dissenting:

Although I personally approve of the approach outlined by Judges Rubin and Frank *126M. Johnson, and would recommend it to district courts in this circuit, I agree with Judge Brown that the Supreme Court in Jones & Laughlin Steel Corp. v. Pfeifer, -U.S.-, 103 S.Ct. 2541, 76 L.Ed.2d 768 (1983), has disapproved of that approach. Accordingly, I join Judge Brown’s dissent.

dissenting:

In Culver, No. 79-3985, I continue to adhere to my view that appellants are not entitled to a new trial. Culver v. Slater Boat Co., 688 F.2d 280, 315 (5th Cir. 1982) (en banc) (Culver I) (Tjoflat, J., dissenting).

In Byrd, No. 78-3064, I join in Judge Brown’s dissenting opinion.

ON PETITIONS FOR REHEARING

In Johnson v. Penrod Drilling Co., 510 F.2d 234 (5th Cir.) (en banc), cert. denied, 423 U.S. 839, 96 S.Ct. 68-69, 46 L.Ed.2d 58 (1975), this court held that juries “should not be instructed to take into account future inflationary or deflationary trends in computing lost earnings ...” Id. at 241. Reviewing that decision in our first en banc consideration of these two cases, we overruled Penrod and held admissible evidence of inflation’s probable effect on damage awards. Culver v. Slater Boat Co., 688 F.2d 280 (5th Cir. 1982) (en banc) (Culver I). Concluding that the jury should resolve the issue on a case-by-case basis, we disclaimed *117any intention to establish a “single method” ior considering future economic conditions. Id. at 299 n. 23. Instead, we discussed several permissible methods the district courts and parties could use. Id. at 305-06.

While we were considering an application for rehearing in Culver I, the Supreme Court decided Jones & Laughlin Steel Corp. v. Pfeifer, — U.S.-, 103 S.Ct. 2541, 76 L.Ed.2d 768 (1983). The Court’s opinion in Pfeifer cites Culver I and confirms our holding that the fact-finder should consider inflation in determining an appropriate damage award. The Court’s opinion also emphasizes, however, a fundamental point that we did not fully consider in Culver I: that courts must not allow the adjustment for inflation to convert “ ‘[t]he average accident trial ... into a graduate seminar on economic forecasting.’ ”1

Reconsideration of Culver I in light of Pfeifer has convinced us that our failure to identify a single method as the one trial courts should use in adjusting damage awards for inflation, particularly in jury trials, would extend an invitation to litigants to engage in just such a seminar. We, therefore, withdraw the opinion in Cul-ver I insofar as it goes beyond overruling Johnson, and hold that, in the absence of a stipulation by the parties concerning the method to be used, fact-finders shall determine and apply an appropriate below-market discount rate as the sole method to adjust loss-of-future-earnings awards to present value to account for the effect of inflation. While expert testimony and jury instructions must be based on this method, juries may be instructed either to return a

1. Jones & Laughlin Steel Corp. v. Pfeifer,-U.S.-, 103 S.Ct. 2541, 2556, 76 L.Ed.2d 768 (1983) (quoting Doca v. Marina Mercante Nicaraguense, S.A., 634 F.2d 30, 39 (2d Cir. 1980)). general verdict or to answer special interrogatories concerning the computation of damages.

I.

The calculation of damages suffered either by a person whose personal injuries will result in extended future disability or by the representatives of a deceased person involves four steps: estimating the loss of work life resulting from the injury or death, calculating the lost income stream, computing the total damage, and discounting that amount to its present value.

In Pfeifer the Court recognized, as we did in Culver I, that calculation of the lost income stream begins with the gross earnings of the injured party at the time of injury.2 To this amount other income incidental to work, such as fringe benefits, should be added. From it, the fact-finder should subtract amounts the wage earner would have been required to pay, such as income tax and work expenses.3

Even in a non-inflationary economy, earnings of a worker “tend to inflate” from the operation of a number of factors, “some linked to the specific individual and some linked to broader societal forces.”4 The operation of both kinds of influences should be considered to reach the first stage in calculating an appropriate award, an estimate of what the lost stream of income would have been “as a series of after-tax payments, one in each year of the worker’s expected remaining career.”5

A lump-sum award may be invested in a wide variety of securities, yielding varying *118rates dependent in part on their relative safety. Both Pfeifer and Culver I require that rate to be based on the return available from “the best and safest investments,”6 and to be computed after considering the effect of income tax on the interest received.7

The rate of interest available even on the safest investments varies depending on the time at which the investment matures. Therefore, in computing the rate to be used, it is essential to determine how the award will be invested. In Pfeifer, the Court recognized that the plaintiff might invest the award “exclusively in safe short-term notes, reinvesting them at the new market rate whenever they mature,” or in “a mixture of safe short-term, medium-term, and long-term bonds, with one scheduled to mature each year of his expected worklife.” It then stated, “We perceive no intrinsic reason to prefer one assumption over the other.” 8

Inflation, which has been a permanent fixture in our economy for many decades, affects both the income stream and the market interest rate. In this opinion, we consider only the method of taking general economic inflation into account. In accomplishing this, as we said in Culver I, “[t]he goal is not ... to protect the lump-sum award from the effect of inflation.” It is, instead, “to assure the plaintiff the equivalent of the total of all of his future wages,” including those earnings likely to be affected by inflation.9

As the Court noted in Pfeifer, three methods are available for adjusting damage awards to account for the effect of inflation. In the case-by-case method, the fact-finder is asked to predict all of the wage increases a plaintiff would have received during each year that he could have been expected to work, but for his injury, including those attributable to price inflation. This prediction allows the fact-finder to compute the income stream the plaintiff has lost because of his disability. The fact-finder then discounts that income stream to present value, using the estimated after-tax market interest rate, and the resulting figure is awarded to the plaintiff.10

In the below-market-discount method, the fact-finder does not attempt to predict the wage increases the particular plaintiff would have received as a result of price inflation. Instead, the trier of fact estimates the wage increases the plaintiff would have received each year as a result of all factors other than inflation. The resulting income stream is discounted by a below-market discount rate. This discount rate represents the estimated market interest rate, adjusted for the effect of any income tax, and then offset by the estimated rate of general future price inflation.11

The third method is the “total-offset” method. In this calculation, future wage increases, including the effects of future price inflation, are legally presumed to offset exactly the interest a plaintiff would earn by investing the lump-sum damage award. Therefore, the fact-finder using this method awards the plaintiff the amount it estimates he would have earned, and neither discounts the award nor adjusts it for inflation.12

*119II.

The Culver I opinion would have permitted the parties to pursue any of these methods in order to demonstrate to the fact-finder what adjustment, if any, should be made to compensate for increases in the plaintiff’s future wages due to the economic factors summed up by the term inflation. Indeed, we endorsed the first method, involving specific forecasts of inflation’s effect on a particular plaintiff’s wages, as “simpler and more accurate” than the below-market discount approaches.13

The procedures necessary to determine the impact of future price inflation on a particular plaintiff on a year-by-year basis are complex and time consuming. Elaborate expert testimony is required. The battle of economic experts is apt to shift the trial’s emphasis from the determination of liability and the estimation of basic damages into the formulation of predictions concerning the national, indeed, global, economic future, including the likelihood of future increases or decreases in oil prices and the stability of foreign governments. Under Culver I, defendants were.free not only to controvert the plaintiffs’ damage calculations; they could also dispute the validity of the method by which the plaintiffs’ experts calculated damages. If such a dispute evolved, the fact-finder would be required to decide which of at least two competing methods, each avouched by an expert, it would use to determine discount before applying the method it chose to determine the rate of discount, yet another determination to be based on contradictory expert testimony.

Different formulae, each thought to be theoretically accurate, might be applied in literally hundreds of individual cases because of the conclusions reached by different fact-finders. The results in otherwise similar cases would vary widely depending on the particular expert witnesses called, on the fact-finders’ agreement or disagreement with the methods they advocated, and on the different fact-finders’ evaluations of their testimony. The voyage in search of mathematical certainty would discover instead a continent of conflict and conjecture.

Anticipating these problems to some degree in Culver I, we expressed hope that the parties would, by agreement, eliminate the need to litigate many of these confusing issues:

In the great majority of cases, we believe, the parties can and will be able to stipulate to the methodology, discount rate, the admissibility of economic data, tables, etc. and other technical aspects, as well as to any particular issues of fact underlying the calculations.

Culver I, 688 F.2d at 311. We continue to express that hope. Nothing in this opinion prevents the parties from stipulating to any of these matters or to anything else that may affect the determination of damages. We cannot ignore the possibility, however, that in many, perhaps most, cases such stipulations will not be possible. This opinion determines only the method to be followed when the parties cannot agree.

In such cases, litigating in every instance the amount and the effect of future price inflation is not likely to produce the result that would have been obtained had the plaintiff not been injured.14 No one can accurately predict the course of future inflation. A survey of the general literature for the past several years illustrates a sorry tale of repeated confusion, contradiction *120and uncertainty in economic forecasts.15 Current complaints about economic forecasting generally concern relatively short-term forecasts. Over a longer term, events as diverse — and unpredictable — as spending to finance a Southeast Asian war, an oil embargo, disagreement between members of the oil cartel, or a drought causing widespread agricultural failures could profoundly increase inflation.16 A major depression or the discovery of significant and inexpensive fuel resources might reduce its present impact.

Whether or not the science of economics continues to be dismal,17 it is assuredly in this regard conjectural. The case-by-case method sacrifices efficiency and simplicity for pursuit of a “delusive exactness.”18 As the Court noted in Pfeifer: “It is perfectly obvious that the most detailed inquiry can at best produce an approximate result.”19

The inherent inaccuracy of specific forecasts of future inflation, coupled with the length and complexity of the proceedings necessary to engage in such forecasts, convinced the Supreme Court that the case-by-case method of adjusting damage awards for inflation “should be discouraged.”20 “[Sjince specific forecasts of future price inflation remain too unreliable to be useful in many cases, it will normally be a costly and ultimately unproductive waste of [plaintiffs’] resources to make such forecasts the centerpiece” of accident litigation.21 Recognizing the Court’s disapproval of the case-by-case method, we hold that it should not be used in this circuit to adjust damage awards for inflation.

III.

The paramount concern of a court awarding damages for lost future earnings is to provide the victim with a sum of money that will, in fact, replace the money that he would have earned. Arriving at a reasonable estimate of anyone’s financial future-involves estimates of a whole spectrum of factors. We commonly exclude many relevant factors from consideration on the basis that they are so speculative that they cannot accurately be determined. For example, we consider only work-life expectancy *121and do not take into account the possibility that a worker will change to work that is more pleasurable but pays less. When considering the loss suffered as a result of the death of a wage-earner, we do not consider the likelihood that a widowed spouse may remarry.22 Nor do we take into account the stability of an already accomplished remarriage, or the age, appearance or personality of the surviving spouse23

While factors such as these do in fact affect the loss of future income, we place a higher value on a reasonable and workable system for establishing damages than we do on detailed precision in arriving at an award. “[W]e must remember that the ultimate total damage figure awarded is the sum of a series of predictions, none of which involves mathematical certainty, and that it is the reasonableness of the ultimate figure that is really in issue in such a case as this.” Barnes v. United States, 685 F.2d 66, 70 (3d Cir. 1982) (quoting United States v. Furumizo, 381 F.2d 965, 970 (9th Cir. 1967)).24

“[I]f forecasts of future price inflation are not used, it is necessary [for the fact-finder] to choose an appropriate below-market discount rate.”25 Adoption of this method guards against the wide disparity in results, the extended duration in trial time, and the increased cost to the parties that may be anticipated if a specific forecast of future price inflation is made anew in each case involving loss of future earnings. It achieves a no more uncertain approximation of inflation’s effect on the damage award. Because it does not attempt a specific forecast of how inflation will affect the particular plaintiff before the court, however, it is less complex and time consuming.

Even establishing an appropriate below-market discount rate, however, is difficult. Economists do not yet fully understand the relationship between inflation and interest. Recent studies discredit the received wisdom, voiced a decade ago, that there is a constant real rate of interest that can be determined merely by filtering inflationary expectations from nominal interest rates.26 However, while the studies find that the real rate varies, estimates uniformly fix its amount over any fairly lengthy period as falling into a range that runs from three percent to a negative rate of -1.5%.27

*122We hold that fact-finders in this Circuit must adjust damage awards to account for inflation according to the below-market discount rate method. The parties may, if they wish, stipulate the below-market discount rate, as they may stipulate any other disputed issue. If they are unable to do so, they may introduce expert opinion concerning the appropriate rate. Other evidence about the effect of price inflation is inadmissible. Evidence about the likelihood that the earnings of an injured worker would increase due to personal merit, increased experience and other individual and societal factors continue, of course, to be admissible.28 We recognize that the Supreme Court declined in Pfeifer to select a single method of accounting for inflation. We are confronted, however, with the need to adapt that opinion to jury trials. We also think it desirable to afford litigants and the courts the opportunity to determine the actual operation of this less complex method in order that its efficacy for national use can be determined.

As we have noted, the discount rate may be affected by the fact-finder’s assumption about the type of investment the plaintiff will choose, for long-term investments usually yield higher nominal interest rate returns than short-term investments of the same quality. The Supreme Court having said in Pfeifer that it perceives “no intrinsic reason to prefer one assumption over the other,” we mandate neither.29 However, the fact-finder should not consider the plaintiff’s possible need for emergency funds as a factor in favor of short-term investments; the injured wage-earner should have no greater right to a resource against future emergencies than he would have had if he had continued to work.

In judge-tried cases, a trial court adopting a pre-tax discount rate between one and three percent will not be reversed if it explains the reasons for its choice.30 This guideline, however, goes only to the reasonableness of the correlation between the pre-tax market rate of interest and the inflation rate. As discussed above, this pretax discount rate must then be adjusted for tax effects. If supported by appropriate expert opinion, the trial judge might make no discount or even adopt a negative rate not to exceed -1.5% before adjusting for tax effects. In jury trials, the jury should be instructed in the usual fashion concerning the weight to be given expert opinion evidence. The jury may then be permitted to return a single-figure award for damages or it may be required to answer interrogatories stating, among other items, the amount of loss of future earnings for each year for which it makes an award, and the discount rate it chooses to apply. The court will then be able to compute the total award or to require the parties to complete the arithmetic.

IV.

Just as we cannot ignore the reality of inflation, we cannot ignore the fact that long term economic conditions may change or that economic forecasting may become more certain. We recognize too that, although both the Fifth and Eleventh Circuits will be bound by this decision, Congress has declared that we are no longer a single court.31 It fulfills rather than ig-*123ñores the will of Congress to acknowledge that, in the future, either circuit en bane will be free to reconsider this decision. Should counsel for any party in any future case think this decision no longer reflects long-term economic facts or contemporary economic theory, a proffer of evidence may be made in the trial court to preserve the issue for an appellate panel and, thereafter, for en banc consideration.

Although we adopt a single method for adjusting future damage awards, our ruling, made on application for rehearing, shall not apply to any case in which, before the date this opinion was published, either a jury returned a verdict after being instructed on the basis of the principles set forth in Culver I, or a judge made findings of fact fixing damages pursuant to those principles.

For these reasons, Parts I through VI of our prior en banc decision, ■ 688 F.2d 280, 283-95, are reaffirmed. The rule announced in Johnson v. Penrod Drilling Co., 510 F.2d 234 (5th Cir.) (en banc), cert. denied, 423 U.S. 839, 96 S.Ct. 68-69, 46 L.Ed.2d 58 (1975), is overruled. We withdraw, however, Parts VII through XII of our decision in Culver I. 688 F.2d at 295-311. The views expressed in this opinion shall govern the adjustment of federal damage awards for future lost earnings in this circuit.

Courts are not prophets and juries are not seers. In making awards to compensate injured plaintiffs or the dependents of deceased workers for loss of future earnings, however, these fact-finders must attempt, in some degree, to gauge future events. Absolute certainty is by the very nature of the effort impossible. It is also impossible to take into account every bit of potentially relevant evidence concerning the tomorrows of a lifetime. The approach we adopt attempts to assure plaintiffs a fair measure of damages, to give defendants a reasonable adjustment for reducing future losses to present value, and to avoid making trials even more complex and their results even more uncertain. It is the product of a balancing of competing values. Ultimately, however, that is the root of all justice.

The judgments in Culver, No. 79-3985, and Byrd, No. 78-3064, are REVERSED, and the cases are REMANDED to the district court for further proceedings on the issue of damages. The panel opinions are adopted in all other respects.

with whom FAY, POLITZ, R. LANIER ANDERSON, III, RANDALL, TATE, SAM D. JOHNSON and JERRE S. WILLIAMS, Circuit Judges, join, dissenting:

With the ink scarcely dry on Pfeifer (Jones & Laughlin Steel Corp. v. Pfeifer, -U.S.-,-, 103 S.Ct. 2541, 2554, 76 L.Ed.2d 768, 788 (1983)), in which, with remarkable unanimous clarity, the High Court chose not to pick one of the several methods in its collective economic hat it considered legally sound for calculating lost future wages in civil damage actions and instead chose to leave that crucial decision to the trial judge in that and future cases, a majority of this Court elevates one approach beyond any choice to be the only way, for mandatory application in eighteen federal judicial districts across a half-dozen Southern states,1 for a broad group of litigants to arrive at fair sums of money to compensate injured parties. Guided by the decision of Pfeifer, its approval of Culver I, and the direction that the choice of methodology should be left to the district court, not this Court, I respectfully dissent.2

*124I.

In overruling Penrod (Johnson v. Penrod Drilling Co., 510 F.2d 234 (5th Cir.) (en banc), cert. denied 423 U.S. 839, 96 S.Ct. 68-69, 46 L.Ed.2d 58 (1975)) which forbade proof, argument, or jury instruction concerning inflationary factors in damage awards in actions under maritime or federal law, “it [was] not our purpose to establish a single methodology” for all federal courts in the Old Fifth to thenceforth account for such factors in determining future lost earnings. Culver v. Slater Boat Co., 688 F.2d 280, 305 (5th Cir. 1982) (en banc) (Cul-ver I). Rather, we adopted a “flexible approach to the problem which [w]ould result in fairness regardless of the economic circumstances that exist at the time of trial” and left open the opportunity for each party to offer relevant evidence supporting or refuting any of the several discussed theories of calculating lost earnings. Id. at 286. “In the dynamic and ever-changing world of finance and economics,” we felt that “any standard which is inflexible . . . will likely be unable to cope with the problem of preventing windfalls either to plaintiff or defendant.” Id. at 300. Our goal was “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.” Id.

Pfeifer, on the heels of Culver I, clearly follows this flexible approach. The Supreme Court unequivocally refused to establish anything but “the general boundaries within which a particular award will be considered legally acceptable,”3 U.S. at -, 103 S.Ct. at 2555, 76 L.Ed.2d at 789, simply because no one method of calculating lost earnings was above some legitimate criticism and thus appropriate for all cases.

The Court questioned certain aspects of each of the three principal proposed methods, making it clear that under some circumstances one or more of the three could be inappropriate, (i) It thought that methods employing specific forecasts of future inflation on a case-by-case basis “remain too unreliable to be useful in many cases,” - U.S. at -, 103 S.Ct. at 2556, 76 L.Ed.2d at 790 (emphasis added), and discouraged their general use. (ii) As to the real interest (or below-market) discount rate approaches, the Court found the “economic evidence distinctly inconclusive regarding an essential premise of those approaches,” id. — the “key premise ... that the real interest rate remains stable over time.” Such a rate is “obviously not perfectly stable,” and, the Court observed, “whether it is even relatively stable is hotly disputed among economists.” Id. at -n. 30, 103 S.Ct. at 2556 n. 30, 76 L.Ed.2d at 790 n. 30; accord Culver I, 688 F.2d at 302 (“As evidence becomes available that the next 20 or 30 years will not be like the past 20 or 30 years, then the Feldman [real interest rate] approach is less helpful.”). Even the majority recognizes this.4 (iii) Lastly, it “was not prepared to impose [the “total offset” approach] on unwilling litigants,” as had the Third Circuit below, for such approaches rely on at least one assumption the Court could not unquestioningly accept — that “the national patterns of wage growth are likely to reflect the patterns within any given industry.”-U.S. at -, 103 S.Ct. at 2557, 76 L.Ed.2d at 791-92; cf. Culver I, 688 F.2d at 299 (“penalizes defendants because . .. interest *125rates on relatively safe investments will typically ride several percentage points above the rate of inflation”). To the Court, “[t]he legislative branch of the federal government [was] far better equipped than [it] to perform a comprehensive economic analysis and to fashion the proper general rule.” -U.S. at-, 103 S.Ct. at 2557, 76 L.Ed.2d at 792. Importantly, though, the Court characterized each of the three methods as “legally acceptable.”

Consequently, the Court chose to allow to the trial judge on remand discretion to apply in a deliberate choice any one or more of the “legally acceptable” methods the Court discussed — those involving specific forecasts of future inflation, those using a below-market or real interest discount rate, or those based on the notion of total offset. That is precisely what Culver I ordained: no one method was either prescribed or forbidden. And Pfeifer approved Culver I5

II.

Despite the clear holding in Pfeifer and its approval of Culver J’s approach, the majority today declares that, though “the Supreme Court declined in Pfeifer to select a single method of accounting for inflation,” unless the parties can stipulate to the contrary, all “fact-finders in this Circuit must adjust damage awards to account for inflation according to the below-market discount rate method.” See supra at 16123. The Court takes from the district court— and puts squarely in the hands of the Court of Appeals — the responsibility of choosing, from among competing economic theories, the methodology or methodologies which could be used to most equitably and adequately determine future lost earnings under the circumstances of that case. Apparently this is done to spare jurors from the rigors of listening to presentations on such competing methods (which the Court seems “convinced” will invariably turn into “graduate seminar[s] on economic forecasting,” see supra at 117), and from the difficulty of making such reasoned choices. In doing so, the Court makes choices of economic fact “hotly disputed [even] among economists.”

Pfeifer, like Culver I, however plainly evidences a confidence that trial courts6 are fully capable and actually better able to make such difficult decisions on the facts of individual cases, at least until Congress declares otherwise.

I thus must dissent.7