concurring in part; dissenting in part.
I concur in the majority’s disposition of all issues except its limitation of the remand on the fraud claim to the question of punitive damages. I would follow the rule of Maxwell v. Port Terminal RR Co., 253 Or 573, 456 P2d 484 (1969), and would remand for a retrial on all factual issues, including liability for fraud as well as punitive damages.1
The court said in Maxwell:
“In the ordinary two-party personal-injury case, however, evidence of fault can influence the jury’s measurement of damages; and the kind and degree of injuries may influence some jurors in their evaluation of the evidence on liability. Whatever logical problems these elements of lawyer forklore may suggest, we believe that neither side in this type of case should be encouraged to manipulate errors in one trial to gain *790tactical advantage in a new trial before a new jury. Accordingly, we hold that the new trial in a personal-injury case ordinarily should be a new trial on all contested factual issues, regardless of the ability of the parties on appeal to pinpoint error so as to show that the error, if any, may have affected only one issue. There will of course, be exceptional cases in which the trial court, in the exercise of judicial discretion, properly will limit the issues for a new trial. But the standard to be applied in the exercise of this discretion is reasonable certainty that the issue or issues to be eliminated from the second trial are no longer viable issues in the case and that their removal will not prejudice the right of either party to the kind of jury trial to which he would have been entitled but for the error or errors necessitating the new trial.” 253 Or at 577. (Citation omitted.)
Stripped to its essentials, the majority opinion offers two reasons for a limited remand in this case: first, that later cases have departed from Maxwell and render it unauthoritative; and, second, that it is somehow unfair to deprive plaintiffs of the judgment they have obtained on the liability question, when the error necessitating reversal pertains only to the punitive damages question.
The majority’s first proposition rests on a random collection of post-Maxwell decisions by the Supreme Court and this court which have ordered partial remands. Most of those decisions offer little or no explanation for their dispositions, and some of them do not even cite Maxwell, let alone overrule or distinguish it. Wolf v. Nordstrom, 291 Or 828, 637 P2d 1280 (1981), is an exception. There, the court said:
“Since the trial court’s directed verdict was proper, we find this case to be exceptional and not within the general rule of Maxwell that retrial should be on all factual issues.” 291 Or at 835.
Although it does so cursorily, Wolf offers some explanation for its limited remand. At the first trial, a verdict had been directed against the defendant on its affirmative defense to the claim of liability, because the evidence was insufficient to support the defense. Hence, there was no factual issue to retry with respect to liability, because the defendant lacked the evidence to create a question for the factfinder. Here, conversely, there was evidence at the first trial from which the jury could have found that defendants were not liable on the *791fraud claim. It is also noteworthy that, in Wolf, the error which necessitated the remand was raised by the cross-appeal of the defendant which had been found liable. Here, the error requiring that we remand is raised by the cross-appeal of plaintiffs from the judgment which is favorable to them on the liability question.
Most saliently, the court expressly stated in Wolf that the Maxwell rule remains the general rule. Given that statement in Wolf, I find it somewhat dumbfounding that the majority seems to regard Wolf as having overruled sub silentio a case about which it was not silent and the general rule of which it endorsed.
The only other recent case I have found that deals with the question of plenary or limited remands with anything more than a conclusory dispositional statement is Wilson v. B.F. Goodrich, 52 Or App 139, 627 P2d 1280 (1981), aff’d 292 Or 626, 642 P2d 644 (1982). There, we adverted to the tendency of other recent cases to disregard Maxwell, and we explained why we followed Maxwell and why we should:
“In decisions more recent than Maxwell, the Supreme Court and this court have remanded cases for retrial on damages alone, where the circumstances were hardly ‘exceptional.’ See, e.g., Weiss v. Northwest Accept. Corp., 274 Or 343, 546 P2d 1065 (1976); Criqui v. Pearl Music Company, 41 Or App 511, 599 P2d 1177, rev den 288 Or 173 (1979). The argument can be made that the quoted language from Maxwell presupposes that juries will not follow the court’s instructions. In our view, however, a more convincing argument can be made that Maxwell recognizes a predictable reality about jury behavior to which courts should not be blind in fashioning remedies.
“In any event, because the measure of damages in this case cannot be ascertained without evidence of liability on the negligence count and contributory negligence defense being produced, our remand is for a new trial of all issues.” 52 Or App at 153. (Emphasis supplied.)
Similarly, in this case, before the question of punitive damages can be reached, the jury must consider whether defendants’ fraud, if any, was intentional or reckless and, if the latter, whether it was accompanied by aggravating factors. In other words, the issue on which the remand is based cannot be retried without a presentation of all evidence about the *792conduct of defendants on which plaintiffs base their assertion that defendants are liable.
It is true, as the majority notes, that Maxwell itself states that there will be “exceptional cases” to which the general rule does not apply, but, as we suggested in Wilson, not every case can be an exceptional one. This case certainly is not. It involves a garden variety cross-appeal by plaintiffs from a judgment with which they are only partially satisfied. They have persuaded us that there was error, and the effect of the error is that all evidence germane to their fraud claim must be presented to a new trier of fact. If, contrary to Maxwell, that trier of fact is not to find all of the facts, the time has come for the Supreme Court to give Maxwell a proper burial, rather than leave a stated rule in place which the two appellate courts have come to follow or disregard on a completely ad hoc basis. The Supreme Court may overrule Maxwell. We cannot. Were I a member of the Supreme Court, I would vote to leave Maxwell intact.
As I have noted, part of the majority’s basis for concluding that the new factfinder should not be allowed to try all of the facts is that the majority perceives it as unfair for plaintiffs to have to prove liability again when liability was found at the first trial. I am at a loss to understand the majority’s premise that plaintiffs are entitled both to eat and have their cake. It seems to me to be a novel propositon that a party who appeals and demonstrates error is unfairly burdened by having to retry his or her case. That is a normal consequence of a successful appeal. Stated otherwise, a party who appeals runs the inherent risk of winning.
The point of Maxwell which apparently escapes the majority is that any unfairness that enters the picture arises when there is a limited remand, and the party on whom the unfairness falls is the unsuccessful respondent or cross-respondent. That is so for two reasons. The first is that the trier of fact’s findings on liability, damages and punitive damages are often interrelated and influence one another. The second is that seriatim trials and appeals through which one party gradually prevails, inch by inch, are inherently unfair to the other party. It is simply not a value of our system to structure things so that one party to a civil appeal risks everything and the other nothing, or to enable a party — in the *793words of Maxwell — “to manipulate errors in one trial to gain tactical advantage in a new trial before a new jury.” 253 Or at 577.1 do not suggest that that was plaintiffs’ objective here. I do suggest that the majority’s thesis, that an appealing party is entitled to put the favorable part of a judgment in the bank and seek reversal only of the unfavorable part, is rife with the potential for that kind of abuse of the system and the resulting potential for prejudice to the opposing party.
No case could better illustrate the extreme prejudice that a limited remand can achieve than this one. Maxwell states that liability and compensatory damages are interrelated in a jury’s deliberations and its findings; it is even more clear that liability and punitive damages are interrelated. That biasing factor is particuarly pronounced in this case, where the question of compensatory damages cannot go to the jury. See n 1, supra. In effect, the majority’s remand would permit that the jury be told (1) that defendants are perpetrators of a fraud, (2) that it not be told that the actual damage the fraud has occasioned has already been compensated by the first verdict and (3) that the only redress it can provide is an award of punitive damages. Compare the separate opinion of Peterson, J., in Wolf v. Nordstrom, supra.
A final concern that warrants consideration in determining whether remands should be limited or plenary is the preservation of judicial resources. I have noted earlier (and the majority does not assert otherwise) that, in this case, there is no possibility that the evidentiary showing the parties must make to resolve the punitive damages issue will not entail all of the evidence that would have to be presented if liability on the fraud claim were in issue. The trial court’s time will not be saved by letting the jury hear and see everything germane to liability but not letting the jury decide whether defendants are liable. On a more abstract plane, the kind of manipulative use of the process which partial remands encourage is hardly conducive to judicial economy.
For the foregoing reasons, I would hold that the remand on the fraud claim should be for a retrial on the questions of both liability and punitive damages, and I respectfully dissent from the portion of the majority opinion that holds otherwise.
Warren and Van Hoomissen, JJ., join in this opinion.would also include the issue of compensatory damages in the remand, were it not for the fact that the present judgment — if not the verdict — awards plaintiffs of damages alternatively on the breach of contract and the fraud claims in the full amount they sought.