Plaintiff was injured when a tire manufactured by defendant exploded while plaintiff was inflating it. The tire was a so-called “space saver” spare tire, different from tires used on regular automobile wheels, and instructions affixed to it stated that the tire and wheel should be fixed to the car or to a tire mounting stand before being inflated. There was evidence that plaintiff did not follow this and other precautions in attempting to use the tire.
The present action against defendant manufacturer, along with other defendants, was brought on separate allegations of negligence and marketing a dangerously defective product. Defendant’s answer alleged that plaintiffs injuries were caused by his “knowing, voluntary, and unreasonable express assumption of the risk” and his own negligence in certain specified respects. In a form of verdict containing nine separate questions, the jury found that defendant B.F. Goodrich was at fault for the design, manufacture, or sale of a dangerously defective tire, that it was negligent, that plaintiff was at fault “in misusing the tire which caused injury to plaintiff as contended in defendants’ answers,” that plaintiff was negligent, that the “percent of each of the parties’ fault which caused damage to plaintiff’ was 50 percent each, and that plaintiffs “total money damages” were $60,000. The jury found in favor of the other defendants.
Plaintiffs major assignments of error on appeal were that his alleged contributory negligence should not have been submitted to the jury in defense against his products liability count and that the trial court erred in striking expert testimony on plaintiffs loss of earning capacity. Agreeing with both contentions, the Court of Appeals reversed and remanded the case for retrial, 52 Or App 139, 627 P2d 1280 (1981), and we allowed review. We affirm the remand for a new trial but without the restriction imposed on the contributory negligence defense.
Claimant’s negligence under ORS 18.470.
We have dealt with the question whether an injured party’s negligence can diminish or defeat recovery on a products liability claim in Sandford v. General Motors, *629292 Or 590, 642 P2d 624 (1982), also decided today. As Sandford sets forth in detail, we understand the references in ORS 18.470 to the “fault” of the several parties as intending such an effect under certain circumstances.1 To summarize, ORS 18.470 applies to products liability claims like Sandford and the present case when both the product defect and the claimant’s conduct that is alleged as “fault” in fact joined to cause an unsegregated injury. Each must be found to have been a necessary cause of the injury, assuming that this is open to dispute. See Sandford, supra, 292 Or at 601-602. A claimant’s conduct can be “fault” for being negligent as well as in other respects, except that it will not reduce or defeat a products liability claim when unobservant, ignorant, or awkward failure to discover the defect or to guard against it is the kind of user conduct that is considered in finding the product dangerously defective in the first place. See Sandford, supra, 292 Or at 598. When ORS 18.470 applies, it calls for the factfinder to assess the relative magnitude of the fault charged against each party as measured against the respective governing norm, the degree to which a product is defective against what would be an adequately safe product, and the plaintiffs misconduct against what would be faultless conduct. See Sandford, supra, 292 Or at 607.
In this case, there was evidence from which a jury could find that the tire was dangerously defective in design or in manufacture and that but for this defect, the tire would not have exploded or caused the same injuries when plaintiff inflated it as he did.2 The jury also could find that *630plaintiffs attempts to install and inflate the tire were negligent in a number of respects and that the defect otherwise would not have caused plaintiffs injuries. This therefore was a proper case for instructions under ORS 18.470 and ORS 18.480.3
Expert testimony on loss of earning capacity.
At the time of his injuries plaintiff was just short of 20 years old. He had left high school with poor grades at the beginning of his third year, worked briefly in a McDonald’s restaurant, and served two years and four months in the Army. After returning home, he found employment as an assembly worker in an aluminum door and window plant at a wage of approximately $2.30 an hour. He -had worked for three days when he was disabled by his accident with the tire.
*631There was evidence that plaintiff suffered brain injuries that would permanently disable him from gainful employment. Plaintiff called an economist, Dr. Russell Dawson, as an expert witness on the financial measure of his consequent loss of earning capacity. Over defendants’ objections Dr. Dawson, upon certain stated, assumptions, placed the value of plaintiffs lost lifetime income at $597,866.00. He also placed a present value of $105,576.00 on the cost of future nursing assistance to plaintiff. At the conclusion of the trial, the court granted defendant’s motion to strike the economist’s testimony and exhibits and instructed the jury that this evidence was withdrawn from their consideration.
The trial court considered itself bound to strike Dr. Dawson’s testimony on plaintiffs loss of earning capacity by this court’s decision in Plourd v. Southern Pac. Transp. Co., 266 Or 666, 513 P2d 1140 (1973). The parties and the Court of Appeals have analyzed the ruling in the light of that decision, its contemporary, Conachan v. Williams, 266 Or 45, 511 P2d 392 (1973), and its sequel, Plourd v. Southern Pac. Transp. Co., 272 Or 35, 534 P2d 965 (1975), as well as language found in earlier cases. There is no need to repeat the analysis at length here. The question is how it applies to evidence adduced to estimate the earning capacity of a young person with no prior employment history and no special educational background or skills pointing toward any specific future occupation or employment.
The general principles are not disputed. The testimony of an expert is admissible if it can be of appreciable help to the factfinder. Koch v. Southern Pacific Co., 266 Or 335, 341, 513 P2d 770 (1973); see also State v. Stringer, 291 Or 527, 633 P2d 770, aff’d on reh 292 Or 388, 639 P2d 1264 (1982). In principle, expert testimony is admissible to assist the factfinder in placing a present value on future earning losses. In doing so, an expert may testify to economic assumptions about the future, such as inflation, wage levels, and interest rates. Plourd v. Southern Pac. Transp. Co., 266 Or at 676-679. These assumptions necessarily rest on estimates and predictions of uncertain future events, whether made by a witness or by a factfinder without expert assistance, that weakness can be explored by cross-examination or contrary evidence; it does not make the expert testimony inadmissible. Id. at 678-679.
*632The problem lies in relating general economic data and forecasts of future trends to the prospects of the unique individual in a given case. When the evidence offered to show the probable earning capacity of one individual was the earning records of other individuals in similar employment, the prior decisions have been concerned that there be a showing of substantial similarity of earning capacity between the respective individuals. Thus Conachan v. Williams, supra, found error in the admission of evidence of the average earnings of other division managers of an insurance company as proof of the earnings capacity of an injured manager, when these earnings depended on the production of salesmen working under each manager and there was no evidentiary showing that the situations of the several division managers were substantially similar. Similarly a majority in Plourd v. Southern Pac. Transp. Co., supra, found error in admitting a computation of a railroad brakeman’s loss of earning capacity based on the earnings of another brakeman, in view of an unexplained history of lower earnings by the injured brakeman. 266 Or at 687. On retrial, plaintiff offered explanatory evidence on that issue, and this court held that the computations then had a sufficient foundation of similarity of circumstances. Plourd v. Southern Pac. Transp. Co., 272 Or 35, 40-42, 534 P2d 965 (1975).
The problem in the present case is not like that in Conachan or Plourd. The occupations of those men were well established. The computations in their cases were offered to prove what they might have earned in those occupations by showing the actual earnings of other individuals in the same occupations. They did not purport to estimate the injured men’s potential earning capacities in other lines of work. Cf. Conachan v. Williams, 266 Or at 61-62.4 In making such direct comparisons with the *633earnings of other individuals, it was deemed necessary to support the proffered comparison with evidence that the individuals being compared in fact were comparably skilled, diligent, or otherwise similarly situated.
This question of comparability between identified individuals, however, often will be distinct from the inferences that an expert may draw from the comparison and not itself be a matter of expertise. But in the present case no comparison with the actual earnings of specific persons was offered. Steven Wilson had been employed for only three days at his first regular job, before he turned twenty years old. Evidence of his earning capacity could not well be limited to projecting a lifetime wage for elementary assembly work because that happened to be his work at the moment. It might indeed be less as well as more. But when assessing the earning capacity of a young person who has not prepared for or settled into even an initial line of work, the factfinder necessarily must draw some inferences from the available evidence as to the individual’s likely level of earnings and continuity of employment over many years. The factfinder must do so whether those inferences are aided by expert testimony or not. Of course proof of lost earning capacity is part of the plaintiffs burden of proof, but the adequacy of evidence to support these inferences is a different question from whether expert testimony is admissible to aid them.
The Court of Appeals cited the decision of the Montana Supreme Court in Krohmer v. Dahl, 145 Mont 491, 402 P2d 979 (1965), which dealt with the earning capacity of a deceased college student, for the proposition that in the absence of an established record of actual employment expert testimony could be based on the possible earnings of “classes of people of the same type as the decedent if they had survived through their normal life expectancy.” The Montana court saw the issue before a trial court as being “whether the testimony of [the expert] should be allowed, in order to give the jury some basis upon which to reach a conclusion in regard to possible future earnings of the decedent, or whether to leave the jury unguided and hope that by their common knowledge and sense of justice they might arrive at a more accurate estimation of damages,” and it concluded that the expert’s *634testimony could only reduce the element of conjecture. 145 Mont at 495, 402 P2d at 981.
For the same reasons, Judge Richardson’s opinion for the Court of Appeals concluded that in a case like the present, “any failure by the offering party to demonstrate substantial similarity between himself and his expert witness’s statistical model should go to weight rather than admissibility.” It continued:
“The substantial similarity of circumstances requirement for comparative statistical evidence may make little sense in cases involving minors or young plaintiffs without work histories. That requirement presupposes that specificity, which is a necessary prerequisite to substantial similarity, is either essential to or a measure of relevance. That presupposition is logical enough when a plaintiffs work history, education and training level and age level make it probable that his future employment, had he not been injured, would be identical or similar to his employment when injured. The presupposition is not logically compelling when the plaintiff is still in a formative stage of life and has not committed himself to any particular line of work; indeed, the less definable a young plaintiffs future is, the less similar any specific model can be and, therefore the less relevant any such model can be to predicting his future earning capacity.
* * * *
“The better course, in our view, would be to require the assumptions underlying a young plaintiffs statistics to be presented to the factfinder, along with evidence from which the factfinder can test the validity of those assumptions. In other words, whether a young plaintiffs statistical evidence bears sufficient resemblance to his circumstances to be entitled to weight in the factfinding process — unless the evidence is not even colorably related to the plaintiffs future prospects — should be a question for the trier of fact to decide.”
52 Or App at 147-8. We agree with these statements.
Problems predictably can arise in determining what are “classes of people of the same type” as the injured person, in the words of the Montana court. Should “college student,” as in Krohmer v. Dahl, supra, be regarded as a “class” or “type” without regard to field of study, academic record, or other indications of probable occupational *635qualifications? Should an expert testifying about “classes” or “types” take into account statistical employment expectations by sex or ethnic origins? The college student may be a basketball player or a basket weaver, a future unemployed architectural historian or a millionaire builder. A skilled as well as an unskilled worker may find himself in an industry with steady employment, good wages and benefits, and opportunities for further training or in a shrinking industry beset with seasonal or cyclical unemployment or geographic dislocation.
The witness in this case made clear that his testimony rested on very simple and undifferentiated premises. He postulated the national average hourly wage rate, including fringe benefits, for factory production work not involving special skills. He explained that an individual worker’s actual wage would be lower than this average when he first begins work and higher than the average in the later years of a working career. He testified that, over extended periods of time, the rate of increases in this average wage have approximated the interest rate by which lifetime earnings must be discounted to arrive at their present value. Accordingly the discount rate could be disregarded, and the present loss of earning capacity could be reasonably approximated as equal .to the actual loss of lifetime earnings. The witness then multiplied the postulated hourly wage rate by the average expected working life of a man of plaintiffs age, reducing this by allowing three years for periods during which such a person might be unemployed for one or another reason. The multiplication produced an estimate of lost lifetime earnings for a factory worker of plaintiffs age in the amount of $707,533, reduced by estimated federal and state income taxes to a total after tax income of $597,866. As already mentioned, Dr. Dawson equated this to lost earning capacity on the premise that average wages historically have risen at a rate equal to the discount rate.
Obviously this testimony is very different from the evidence of the income of specific individuals offered for comparative purposes in Conachan and Plourd, supra. The witness clearly stated that he was presenting general national statistics for factory workers of all kinds. Defendants’ counsel subjected his assumptions and methodology *636to searching cross examination. Any weaknesses in these respects would call for argument and opposing evidence, not for exclusion of his testimony. The crux of the dispute is whether calculations based on average factory wages were relevant to this plaintiffs situation.
Again, Dr. Dawson disavowed any intention to evaluate Steven Wilson as an individual future worker. He stated that his testimony was based on the general occupational potential of a 20-year old with two years of high school attendance and prior military service. He had no personal knowledge of plaintiffs past record or wage rate when injured. On cross-examination, the witness was asked whether certain assumed facts would affect his estimate, for instance, if the individual had a record of drug or alcohol abuse or of poor performance in school or in the military. He answered that this might indeed reduce the individual’s earning capacity below the statistical assumptions.
In support of the withdrawal of Dr. Dawson’s testimony from the jury, defendant argues that the role of the economist is to provide “statistical assistance” but not to fill factual gaps in the foundation for his calculations. In the calculation of lost earnings capacity from projected future earnings, that distinction does not draw a clear and simple line. The assumption that someone with ten years of school ordinarily can qualify at least for nonsupervisory, low-skilled factory work was itself presented as a statistical datum of labor economics, not as an assertion about Steven Wilson. The need for more individualized evidence increases as the expert’s testimony is focused more narrowly, for instance, if he were asked to assume that an individual would work in a plywood plant instead of merely in some kind of production work, or that an employee might shift from an existing occupation into a specific different one or succeed in a narrow and competitive specialty such as the performing arts, as in Weinstein v. Wheeler, 127 Or 406, 257 P 20, 271 P 733 (1928). It is the very generality of this economist’s statistical assumption that saves its admissibility here. For in the case of a young person disabled before establishing any regular work experience it surely is not less relevant to inquire in the most general terms what kind of work and earnings others of *637similar background statistically obtain over their working lives than to assume that such a person would work a lifetime at the statutory minimum wage, if at all.
We speak only of admissibility; of course the fact-finder need not believe that a particular disabled person’s experience would approximate the general norm. If there are reasons why it probably would not, they can be brought out. The homogenization of statistics is as apt to understate as to overstate an individual case. The distribution of actual numbers around the mean can be explored in more informative detail than happened in this case. None of this makes the expert’s testimony inadmissible under the test whether it might be of appreciable help to the factfinder in projecting the value of a young person’s earning capacity over his working life. The answer to the charge of speculativeness is that the factfinder must arrive at such a conclusion with or without help, see Conachan v. Williamson, supra, 266 Or at 64; the question is whether the offered evidence is more help than it is misleading or confusing.
This court has said that the trial court has some latitude of judgment in determining whether any given expert’s testimony is of a kind that will meet this test. State v. Stringer, supra, 292 Or at 394; Cooney v. McGee, 268 Or 521, 525, 521 P2d 1051 (1974); Yundt v. D & D Bowl, Inc., 259 Or 247, 259, 486 P2d 553 (1971). If the trial court had excluded Dr. Dawson’s testimony for that reason, we would face the question of the scope of appellate review of that latitude. In this case, however, Judge Unis initially admitted the testimony and later withdrew it from the jury’s consideration only under what he believed to be the compulsion of the prior holdings of this court, specifically in the Plourd cases, supra. After searching for a distinction he ultimately decided “reluctantly” that Dr. Dawson’s testimony and exhibits had to be stricken. He did not reach that conclusion by exercising his own judgment on the relevancy and potential helpfulness of the evidence.
This case differs from Plourd for the reasons set out herein and in the opinion of the Court of Appeals. We do not deal with the admissibility of every detail of the offered testimony and exhibits. We hold only that it was *638error to withdraw it from the jury’s consideration in its entirety, and that this withdrawal of evidence relevant to plaintiffs loss of earning capacity requires a new trial. The decision of the Court of Appeals to this effect is affirmed.
ORS 18.470:
“Contributory negligence shall not bar recovery in an action by any person or his legal representative to recover damages for death or injury to person or property if the fault attributable to the person seeking recovery was not greater than the combined fault of the person or persons against whom recovery is sought, but any damages allowed shall be diminished in the proportion to the percentage of fault attributable to the person recovering. This section is not intended to create or abolish any defense.”
The products liability count of the complaint also alleged that defendant knew that the “space saver” tire had caused serious injuries but “continued to use instructional labels rather than warning labels.” Possibly this might intend to allege that the tire was dangerous by reason of defective warning, but the allegation appears only in a section of the complaint demanding punitive damages. By contrast, the second count alleged inadequate warning as a specification of negligence. The verdict on the products liability count only referred back to the complaint without specifying the alternative allegations of defect.
ORS 18.480:
“(1) When requested by any party the trier of fact shall answer special questions indicating:
“(a) The amount of damages to which a party seeking recovery would be entitled, assuming that party not to be at fault;
“(b) The degree of each party’s fault expressed as a percentage of the total fault attributable to all parties represented in the action.
“(2) A jury shall be informed of the legal effect of its answer to the questions listed in subsection (1) of this section.”
Plaintiff objected only to submitting contributory negligence in a products liability case; neither party objected to the form of the verdict. Cf. Sandford, supra, 292 Or at 593.
Defendant also pleaded that plaintiffs injuries were caused by his “knowing, voluntary and unreasonable express assumption of the risk that the tire could cause injury if inflated” as alleged (emphasis added), and by “plaintiffs misuse of a Space Saver Tire” in specified respects. The verdict form submitted to the jury a question on “misusing the tire” separate from a question on plaintiffs negligence.
It is not clear whether “express assumption of the risk” was meant in a sense other than that plaintiff expressly stated to a defendant his willingness to assume the risk of any defect in the tire, cf. Restatement (Second) of Torts § 496 B, nor whether “misuse” here was meant in any sense different from negligence in attempting to use the tire for its intended purpose. Cf. Anderson v. Klix Chemical, 256 Or 199, 208-209, 472 P2d 806 (1970), recognizing “misuse” as a defense to a products liability claim before the enactment of ORS 18.470 and 18.475, at a time when the status of contributory negligence as a defense was unsettled. Voluntary disregard of cautionary instructions might raise an issue of causation if a product was alleged to be defective only in having inadequately prominent warnings. Cf. McEwen v. Ortho Pharmaceutical, 270 Or 375, 407, 528 P2d 522 (1974). No issue about the quoted terms is before us, but because the case may be retried, we do not want to be understood as having endorsed the specific details of the manner in which these issues were submitted.
See also Brown v. O. W.R. & N. Co., 63 Or 396, 409, 128 P 38 (1912), quoted in Conachan v. Williams, 266 Or at 59-60:
“A fair rule would seem in cases of this character to be that any evidence which would indicate fairly the capacity of the plaintiff to earn money in his usual vocation, and the probability of his being able to do so in the future should be admitted; but, where such evidence consists of mere guesswork and speculation upon what might happen in the future, it should be excluded. Such testimony in any court is seldom, or never, conclusive, and merely furnishes one factor in solving the equation of a man’s earning capacity.”