*419OPINION OF THE COURT

Lazer, J. P.

Directly before us at last is the “identical five” issue — whether in a special verdict all answers approved by a five-sixths vote must have the concurrence of the identical five jurors. The appeal also presents the related question of whether a juror who has dissented from the answer to a special verdict question is bound by the answer as further questions are considered. Finally, we must decide whether it was error for the trial court to respond to the jury’s inquiry during damage assessment by explaining how the award would be affected by their earlier apportionment of liability.

The action has its genesis in an injury suffered by Jennifer Schabe, a 12-year-old junior high school student who was struck by a school bus while running across the driveway of the Hampton Bays Jr.-Sr. High School. Although the ensuing lawsuit named the Hampton Bays Union Free School District, the Hampton Bays Jr.-Sr. High School, the East End Student Transportation Corp. and the bus driver as defendants, the plaintiffs subsequently settled their action against the bus company and the driver, leaving the school district and the school as the sole defendants. Despite the bus company’s absence from the trial, its negligence remained at issue because section 15-108 of the General Obligations Law provides that the settling tortfeasor’s settlement will serve to reduce the remaining tortfeasor’s liability to the extent of the monetary settlement or the settling tort-feasor’s proportion of fault, whichever is greater. The bus company’s proportion of fault, if any, thus became a matter for resolution by the jury.

The trial revealed that upon dismissal of their classes on November 5,1976, some of the junior high school students congregated at the front of the school in anticipation of boarding buses then parked in the school driveway. Although additional supervision of the departure process had been furnished on other days, on November 5 only one teacher was assigned to that task. Jennifer was standing near a small shuttle bus parked behind the other buses when yet another bus approached from the rear and was *420waved on by the shuttle bus driver. At that point, Jennifer ran into the driveway, slipped while trying to avoid the oncoming bus, and was pinned beneath its right front wheel.

The liability issues were submitted to the jury in the form of a special verdict containing seven written questions, the first six of which dealt separately with the issues of negligence and proximate causation relating to the conduct of the school district, the bus company and Jennifer. The seventh question asked the jury to apportion fault between these three named participants in the events at issue. In its charge, the trial court declared that at least five jurors would have to agree before any question could be answered but it was unnecessary that the same five agree on each answer. The jury was soon back with a request for guidance because different questions were drawing different dissenting jurors. After repeating its earlier instruction — that the law did not require that each answer be approved by the same five jurors — the court added that the dissenting juror “has to abide by the decision of the other five under our system”, that the dissenter “has to go along with what the others do because five of the others are in accord” and that “once you decide one question on a five-sixths basis, the other dissenting juror must regard that as having been determined since five out of the six have spoken”. This additional instruction drew a prompt exception from the defendants which argued that under the instruction a juror who disagreed as to the negligence of one of the participants in the events of November 5 would be unable to apportion liability between that participant and the others. The court denied the defendants’ further request that the jury be told that in dealing with the separate questions of negligence and proximate causation as to any particular participant it was necessary that the same five jurors agree on both answers.

Upon return of the liability verdict, polling disclosed that on four questions answered by five-sixths vote, the majorities had not been comprised of the identical five jurors. In its answer to question 1, the jury unanimously found the school district negligent, but in answering question 2, juror number 2 abstained from his colleagues’ finding that there was proximate causation. Responding to *421question 3, the jury found the bus company free of negligence, with juror number 4 dissenting; the next question relative to proximate causation was skipped because of this finding. By their fifth and sixth answers, the jury found Jennifer’s conduct negligent and a proximate cause of the injury, but juror number 1 dissented. In their last answer, the jury apportioned 59% of the negligence to the school district and 41% to Jennifer, with juror number 2 dissenting.

After the trial proceeded through its damage phase, the jury was instructed to arrive at “a one hundred percent [dollar amount] and allow the Court to apply the necessary mathematics”. Following brief deliberation, the jury inquired: “Is the dollar amount we agreed upon the exact amount Jennifer Schabe will be awarded or only a percentage?”. Replying that the figure arrived at would be multiplied by 59%, the court reiterated its exhortation that the jurors find a “.one hundred per cent valuation”. The defendants then requested that the jury be instructed not to concern themselves with “any computations” in order to erase any feeling that the verdict should be adjusted to compensate for Jennifer’s 41% responsibility for her injury. The request was denied. Still confused, the jury was soon back with its foreman informing the court that: “The concern of the jury was, sir, that the forty-one percent that’s attributed to the Plaintiff in negligence would be subtracted from the fifty-nine percent”. The court explained that Jennifer would receive 59% of the verdict and denied the defendants’ request to charge that she would receive the entire amount awarded by the jury.

The jury subsequently returned a unanimous verdict of $750,000, from which the court deducted $225,000 (the value of the bus company’s structured settlement), multiplied the remainder by 59%, and entered a judgment of $309,750 in favor of the infant plaintiff. Although the formula applied seems to be erroneous, it remains unchallenged on this appeal by the defendants, for they focus entirely on the significant issues we now discuss.

The Court of Appeals has never decided whether every nonunanimous answer in a special verdict must be ap*422proved by the identical five jurors. The Appellate Division, First Department, has held that any five jurors can answer questions in a general verdict accompanied by interrogatories (CPLR 4111, subd [c]), but even that issue was not before the Court of Appeals when it affirmed the judgment in that case (see Bichler v Lilly & Co., 79 AD2d 317, affd 55 NY2d 571). The trial courts have divided on the question (compare Aiello v Wenke, 118 Misc 2d 1068; Forde v Ames, 93 Misc 2d 723; Reed v Cook, 103 NYS2d 539; PJI 1:96, 1:97 [any five], with Cohen v Levin, 110 Misc 2d 464; Murphy v Sherman Transfer Co., 62 Misc 2d 960 [the identical five]) as have jurisdictions across the country (compare McChristian v Hooten, 436 SW2d 844 [Ark]; Juarez v Superior Ct., 31 Cal 3d 759; Tillman v Thomas, 99 Idaho 569; Ward v Weekes, 107 NJ Super 351; Naumburg v Wagner, 81 NM 242; Bullock v Yakima Val. Transp. Co., 108 Wash 413 [any five], with Baxter v Tankersley, 416 SW2d 737 [Ky]; Ferguson v Northern States Power Co., 239 NW2d 190 [Minn]; Plaster v Akron Union Passenger Depot Co., 101 Ohio App 27; Clark v Strain, 212 Ore 357; McCauley v International Trading Co., 268 Wis 62 [the identical five]).

Majority verdicts and special verdicts are hardly novel to the judicial process, but these days their joinder complicates a litigation scene in which the issues are complex enough without additional problems stemming from the identity of the majorities who voted in favor of certain answers (see Comment, Vote Distribution in Non-Unanimous Jury Verdicts, 27 Wash & Lee L Rev 360). Since the requirement for unanimity in civil cases was abandoned in this State in 1937 (see former Civ Prac Act, § 463-a, added by L 1937, ch 120; NY Const, art I, § 2; Note, 37 Col L Rev 1235) and special verdicts have been with us since early common law, it is remarkable that the identity issue has never received final resolution in New York. CPLR 4113 (subd [a]) tells us that “[a] verdict may be rendered by not less than five-sixths of the jurors constituting a jury” and CPLR 4111 (subd [b]), authorizing special verdicts, also is barren of any hint that might help resolve the issue. It would seem apparent that the drafters of the two sections were oblivious to the problems that confront us now. In the *423absence of indications of intent in the statute relevant to the current issues, we are left to divine legislative purpose through the use of other construction devices.

One such device of universal acceptance is the rule that statutes be interpreted in light of the legislative reasons that brought about their enactment and that they be given meanings that serve rather than defeat those reasons (see Matter of Allstate Ins. Co. v Shaw, 52 NY2d 818, 820; Lincoln First Bank v Rupert, 60 AD2d 193, 197; McKinney’s Cons Laws of NY, Book 1, Statutes, §§ 95, 96). The policy considerations underlying the abolition of the unanimity requirement are not difficult to discern. Nonunanimous verdicts decrease the number of mistrials and retrials and thus reduce court congestion, delay and the cost of maintaining the judicial system. They also reduce the number of unjust verdicts deriving from juror obstinacy or dishonesty and discourage compromise verdicts (Winters, Majority Verdicts in the United States, 26 J Amer Judicature Soc 87; Note, Civil Juries: Recent Legislation Allowing Nonunanimous Verdicts, 18 Washburn LJ 269; 4 Weinstein-Korn-Miller, NY Civ Prac, par 4113.02).

The policy imperatives advanced by the abolition of the unanimity requirement are also furthered when special verdict questions can be answered by any five jurors rather than the same five that have approved all other answers. The “any five” principle reduces the number of mistrials and retrials while diminishing confusion for both court and jurors and it does not interfere with the operation of the jury or sacrifice fairness (Juarez v Superior Ct., 31 Cal 3d 759, supra). There is no reason why a juror in an “any five” jurisdiction cannot decide a case just as conscientiously and well as a juror required to abide by an “identical five” mandate (Ward v Weekes, 107 NJ Super 351, supra). The identical five rule has been characterized as “mechanistic” and as not assuring additional fairness while substantially increasing the risk of hung juries and seriously undermining the usefulness and viability of laws that authorize majority and special verdicts (see Tillman v Thomas, 99 Idaho 569, 572, supra).

Although the arguments most often proffered to support the any five concept remain the strong public policy factors *424we have mentioned, there is yet another consideration that we find quite compelling. The identical five concept tends to alter a fundamental premise of the jury system — that all members of a jury panel partake meaningfully in disposition of the case. Under the identical five principal, the casting of a dissenting vote on any question reduces the dissenter’s influence to a state of practical impotence and creates a mandate for continued unanimity among the other jurors on the remaining questions if the verdict is to survive. The dissenter is then bereft of real voting power, for his vote on the remaining questions can no longer affect the verdict (see Juarez v Superior Ct., 31 Cal 3d 759, supra; Fields v Volkswagen of Amer., 555 P2d 48 [Okla]). With the dissenter stripped of the power to affect further answers, the six-person jury selected to decide the issues becomes for all practical purposes a jury of five because one member’s opinion cannot be backed with a meaningful vote (see Fields v Volkswagen of Amer., 555 P2d 48, supra; Ward v Weekes, 107 NJ Super 351, supra).

Adherents of the identical five principle argue, however, that a verdict, whether general or special, is a nonfragmentable totality representing one ultimate finding, and jurors cannot be permitted to agree with only one part of the dispositions essential to the verdict (see Earl v Times-Mirror Co., 185 Cal 165; Clark v Strain, 212 Ore 357, supra; Comment, Vote Distribution in Non-Unanimous Jury Verdicts, 27 Wash & Lee L Rev 360, 363-364). They illustrate by noting that if two answers draw different dissenters, only four jurors have agreed on the ultimate conclusion (see Dick v Heisler, 184 Wis 77; Note, 3 Wis L Rev 51). Under this theory, a juror who dissents as to any of the answers cannot be counted in favor of the verdict and the fact that its form is that of a special verdict is immaterial (see Plaster v Akron Union Passenger Depot Co., 101 Ohio App 27, supra).

Although application of the any five rule to verdicts that contain interrogatories is rooted in strong public policy considerations that transcend the origin of special verdicts, the historic basis for treating general and special verdicts differently still retains significance. Special verdicts were introduced to the judicial process early in common law by *425jurors fearful that their general verdicts might offend the authorities. By answering factual questions without deciding the case directly, these early jurors sought to avoid the charge of attainder for a false verdict — a matter triable before an attaint jury whose members were of higher rank than ordinary jurors (Morgan, A Brief History of Special Verdicts and Special Interrogatories, 32 Yale LJ 575, 576; Clementson, Special Verdicts & Special Findings by Juries, p 5). From the time of its creation, a special verdict was not merely a reflection of a general verdict split into parts, but was a device for returning the facts only — leaving the legal consequences to the Judge (Clementson, Special Verdicts & Special Findings by Juries, p 5; Sunder-land, Verdicts, General and Special, 29 Yale LJ 253; Comment, 74 Yale LJ 483). The total focus of a general verdict is the outcome of the case, but the focus of a special verdict is the resolution of specific factual questions (see Green, Blindfolding the Jury, 33 Tex L Rev 273; Wright, The Use of Special Verdicts in Federal Court, 38 FRD 199).

While a general verdict is the “merger into a single indivisible residuum of all matters, however numerous, whether of law or fact” and is “incapable of being broken up into its constituent parts” (Sunderland, Verdicts, General and Special, 29 Yale LJ 253, 258), a special verdict is separated into distinct parts (Sunderland, Verdicts, General and Special, 29 Yale LJ 253; Comment, 18 U of Chi L Rev 321), each of which can be treated as a separate and distinct issue (Tillman v Thomas, 99 Idaho 569, supra). The courts have thus been free to set aside unsound portions of special verdicts while saving others (see, e.g., Kojic v City of New York, 76 AD2d 828; Gannon Personnel Agency v City of New York, 57 AD2d 538; see, also, Davis v Caldwell, 54 NY2d 176). Each issue to be decided in a special verdict requires the concurrence of at least five jurors as the Legislature has mandated, but nothing in statute or public policy so thrusts a special verdict jury toward outcome resolution that the same analysis that applies to the validity of a general verdict must apply to a special verdict (see Note, 37 Col L Rev 1235). Special verdicts do not exist to test the accuracy of a hypothetical general verdict (see Gibbons, Rule 49: Special Verdicts and *426Interrogatories, 7 [No. 3] Litigation 34), for their function is a different one. There is no reason, therefore, to mandate that special verdicts conform with what might be perceived to have been the likely general verdict.

We also reject the defendants’ final challenge to the any five concept — that it permits jurors to vote inconsistently. Neither statute, common law nor even public policy requires that the vote of individual jurors be scrutinized to assure consistency. The validity of a special verdict may depend on the jury’s answers being consistent enough for the entry of judgment (see Nallan v Helmsley-Spear, Inc., 50 NY2d 507, 518, n 5; Pfeil v Elkcom Co., 51 AD2d 553; 8 Carmody-Wait 2d, NY Prac, § 58:10) but it does not depend upon the consistency of individual juror voting patterns. A requirement that individual jurors vote consistently would sharply minimize the usefulness of special verdicts (Aiello v Wenke, 118 Misc 2d 1068, supra) by increasing the number of mistrials and retrials. It would also create an absurd necessity for counsel and the court to conduct a consistency review of each juror’s votes before the jury was discharged in order to avoid waiver of the issue (see People v Satloff, 56 NY2d 745; Resch v Volkswagen of Amer.,_ Cal 3d_ [Aug. 27, 1984]). Although some courts have theorized that an apparently inconsistent vote by a juror might not actually reflect inconsistency because the juror might have decided to abide by his colleagues’ determination on earlier questions (see Aiello v Wenke, 118 Misc 2d 1068, supra; Forde v Ames, 93 Misc 2d 723, supra; Resch v Volkswagen of Amer.,_Cal 3d_, supra; Juarez v Superior Ct., 31 Cal 3d 759, supra), our conclusion does not depend on such a rationalization. Although any number of variables may impel a juror to cast a vote that appears to be inconsistent, nothing in our law mandates consistency in the votes of individual jurors, and we see no purpose in imposing such a requirement that would reduce juror flexibility in voting in order to achieve some greater symmetry in individual juror voting patterns.

While our analysis of the identical five issue has focused some attention upon the historic roots of special verdicts, it is important to emphasize that the purpose of verdicts that *427utilize interrogatories is profoundly different today. Special verdicts (and even general verdicts with interrogatories) are helpful in the trial of cases involving multiple parties, multiple issues and third-party practice (1 NY PJI 2d 35-36 [1983 Supp]), to aid juror understanding (Finz, Does the Trend in Our Substantive Law Dictate an Expanded Use of the Special Verdict?, 37 Alb L Rev 229), to avoid retrial of multiple theory cases since a defective theory will destroy a general verdict even if the other theories proffered have support in the record (Davis v Caldwell, 54 NY2d 176, supra), and to determine issue preclusion questions (Siegel, Supplementary Practice Commentaries, McKinney’s Cons Laws of NY, Book 7B, CPLR 4111, 1964-1983 Supp Pamph, pp 102-103). Our rather obvious conclusion is that public policy and fairness are best served when the votes of any five jurors suffice to answer any interrogatory associated with a verdict. The trial court’s charge in this respect was correct.

Ill

The second issue concerns the rights and obligations of a dissenting juror once he has been outvoted. The jury was told that when any answer was approved by five persons, the dissenter was bound by it in considering further questions. The instruction can be correct only if special verdicts alter the traditional rights and obligations of jurors.

A juror who disagrees with his colleagues on a general verdict disagrees with the entire verdict, and the result is either a dissenting vote if unanimity is not required or a hung jury if it is. The freedom to disagree derives from the respect that must be accorded to a minority juror and the valid role the hung jury plays in our system (Kalven & Zeisel, The American Jury, p 453). Whether in accord or disaccord with their colleagues, all jurors are entitled to participate fully in deliberations by attempting to alter their colleagues’ votes at any time before the verdict is finalized and by rendering their own decisions in accordance with their honest convictions (Mendelson, Beyond Allen: from Dynamite to Nukes, 54 NY St Bar J 91; Note, 47 NYU L Rev 296). The paramount importance of maintaining the independence and intellectual integrity of each *428juror is underscored by cases holding that all of the statutory number of jurors must participate in the decision-making process so that nonparticipation by a juror — whether due to illness or other cause — is destructive to a verdict, even where unanimity is not a requisite (see, e.g., Measeck v Noble, 9 AD2d 19; Johnson v Holzemer, 263 Minn 227; City of Flat River v Edgar, 412 SW2d 537 [Mo]). These cases illustrate the principle that participation by less than all of its members deprives the jury of the reflections and judgment of an individual who might have opposed the verdict and might have persuaded one or more of the other jurors of the wisdom of his position (see Comment, Courts: Jurors Dissenting on Special Verdict Issue Excluded From Subsequent Deliberations, 61 Minn L Rev 151, 161; 4 Weinstein-Korn-Miller, NY Civ Prac, par 4113.03).

The fact that each element of a special verdict represents a separate determination in no sense implies that dissenters must abandon their intellectual or factual perceptions as the jury proceeds from question to question. It is bedrock in our system of jurisprudence that no juror can be required to surrender honest convictions solely to accommodate the opinions of fellow jurors or merely to force a verdict (see ABA Standards Relating to Trial by Jury, § 5.4 [a] [v]). A special verdict juror is entitled to no less respect than his general verdict counterpart (see Juarez v Superior Ct., 31 Cal 3d 759, supra; Ward v Weekes, 107 NJ Super 351, supra) and a verdict based on an instruction that requires dissenting jurors to be bound by a factual conclusion with which they disagreed is fatally deficient because some jurors may have been impelled to vote contrary to their own judgment in order to adhere to the decisions of others with whom they disagreed (see Comment, 36 Tenn L Rev 749; Note, 22 Loyola L Rev 667). While the public policy considerations that underlie the acceptance of non-unanimous verdicts and the adoption of the any five rule are compelling indeed, they are not powerful enough to require any juror to surrender intellectual integrity in the interest of expeditious dispute resolution. An instruction that each answer arrived at is conclusive on all the jurors gives inordinate importance to the order in which the questions *429are answered, because it reduces the parameters of the debate on later questions.

Because the incorrect instruction may have limited the ability of some jurors to dissent, the current verdict may not represent the free and untrammelled will of the jury. Juror number 4 voted that the bus company was negligent and yet apportioned negligence solely between the school district and plaintiffs. Juror number 1 found Jennifer free of fault but later voted that she was 41% at fault. Although we have already determined that consistency in individual juror voting patterns is not requisite to a valid verdict, the votes we have mentioned may have been compelled by the instruction that the jurors were bound by answers with which they disagreed.

We hasten to note, however, that once liability has been decided in a bifurcated trial, the determination binds the jury as well as the parties, and we see no obstacle to the same jury assessing damages. A jury in such a trial is instructed that the defendant is liable as a matter of law (see PJI 1:2B) and that the damage issues must be decided based on the premise of liability (see Comment, Vote Distribution in Non-Unanimous Jury Verdicts, 27 Wash & Lee L Rev 360). Since a liability verdict has sufficient force of law to warrant the entry of judgment on it (see Matter of Parker Constr. Corp. v Williams, 35 AD2d 839; Hacker v City of New York, 25 AD2d 35), there is nothing coercive about a charge that tells the jury that liability has been determined as a matter of law, and we see no reason why a juror who has voted against liability will be unable to assess damages fairly if properly instructed (see 22 NYCRR 699.14).

Finally, we conclude that no error was committed when, in response to the jury’s inquiry, the court told the jurors how their liability apportionment affected the damage verdict under consideration. Whether a special verdict jury should be told how the outcome of the case will be affected by its answers remains a matter of controversy (see Wicker, Special Interrogatories to Juries in Civil Cases, 35 Yale LJ 296, 303; Note, 58 Minn L Rev 903; Note, 43 Minn L Rev 823; Note, 61 Mich L Rev 1165; Smith, *430Comparative Negligence Problems with the Special Verdict: Informing the Jury of the Legal Effects of Their Answers, 10 Land & Water L Rev 199; Ann., 90 ALR2d 1040; 4 Weinstein-Korn-Miller, NY Civ Prac, par 4111.07; PJI 1:97), although only one New York court seems to have dealt with the issue and then only in dictum (see Kennard v Housing Assoc., 26 Misc 2d 1000, affd without consideration of issue sub nom. Kennard v Welded Tank & Constr. Co., 27 AD2d 578, affd 25 NY2d 324). While the arguments against outcome instructions sometimes refer to the historical basis for special verdicts, they are more often founded on the contention that juror objectivity is furthered by limiting deliberations to the specific questions submitted and by withholding information as to consequences (see Skidmore v Baltimore & Ohio R.R. Co., 167 F2d 54; Wicker, Special Interrogatories to Juries in Civil Cases, 35 Yale LJ 296, 303; Note, 58 Minn L Rev 903, 906). The partisans of juror awareness respond that in most cases jurors already are aware of the effects of their answers (see Green, Blindfolding the Jury, 33 Tex L Rev 273), but if not, it is inevitable that they will speculate concerning the meaning and effect of the law and occasionally guess wrong (Smith, Comparative Negligence Problems with the Special Verdict: Informing the Jury of the Legal Effects of Their Answers, 10 Land & Water L Rev 199, 210). They thus conclude that refusal to provide information presents an undesirable barrier to the jury’s exercise of its “common sense wisdom” to apply the law fairly to the individual case (Note, 58 Minn L Rev 903, 912).

Although the use of outcome instructions in comparative negligence actions has been rejected in some States (see, e.g., Argo v Blackshear, 242 Ark 817; McGowan v Story, 70 Wis 2d 189; see Ann., 90 ALR2d 1040), an increasing number of jurisdictions now permit such instructions either by statute (see, e.g., Col Rev Stats, § 13-21-111, subd [4]; Conn Gen Stats, § 52-572h, subd [b]; Minn Rules Civ Pro, rule 49.01; ND Cent Code, § 9-10-07; Tex Rules Civ Pro, § 277; Wyo Stats Ann, § 1-1-109, subd [b], par [iii]) or judicial holding (see, e.g., Seppi v Betty, 99 Idaho 186; Scales v St. Louis-San Francisco Ry. Co., 2 Kan App 2d 491; Porche v Gulf Miss. Mar. Corp., 390 F Supp 624; *431Roman v Mitchell, 82 NJ 336; see Heft & Heft, Comparative Negligence Manual, § 7.40). The trend toward permitting outcome charges in comparative negligence cases springs from the view that deliberations concerning the percentages of negligence ought not take place in a vacuum or be based on a mistaken notion of how the comparative negligence statute operates (Roman v Mitchell, 82 NJ 336, 345, supra).

Here, the instruction followed an inquiry from the jury that strongly implied confusion and speculation. Absent the information provided in response to the inquiry, speculation along erroneous lines might have led to an erroneous verdict (see Smith, Comparative Negligence Problems with the Special Verdict: Informing the Jury of the Legal Effects of Their Answers, 10 Land & Water L Rev 199, 226). Since we decline to assume that the jury disregarded its instructions, we cannot agree with the defendants’ contentions that the jury adjusted the verdict to compensate for the reduction caused by Jennifer’s degree of fault. The real danger was that an uninformed jury might have reduced the damage award to compensate for Jennifer’s fault and thus created the potential for an erroneous double reduction when the Judge reduced it again (see Nixon, The Actual “Legislative Intent” Behind New Hampshire’s Comparative Negligence Statute, 12 NH Bar J 17, 29). While some information obviously must be kept from a jury (see PJI 2:275.2), there was no prejudice in the court’s effort to eliminate speculation by telling the jury that the total amount of damages would be reduced by the degree of Jennifer’s fault.

Although a new liability trial is necessary, there is no need for a new damage trial. In the past, most partial retrials have been limited to damages and have occurred where liability and damages were “neither intertwined nor the result of a trade-off of a finding of liability in return for a compromise on damages” (Figliomeni v Board of Educ., 38 NY2d 178, 182; see, also, Hogue v Wilson, 51 AD2d 424; Mercado v City of New York, 25 AD2d 75). Recently, however, the Court of Appeals has permitted damage verdicts to stand while ordering that the liability issues be retried (see, e.g., Trimarco v Klein, 56 NY2d 98, 105; Ferrer v Harris, 55 NY2d 285, 295, mod 56 NY2d 737). Liability *432retrials while damage verdicts are permitted to stand are not unusual in other jurisdictions (see, e.g., Smith v Lumbermen’s Mut. Cas. Co., 360 So 2d 1098 [Fla]; Begin v Weber, 305 Minn 441; Crawford v Miller, 18 Wn App 151; see, also, Sturm, Ruger & Co. v Day, 615 P2d 621 [Alaska]). The error that overthrows the instant liability verdict did not taint the damage finding, for that related to Jennifer’s injuries and not her conduct.

Accordingly, the judgment should be reversed insofar as appealed from and a new trial of the first cause of action asserted in the complaint as against the appellants should be granted, limited to the issue of liability only, with costs to abide the event. The jury’s finding on the issue of damages should be affirmed.

Gibbons, Weinstein and Boyers, JJ., concur.

Judgment of the Supreme Court, Suffolk County, entered June 28, 1982, reversed insofar as appealed from, on the law, and new trial granted as to the first cause of action asserted in the complaint as against appellants, limited to the issue of liability only, with costs to abide the event. The jury’s finding on the issue of damages is affirmed.