*283ON REHEARING EN BANC
Before GODBOLD, Chief Judge, BROWN, CHARLES CLARK, RONEY, GEE, TJOFLAT, HILL, FAY, RUBIN, VANCE, KRAVITCH, FRANK M. JOHNSON, Jr., HENDERSON, REAVLEY, POLITZ, HATCHETT, ANDERSON, RANDALL, TATE, SAM D. JOHNSON, THOMAS A. CLARK and WILLIAMS, Circuit Judges.** JOHN R. BROWN, Circuit Judge:This case comes before us on rehearing en banc to consider whether the holding of this Court in Johnson v. Penrod Drilling Co., 510 F.2d 234 (5th Cir. 1975) (en banc), that neither proof, nor argument, nor jury instructions concerning inflationary factors may be considered or used in maritime, Jones Act, and FELA personal injury and wrongful death actions, should be overruled.1 After careful consideration of this singular issue, we overrule Penrod and remand this case to the District Court.
I.
Bound by Penrod
The facts leading up to this appeal have already been outlined by this Court in the panel’s opinion, 644 F.2d 460, 462-63 (5th Cir. 1981). Briefly, Curtis Culver was killed while working on a vessel owned by Slater Boat Company. The vessel upon which the accident took place was engaged in moving the drilling barge OCEAN QUEEN from its location on the Outer Continental Shelf to a new location. The fatal injury occurred before the barge was actually moved. Culver’s widow and children brought suit under the Jones Act, 46 U.S.C. § 688, the Death on the High Seas Act (DOHSA), 46 U.S.C. § 761 et seq., and the general maritime tort and negligence theory. The jury found negligence on the part of Gulf Overseas Marine Corporation (Culver’s employer), Euro-Pirates International (the charterer of the vessel), and Ocean Drilling & Exploration Company (the owner of a barge involved in the fatality), and no contributory negligence on the part of Culver.
On the issue of damages, the District Court allowed testimony concerning discount rates and the earning power of money invested in low risk bonds. The jury was instructed to “discount the total amount” of any award by a percentage that represented an appropriate rate of interest. In answer to the special interrogatory submitted asking what discount percentage rate was applicable, the jury filled in “25%”. The trial judge, on the basis that the jury obviously misunderstood the interrogatory, substituted 9.125%, the only other rate put into evidence by the defendant. In accordance with Penrod, the District Court did not allow testimony, charges, or interrogatories to be submitted on the effects of inflation on probable loss of future income.
In short, Culver was not permitted to show any likely increase in future earnings due to inflation. But the award was to be discounted by an interest factor reflecting anticipated inflation.
Culver initially appealed the judgment, raising five issues: (i) should Penrod be overruled?; (ii) if Penrod is not overruled, should evidence of probable non-inflationary future wage increases (e.g. merit raises) be prohibited?; (iii) can a District Court disregard a jury finding regarding the discount rate and apply one based on opinion testimony?; (iv) was the testimony of Culver’s adverse witness sufficiently clear that the court could apply that witness’ opinion of the discount rate?; and (v) can a District Court enter a final judgment for damages applicable to all beneficiaries in a maritime death action that is incapable of apportion*284ment among the various beneficiaries?2 Cross-appeals were brought by all of the defendants raising several additional issues.3
Oral argument was heard by a panel of this Court, and the District Court judgment was affirmed as slightly modified. Specifically, the panel considered itself bound by Penrod’s holding that “the influence on future damages of possible inflation or deflation is too speculative a matter for juridical determination,” 644 F.2d at 643, quoting Penrod, 510 F.2d at 241. And in accordance with Byrd v. Reederei, 638 F.2d 1300 (5th Cir. 1981) (rehearing en banc granted), the panel rejected Culver’s argument that the Supreme Court overruled Penrod in Norfolk & Western Railway v. Liepelt, 444 U.S. 490, 100 S.Ct. 755, 62 L.Ed.2d 689 (1980).4 Likewise, the panel found no error in the trial judge’s refusal to allow Culver to argue likely future wage increases on the basis of merit, because no evidence was offered to show that such an argument was warranted and, in addition, such evidence was “merely an indirect way of putting inflation factors into evidence before the jury [which is] not allowed under Penrod.” 644 F.2d at 464.5
*285Culver filed a petition for panel rehearing,6 and a separate petition for rehearing en banc, as required by our rules, raising two closely related issues: (i) should Penrod be overruled and (ii) does Penrod prohibit a trier of fact in this Circuit to receive evidence of non-inflationary factors, such as probable merit raises and productivity increases, in arriving at future losses? This Court, voting for rehearing en banc, determined that the time was ripe for reconsideration of the rule in Penrod pertaining to the “inflation factor” in damages awards. Although Culver’s brief to this court on rehearing en banc addressed primarily issue (i), whether proof and argument concerning inflationary factors should be permitted in this Circuit, issue (ii) is also involved since automatic exclusion of evidence of probable merit raises has resulted from a misreading of Penrod. The defendants argue that Pen-rod should be upheld, and, in addition, contend that Culver waived the right to relief on the inflation issue by failing to make an offer of proof regarding the likelihood of inflation.
A.
Briefly, we will address the defendants’ claim that Culver waived the right to raise the issue of inflation by his failure to make a formal proffer at trial. The defendants called as a witness an investment banker who specialized in bonds. On cross-examination, Culver’s attorney attempted to ask whether people earned more money over the years of their employment due only to productivity. An objection to the question was sustained. After several more questions, all counsel approached the bench where the judge made it clear that the type of evidence excluded by Penrod would not be permitted in court. The jury was then temporarily excused, and Culver’s attorney told the judge that he wanted to ask the expert whether the principal of a bond would be worth less in the future. The trial judge replied that deflation was as likely as inflation, and any testimony as to such factors would be pure speculation. The judge concluded that “I will not permit you to introduce any evidence before the jury relative to inflation in view of the present law that I am bound by which is the Penrod case.” Before closing argument, and before the jury returned, the judge explained that:
Plaintiff’s counsel will not be permitted to argue that the jury should take into consideration inflation, nor will it be permitted to make an indirect argument which would achieve the same purpose of permitting him to argue inflation by attempting to indicate that it could be expected that plaintiff would receive job promotions or merit increases.
Again, the judge mentioned the prohibitions of Penrod. Given this flat prohibition against introducing testimony (or making argument) as to future inflationary trends on the basis of Penrod, it makes no sense to argue that Culver may not raise this issue on appeal simply because he did not proffer an expert on inflation. Reading the trial transcript, it is obvious that a record was being created to challenge or at least find an exception to Penrod. We find that the matter was sufficiently presented to the judge at trial and that all parties understood the dilemma — Culver wanted the jury to consider inflation, and Penrod stood in the way.
B.
The most problematic issue on this appeal remains: Should Penrod be overruled so as *286to allow parties to present proof and argument concerning inflationary factors to the trier of fact? Our analysis will begin with a brief review of the Penrod decision. In order to illustrate graphically the implications of Penrod and why it has to be overruled, we will discuss many of the cases in this Circuit where Penrod’s prohibitions were effective. Next, we will summarize the criticisms of Penrod which come from other circuits as well as commentators. As our analysis proceeds, it will become clear that the problem is one of fairness to plaintiffs as well as defendants in the trial process. Personal injury awards, once they have been calculated on the basis of projected income and life expectancy, are paid immediately to the plaintiff. It is well-known that if a plaintiff (or his beneficiaries) receives a lump-sum award, totaling the income that the plaintiff would have received throughout the rest of his work life, that money can be invested so as to yield a far greater amount than the initial award. This would be unfairly liberal to the plaintiff. Consequently, the law traditionally permits introduction of testimony, usually by an expert in financial matters, that the lump-sum should be discounted by a factor equal to the interest rate which could likely be earned on a relatively safe investment by an unsophisticated investor. The total projected earnings are thus reduced to the present cash value. However, it will be demonstrated that in an inflationary economy, discounting an award in this manner results in unfairness to a plaintiff or his beneficiaries. When a discount rate is applied to an award in an economy where wage-earners typically receive cost of living increases each year due to inflation, the plaintiff will not be adequately compensated for his loss of future income — inflation will erode the value of the award, and no adjustment for cost of living can be made even though wages continue to increase throughout the economy. On the other hand, if courts respond to this dilemma by refusing to apply a discount rate, as some have, the result would be unfairness to defendants. Quite simply, plaintiffs are to be compensated, by a culpable tortfeasor, but should not be over- or under-compensated. Likewise, defendants found liable should pay no more or less than the amount a plaintiff lost because of the injury. Keeping in mind the goal of fairness to both sides of a controversy, this opinion will consider alternatives and possible guidelines for use in this Circuit. We will establish a flexible approach to the problem which should result in fairness regardless of the economic circumstances that exist at the time of trial. In the process, we overrule Penrod.
II.
Prolegomena
It is crucial to keep in mind that this is really an economic, and not a legal, problem. The likelihood of future wage increases, whether given on the basis of productivity, merit advancement, or inflation, or a combination of them, does not have anything to do with the law. The legal question whether such matters should be considered by the courts is a foundational policy question, the answer to which will be influenced by recourse to the financial and economic community. At trial, the issue may well be a legal credibility of witnesses or competency of evidence, but a pronouncement that inflation is speculative or that workers will or will not receive productivity and merit raises is essentially an economic statement, not a declaration of law.
If we consider, as an example, a person totally and permanently injured by a defendant’s negligence, the injured worker is entitled to the equivalent of his lost future earnings. The law decides that much, and the question remains how much money should be given to him now. Once a lump-sum determination- is made, it is clearly unfair to give him the whole amount, because even an unsophisticated investor could earn a great deal in interest. However, in predicting how much money this worker will likely receive throughout the remainder of his work-life, we must consider the likely increase in wages he would have received. These will include possible *287advancements from one position to another, any raises due to company productivity, or other merit increases. In addition, in an inflationary economy, we must consider projected cost of living increases which are given by an employer to offset in part the effects of inflation. Each of these factors must be carefully distinguished, since they arise from wholly different causes. Although these factors may not all be present in a particular case, each must be considered in computing the total amount of likely earnings before a discount rate is applied to reduce that amount to its present cash value because of investment potential. The goal, albeit ideal, is that if the injured worker has a 20-year work-life expectancy, and he invests his lump-sum award in relatively risk-free investments, he will receive in the nineteenth and in the twentieth year roughly the same amount as if he had worked each of those years. The goal is not, it must be made clear at the outset, to protect the lump-sum award from the effect of inflation. Rather, before determining how much now needs to be paid, the goal is to assure the plaintiff the equivalent of the total of all' of his future wages, including those likely to be given/received in the form of cost of living increases in response to inflation.
III.
The Prohibitions of Penrod
In Johnson v. Penrod Drilling Co., 510 F.2d 234 (5th Cir. 1975) (en banc), we recognized “the likelihood that inflation could become a predictable condition for the future,” but we were not able to “so surely discern the shadow of inflation as a coming event as to warrant requiring its inclusion in a present rule for calculating future damages.” 510 F.2d at 236. Quoting the full initial panel opinion in Penrod, we expressly disapproved “the district court’s attempt to take into account, in computing the plaintiffs’ future lost earnings, inflationary trends in this nation’s economy for the next several decades.” 510 F.2d at 241. The “influence on future damages of possible inflation or deflation is too speculative a matter for judicial determination.” Id. Therefore, triers of fact “should not be instructed to take into account future inflationary or deflationary trends in computing future lost earnings, nor should the jury be advised to consider such alternative descriptions of inflationary and deflationary trends as the purchasing power of the dollar or the consumer price index.” Id.
The defendants and others have suggested that Penrod could be “modified”, “explained”, or watered down by this Court, but not overruled, so that adjustments could be made in individual cases where the influence of inflation was not speculative. However, if fairness requires that we allow some evidence of inflation properly to compensate plaintiffs and avoid windfalls to defendants, any hope of “distinguishing” or “explaining away” Penrod crumbles in the face of Penrod’s own words — inflation and deflation are too speculative for judicial determination, triers of fact should not take into account future inflationary or deflationary trends, juries should not consider changes in the purchasing power of dollars or the consumer price index, and judges should not undertake to instruct juries on such ideas. The prohibition on such evidence could not be clearer.
To argue, as the defendants do, that Pen-rod forecloses only consideration of future economic trends, while it allows evidence of past inflation, is to miss the point. The significant basis for predicting future inflation or deflation is past inflation or deflation. And the effect of past inflation alone, under a rule forbidding consideration of likely future inflation, leaves economically eclipsed a significant portion of the amount of money the wage-earner may receive in the future.
Even if it were possible to isolate Penrod and analyze its terms in an academic vacuum, to reach the conclusion that evidence of inflation is permissible in the Fifth Circuit so long as the jury is not permitted to “speculate” as to the “influence” on future damages of “possible” inflation or deflation, the result of this exercise would be unworkable as a practical matter. For in practice, *288Penrod has placed a prison-like wall between any evidence of likely inflationary or deflationary trends (and argument or instruction thereon) and the courtrooms of the Fifth Circuit. Our decisions since 1975 on this issue have merely served to place barbed wire on top of the wall. Moreover, it is likely that in other cases, as in this one, evidence of probable increases in wages due to merit or productivity has been kept out of the jury room unwittingly because of Penrod’s prohibitions. Economic data is readily available concerning past average annual wage increases of workers in the United States, and such data is often broken down into particular professions and geographic areas. But it is difficult to determine the percentage of those increases given for merit-productivity and distinguish the percentage representing cost of living increases in response to inflation. Many plaintiffs have thus been unable to scale Penrod’s wall and introduce evidence of non-inflationary wage increases, even though such increases are theoretically unaffected by inflation, and thus evidence of such increases should not have been prohibited.
IV.
Penrod in Practice
In our determination of whether Penrod ought to be overruled, it is important not only to consider what Penrod says, but also what this Court through the years since 1975 has said it says. We will look first to the recent cases arising in this Circuit applying Penrod. In Davis v. Hill Engineering, Inc., 549 F.2d 314 (5th Cir. 1977), this Court reduced the District Court’s award of $628,991 to $425,321 for the total loss of the plaintiff’s future earning capacity, because the District Court increased the basic damage award, prior to discounting to present value, to offset wage and price inflation. The plaintiff’s expert was unimpeached, and the defendants had no opposing expert. The trial judge was thus free to rely on the expert’s testimony, and was apparently convinced of the need to allow over $200,000 to offset expected wage increases due to inflation. 549 F.2d at 331. Nevertheless, we held that until Penrod is overruled,
an inflation element cannot be included in damage computations, either in the form of [i] calculating loss of future earnings without discounting to present value or of [ii] increasing the basic damage award, prior to discounting, with a figure representing the projected inflation rate. The former [i] is the Beaulieu [v. Elliott, 434 P.2d 665, 671 (Alaska 1967) ] method, the latter [ii] the District Court’s method in this case.
549 F.2d at 332 (brackets added). Judge Wisdom acknowledged that other circuits consider inflation as a factor in computing future earnings damages, an approach that “more accurately [reflects] loss of future earnings than does a discounted figure because it prevents inflation from eroding the damage award.” 549 F.2d at 332. Alas, he conceded, his hands were tied by Penrod. Id.
In In the Matter of S/S HELENA, 529 F.2d 744, 753 (5th Cir. 1976), this Court remanded for recalculation the District Court’s damages award because an inflationary rate was assumed and applied. In determining the damages for loss of support payable to the wives of deceased crew members of a vessel, the trial judge stated:
I have to some degree offset discount to present value by the effects of inflation and a likely rise in decedent’s earnings resulting from increases in the general wage level. I have assumed a discount rate of 5%, and a three per cent inflationary level thus making the net discount rate 2%.
529 F.2d at 753, quoting 329 F.Supp. 652 at 660. Other calculations of the damages for loss of support also “seemed to be based on the inflation and present value factors, the expert testimony of an economist as to expected future earnings, and the present cost of annuities that would pay monthly amounts increasing at the estimated rate of inflation.” 529 F.2d at 753. Quoting Pen-rod, we remanded to District Court for a more detailed calculation of the damages. On remand, the District Court filed supple*289mental responses heeding our instructions, and Penrod was again faithfully followed. Matter of S/S HELENA, sub nom. Sincere Navigation Corp. v. United States, 547 F.2d 255, 256 (5th Cir. 1977). The element of inflation was eliminated and the damages recomputed. Id.
In Higginbotham v. Mobil Oil Corporation, 545 F.2d 422, 433-35 (5th Cir. 1977), rev’d and remanded on other grounds, 436 U.S. 618, 98 S.Ct. 2010, 56 L.Ed.2d 581 (1978), this Court reviewed the District Court’s use of a 5% annual straight line estimated salary increase to calculate the probable future earnings of the deceased. Although the trial judge stated that he did not take into consideration “the decreasing purchasing power of the dollar,” we held that under Penrod “the plaintiff must bear the difficult burden of proving what portion of the increases would have been given other than as an automatic hedge against inflation.” 545 F.2d 434-35. The defendant, wielding the mighty sword of Penrod, made the factually unsupported argument that “it is common experience that annual ‘raises’ are for the most part cost of living wages,” therefore a 5% straight line annual increase amounted to a hidden inflation-based award. 545 F.2d at 434. While common experience likewise tells us that some portion of pay raises reflect performance and experience, the plaintiff’s expert failed to distinguish between cost of living or inflationary increases and rewards for experience or productivity. 545 F.2d at 435. Thus we rejected the District Court’s award based on the 5% future earnings factor, and remanded to give the plaintiff an opportunity to walk the straight and narrow of Penrod. Judge, now Chief Judge, Godbold, in his partial dissent, observed that in “innumerable other contexts we accept the events of the past as the basis for an inference as to what the future will hold.” 545 F.2d at 437 (Godbold, J.). “Yet we would not permit consideration of evidence that a particular plaintiff had received a $200 per year increase in pay every year for 20 years, unless he could divide it up into ‘productivity’ pay and ‘economic change pay.’ ” Id. (Godbold, J.).
In this single area of projecting future earnings we deny ourselves the best evidence available on the asserted ground that it is not sufficiently reliable, and, in the name of reliability we mandate the artificial conclusion that one will earn the rest of his life what he is earning on the day he is killed or injured. The only thing certain about this is that it is certain to be wrong.
Id. (Godbold, J.).
In Menard v. Penrod Drilling Company, 538 F.2d 1084, 1089 (5th Cir. 1976), the plaintiff’s economic expert estimated that, apart from the possibility that the plaintiff might advance through the ranks of his employer, a worker “would have to get at least 2% per year, on the average, increase to maintain the same buying that he has to-day.” A discount rate of 4}A% was selected by the expert. Id. Because no objection was made, the jury was actually permitted to consider future inflation as a factor in calculating damages. One member of this Court’s panel observed that although this was not plain error, it would have been reversible error under Penrod had an objection been made. Id. (Gee, J., concurring).
Continuing our historical journey into the past, this Court in Lacaze v. Olendorff, 526 F.2d 1213, 1222 (5th Cir. 1976), reh’g en banc granted, 526 F.2d at 1223, found error in the trial court’s overruling an objection to expert testimony on future lost earnings because the expert included a 3% inflation rate in his calculations. On the basis of Penrod the case was reversed and remanded on the issue of damages. 526 F.2d at 1223.
In Weakley v. Fischbach & Moore, Inc., 515 F.2d 1260 (5th Cir. 1975), a diversity suit arising out of an electrical explosion in Texas, we affirmed the judgment of the District Court even though the jury considered both future productivity and future inflation in assessing a $300,000 award for lost earning capacity. For this Court, Judge Wisdom observed that although this “Court has writ large its disapproval of calculating future damages by reference to *290predictions of future inflation,” Texas law applied, and juries in Texas are permitted to weigh evidence of future inflation in setting awards for future damages. 515 F.2d at 1266-67, citing Penrod. In a similar federal cause of action we may be certain that this Court would have reversed a judgment based on such a “speculative” jury finding.
Again, specifically on the basis of Penrod, we concluded in Petition of M/V ELAINE JONES, 513 F.2d 911, 912 (5th Cir. 1975) cert. denied, 423 U.S. 840, 96 S.Ct. 71, 46 L.Ed.2d 60 (1975), that the District Court erred by including a 2% per year cost of living increase in the computation of loss of future earnings. We announced that “the trier of fact should not be instructed to take into account future inflationary or deflationary trends in computing future lost earnings.”
In Standefer v. United States, 511 F.2d 101 (5th Cir. 1975), the District Court, sitting without a jury, awarded damages to a plaintiff for loss of future earnings and included an inflationary factor of 5.5%; in addition, the award of future medical expenses included an inflationary factor of 4.5%. Because Penrod foreclosed any consideration of inflation in computing such damages, the District Court’s decision was reversed and remanded for a recomputation of damages.
Robertson v. Douglas Steamship Company, 510 F.2d 829 (5th Cir. 1975), stands as yet another example of how we applied the Penrod strait-jacket. In accordance with the trial court’s instructions, the jury divided the plaintiff’s award into (i) the amount of basic damages and (ii) the amount of damages sustained “as a result of loss of future wage increases or inflation or decrease in the purchasing power of money.” 510 F.2d at 836-37. Of course, the second part of the award was reversed, even though it is possible that a portion of that element was awarded due to likely merit or productivity increases not based upon inflation or cost of living adjustments. Thus we see that although Penrod was not intended to eliminate evidence of future wage increases on the basis of merit or productivity, the practical effect of Penrod is often to throw the baby of future merit increases out with the inflationary washtub waters.
Finally, on the very day that Penrod was published, this Court, in Law v. Sea Drilling Corporation, 510 F.2d 242, 251-52 (5th Cir. 1975) on reh’ing, 523 F.2d 793 (5th Cir. 1975), held that the District Court’s allowance for additional inflationary cost of living increases was improper. Again, the increases in the decedent’s income over the years before his death, upon which the trial judge based his reasonable expectation that the decedent’s income would have continued to rise, probably reflected increases due to both merit and productivity and not just cost of living increases. Nevertheless, because an element of inflation was considered, Penrod was dispositive. .
In all of the above discussed cases, this Court, by both its actions and its words, flatly prohibited any consideration of inflation, whether in evidence presented, in jury instructions or in final argument, without regard to the cogency or persuasiveness of the evidence offered to prove that inflation, to some extent, will likely occur. Swept unthinkingly before Penrod also were merit-productivity increases. Thus stands the law in this Circuit, until today.
In our attempt to assess the impact of Penrod in this Circuit, not only is it necessary to consider our own appellate opinions, as we have done above, but it is also helpful to review some of the District Court opinions where the tentacles of Penrod were perceived and predictably effective. In Complaint of Metcalf, 530 F.Supp. 446 (S.D. Tex. 1981), the District Court stated emphatically that in its award for loss of earnings in the future, any “future inflationary or deflationary effect on earnings is not to be considered.” 530 F.Supp. at 458, citing Penrod. Likewise, in Kratzer v. Capital Marine Supply, Inc., 490 F.Supp. 222 (M.D. La. 1980), aff’d 645 F.2d 477 (1981), the plaintiff produced an economist who testified as to lost wages predicated upon an inflation factor and an increased productivity factor. The District Court simply cited *291Penrod and stated that there “was no evidentiary justification for these factors, and the Court declines to accept them, and as noted, utilized a discount figure of 6 percent with no other factors in computing lost wages.” Again, we see how Penrod often operates to prohibit not only evidence of inflation, but also evidence of likely merit or productivity wage increases. In McLean v. United States, 446 F.Supp. 9 (E.D. La. 1977), the plaintiff carefully met his burden of proof by establishing a “productivity” factor of 5.7% on the basis of past merit and change in job title increases over his previous years of employment. Because that factor was in no way compensation for inflation, evidence thereof was admissible under Penrod. 446 F.Supp. at 13, 14. In Thompson v. Offshore Co., 440 F.Supp. 752 (S.D. Tex. 1977), the District Court announced that in calculating a decedent’s future earnings, any future inflationary effect on wages is not to be considered. 440 F.Supp. at 762. Finally, in Hamilton v. Canal Barge Co., 395 F.Supp. 978 (E.D. La. 1975), allowance for inflation or increased cost of living was denied in light of Penrod, and for the additional reason that the diminution in the purchasing power of an award, caused by inflation, is not an item of damage caused by death or by prepayment of future wages. 395 F.Supp. at 987. However, the District Court did allow a 4% annual increase factor on the basis of skill and productivity. Id.
Nesmith v. Texaco, Inc., 491 F.Supp. 561 (W.D. La. 1980), stands out as an anomaly to the law of Penrod. In startling contrast to the usual acceptance of Penrod’s rejection of evidence of inflation, the District Court interpreted Norfolk & Western Railway v. Liepeit, 444 U.S. 490, 100 S.Ct. 755, 62 L.Ed.2d 689 (1980), to say that “in arriving at an award for future loss of earnings the very factors disapproved in [Penrod] should now be considered.”7 491 F.Supp. at 564. While the court was “mindful of the fact that this interpretation of Liepeit may be erroneous, and [Penrod] may still bar the door to speculative damages as to factors other than income taxes in a case of future loss from personal injuries,” the plaintiff was permitted to present expert testimony that the growth rate for wages based upon inflationary factors in the plaintiff’s job and geographic area will be approximately 6% over the next 30 years. 491 F.Supp. 564, 565 n. 4. This attempt to break the bonds of Penrod resulted in an appeal, disposition of which is being held pending the outcome of the present opinion and Byrd, see note 1.
The cases discussed in this section demonstrate the absolute prohibition on any consideration of inflationary or cost of living trends in the courts of this Circuit. District Court judges have found themselves under a duty, imposed by Penrod, to prevent experts in their testimony, and attorneys in their arguments, from even mentioning economic trends that would effect future wage increases. Likewise, in their instructions to the jury, District judges have been careful not to mention inflation as a factor in computing damages, or, where inflation was mentioned, the purpose was to explain to the jury that they should not consider inflation in their computations. In those cases where the jury was permitted to consider the effect of future inflation on damage awards, either because they heard testimony by an expert or because the trial judge instructed them to consider inflation, we have reversed the judgment and, in most *292cases, remanded for a new trial. In non-jury cases, the District Court judges have not been permitted to consider inflation in their assessment of damages. Even where plaintiffs have clearly limited their evidence (as to inflation) to likely cost of living increases, thereby carefully avoiding any inference that the total award should be somehow “protected” from inflation irrespective of wage increases, the evidence has been rejected. In short, Penrod in practice has been a firm, inflexible standard that we have been unable to explain or interpret away. The standard is simple: no evidence of inflation may be presented to the trier of fact, regardless of how expert the testimony, how understandable the presentation, or how fair to the parties. Argument of counsel is also forbidden, as is jury instruction by the Court. This standard has at times been so overwhelming that it has prohibited evidence that should have been allowed, such as evidence of likely wage increases based upon merit or productivity, either on a misreading of Penrod or a perceived (and sometimes actual) impossibility of separating out inflationary elements from admissible merit-productivity increases. Quite possibly, an employer may take into account rising inflation in his decision to grant promotions, and it would be impossible for anyone, expert or otherwise, to distinguish that part due to merit only and determine the percentage amount of the raise due wholly to non-inflationary factors.
V.
Penrod and Other Circuits
Worthy of our consideration is how our sister circuits perceive Penrod. In 1975, the Eighth Circuit, citing Penrod, was able to say that the
federal circuits that have faced the issue in cases involving federally created claims, governed by federal rather than state law, have for the most part rejected testimony, jury instructions or trial court consideration of future inflationary trends in damage assessment.
Johnson v. Serra, 521 F.2d 1289, 1295-96 (8th Cir. 1975), aff’d on remand, 586 F.2d 1291 (8th Cir. 1978). More recently, however, in the trend toward increasing consideration of inflation, Taenzler v. Burlington Northern, 608 F.2d 796, 800 n. 9 (8th Cir. 1979), cited Penrod but declined to follow its approach. Expert testimony on future wage increases, when limited to future trends in earnings of a particular group of employees, thus avoiding the excessively general national inflation rate, was found to be acceptable in trial courts.8 608 F.2d at 801.
Feldman v. Allegheny Airlines, Inc., 524 F.2d 384 (2d Cir. 1975), a case arising under Connecticut law, cited with approval Judge Friendly’s earlier assessment of inflation:
There are few who do not regard some degree of continuing inflation as here to stay and would be willing to translate their own earning power into a fixed annuity, and it is scarcely to be expected that the average personal injury plaintiff will have the acumen to find investments that are proof against both inflation and depression — a task formidable for the most expert investor.
524 F.2d at 388, citing McWeeney v. New York, New Haven and Hartford Railroad, 282 F.2d 34, 38 (2d Cir. 1960) (Friendly, J.). The computation by trial Judge Blumenfeld of the discount rate by offsetting the anticipated rate of earnings on a prudent unsophisticated investment by an inflation factor was approved.9 This computation, sub*293tracting the assumed inflation rate from the rate of interest on a safe investment (the traditional discount rate), yields an “inflation-adjusted discount rate.” Judge Friendly, a member of the panel in Feldman, was not completely satisfied with this result — Penrod’s emphasis on “the plethora of uncertainties” that accompany testimony regarding inflationary trends was contrasted with the trial court’s apparent construction of “an iron-clad guaranty against the unknown and unknowable future effects of inflation.” 524 F.2d at 392 (Friendly, J. concurring). Such protection against inflation is not enjoyed by the millions of Americans on fixed pension or investment income, or even by the average worker whose wages do not keep up with inflation. This highlights a weakness in the Feldman approach: it assumes that wage increases will mirror inflation. Id. Judge Friendly thus predicted that Feldman “will not constitute a precedent on the inflation problem in a case arising under federal law.” 524 F.2d at 393, (Friendly, J. concurring).
More recently, in a suit brought by an injured cargo checker against a stevedore and a shipowner (neither was his employer), pursuant to the LHWCA, 33 U.S.C. §§ 901 et seq., the Second Circuit carefully considered the question whether and to what extent an award for lost future wages should be adjusted because of inflation. Doca v. Marina Mercante Nicaraguense, S.A., 634 F.2d 30 (2d Cir. 1980), cert. denied, 451 U.S. 971, 101 S.Ct. 2049, 68 L.Ed.2d 351 (1981). Penrod as well as similar cases in the Third 10 and First11 Circuits were identified as unsound minority views. 634 F.2d at 36. Broad agreement was perceived among economists that inflation “is a dominant factor on the current economic scene and, despite episodic recessions, is likely to be so for the forseeable future.” 634 F.2d at 37.
The District Court in Doca, 474 F.Supp. 751, 758 (S.D.N.Y. 1979), found that the plaintiff was entitled to recover $352,560, “representing his future loss of earnings considering both probability of continued inflation and a discount to present value.” The extent to which inflation was considered was unspecified, and no computation was provided by the District Court. 634 F.2d at 34. On appeal, the plaintiff urged that the trial judge used a discount factor of 1%, an inference drawn by comparing the difference between the award given and the plaintiff’s own projection of future lost wages. 634 F.2d at 34 & n. 4. The defendants objected to the court’s computation on the grounds of lack of evidence to support an estimate of future inflation and argued that any adjustment for inflation is impermissible. 634 F.2d at 34-35.
The Second Circuit in Doca described two basic approaches as to how inflation could be considered. First, the question of inflation rate may be submitted to the fact finder in one of three ways: (i) by allowing expert opinion, (ii) by requiring the fact finder to apply its own knowledge, or (iii) by permitting the fact finder to apply its own knowledge together with expert testimony. Each of the three contemplate that a separate determination as to inflation is to be made in each case. Second, a rule of law may be adopted that focuses on the somewhat constant relationship between inflation and interest rates, as in the Alaska Supreme Court’s view12 that the inflation rate should be assumed to equal the interest rate, thereby eliminating the discount to present value. 634 F.2d at 38-39. The Doca Court was not prepared to specify any particular methodology:
If litigants prefer to offer evidence as to future rates of both inflation and interest, they are entitled to do so... .
*294We emphasize that we are not requiring the use of an inflation-adjusted discount rate. . . . Litigants are free to account for inflation in other ways, or, if they use the adjusted discount rate approach, to offer evidence of a rate [other] than 2%.
634 F.2d at 39-40. The District Court’s award was remanded for reconsideration for two reasons: (i) if a 1% adjusted discount rate was used, it was too low since 2% is the “true cost of money appropriate for use in a computation to determine the present value of lost future wages,” 13 and (ii) the District Court took into account a post-trial wage increase based upon cost of living, therefore duplicative consideration was given to the impact of inflation.14 634 F.2d at 40.
Feldman, supra, approved the use of an “inflation-adjusted discount rate” of 1.5%. 524 F.2d at 387. The purpose of the method used in Feldman, and approved as one possible technique in Doca, was to determine the “real yield” of money — that portion of interest gained on virtually risk-free investments which represents only the real growth of the investment and not the losses due to future inflation.15 The theory is that the “real yield” of money is roughly 2% in any year, because in periods of high inflation, interest rates will be a bit higher, and in low inflationary periods, the interest rates will still be a little above the rate of inflation.
Doca recognized the disagreement among economists over the validity of the assumption that the real rate of interest is constant and consequently independent of inflation, but noted that various economic studies of similar time frames have estimated that the real rate of interest was between 1.5% and 3%. 634 F.2d at 39 n. 10.
VI.
Penrod and Some Commentators
In 1977, one commentator — with an error in timing — predicted that “the Fifth Circuit soon could decide . . . that Penrod is ‘substantively indefensible.’ ” 16 This was precipitated by Judge Wisdom’s concurring opinion in Freeport Sulphur Co. v. S/S HERMOSA, 526 F.2d 300, 311 (5th Cir. 1976) (Wisdom, J., concurring). Judge Wisdom there expressed his dissatisfaction with Penrod and identified the inconsistency in *295this Circuit’s insistence upon discounting while ignoring inflationary effects. Id. The policy concerns of Penrod, (i) achieving complete compensation, (ii) preventing speculation, and (iii) simplifying trial procedures, would be served better by adopting an approach that considered the effect of inflation on damage awards. Id.
Penrod has also been labeled a source of serious judicial error. Kane, Inflation, tight money, and Penrod: the cost of judicial error, Texas Trial Lawyers Forum, April-June, 1980, at 3-5. In Penrod, we stated that “if future inflation does cause higher wages, experience predictably demonstrates that higher interest rates on investments which have always accompanied inflation will also occur and this factor will mitigate the failure to include an inflationary surcharge in wage rate calculations.” 510 F.2d at 236.
The obvious flaw in this line of reasoning lies in the fact that while interest rates do rise in response to accelerating inflation, a higher interest rate only reduces the present value of a future loss, thus further penalizing the plaintiff. The problem for the plaintiff is compounded when the series of future income payments cannot be proportionally increased for the same inflation that drove up the interest rate.
Kane, supra, at 3-4.
Spiraling inflation during the years since Penrod has indirectly led to much of the criticism of our 1975 opinion.17 Although consumer prices have fallen in the early months of 1982, which might signal the beginning of the end to “spiraling” inflation,18 this does not demonstrate that Pen-rod is not so bad after all. The problem with Penrod is not primarily its spectacular unfairness in periods of extremely high inflation. To the contrary, the danger of Penrod lies in its unwillingness to consider the effect of inflation on future wages at all. Penrod stands for the inflexible proposition that triers of fact should not consider inflation or deflation at all, whether high or low or nonexistent, in predicting wage loss. At the same time, defendants are freely allowed to show the highest inflation-induced interest rates available on relatively safe investments. Unfortunately, any evidence of the fact that wages will likely increase to combat the eroding effects of inflation remains eclipsed by the Court-declared spectre of speculation.
What we seek, in response to criticism from both within and without the Circuit, is fairness with regard to the presentation of economic data by either side in the legal controversies that our federal courts face. The principles that we adopt to facilitate this goal must be flexible enough to remain workable in any economic climate. To this end, the remainder of this opinion will explore the alternatives and adopt those that meet the standards of economic flexibility and fairness to plaintiffs and defendants alike.
VII.
The Real Rate of Interest: A Possible Solution
In layman’s terms, the real rate of. interest mirrors the rate of interest that lenders would charge in an inflationless society. Interest rates are much higher because of the expectation of inflation and the uncertainty as to its extent. Thus, the real rate of interest becomes more difficult to calculate. For example,
You lend [or invest] $100 for ... 30 years, at a rate of interest of 5 percent *296per year. The $5 interest paid to you annually will rise or fall in purchasing power depending upon whether or not the general price level of various goods and services falls or rises. If the price level should increase during the first year by 3 percent, this means that of the 5 percent interest ($5 each year) about 3 percentage points (or $3 in real terms) is eroded away by the higher prices you must pay for goods and services. In effect, in real terms you have gotten only about 2 percent interest on your $100 loan [or investment].
A. Alchian & W. Allen, University Economics 193 (3d ed. 1972). Thus, the lender who correctly anticipated the rise in price level would have set a higher nominal interest rate on the loan, perhaps 8%, to allow 3% for the anticipated rise in prices and 5% for interest in terms of real purchasing power. Id. Accordingly, the nominal or advertised interest rate reflects both the basic or real rate of interest (which would exist absent any inflation anticipations) and an adjustment for the anticipated rise in price levels.19
In theory, the real rate of interest represents a possible standard for determining the proper and fair discount rate to be applied to damages resulting from the loss of future income. For courts using this method, the rate of inflation, however calculated, is subtracted from the interest rate for some relatively risk-free investment, and the remainder is the real rate of interest, or discount rate, to be applied to the damages award.20 In practice, of course, the problem becomes more difficult. Even *297where parties to a controversy are able to establish both a projected inflation rate and an interest rate on some risk-free investment, using legally acceptable indexes and tables, the fact remains that the goal of the exercise is to compensate the plaintiff for the income the worker would in all probability actually receive in future years. It is well-recognized that the average wage increases of many workers in the United States have not kept up consistently with inflation.21 This factor, then, must be ap*298plied to increase (e.g., from 2% to 4%) the discount rate so that the plaintiff, whose wages have not kept up and likely will not keep up with inflation, will not receive increases based upon inflation.22 On the other hand, if the average annual increase in wages in a particular profession has risen above the inflation rate, the plaintiff should not be limited to increases based upon the inflation rate. This factor highlights the major weakness of the “real rate of interest” approach.
A simpler and more accurate approach to the problem is found where the plaintiff shows, by expert testimony or otherwise, all of the increases in wages he is likely to receive during his work-life expectancy. Likely increases in wages due to cost of living increases, merit or productivity increases, or promotion increases could all be separately established by the plaintiff in computing the total amount of lost wages. For most occupations, it is easy to find economic data of general wage increases, which include increases due to cost of living, merit, and productivity. The more difficult task of breaking down the data into the reasons for the increase, e.g. cost of living or merit increases, is not necessary in this simpler method. Then, the defendant has the opportunity to establish, by expert testimony or otherwise, the discount rate which reflects the rate of interest in a relatively safe investment, which is applied to the award to reduce it to present value. If this methodology is followed, the rate of inflation itself is not directly involved — inflation is only indirectly reflected (i) in the cost of living increases projected by the plaintiff (which may be more or less than the rate of inflation), and (ii) in the discount rate established by the defendant (which includes both the real rate of interest and the predicted effect of inflation). The result of such a calculation should not differ greatly from the result obtained *299through the Feldman approach. However, unlike the Feldman approach, this alternative method would more accurately represent the future wages lost by a particular plaintiff, in a particular occupation, in a particular geographic area.23
VIII.
Economic Predictions and the Courts’ Serbonian Bog24
We agree with the critics that Penrod represents an idea whose time has passed. In an era of inflation, which has been with us for over forty years,25 it is quite clear that plaintiffs are unfairly penalized by their absolute inability to present evidence of a historical fact — inflation. The likely effect of inflationary trends upon future wages is forbidden even though Penrod allows evidence from defendants regarding the applicable discount rate which has been increased by the very inflation so roundly excluded.
However, we are less than satisfied with the so-called Alaska Rule,26 which by assuming that the discount rate and the inflation rate are virtually identical, unnecessarily penalizes defendants because, as noted above in our discussion of the real rate of interest, interest rates on relatively safe investments will typically ride several percentage points above the rate of inflation.27 In addition, 'tied as it is to changes in the Consumer Price Index (CPI), the result would unfairly award the plaintiff the difference between the changes in the CPI and the actual or average increase in wages.
*300We are much impressed with, but certainly not willing to embrace uncritically, Feldman ’s approval of an inflation-adjusted discount rate in the neighborhood of 1.5%.28 524 F.2d at 387-88. This view represents a compromise between the Alaska rule’s penalizing of defendants and Penrod’s penalizing of plaintiffs. However, fixing the inflation-adjusted discount rate at 1.5%, or even 2% or 3%, would subject this Court to criticisms not unlike those aimed at Penrod and even the Alaska rule. In the dynamic and ever-changing world of finance and economics, we as judges cannot rule out that the interest rate on risk-free investments might equal the inflation rate during a certain period such that the Alaska Rule is vindicated. Indeed, the logic behind the Alaska Rule is that in an economy of ups and downs, “it will all come out in the wash” — the interest rate and the inflation rate being, on the average, roughly equivalent.29 However, any standard which is inflexible in a dynamic economy will likely be unable to cope with the problem of preventing windfalls either to plaintiff or defendant. Although a perfect method may never be found, we must attempt to create standards that are fair to both sides of the controversy, with the trier of fact being allowed to receive and act upon credible evidence of the economic facts bearing, pro and con, on the competing theories.
The methodological basis of Feldman must be clearly understood as our analysis proceeds. Quite simply, and perhaps even simplistically, Feldman is a backward-looking, past-performance approach:
(1) It begins by considering the effective annual interest rate payable on a certain relatively risk-free investment each year during a particular past period of years, for example, 1940-1980. For each of those years during the period chosen, the average annual percentage change in the Consumer Price Index (CPI) (see note 25, supra) is determined using historical data.
(2) Subtracting the average annual percentage change in the CPI for each year from the effective annual interest on the investment for that same year yields the actual or “real” rate of interest that an investor would gain during that year in terms of the buying power of the dollars invested. Because a single year in the past several decades might be unreliable, all of the real rates of interest are averaged. The result of such an analysis, for almost any relatively risk-free investment during any 10 or 15 year period in the past several decades, will be approximately 1-3%.
(3) This inflation-adjusted discount rate, the result of subtracting changes in the CPI from safe investment interest rates, is then projected over the plaintiff’s expected work-life by applying the inflation-adjusted discount rate to each year’s estimated salary to reduce the future annual salaries to their present value. The discount factor applied to each year would be different because the present value of a dollar received in the future, e.g., ten or twenty years from trial-date, is much less than the *301present value of a dollar received on the date of trial. Indeed, one dollar received even one year from now is worth less than one dollar today.30
(4) If wages will increase due to promotion, merit, or productivity, then instead of using the plaintiff’s current wages as the amount to be discounted to present value for each future year of expected work-life, each future year’s wage is computed by finding what a person in the promoted position would likely be earning in the respective year. All the years are then added together to reach the lump-sum award. This computation is not limited to the single promotion or merit increase that the plaintiff would have received during the year of his injury, but may involve several increases during the plaintiff’s work-life expectancy. For example, if the plaintiff is a stock-boy with a 30-year work-life expectancy, and he likely would be promoted in 10 years to stock manager, and in 10 more years to store manager, which is the highest position he would predictably attain, then the trier of fact would compute the trial-date salary of stock manager and store manager. The stock-boy’s salary for each of the first 10 years would be reduced to its present value, the stock manager’s salary for each of the second 10 years would be reduced to its present value, and the store manager’s salary for each of the third 10 years would be *302reduced to its present value. All these results are added to find the lump sum award.31
In light of our discussion, supra, of the theory of the “real rate of interest,” several critical remarks are appropriate. First, the purpose of an inflation-adjusted discount rate, which simply means that the discount rate (proposed on the basis of safe investment) should be reduced by the rate of inflation eroding the profit on such an investment, is fully to compensate the plaintiff such that if the lump-sum award given in court is invested, the plaintiff will be able to receive, as nearly as possible, the amount of assumed income that would have been received each year in the future if no injury had occurred.
However, the Feldman approach assumes that the plaintiff’s past and future income would increase, as it has, roughly, in the past forty years, along with the CPI. The difficulty with this assumption is that annual wage increases may not keep up with inflation as reflected in the CPI, and, to the contrary, such increases in average earnings in some occupations might be much greater than inflation. Second, the American economy may change radically in the next several years. As evidence becomes available that the next 20 or 30 years will not be like the past 20 or 30 years, then the Feldman approach is less helpful. Nevertheless, despite these criticisms, the Feldman approach represents a fair and flexible alternative that' is clearly superior, and more economically sound, than either Penrod or the Alaska Rule.
Lest we lose sight of the forest of practice in these trees of theory, our analysis must move to a practical example.
IX.
A Helpful Hypothetical
“The war against inflation is a grim affair.” Justice Douglas, Davies Warehouse Co. v. Bowles, 321 U.S. 144 at 158, 64 S.Ct. 474 at 482, 88 L.Ed. 635.
We may now consider two hypothetical situations illustrating that the process of computing likely wage increases, due to the effects of inflation and other factors, need not depend upon suspect speculation, the use of strange formulae, or intricate expert prognoses.
First, assume that a plaintiff is totally and permanently disabled and that after allowable deductions for income taxes, social security taxes, and union dues, and after adding various fringe benefits, where applicable, his/her loss of future earnings would be $10,000 per year for the balance of a work-life expectancy of 18 years, or a gross expectancy of $180,000. This plaintiff could purchase for approximately $78,000 enough United States Bonds to allow him or her to receive $10,000 each year from 1981 through 1998.32 Government Bonds *303are relatively risk-free investments for an unsophisticated investor, because they typically require no reinvesting of funds over the next 18 years provided that no bonds are redeemed in advance of their maturity. This rough computation takes into account the interest that would be paid on the bonds, as well as the estimated income taxes on the interest received each year by the plaintiff, such that the plaintiff receives $10,000 spendable dollars each year. The economic fact that the plaintiff needs only $78,000 to ensure that he or she will receive the assumed $180,000 in lost future wages is, of course, the legal justification for applying a discount rate. Significantly, this model assumes that the plaintiff would have received no wage increases, for whatever reason, in the next 18 years.
Second, assume a projected 5.12% annual increase in wages based upon past increases in the national average weekly wage, such that this plaintiff’s salary increases from $10,000 in the first year, to $10,512 in the second year, to $11,050 in the third year, and so on until an annual salary of $23,314 is reached in 1998.33 The total amount *304earned by the plaintiff, assuming this increase in wages, would be $284,494. Of course, this estimate assumes that the plaintiff would experience wage increases, either on the basis of performance and experience or due to cost of living increases, at the national average (5.12%) of workers over the past several decades.34 Using the same calculations as above for a risk-free U.S. Bond portfolio,35 it can be estimated that approximately $120,000 would produce the $284,000 in wages lost with the 5.12% per year increase considered.36
In summary, in the first hypothetical situation above, the plaintiff would introduce evidence that his/her annual take-home pay is $10,000 and that he/she would probably have worked for 18 more years. The defendant would simply prove that it would take about $78,000, safely invested, to assure that projected loss of income. In the second hypothetical situation, which is much more realistic in terms of our national economy, the plaintiff not only shows present income and work-life expectancy, but also presents evidence that his/her salary will likely increase, just like average wages have increased in the past, at the annual rate of 5.12% due to inflationary (cost of living increases) and non-inflationary (merit, productivity, etc.) factors. The defendant in its response again can establish that approximately $120,000, safely invested, would allow the plaintiff to spend each year an amount roughly equivalent to his/her projected annual salary.
Comparing the above two hypothetical situations with the approach in other jurisdictions, it is clear that under the Alaska Rule, the plaintiff would receive a lump sum of $180,000 in both situations, because inflation (and cost of living increases) are presumed equal to the discount rate (i.e., safe investments). No discount rate is applied; an obvious windfall to the plaintiff who could, using safe investments, produce $414,000 over the next 18 years.37 Under Penrod, the first situation was common: lump-sum awards were discounted while no evidence of cost of living increases, which account for most of the 5.12% figure in the second situation, was permitted. Under Feldman’s inflation-adjusted discount rate (real rate of interest) approach, past average wage increases are not considered, but past annual inflation rates are subtracted from past annual interest rates on safe investments to determine the discount rate. The plaintiff in this situation would receive $156,725.26.38
*305 Given the likely availability of relatively risk-free investments to the unsophisticated investor, and given a dynamic economy, it is not our purpose to establish a single methodology. Nor are we attempting to decide today what precise types of economic data should be found competent by the courts as parties attempt to prove up, or discount, probable future lost wages. We do hold, however, that Penrod’s prohibition of any consideration of inflationary factors is unfair to plaintiffs and is therefore overruled. The basis for that holding is not that “inflation is here to stay,” for we are not capable of making such an economic prediction. The critical error of Penrod was its failure to recognize the effects of inflation on wages in this country over the past several decades, as they would bear upon the actual dollar amount of money a person would receive over a future period of time. Given that wage-earners have typically received cost of living wage increases on the basis of inflation, plaintiffs should be permitted to establish by factual economic and labor data, and expert testimony, that their income would probably continue to increase in response to inflation in future years if they continued to work.39 Likewise, using economic and labor data and expert testimony, the defendant should be permitted to rebut such evidence.
Significantly, economic and labor data regarding increases in wages and inflation are not always complex and mysterious so as to require highly skilled economists to interpret them. On the contrary, most of the tables prepared by the Department of Labor, Bureau of Labor Statistics, are readily available and easy to understand, and, more importantly, are considered accurate for most purposes.40 The availability of such statistics is important because an injured plaintiff (or his beneficiaries) is never to recover damages simply because inflation is likely to erode an award, but only lost wages, and insofar as wages will likely increase in the future, evidence of inflation is permitted to show what future wages will be.
Various methods are available that allow the plaintiff to ensure consideration of the effects of future inflation on lost wages. We have already discussed the inflation-adjusted discount rate (or real rate of interest) used in Feldman. Although we find such an approach acceptable, and certainly superior to the Penrod or Alaska Rule approach, we must emphasize that defendants must be permitted to demonstrate the likelihood that the plaintiffs wages have not kept up, nor will they likely keep up, with inflation. Given this additional factor, the Feldman approach becomes less simple. Another approach, illustrated in our hypothetical above, allows the plaintiff to present evidence of average annual national, local, or occupational wage increases over the past several decades, and thereby to present an average annual increase in wages due to merit, productivity, promo*306tion, or cost of living increases. This may, in some cases, be a simpler approach. Other approaches will likewise be found acceptable, insofar as they permit the plaintiff to show likely wage increases, due to inflation or any other reason, and at the same time allow defendants to present evidence of relatively risk-free investments in the economy that would allow a plaintiff to replace lost income in future years.
X.
Jury Instructions
Where the case is tried to a jury, instructions are required. Without attempting to write or construct a suggested charge,41 we point out that the instructions may take several different forms depending on the permissible methodology used.
If the Feldman approach is used with its inflation-adjusted discount rate (real rate of interest), the jury must make three separate determinations.42 First, they must consider and determine all of the future income losses on the basis of the plaintiff’s present income and, in addition, any increases that the plaintiff would likely have received due to merit, productivity, or promotion, but not inflation. In determining future merit raises or promotions, it should be made clear to the jury that any portion of such increases due to inflation or cost of living raises should not be added into the aggregate sum. Second, the jury must determine, on the basis of the evidence presented to them, the likely increase in the CPI (the acceptable inflation indicator under Feldman). Third, the rate of interest available on some safe investment must be determined. On the basis of the second and third findings, the trial judge will then be in a position to determine the inflation-adjusted discount rate (real rate of interest) by subtracting the projected change in the CPI from the investment interest rate. The result is the discount rate to be applied to reduce the lost future income to its present value to determine the amount of the award.
On the other hand, if the plaintiff shows all likely wage increases due to inflation and other factors (as in the hypothetical above where a 5.12% annual increase was used), which allows the defendant to show what amount of money is necessary, using a relatively risk-free investment, to ensure that the plaintiff will receive those projected future wages losses, the jury should be instructed to answer two basic questions.43 First, the amount of lost future earnings must be determined and must include all likely increases due to inflationary factors such as cost of living increases as well as non-inflationary factors such as merit or promotion raises. It must be made clear that the jury is to project only those increases that the plaintiff would actually have received, with reasonable likelihood. All projected future lost earnings are then aggregated by the jury without considering any discount to present value. Second, the jury must determine the likely earning capacity of an invested award. The answer to this question may take two forms: (1) Where the parties have introduced evidence of the probable availability of particular interest rates, or rates of return, on reasonably safe investments, the jury should find and fix the particular rate of interest available to the plaintiff. This finding would then serve as the discount *307rate to be applied by the judge to the aggregated lost earnings to reduce that sum to its present value. (2) If the parties wish to show, instead of an interest rate, an amount of dollars that, if invested in government bonds or some other safe investment, would fully compensate the plaintiff for all projected lost future wages, then the jury should consider that evidence and find a particular amount of money that they believe could produce the aggregate lost earnings, which they previously determined. It would then be possible for the judge to determine the amount to be awarded by either (1) applying the discount rate found by the jury to the aggregate lost earnings to reduce that amount to its present value, or (2) adopting the jury’s finding as to the amount of money which, if invested, would produce in future years the plaintiff’s projected lost earnings.
In the jury instructions, and, of course, in the final arguments, it should be clear that the purpose of the award for future lost wages is not to protect the plaintiff from future inflation. The goal is simply to replace for future wages actually lost. If the plaintiff’s income in future years will be greater due to likely cost of living increases, then the plaintiff is entitled to those increases. But if the plaintiff’s wages are likely to increase at a rate less than inflation (cost of living) the plaintiff is only entitled to such wage increases and not an increase based on the rate of inflation. Jurors are thus entitled to consider and determine the actual likely wages that the plaintiff would have received but for the disabling event.
XI.
The Search for Federal Uniformity
The Third Circuit recently held, in Pfeifer v. Jones & Laughlin Steel Corp., 678 F.2d 453, 461 (3d Cir. 1982) that the “total offset method” of measuring damages for loss of future earnings, in which the discount factor used to reduce future earnings to present value is presumed offset by future inflation, applies in negligence actions against a vessel owner under the L.H.W. C.A., 33 U.S.C. §§ 901 et seq. As discussed above, this is an adoption of the Alaska Rule, although the Court did not describe it as such. Recognizing that inflation has become an established phenomenon in our economy that must be considered in awarding damages for future lost earnings, the court stated that a uniform federal rule must be established and applied in maritime cases. 678 F.2d at 457. Toward this end, the Alaska Rule was embraced because it contributes to judicial efficiency and eliminates the necessity for economic speculation, thereby introducing greater certainty as well as facilitating settlement of personal injury claims.
The Seventh Circuit recently chimed in, holding that the trier of fact could take inflation into account in computing damage awards. In O’Shea v. Riverway Towing Co., 677 F.2d 1194, at 1200 (7th Cir. 1982), Professor, now Judge Posner, citing with disapproval the panel opinion in Culver, declared that “it is illogical and indefensible to build inflation into the discount rate yet ignore it in calculating the lost future wages that are to be discounted. That results in systematic undercompensation.” We can but agree.
We share the concern for uniformity in the federal law of damages in maritime cases. Although we find it imprudent to adopt the Alaska Rule, because it is fraught with the same inflexibility that Penrod exhibited, we approve the use of any of the methods outlined for calculating future wage losses that results in fairness to plaintiffs and defendants. We see no reason to make the economic judgment, as did the Third Circuit in Pfeifer, that the rate of future inflation will be equivalent to future interest rates.
One important aspect of the problem of uniformity in federal maritime law is illustrated in Gulf Offshore Co. v. Mobil Oil Corp., 453 U.S. 473, 101 S.Ct. 2870, 69 L.Ed.2d 784 (1981). It was held that (i) federal courts do not have exclusive jurisdiction over personal injury and indemnity cases under the Outer Continental Shelf *308Lands Act (OCSLA), 43 U.S.C. §§ 1331 et seq., and (ii) the issue of whether the jury should be instructed that personal injury damages are not subject to federal income taxation required remand to the state courts for determination under Louisiana law — agreed to be controlling in the case. 453 U.S. at 483-488, 101 S.Ct. at 2877-2880, 69 L.Ed.2d at 795-98. On remand, the Texas Court of Civil Appeals held that Louisiana law does not require a jury instruction that damage awards are not subject to income taxation. 628 S.W.2d 171, 174 (Tex. Civ. App. — Houston [14th] 1982). Moreover, the Texas court found that the Supreme Court’s holding in Liepelt, supra, that a defendant in a FELA death case is entitled to an instruction that damage awards are not subject to federal income taxation, did not displace the Louisiana rule in this OCSLA case.
Although these precise issues are not presented in this case, we are faced with the need for uniform federal law with regard to the admissibility of evidence on inflation in the determination of future damage awards. Because the opinions in Gulf Offshore, supra, demonstrate (i) that state courts have concurrent jurisdiction over nearly all maritime in personam suits, (ii) that various state standards with regard to damages are possible, then to the extent it will be favorably received, we view our opinion as minimizing if not eliminating one more source of friction between federal and state courts. Permitting the introduction of inflationary factors in projecting future wage losses brings this Circuit, we believe, more in line with the prevailing view throughout the nation.
XII.
Summary and Conclusions
Our goal in this opinion is to formulate a simple principle, without being simplistic, that will permit the determination of damages caused by future loss of wages such that neither the plaintiff nor the defendant is penalized (or given a windfall) by economic theory or reality. To begin with, we reject the suggestion that the discount rate, based upon safe investments, is roughly equal to the wage increases that an individual will receive. On any weekday in trial courts throughout the nation, a discount rate can be established using government securities or savings certificates, as examples of safe investments, for any amount of damages. However, the average wage increases for a particular plaintiff can be specified for a particular occupation and, if appropriate, in a particular area of the country. Thus it would be unfair and unreasonable to rule as a legal matter in advance that the discount rate equals wage increases, or, likewise, to say as a legal matter that the wage increases of a particular plaintiff are equal to the rise in the consumer price index or the average economy-wide wage increases for all workers. Our first conclusion, therefore, is that parties should be allowed, in future earnings damages cases, to present evidence not only of the interest rates available on safe investments, but also the likely wage increases that would have been obtained by the particular plaintiff in his occupation, whether these likely increases are due to cost of living, promotions, merit raises or productivity.44 Significantly, this is only part of the solution, because the more difficult question is how the likely wage increases are to be established or determined.
We have discussed several acceptable methods that are useful in the consideration of the likely effect of inflation on future wages in a case involving total and permanent disability. Two may be summarized. In the first, the present-day value of the earnings that the plaintiff would likely have received may be calculated in three steps:
*309(1) Using the average annual rate of increase in the plaintiffs own salary in the years prior to the incapacitating event, or in the alternative, using the average wage increase of workers nationally or in the decedent’s occupation and geographic area over, for example, the ten years prior to his injury, the parties can project the annual earnings for the remainder of the plaintiff’s estimated income-generating years. A lump sum of likely lifetime earnings is the result of these calculations.45 This total would include wage increases due to cost of living increases, merit, or productivity, as they are received by the average worker or, if the plaintiff’s own past wages are used, by the plaintiff.46
(2) The above lifetime earnings are converted to an average annual income by dividing the lump sum by the number of income-generating years.
(3) The present value of the plaintiff’s average annual income is then computed by determining how much money must be invested at the present time to yield each year the average income for the remaining income-generating years 47 This calculation can take the form of applying a traditional discount rate, and it will be based upon relatively safe investments such as Treasury Bills or bonds, or similar instruments.
A second approach is the Feldman inflation-adjusted discount rate, which is based upon the real rate of interest. The steps followed in Feldman to ensure that the plaintiff was compensated for future wage losses due to inflation are as follows:
(1) Using historical economical data, the court establishes the effective annual interest rate payable on some safe investment for each year during a particular period of years, e.g. 1940-1980.
(2) For each of the years during that period, the average annual percentage change in the CPI is established.
(3) The average annual percentage change in the CPI for each year is subtracted from the effective annual interest rate on the chosen investment for that same year, thus establishing a series of real rates *310of interest, or the inflation-adjusted discount rates.
(4) All of the real rates of interest are averaged, and the resulting percentage rate represents the inflation-adjusted discount rate to be applied to the plaintiff’s lump-sum award.
(5) In computing the lump-sum total, the plaintiff is permitted to show all likely future wage increases due to promotions,, merit or productivity raises, or any other non-inflationary factor.48 If promotions are likely, then the annual salary of a worker presently in that promoted position should be used to compute the plaintiff’s projected income for the years he/she will be in the promoted position. If the plaintiff can show no likely promotions or raises, then in the number of years remaining in his/her work-life expectancy, the plaintiff’s present income is assumed as the projected salary for each year.
(6) The inflation-adjusted discount rate is then applied to reduce each projected annual salary to its present value, and the total of these future discounted annual salaries is the amount to be awarded.49
The above methodologies are only suggested approaches, and not strait-jackets, for courts to use in determining future earnings. The methodologies do illustrate, however, the issues upon which evidence may be presented in this Circuit now that Penrod is overruled. In response to the plaintiff's evidence on inflation, defendants will continue to be permitted to introduce evidence of the interest rate available on risk-free investments, and plaintiffs will continue to be able to rebut that evidence. Plaintiffs will now also be able to introduce evidence, not only on inflation, but more importantly on the likely wage increases in the decedent's or injured party's occupation, basing their calculation on past average wage increases and future inflation, with *311all parties and trial courts keeping in mind that wage increases are influenced by, but not necessarily dependent upon or identical to, inflation.
Postlude
Having devoted substantial consideration to the economic problems inherent in properly taking inflation into account, we now emphasize that in most eases the apparent difficulties should not arise. As complicated as this subject appears to be, it is a place for vigorous pre-trial discussion and handling between the trial judge and the attorneys. In the great majority of cases, we believe, the parties can and will be able to stipulate to the methodology, discount rate, inflation rate, the admissibility of economic data, tables, etc. and other technical aspects, as well as to any particular issues of fact underlying the calculations. In a jury trial, this should include, to the maximum extent possible, agreement as to specific issues, the form and manner of their submission, appropriate jury instructions and interrogatories and any objections thereto.
And, to the extent complete stipulation is not reasonably possible the formalized pretrial effort should assure that the areas of dispute are considerably reduced and certainly well defined.
If the trial court can bring the parties together in this fashion, economic technicalities should not trouble the jury and, we fervently hope, the issues, if any, for appellate review will be sharply presented on an adequate evidentiary record preserving identifiably distinct legal problems.
To eliminate doubts we also declare that this decision is immediately effective to control cases (i) now being tried, (ii) tried hereafter and (iii) those heretofore tried and now on, or subject to, appeal in which the issue has been properly and adequately raised. And while not fundamentally a Penrod problem, the question whether to allow pre-judgment interest may well arise in some cases. In those instances where the substantive law permits a court to award pre-judgment interest,50 the court must discount the damage figure back to the date of the event, i.e. injury or death, and may award pre-judgment interest for the period between the event and judgment. See Havis v. Petroleum Helicopters, 664 F.2d 54 (5th Cir. 1981). See generally Linke, Assessing the Pecuniary Value of Human Capital, supra; Annot., Award of Prejudgment Interest in Admiralty Suits, 34 A.L.R. Fed. 126, 228-38 (1977).
On the basis of the above analysis, we overrule Penrod, and remand the present controversy to the District Court on the issue of damages. The panel opinion is adopted in all other respects.
REVERSED AND REMANDED.
. Another case involving the same issue, Byrd v. Heinrich Schmidt Reederei, 638 F.2d 1300, on reh’g en banc, 688 F.2d 324 (5th Cir. 1982), was consolidated with the present case on rehearing en banc. The Byrd opinion appears on the pages immediately following this opinion.
. Briefly, Culver objected to the special interrogatory, which asked the jury to find the amount that “will fairly and adequately compensate Mrs. Culver and her children for their damages,” on the ground that it was impossible for the jury to separate the losses suffered by the children from those suffered by Mrs. Culver. Because a wrongful death action in Louisiana is a new cause of action by the survivors, and not a part of the decedent’s estate, La. Civ. Code Art. 2315, Culver argued that damages must be apportioned outside the probate courts. This argument raised the issue of whether Louisiana law applied at all, in light of Culver’s Jones Act and DOSHA claims. DOH-SA, for example, specifically provides for apportionment by the Court. 46 U.S.C. § 762. The defendants argued that a new trial on this issue could be avoided by permitting the District Court or a state court to apportion the award. The panel found that the record contained no evidence on which a jury could have based an apportionment, and that this maritime action was an improper vehicle to apportion the award. 644 F.2d at 465. We need not decide this issue because of our remand for a new trial on damages, at which time the District Judge may properly frame the issue for Culver’s various claims. See note 6, infra.
. The following issues were raised on cross-appeal: (i) was it proper for the District Court to submit a liability question under the DOHSA, 46 U.S.C. § 761, to a jury?; (ii) was there substantial evidence to support the jury verdict against the defendants?; (iii) did the District Court fail to find a right of indemnification among the defendants?; and (iv) was there sufficient evidence to establish William Culver as a child of the decedent?
. In Liepelt, the Supreme Court stated that: [Fjuture interest rates, and future inflation are . .. matters of estimate and prediction. Any one of these issues might provide the basis for protracted expert testimony and debate. But the practical wisdom of the trial bar and the trial bench has developed effective methods of presenting the essential elements of an expert calculation in the form that is understandable by juries that are increasingly familiar with the complexities of modern life.
444 U.S. at 494, 100 S.Ct. at 758, 62 L.Ed.2d at 694. This Court in Byrd, 638 F.2d at 1308, held that “Liepelt’s favorable dicta is only that.”
. In addition, the panel found no error in the District Court’s substitution of 9.125% for 25% as the discount rate to be applied to the jury’s award, because Culver did not offer discount rate evidence and the court’s adoption of the rate suggested by the defendants’ expert was a proper exercise of discretion. 644 F.2d at 464-65. As to apportionment of damages, the panel found this maritime action an improper vehicle to apportion the award, especially since the record contained no evidence on which the jury could have based its apportionment. 644 F.2d at 465. The panel also found sufficient and adequate evidence to support the verdict on liability. The defendants’ arguments as to indemnity on the basis of the active/passive negligence theory, was likewise found to be without merit. 644 F.2d at 465-66. Finally, the claim by one defendant for a reduction in the damage award based upon the trial judge’s finding of comparative negligence on the part of Culver was found by the panel to be unsupportable and that portion of the judgment was modified. This claim for reduction was based on the finding, in the indemnity phase of the case, that Culver’s employer was actively negligent because of the failure of its employee (Culver) to stay in a position of safety. Yet the jury found Culver free from negligence, and the panel held that the jury determination should prevail. 644 F.2d at 465.
Except as to those issues concerning the damages award which will be decided on remand, we adopt all of the panel’s findings and conclusions.
. The petition for panel rehearing asserted three errors: (i) it was improper for the trial judge to adopt the opinion testimony of the defendants’ witness as a basis for discounting the jury award, (ii) it was statutorily necessary that the awards be apportioned by the trier of fact, and (iii) sufficient evidence was presented such that the jury could have apportioned damages.
These issues, all relating to the damages award, need not be reached because of our decision to reverse and remand the case for a new trial on the question of damages. Specifically, we leave to the District Court initially the proper handling of the apportionment issue, bearing in mind the respective beneficiaries, adult and minor, and the respective bases for liability: Jones Act, DOHSA, Maritime law and Louisiana Law.
. In Liepeit, the Supreme Court said:
[There] are many variables that may affect the amount of a wage earner’s future income tax liability. The law may change, his family may increase or decrease in size, his spouse’s earning may effect his tax bracket, and extra income or unforeseen deductions may become available. But future employment itself, future help, future personal expenditures, future interest rates, and future inflation are also matters of estimate and prediction. Any one of these issues might provide the basis for protracted expert testimony and debate. But the practical wisdom of the trial bar and the trial bench has developed effective methods of presenting the essential elements of an expert calculation in a form that is understandable by juries that are increasingly familiar with the complexities of modern life.
444 U.S. at 494, 100 S.Ct. at 757, 62 L.Ed.2d at 694.
. The Court tempered its acceptance of limited expert testimony on inflation with the caution that expert testimony predicting future inflation at a specified rate (i) may mislead the trier of fact by its pretended exactness, (ii) may distort the effect of inflation by not including “concomitant increases for countervailing effects of inflation such as higher rates of return on investment,” and (iii) may open up collateral matters that waste time and confuse the issues such that its probative value would be outweighed. 608 F.2d at 801. As will become clear in our discussion, we share these concerns.
. 524 F.2d at 387. The plaintiffs expert at trial calculated the discount rate by establishing 4.14% as the average annual earning power of a safe investment, and then subtracted 2.87% as the average yearly inflation rate over an *29318-year period, which yielded a 1.27% difference which was rounded up to 1.5%. Id.
. Magill v. Westinghouse Electric Corp., 464 F.2d 294, 299-301 (3d Cir. 1972). But see Pfeifer v. Jones & Laughiin Steel Corp., 678 F.2d 453 (3d Cir. 1982) (“total offset method” adopted: no discount rate to present value is applied, because future inflation will offset the plaintiff’s ability to draw interest on the award), discussed in section XI infra.
. Williams v. United States, 435 F.2d 804 (1st Cir. 1970).
. Beaulieu v. Elliot, 434 P.2d 665 (Alaska 1967).
. This figure, 2%, “lies within the narrow range bracketed by many economists as representing the true yield of money.” 634 F.2d at 39. There “is substantial opinion that during periods of stable rates of inflation, the real yield of money, whether constant or slightly fluctuating, is approximately 2%....” 634 F.2d at 39 n. 10. This theory of the real rate of interest will be discussed in detail later in this opinion. It may be noteworthy that the Second Circuit was dealing with findings of a District Judge, not a jury verdict.
. The court is here making a critical point with regard to adjusting a discount rate for inflation. There are basically two methods of calculation: In the first, if the sum total of a plaintiff’s future earnings is computed by multiplying his annual income for the year he was injured times the number of years of his work expectancy, then it is fair to both plaintiff and defendant to apply a discount rate that has been adjusted for inflation. For example, if a relatively safe investment yields 6%, and the inflation rate is predicted to be 4% over the period of that investment, then the adjusted discount rate would be 2%, and this would be separately applied to each year’s projected future earnings, based upon projected actual earnings not increased by inflation. Under the second method, if the sum total of the plaintiff’s future earnings is computed by adding up the annual earnings of the plaintiff based upon likely cost of living increases (j'.e., $10,000 in 1981, $11,000 in 1982, $12,000 in 1983, etc.), then the full discount rate, e.g. 6% in the above example, should be separately applied to each year’s figure of the plaintiff’s projected future earnings.
. 634 F.2d at 35. Under this approach, the adjustment to account for future inflation is accomplished by decreasing the traditional present value discount. This rate is determined by subtracting the average annual change in the consumer price index over each of several years from the effective annual interest on government bonds held during each of the same years, and then computing the average of the rates resulting from these subtractions. Feldman, supra, 382 F.Supp. 1271 at 1306-12.
. Comment, Future Inflation, Prospective Damages, and the Circuit Courts, 63 Va. L. Rev. 79, 115-16 (1977).
. The Statistical Abstract of the United States, 1980 (101st Ed., U.S. Dept, of Comm.), indicates the following annual Consumer Price Index increases:
1975 9.1%
1976 5.8%
1977 6.5%
1978 7.7%
1979 11.3%
1980 14.4%
Id., Table No. 807, at 486. In addition, the annual percent increase in Consumer Prices for 5-year periods reveals the following changes:
1970- 1975 6.6%
1971- 1976 6.9%
1972- 1977 7.3%
1973- 1978 8.0%
1974- 1979 7.9%
Id., Table No. 706, at 425.
. “Why Prices Are Falling,” Newsweek, Feb. 22, 1982, at 59.
. Theoretically, therefore, the components of the legal discount rate applied to future wages awards are (i) inflation and (ii) real interest. For example, if a discount rate of 8% is established by usual acceptable evidence, that figure already reflects a rate of inflation. If a bank is willing to pay 8% interest on a savings account (because that bank can lend or invest the money at a higher rate of interest) and the inflation rate is 6%, the real rate of interest paid on the 8% savings account is only 2%, i.e., 6% of the buying power of the invested dollars is eroded away by inflation.
. This was the approach of the District Court in Feldman, with the rate of inflation calculated by averaging the average annual percentage changes in the CPI over a certain number of years, and subtracting that average “inflation” rate from the effective annual interest rates of some safe investment over the same period of years. Another method for determining the real rate of interest, where it is to be used as the adjusted discount rate applied to a future damages award, would be to predict the average annual rate of inflation in coming years on the basis of past increases in the CPI and other reliable economic indicators, and subtract that projected rate of inflation from the interest rate available on relatively safe investments available in today’s market. — for example, U.S. Government Bonds. As in Feldman, this method assumes that the plaintiff’s future wages would increase at the same rate as inflation.
Again, it must be kept in mind that this inflation-adjusted discount rate, which is a prediction of the real rate of interest in future years, is to be applied in the following fashion:
(i) the plaintiffs annual income at the time of the injury is computed, e.g. $10,000 per year; (ii) the plaintiff’s work-life expectancy is estimated, e.g., 20 years; (iii) in trial-date dollars, any likely merit increases in wages due to promotions or productivity, or other types of merit increases such as likely bonuses which are uninfluenced by intervening inflation must be estimated and added to the $10,000 for the years that these increases are likely; (iv) finally, the inflation-adjusted discount rate is applied to reduce each future year’s annual income to its present value, and the total of all of the future years’ income is the amount to be awarded. See note 31 infra. Two potential errors in this somewhat complicated method are identifiable. First, the plaintiff must be permitted to show likely future increases in wages due to non-inflationary factors; these increases must be added to the lump-sum award before discounting. Second, the defendant must be permitted to show that the plaintiff’s wages in the past have not kept up with the inflation rate, or that the plaintiff’s wages in the future will likely not keep up with inflation.
The gravest potential error is likely where the plaintiff, by expert testimony or other evidence, projects his future income on the basis of past cost of living increases. This is accomplished by showing that the plaintiff’s income has increased at an average rate of, for example, 6% over the past 20 years, and then predicting that his future income will also increase at the same rate. If each year’s salary is projected for the remainder of the plaintiff’s expected work-life, and all of the years are added together to determine the lump-sum award, then it would be erroneous to apply an inflation-adjusted discount rate to that figure. *297Because wage increases from inflationary factors were used to compute the lump-sum award, the full or traditional discount rate would have to be applied to reduce the award to present value. See also Linke, Assessing the Pecuniary Value of Human Capital: How To Use an Economist To Establish Damages in a Wrongful Death Case, in Proof of Damages, Chapter 5 (Illinois Institute for Continuing Legal Education 1981).
. The following table compares the average annual increase in the CPI with the annual percent change of the average weekly earnings in selected private nonagricultural industries.
ANNUAL % YEAR AVG ANNUAL % CHANGE: CPI CHANGE IN AVG EARNINGS
1970 5.9 4.6
1971 4.3 6.2
1972 3.3 7.5
1973 6.2 6.2
1974 11.0 6.4
1975 9.1 5.7
1976 5.8 7.3
1977 6.5 7.7
1978 7.7 7.8
1979 11.3 7.8
1980 14.4 7.2
See Statistical Abstract of the United States, 1980 (101st. ed., U.S. Dept. Comm.), Table No. 807, at 486 (CPI percentage changes by years); Dept, of Labor, Bureau of Labor Statistics, Table B-37 (“Average Weekly Earnings in Selected Private Nonagricultural Industries, 1947-80”) (in Economic Report of the President, 1981).
The following table, based upon 5-year periods, illustrates that the average wage increases in some occupations are higher than the inflation rate as reflected in the CPI, whereas many occupations do not keep up with inflation.
OCCUPATION A. B. C. D. E. 1970-75 1971-76 1972-77 1973-78 1974-79
1. Private Industry Clerical/Beginning Technicians 6.7 6.9 7.1 7.5 7.7
2. Private Industry Professional Entry/Advanced Technical Levels 6.1 6.1 6.3 7.0 7.4
3. Private Industry Experienced Professional/ Supervisors, etc. 6.5 6.6 7.0 7.6 8.0
4. Private Industry Production/ Nonsupervisory Workers (nonagricultural) 7.1 7.2 7.3 7.7 8.1
5. Police and Firefighters (minimum annual salary scales) 6.0 6.3 6.0 6.1 6.1
*298OCCUPATION A. 1970-75 B. 1971-76 C. D. E. 1972-77 1973-78 1974-79
6. Police and Firefighters (maximum annual salary scales) 7.1 7.3 6.9 6.6 6.5
7. Urban Public School Teachers * 7.0 7.0 6.9 6.9 6.3
8. CPI Change for Urban Consumers 6.6 6.9 7.3 8.0 7.9
Statistical Abstract ... 1980, Table No. 706, at p. 425. It is clear from the above table that production and nonsupervisory workers (non-agricultural) in private industry [line 4] have, in most 5-year periods [Col. A, B, C, & E], kept up with inflation [line 8] due to pay increases. Clerical workers and beginning technicians in private industry [line 1], on the other hand, have not fared so well. Likewise, the minimum annual salary scales of police and firefighters [line 5] have fallen well below the inflation rate as reflected in the CPI [line 8).
. This points up a difficulty in the Feldman and similar approaches. The real rate of interest does not reflect the fact that a plaintiff’s future wage increases might have been well below or well above the rate of inflation. Once the real rate of interest or the inflation-adjusted discount rate has been computed, it is cumbersome then to attempt to reduce or increase the inflation-adjusted discount rate to account for the likelihood that wages will not increase at the same rate of inflation. Nevertheless, in order to make the prediction as to future wage losses as accurate as possible, this factor must be considered.
For example, if the average annual rise in inflation (reflected in the CPI) in the last 30 years was 5%, and the average annual interest rate of government bonds was 7% during that same period, the real rate of interest, or inflation-adjusted discount rate in Feldman, is 2% (7% less 5%). However, if the annual wage increase in the plaintiff’s occupation was only 3%, then the defendant could properly argue that the proper “wage-increase-adjusted” discount rate is 4% (7% less 3%). This method more accurately accounts for lost wages since the plaintiff’s occupation has not kept up with inflation.
. We do not intend to establish, in the present opinion, any single method to the exclusion of others. However, as will be seen, some of the methods of considering inflation in determining awards for loss of future wages more accurately account for the fact that wage increases are not equal to the rate of inflation, nor are such wage increases limited to the rise in the cost of living.
. “A gulf profound as That Serbonian Bog ... where armies whole have sunk.” Milton, Paradise Lost II.
. The Consumer Price Index (CPI) is probably the most relevant indicator of the effect of inflation because it is based upon retail prices for goods and services purchased by urban wage earners and clerical workers. Although it is not completely accurate, it has been deemed sufficiently reliable to be depended upon by the government and by the business community in recent years. The 1974 Economic Report of the President, at 300, Table C-44, cited in the helpful opinion of Judge Blumenfeld in Feldman v. Allegheny Airlines, Inc., 382 F.Supp. 1271, 1306 (D. Conn. 1974), reveals the following average annual percentage changes in the CPI from 1940-1973, with the table being brought up to date with data from the Statistical Abstract, supra note 14, at 486:
1940 6.84% 1961 4.29%
1941 6.74% 1962 4.52%
1942 6.21% 1963 4.76%
1943 5.67% 1964 5.06%
1944 5.53% 1965 5.37%
1945 5.54% 1966 5.62%
1946 5.56% 1967 5.85%
1947 4.51% 1968 6.02%
1948 3.8 1969 5.98%
1949 3.75% 1970 5.73%
1950 3.90% 1971 5.77%
1951 3.71% 1972 6.37%
1952 3.48% 1973 9.02%
1953 3.53% 1974 11.0%
1954 3.66% 1975 9.1%
1955 3.85% 1976 5.8%
1956 4.00% 1977 6.5%
1957 4.04% 1978 7.7%
1958 3.91% 1979 11.3%
1959 3.98% 1980 14.4%
1960 4.14%
. See Beaulieu v. Elliott, 434 P.2d 665 (Alaska 1967) (inflation is presumed equal to the discount rate, thus requiring no expert testimony). Not the least concern is: how do judges know that much? The factors going into this equation are economic facts, and presumably economic facts are proved as any other fact— physical, medical, scientific, etc.
The Third Circuit has just recently adopted the Alaska Rule in Pfeifer v. Jones & Laughlin, supra note 10, for application in LHWCA negligence actions. See discussion in Section XI, infra.
. In the words of one author:
By presuming total offset, under the Alaskan variation the factfinder need not predict future economic trends. However, this ignores rather than enhances predictability. As no sound economic theory supports total offset, the Alaskan variation fails to resolve the speculation inherent in economic forecasting.
Note, Considering Inflation in Calculating Lost Future Earnings, 18 Washburn L.J. 499, 509 (1979) (footnotes omitted).
. The District Court in Feldman subtracted the average annual increase in the Consumer Price Index from average annual yields on U.S. Government securities between 1940 and 1972, and arrived at an estimated discount rate of 1.5%, roughly the difference between the inflationary component (CPI) and the risk free interest rate. A completely different result was reached by Canada’s Supreme Court in four 1978 cases. A 3.5% projected long-term inflation rate was subtracted from the 10.5% rate of return then available on long-term safe investments, yielding a 7% discount damage rate for awards for future earnings and medical expenses. See K. Rosenn, Law and Inñation 231 (1981). This result, however, is subject to far more criticism than the result in Feldman. Id. at 231-32.
. K. Rosenn, in Law and Inñation 232-34 (1981), argues that the Alaska Rule is fairer because it builds “in a margin for error on the side of the innocent tort victim rather than on the side of the tort feasor.” Id. at 234.
In addition, the approach of the Alaskan Supreme Court has the decided advantages of simplicity, predictability, and economy of effort. It may also eliminate opportunities for confusing the jury and courts that are inherent in speculation about both interest and inflation rates.
Id. We are sensitive to the concern for avoiding speculation and confusion, but it is hardly “fairer” to build in a margin of error in favor of plaintiffs or defendants.
. The following table, taken from V. Brudney & M. Chirelstein, Cases & Materials on Corporate Finance (2d ed. 1979) at 38, illustrates the computation to present value:
TABLE 11- 1. Present Value of a Future $1: What a Dollar at End of Specified Future Year Is Worth Today
Year 3% 4% 5% 6% 7% 8% 10% 12% 15% 20% Year
1 .971 .962 .952 .943 .935 .926 .909 .893 .870 .833 1
2 .943 .925 .907 .890 .873 .857 .826 .797 .756 .694 2
3 .915 .890 .864 .839 .816 .794 .751 .711 .658 .578 3
4 .889 .855 .823 .792 .763 .735 .683 .636 .572 .482 4
5 .863 .823 .784 .747 .713 .681 .620 .567 .497 .402 5
6 .838 .790 .746 .705 .666 .630 .564 .507 .432 .335 6
7 .813 .760 .711 .665 .623 .583 .513 .452 .376 .279 7
8 .789 .731 .677 .627 .582 .540 .466 .404 .326 .233 8
9 .766 .703 .645 .591 .544 .500 .424 .360 .284 .194 9
10 .744 .676 .614 .558 .508 .463 .385 .322 .247 .162 10
11 .722 .650 .585 .526 .475 .429 .350 .287 .215 .134 11
12 .701 .625 .557 .497 .444 .397 .318 .257 .187 .112 12
13 .681 .601 .530 .468 .415 .368 .289 .229 .162 .0935 13
14 .661 .577 .505 .442 .388 .340 .263 .204 .141 .0779 14
15 .642 .555 .481 .417 .362 .315 .239 .183 .122 .0649 15
16 .623 .534 .458 .393 .339 .292 .217 .163 .107 .0541 16
17 .605 .513 .436 .371 .317 .270 .197 .146 .093 .0451 17
18 .587 .494 .416 .350 .296 .250 .179 .130 .0808 .0376 18
19 .570 .475 .396 .330 .277 .232 .163 .116 .0703 .0313 19
20 .554 .456 .377 .311 .258 .215 .148 .104 .0611 .0261 20
25 .478 .375 .295 '.232 .184 .146 .0923 .0588 .0304 .0105 25
30 .412 .308 .231 .174 .131 .0994 .0573 .0334 .0151 .00421 30
40 .307 .208 .142 .0972 .067 .0460 .0221 .0107 .00373 .000680 40
50 .228 .141 .067 .0543 .034 .0213 .00852 .00346 .000922 .000109 50
Each column lists how much a dollar received at the end of various years in the future is worth today. For example, at 6 percent per year a dollar to be received ten years hence is equivalent in value to $.558 now, in other words, $.558 invested now at 6 percent, with interest compounded annually, would grow to $1.00 in ten years. Note that $1.00 to be received at the end of fifty years is, at 6 percent, worth today just about a nickel. And at 10 percent it is worth only about .8 of one cent, which is to say that 8 mills (.8 of a cent) invested now would grow, at 10 percent interest compounded annually, to $1.00 in fifty years. Similarly $1,000 in fifty years is worth today $8.52, and $10,000 is worth today $85 — all at 10 percent rate of interest. . . . Why not make that investment? Formula for entry in table is 1/(1 + r).’ (No inflation is involved in this table).
See Feldman, 382 F.Supp. at 1298, Table I, which illustrates the reduction to present value of future annual income using a 1.5% inflation-adjusted discount rate.
. In Feldman, the District Court found that the plaintiff could have been expected to work for 40 more years, from 1971-2011, and would likely have been promoted from her 1971 position of Government Service rank GS-12, step 1, to GS-16, step 7. The court then discounted each year’s salary (e.g., 1971: $15,040; 1976: $17,044; 1986: $17,545; 2011: $33,757) using ¡971 pay scales, to its present value. 382 F.Supp. at 1287. Therefore, only the increases expected on the basis of promotion were considered, with no consideration given to inflation or cost of living increases, in reaching the projected annual salaries. The inflation-adjusted discount rate that was applied to each future year’s salary gave the plaintiff the benefit of expected cost of living increases and, at the same time, gave the defendant the benefit of the fact that the lump sum award could be invested to yield interest to attain the plaintiff’s future wage expectations.
. This approximation is based upon the fact that on April 3, 1981, an arbitrary date chosen for the sake of this example, the plaintiff could have looked at the financial page of any major newspaper, under the heading “U.S. Bonds”, and found U.S. Treasury Bonds for sale that would mature at various times through the year 1995. A portfolio could have been selected that would allow the plaintiff to have an annual income of $10,000, after paying income taxes, at least through 1995, during the first 15 years of his work-life expectancy in the above example. On April 3, 1981, it would have cost $78,245.30 to purchase bonds maturing to allow $10,000 annual income (after taxes) until the year 1995, and in 1995 the plaintiff would receive $40,389.30 on maturity of all of the bonds. At that point, $10,000 would be available for income in 1995, and the remaining $30,-389.30 could then be invested to ensure $10,000 per year in the last three years of his work-life *303expectancy, 1996, 1997, and 1998. The table below illustrates this procedure if the goal is to give the plaintiff $10,000 per year for 18 years:
YEAR CASH FROM MATURING BONDS CASH FROM INTEREST ON BONDS TAXES ON TAXABLE INCOME TOTAL SPENDABLE DOLLARS AFTER TAXES COST OF BONDS AND CASH TO PRODUCE AT LEAST $10,000.00 SPENDABLE DOLLARS
1981 0.00 9.348.75 1.043.26 10,000.00 1,694.51
1982 2,000.00 9.450.00 1,070.86 10.379.10 1.908.90
1983 2,000.00 9.253.75 1,030.72 10.223.00 1,946.78
1984 3.000. 00 8.998.75 1.002.27 10.996.50 2.691.02
1985 3.000. 00 8.644.38 906.80 10.737.60 3.065.03
1986 3.000. 00 8.222.50 831.27 10,391.20 3,003.49
1987 4.000. 00 7.776.25 764.72 11.011.50 3,771.09
1988 4.000. 00 7.371.25 740.12 10.631.10 3,075.45
1989 4.000. 00 7.021.25 665.23 10.356.00 3,206.75
1990 9.000. 00 6.465.00 665.68 14.799.30 6.845.91
1991 0.00 6.093.75 438.88 5,654.88 0.00
1992 5.000. 00 6.093.75 561.73 10.532.00 3,326.69
1993 5.000. 00 5.534.38 454.74 10.079.60 3,415.06
1994 6.000. 00 5.067.50 375.25 10.692.30 4,361.31
1995 38,000.00 2.398.75 9.49 40.389.30 35,933.30
1996 0.00 0.00 0.00 0.00 0.00
1997 0.00 0.00 0.00 0.00 0.00
186,874.00 78,245.30
To obtain the proper mix of bonds, with varying interest rates depending on the term, over the course of future years, is a task requiring the services of one knowledgeable in bond investments.
. This assumed 5.12% average annual increase in wages is computed by averaging the percentage change in national average weekly wages by years from 1948-1980. See Dept, of Labor, Bureau of Labor Statistics, Table B-37 (Economic Report of the President, 1981) (non-agricultural workers). These wage increases reflect not only performance and experience, but also the increase in wages given in reaction to inflation. The following table illustrates the increased wages that, when added together total $284,494.
1981 $10,000.00
1982 10,512.00
1983 11,050.21
1984 11,615.98
1985 12,210.72
1986 12,835.91
1987 13,493.11
1988 14,183.96
1989 14,910.18
1990 15,673.58
1991 16,476.07
1992 17,319.64
1993 18,206.41
1994 19,138.57
1995 20,118.47
1996 21,148.53
1997 22,231.34
1998 23,369.58
TOTAL 284,494.00
Feldman and Linke, supra, agree that the preferred time for determining these figures, in *304order to achieve a fair average, is in the middle of the year rather than either the beginning or the end.
. Refer to note 33, supra. A more precise and thus better approach, but one not always possible due to the limitations of statistical economic and labor data, would be to base the annual increase in wages upon the particular plaintiff’s occupation and locale.
. Refer to note 32, supra.
. For the sake of simplicity, we have reduced the $284,000 award to the present value of $120,000 by assuming a bond portfolio purchase that would yield $15,000 over the 18 years following 1981. It would be much more complicated to ascertain the number and type of government bonds necessary to produce an income that increased at the rate of 5.12% annually.
If the plaintiff were awarded the $284,000 and it was not discounted to present value, the award could yield over $600,000 in principal and interest over the next 18 years, an obviously unfair result.
. Using the same procedure as that used above, buying bonds on April 3, 1981, see note 32, supra, a plaintiff with $180,000 to invest in government bonds would be able to draw an income of $23,000 per year for 18 years, thus giving a total of $414,000. This amount is substantially more than the plaintiff would be entitled to, even assuming an annual increase in his lost future wages of 5.12% as noted above.
. This result is reached by discounting each year’s projected salary by applying an inflation-adjusted discount rate of approximately 1.5%, and then taking the sum of the discounted amounts. The algebraic formula is:
where
PVo = present value of the asset’s future earnings in time zero or today dollars
Ei = income of asset in year 1
Et = income of asset in year t
*305En = income of asset in year n
(1 + i)t = discount factor in year t
i = discount rate of interest
n = number of years asset expected to generate earnings
PV0, the price or present value of an asset, is that amount of money which, when invested at the discount rate (i), would exactly generate from both principal and interest the time configuration of expected annual income. An estimate of the asset’s net present value would be obtained by reducing the annual income estimated (the Et) to account for any costs incurred in generating the income in years 1,2, 3.....n.
. Theoretically, at least, plaintiffs under Pen-rod were permitted in this Circuit to produce evidence of likely merit, promotion, or productivity increases that would occur without regard to inflation or cost of living increases. But as we have shown, see Higginbotham and Robertson, supra, this was frequently denied in practice either because Penrod was misread or, more likely, because “merit-productivity” wage increases are often affected by an inflation factor.
. For example, under the 1972 Amendments to the Longshoremen’s and Harbor Workers’ Compensation Act, 33 U.S.C. § 910(f), the compensation or benefits payable for certain injuries are to be increased each year “by a percentage equal to the percentage (if any) by which the applicable national weekly wage . . . exceeds the applicable national average weekly wage [for the prior year].” The Secretary of Labor utilizes Bureau of Labor Statistics data in making this determination.
. The use of special interrogatories under F.R. Civ. P. 49(a) is generally within the sound discretion of the District Court, see, e.g., Baumstimler v. Rankin, 677 F.2d 1061 (5th Cir. 1982); Central Progressive Bank v. Fireman's Fund Ins. Co., 658 F.2d 377, 381 (5th Cir. 1981); Jones v. Miles, 656 F.2d 103, 106 n. 4 (5th Cir. 1981); but see also, Guidry v. Kem Manufacturing Co., 598 F.2d 402, 404-06 (5th Cir. 1979), cert. denied, 445 U.S. 929, 100 S.Ct. 1318, 63 L.Ed.2d 763 (1980); Thrash v. O’Donnell, 448 F.2d 886, 890 n. 10 (5th Cir. 1971); see also J. Brown, Federal Special Verdicts: The Doubt Eliminator, 1968, 44 F.R.D. 338.
. Making liberal use of F.R. Civ. P. 49(a), these three findings, for the sake of clarity and possible review, should be separately stated. See note 41, supra.
. Again, these issues should be separately stated so that the trial court, or a reviewing court if necessary, could see clearly what the jury determined on the basis of the evidence.
. “Of course, one cannot predict accurately the work-life earnings of each deceased but one can with great precision project the earnings of the statistical group to whom he or she had belonged. Specificity to this extent provides better, more reasonable awards than are available to one who uses a simplified procedure of applying aggregative type data.” Coyne, Present Value Of Future Earnings: A Sensible Alternative to Simplistic Methodologies, Insurance Counsel Journal, Jan. 1982, at 28.
. Assuming without deciding the admissibility under F.R. Evid. 803(8) of tables such as those referred to in note 33 supra, as discussed above in this opinion, the plaintiff might use 5.12% as the projected average annual rate of increase for his/her lost future earnings, because 5.12% represents the national average annual increase in wages for nonagricultural workers from 1948-1980. See note 33, supra. This factor, 5.12%, is multiplied times the plaintiff’s annual income during the year that he/she was injured, e.g. $10,000, and the result will be the projected income for the year following the year of the injury, e.g. $10,512. To compute the projected income for the second year following the year of the injury, 5.12% is multiplied times $10,512, yielding $11,050.21. This computation is continued for each year of the plaintiff’s work-life expectancy. When all of these annual incomes are added together, the total represents the plaintiff’s projected lifetime earnings.
. This method is most useful where a plaintiff’s annual income will likely increase at a somewhat regular rate. If, on the other hand, the plaintiff would likely have experienced several large promotions during his/her work-life expectancy, then the plaintiff may prove up his/her projected earnings on the basis of those promotions, and not some average annual percentage rate of increase. The goal, however, remains the same: the plaintiff should show the grand total of his/her projected lifetime earnings including all increases due to promotions, cost of living increases, or merit or productivity raises, as they would likely be received during the remainder of his/her expected work-life.
. See note 32, supra, and text accompanying.
In a wrongful death suit, the decedent’s maintenance expenses would be determined, in order to adjust downward the damages to reflect personal costs that would have been incurred by the decedent had he remained alive throughout his income-generating years. The Bureau of Census, Technical Paper No. 16 (Present Value of Estimated Lifetime Earnings, U.S. Dept. Comm.) suggests various levels of maintenance expenses for different occupations. The estimated expenses of the decedent could be compounded at the same rate of growth as the projected annual earnings to reflect cost of living increases. The above maintenance figure would also, like the projected annual earnings, be converted to an average annual figure. The present value of the lifetime average annual maintenance is next computed. The present value of average annual maintenance is finally subtracted from the present value of average annual income. Coyne, supra, at 29-31. See also Linke, supra.
. Where the plaintiff projects his/her future income on the basis of likely wage increases due to inflationary factors, then the inflation-adjusted discounted rate should not be used, but rather the traditional discount rate, based upon some safe investment, to reduce the lump-s.um award to its present value. See note 12, supra.
. In enumerating only the bond purchase and Feldman approaches, we do not intend to imply that these are the only possible methods of taking inflation into account. Other methods, so long as they furnish a rational means of ascertaining and discounting inflationary damages, are certainly acceptable. One such method, essentially a variation of the Feldman approach, is illustrated in the following table. Assume the injured or killed worker had an 18 year expected working life, a base salary of $10,000, a 5.12% expected annual increase in money wages (representing merit as well as cost of living increases), and a 10% discount rate. As the table shows, the present value of $284,494.17 is $114,232.99:
YEAR WAGE PRESENT VALUE FACTOR_ PRESENT VALUE
1 $10,000.00 .909 $9090.00
2 10,512.00 .826 8682.91
3 11,050.21 .751 8298.70
4 11,615.98 .683 7933.71
5 12,210.72 .620 7570.65
6 12,835.91 .564 7239.45
PRESENT VALUE PRESENT YEAR WAGE FACTOR VALUE
7 13,493.11 .513 6921.97
8 14,183.96 .466 6609.73
9 14,910.18 .424 6321.92
10 15,673.59 .385 6034.33
11 16,476.07 .350 5766.62
12 17,319.64 .318 5507.65
13 18,206.40 .289 5261.65
14 19.138.57 .263 5033.44
15 20,118.47 .239 4808.31
16 21,148.53 .217 4589.23
17 22,231.34 .197 4379.57
18 23.369.58 .179 4183.15
$284,494.17 $114,232.99
This calculation functions similarly to the Feldman approach — future wages are projected, separately discounted, and then summed to obtain the proper award — except that while the Feldman court explicitly excluded inflationary changes from both the future wages and the discount rate, this method includes inflation in both calculations. The future wage payments include both merit increases and cost of living adjustments, which, of course, attempt to counter inflation. Then, instead of using the 1.5% real interest rate as a discount factor, one substitutes a higher discount rate (e.g. 10%) which takes into account both the real rate of interest and expected inflation.
. Compare National Airlines, Inc. v. Stiles, 268 F.2d 400 (5th Cir.), cert. denied, 361 U.S. 885, 80 S.Ct. 157, 4 L.Ed.2d 121 (1959) (Death on the High Seas Act, 46 U.S.C. § 762, permits pre-judgment interest) with Doucet v. Wheless Drilling Co., 467 F.2d 336, 340-41 (5th Cir. 1972) (in action on law side under the Jones Act, 46 U.S.C. § 688, pre-judgment interest is not permitted).