them, but you shall return a verdict in one single sum against [both defendants/those defendants whom you ñnd to be liable].” (Emphasis in original.)
Thus, the relative degrees of fault attributable to joint tortfeasors is never an issue relevant to the plaintiff. Rather, it is relevant only in an action for contribution which is never initiated by the plaintiff or litigated in the same proceeding with the plaintiff’s claim unless the defendants so agree. For the applicable jury instructions to be used in contribution actions see SJI2d 43.01A, 43.01B, and the corresponding form of verdict reflected in SJI2d 72.01A and 72.01B.
The greater the proportionate fault of the plaintiff toward one hundred percent, the greater the disproportionate difference between the two methods of computation. Defendant-appellant Dresser Industries, Inc., provides the following evaluation which illustrates the effect of the majority’s calculation approach with the only variable being the percentage of the plaintiff’s contributory fault.
Rittenhouse
Gross Jury Verdict $1,500,000
Setoff from settling tortfeasors 395,000
Comparative Fault at Various Plaintiff’s Defendant’s Percentages Calculation Calculation Variance
*20 $884,000 $805,000 $ 79,000
40 663,000 505,000 158,000
50 552,500 355,000 197.500
75 276,250 (20,000) 296,250
90 110,500 (245,000) 355.500
* Actual Jury Determination
Gagnon
Gross Jury Verdict Setoff from settling tortfeasors $200,000 100,000
*206Comparative Fault at Various Plaintiffs Defendant’s Percentages Calculation Calculation Variance
20 $80,000 $60,000 $20,000
40 60,000 20,000 40.000
*50 50,000 -0-50.000
75 25,000 (50,000) 75.000
90 10,000 (80,000)
Actual Jury Determination
California’s Code of Civil Procedure, § 877, provides:
"Where a release, dismissal with or without prejudice, or a covenant not to sue or not to enforce judgment is given in good faith before verdict or judgment to one or more of a number of tortfeasors claimed to be liable for the same tort—
"(a) It shall not discharge any other such tortfeasor from liability unless its terms so provide, but it shall reduce the claims against the others in the amount stipulated by the release, the dismissal or the covenant, or in the amount of the consideration paid for it whichever is greater; and
"(b) It shall discharge the tortfeasor to whom it is given from all liability for any contribution to any other tortfeasor.”
Similarly, Washington’s statutory setoff provision, Wash Rev Code, § 4.22.060(2), provides:
"A release, covenant not to sue, covenant not to enforce judgment, or similar agreement entered into by a claimant and a person liable discharges that person from all liability for contribution, but it does not discharge any other persons liable upon the same claim unless it so provides. However, the claim of the releasing person against other persons is reduced by the amount paid pursuant to the agreement unless the amount paid was unreasonable at the time of the agreement in which case the claim shall be reduced by an amount determined by the court to be reasonable.”
See n 11.
See n 9.
The plaintiffs-appellees in the present cases concede that this argument is speculative, and do not argue that an exact effect can be ascertained. They assert, however, that the contention that the jury’s findings of fact concerning a plaintiff’s comparative negligence is not affected by the absence of a settling defendant is also speculative.
I am persuaded, after reviewing the current Michigan Standard Jury Instructions, that juries are clearly informed of the nature and purpose of determinations concerning whether, and to what extent, a plaintiffs negligence contributed to damages. The general definition of comparative negligence, which should be given when a plaintiffs negligence is at issue, is set forth in SJI2d 11.01 as follows:
"Plaintiffs negligence, if any, does not bar a recovery by the plaintiff against the defendant, but the total amount of damages to which the plaintiff would otherwise be entitled shall be reduced by the percentage that plaintiffs negligence contributed as a proximate cause to [his/her] [injury/property damage].
“This is known as comparative negligence.” (Emphasis in original.)
The note to SJI2d 11.01 states that, when comparative negligence is an issue, the verdict form reflected in SJI2d 66.01 is to be used:
"SJI2d 66.01 Form of Verdict:
Comparative Negligence
"We, the jury, make the following answers to the questions submitted by the Court:
"Question No. 1: Was the defendant negligent?
*215"Answer: _(yes or no)
"If your answer is 'no,’ do not answer any further questions.
"Question No. 2: Was the defendant’s negligence a proximate cause of the injury or damage to the plaintiff?
"Answer: _(yes or no)
"If your answer is 'no,’ do not answer any further questions.
"Question No. 3: What is the total amount of plaintiffs damages?
"Answer: $___
"Question No. 4: Was the plaintiff negligent?
"Answer: _(yes or no)
"If your answer is 'no,’ do not answer any further questions.
"Question No. 5: Was the plaintiff’s negligence a proximate cause of the injury or damage to the plaintiff?
"Answer: _(yes or no)
"If your answer is 'no,’ do not answer any further questions.
"Question No. 6: Using 100 percent as the total combined negligence which proximately caused the injury or damage to the plaintiff, what percentage of such negligence is attributable to the plaintiff?
"Answer: _percent
"Please note that the Court will reduce the total amount of plaintiff’s damages entered in Question No. 3 by the percentage of negligence attributable to plaintiff, if any, entered in Question No. 6. The remainder will be the amount which plaintiff is entitled to recover.”
For cases involving products liability in which the plaintiff’s negligence is an issue, the following jury instruction is provided, SJI2d 25.41:
"SJI2d 25.41 Comparative Negligence—
Burden of Proof
"The defendant has the burden of proof on [his/her] claim that the plaintiff was negligent in one or more of the ways claimed by the defendant as stated to you in these instructions, and that such negligence was a proximate contributing cause of the [injuries/damages] to the plaintiff.
"If you find that the plaintiff was negligent, and that such negligence was a proximate contributing cause of the plaintiffs [injuries/ damages], then using 100 percent as the total combined negligence and/or breach of warranty which proximately caused the plaintiff’s [injuries/damages], you must determine what percentage was due to the plaintiffs negligence. Negligence on the part of the plaintiff does not bar recovery by the plaintiff against the defendant. However, the percentage of negligence attributable to the plaintiff will be used by the Court to reduce the amount of damages which you find to have been sustained by the plaintiff.
"The Court will furnish a Special Verdict Form to assist you in your duties. Your answers to the questions in the Special Verdict Form will provide the basis on which this case will be resolved.” (Emphasis in original.)
None of the applicable Standard Jury Instructions indicate that, in determining what percentage of a plaintiff’s damages is attributable *216to the plaintiff’s negligence (and, therefore, for which the plaintiff will not be entitled to recover), a jury should only consider the plaintiff’s negligence relative to the trial defendants’ negligence, in situations in which less than all the possible tortfeasors are present at trial. Rather, they instruct juries to decide what percentage of the total combined negligence is attributable to the plaintiff.
Nevertheless, if these jury instructions become inadequate or ambiguous in situations in which some party who is potentially at fault for the plaintiff’s injury is not present at trial, then perhaps they should be modified. If, as the majority asserts, the only way that the negligence of a settling tortfeasor will not be attributed to the plaintiff is by requiring the jury to assign a percentage of negligence to each tortfeasor, whether at trial or not, then perhaps adopting this procedure should be considered. See ante, p 182, n 6 (citing the California comparative negligence jury instruction). Short of actually requiring the jury to assign percentages to each party, the current instructions could be modified to clarify the nature of the crucial determination concerning the plaintiff’s percentage of contributory negligence to assure that a settling defendant’s relative fault is not somehow attributed to the plaintiff. It must be emphasized again, moreover, that this Court’s decision in Mayhew does not prohibit a jury from considering the fault of parties not present at trial in determining the amount of a plaintiff’s contributory fault. Nor does Mayhew prohibit a jury from assigning percentages to reflect each party’s relative fault in making this determination. Furthermore, the passage from Mayhew quoted, out of context, by the majority would not be applicable to the present analysis. See ante, p 187, n 11.
The Mayhew Court concluded (correctly, I believe) that if it were to adopt the defendants’ argument in that case (i.e., that the nonsettling defendants’ liability should be reduced by the settling defendants’ percentage of fault rather than the amount of the settlements), plaintiffs would be put in a "unique trial situation” because they would be forced to defend the settling defendants’ relative fault. Because a plaintiff’s recovery against nonsettling defendants would be reduced by the relative fault of settling defendants, a plaintiff would not only have to defend his own conduct but also the relative fault of a settling defendant.
Unlike the situation referred to in Mayhew, a plaintiff defending against an assertion of comparative negligence generally, relative to the total combined negligence that caused the indivisible injury and *217to the total number of possible tortfeasors whether present at trial or not, would only be required to defend his own contributing negligence. The relative fault of the settling defendants would be attributed to the nonsettling defendants and not used to reduce the plaintiff’s recovery. Therefore, the concern expressed with regard to the issue in Mayhew is inapplicable within the present context.
Finally, it should be noted that the current jury instructions will have to be modified, in any event, to reflect the majority’s express holding that "a plaintiff’s percentage of fault [is to be] determined only in relation to the defendants participating at trial,” which in my opinion is a serious error.