Martha Ann Brundage Rozier v. Ford Motor Company

SIMPSON, Circuit Judge: The controlling issue raised by this appeal concerns the post-trial relief available to a party who has lost a civil suit after an adverse party failed to disclose relevant information called for by interrogatories and court order. For reasons not stated in its order, the district court denied plaintiffs motion for a new trial pursuant to Rule 60(b), Fed.R.Civ.P. timely filed when plaintiff’s counsel, after adverse jury verdict and final judgment, learned of the existence of the undisclosed material. Aware of our limited role in reviewing this discretionary function, we nonetheless hold that the unique facts of this case require reversal and remand for a new trial. I. THE FACTS On March 13, 1973, the 1969 Ford Galaxie 500 in which William Rozier was riding as a passenger on a Georgia public highway was struck from behind by a faster moving vehicle driven by Benjamin J. Wilson, Jr. The impact caused the Ford’s fuel tank to rupture, resulting in a fire which engulfed the car and severely burned Mr. Rozier. Within 24 hours, he died as a result of the burns he sustained. On August 26, 1974, his widow, Martha Ann Brundage Rozier, filed suit below against Ford Motor Company (Ford), based on diversity jurisdiction (Title 28, U.S.C., § 1332), alleging that Ford’s negligent design of the 1969 Galaxie’s fuel tank caused the death of her husband. After a one week trial, the jury returned a verdict for Ford. Judgment was entered on March 6, 1976, and Mrs. Rozier timely filed her notice of appeal to this Court, No. 76-2848. During the pendency of that appeal, counsel for Mrs. Rozier learned of the existence of a document prepared by a Ford cost engineer, A. Mancini, in 1971 and arguably covered by plaintiff’s interrogatories in this case, as limited by an order of the trial judge entered on January 6, 1976. Because Ford had failed to produce this document in response to the court’s order, Mrs. Rozier, on February 9, 1977, filed a motion for a new trial pursuant to Rule 60(b)(2), Fed.R.Civ.P., newly discovered evidence, and 60(b)(3), fraud, misrepresentation, and other misconduct. After hearing oral arguments and considering briefs and affidavits filed by the parties, the district court denied the motion on all grounds. Mrs. Rozier’s appeal from this later order, No. 77-1929, has been consolidated with her appeal from the original judgment. II. PLAINTIFF’S RULE 60(b)(3) MOTION The pivotal question in this case is whether the trial judge abused his discretion in denying Mrs. Rozier’s motion for a new trial pursuant to Rule 60(b)(3), Fed.R.Civ.P. We hold that he did and reverse on that basis. Rule 60(b)(3) provides as follows: On motion and upon such terms as are just, the court may relieve a party or his legal representative from a final judgment, order, or proceeding for the following reasons: . ., . (3) fraud (whether heretofore denominated intrinsic or extrinsic), misrepresentation, or other misconduct of an adverse party. Under the express terms of the rule, 60(b)(3) motions must be made within a reasonable time, not more than one year, after the challenged judgment was entered. In this case, Mrs. Rozier moved for a new trial, relying on Rule 60, less than a year after entry of the district court’s judgment for Ford, and only five days after her counsel received the information upon which the motion was based. Because Mrs. Rozier’s 60(b)(3) motion was filed timely, we have no occasion to review Ford’s conduct in light of the more exacting “fraud upon the court” standard also provided for by Rule 60(b), but not subject to *1338 any time limitation. 1 With few exceptions, the cases cited by Ford in support of its argument for affirmance deal with motions made after the one year limitation period had run. 2 Consequently, these cases were decided on the basis of Rule 60(b)’s “saving clause”, and were limited to consideration of whether the challenged conduct amounted to fraud upon the court. Cases in other Circuits make clear that “fraud upon the court” under the saving clause is distinguishable from “fraud . misrepresentation, or other misconduct” under subsection (3). As the district court explained in United States v. International Telephone & Telegraph Corp., 349 F.Supp. 22, 29 (D.Conn.1972), aff’d without opinion, 410 U.S. 919, 93 S.Ct. 1363, 35 L.Ed.2d 582 (1973): Generally speaking, only the most egregious misconduct, such as bribery of a judge or members of a jury, or the fabrication of evidence by a party in which an attorney is implicated, will constitute a fraud on the court. See Hazel-Atlas Glass Co. v. Hartford-Empire Co., 322 U.S. 238, 64 S.Ct. 997, 88 L.Ed. 1250 (1944); Root Refin. Co. v. Universal Oil Products, 169 F.2d 514 (3d Cir. 1948) 7 J. Moore, Federal Practice, ¶ 60.33 at 510-11. Less egregious misconduct, such as nondisclosure to the court of facts allegedly pertinent to the matter before it, will not ordinarily rise to the level of fraud on the court. See Kupferman v. Consolidated Research & Mfg. Co., 459 F.2d 1072 (2d Cir. 1972); see also England v. Doyle, 281 F.2d 304, 310 (9th Cir. 1960). Alternately stated, “[in] order to set aside a judgment or order because of fraud upon the court under Rule 60(b) . . . it is necessary to show an unconscionable plan or scheme which is designed to improperly influence the court in its decision.” England v. Doyle, supra, 281 F.2d at 309. See also United States v. Standard Oil Co. of Calif., 73 F.R.D. 612, 615 (N.D.Cal.1977). The distinction between a 60(b)(3) motion and a motion alleging fraud upon the court is rooted in policies basic to the law of judgments, as the Supreme Court explained prior to the 1946 revision of Rule 60: Federal courts . . . long ago established the general rule that they would not alter or set aside their judgments after the expiration of the term at which the judgments were finally entered. This salutary general rule springs from the belief that in most instances society is best served by putting an end to litigation after a case has been tried and judgment entered. . . . From the beginning there has existed along side the term rule a rule of equity to the effect that under certain circumstances, one of which is after-discovered fraud, relief will be granted against the judgments regardless of the term of their entry. Hazel-Atlas Glass Co. v. Hartford-Empire Co., supra, 322 U.S. at 244, 64 S.Ct. at 1000 (citations omitted). Although Rule 60(b) substitutes a general one year limitations period for the earlier “term rule”, it continues to reflect a strong policy favoring an end to litigation by severely restricting the relief available after the one year limit has run. 3 In the year after a judgment has been entered, however, the district courts have greater discretion to balance the policy of finality of judgments against the other salutary policies embodied in the alternate grounds for relief provided in subsections (1) through (3) *1339 of Rule 60(b). Essentially, this discretion, as guided by the Rule, furnishes an escape valve to protect the fairness and integrity of litigation in the federal courts. Our review in this case must focus on two questions: (1) Did the plaintiff satisfy the threshold requirements for relief under Rule 60(b)(3)? And, if so, (2) would the granting of a new trial in this case effectuate any policy more significant than that of preserving the finality of judgments? 1. Rule 60(b)(3) Requirements: One who asserts that an adverse party has obtained a verdict through fraud, misrepresentation or other misconduct has the burden of proving the assertion by clear and convincing evidence. Saenz v. Kenedy, 178 F.2d 417, 419 (5th Cir. 1949); Gilmore v. Strescon Industries, Inc., 66 F.R.D. 146, 153 (E.D.Pa.1975), aff’d without opinion, Bucks County Const. Co. v. P. Agnes, Inc., 521 F.2d 1398 (3d Cir.). The conduct complained of must be such as prevented the losing party from fully and fairly presenting his case or defense. Toledo Scales Co. v. Computing Scale Co., 261 U.S. 399, 421, 43 S.Ct. 458, 464, 67 L.Ed. 719 (1923); Atchison, Topeka & Santa Fe Ry. Co. v. Barrett, 246 F.2d 846, 849 (9th Cir. 1957); Rubens v. Ellis, 202 F.2d 415, 417 (5th Cir. 1953). Although Rule 60(b)(3) applies to misconduct in withholding information called for by discovery, Petry v. General Motors Corp., 62 F.R.D. 357 (E.D.Pa.1974), it does not require that the information withheld be of such nature as to alter the result in the case. Seaboldt v. Pennsylvania RR. Co., 290 F.2d 296, 299-300 (3d Cir. 1961). See generally, 11 C. Wright & A. Miller, Federal Practice and Procedure, § 2861 (1970). This subsection of the Rule is aimed at judgments which were unfairly obtained, not at those which are factually incorrect. 4 The facts relevant to Mrs. Rozier’s 60(b)(3) motion are as follows. On August 25, 1975, counsel for Mrs. Rozier served her fourth set of interrogatories upon counsel for defendant Ford. Interrogatories 8, 10, 12, and 16 asked whether Ford had conducted “any cost/benefit analyses” with respect to four possible design modifications of fuel tanks for passenger cars, including full-sized sedans and hard-tops. Interrogatory 19 requested similar information not limited to “cost/benefit analyses”: 19. In conducting “its own in-house research and development work on . alternate fuel tank locations” over the last ten years (Stenning Deposition Vol. II, p. 32), has Ford Motor Company prepared any written reports or analyses of the comparative advantages or disadvantages of alternate locations, (e. g. on top of the rear axle or in front of the rear axle) for fuel tanks in full-sized sedans and hardtops, including the 1969 Galaxie 500? R. 434. 5 The interrogatories also asked whether, in the event such documents exist, Ford would make them available for inspection and copying without the necessity of a request for production. In response, Ford objected to these and other interrogatories on the grounds that “the information sought therein does not relate to vehicles of the same size, chassis and fuel system as the 1969 Ford Galaxie 500”. R. 445. On December 11, 1975, counsel for Mrs. Rozier moved in writing to compel Ford to answer the fourth set of interrogatories. R. 484. On January 6, 1976, the district court entered an order directing in part, as follows: *1340 Defendant shall file with the Court and serve upon Plaintiff’s counsel, no later than January 21, 1976, answers to questions numbered 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 19, and 20 of Plaintiff’s Fourth Interrogatories; provided, however, that the aforesaid questions are limited to such written requirements, cost/benefit analyses, and written reports or analyses which are applicable to the 1969 Ford Galaxie 500. The fact that a written requirement, cost/benefit analysis or written report may also be applicable to vehicles other than the 1969 Ford Galaxie 500 does not render it beyond the scope of the aforesaid interrogatory and must be disclosed in response to the aforesaid interrogatories. R. 578-79 (emphasis in original). Finally, on January 22, 1976, in purported compliance with this order, Ford filed amended responses to the fourth set of interrogatories. In answering interrogatories 8, 10, 12 and 16 concerning cost/benefit analyses of alternate fuel tank designs, Ford stated: “There would be no formal cost-benefit analysis with regard to this information”. R. 606. In response to interrogatory 19, concerning “written reports of analyses of the comparative advantages and disadvantages of alternate locations . for fuel tanks”, Ford stated: “Defendant cannot find such written analysis covering the inquiry”. R. 609. Approximately one year after Ford filed these amended responses — and ten months after the jury returned a verdict favorable to Ford — plaintiff’s counsel learned of the document from Ford’s files at issue in this appeal. Dated “2/9/71”, this “Confidential Cost Engineering Report” states as its subject: Trend Cost Estimate — Fuel Tank Proposals (30 MPH Safety Std.) — Prop[osal] I— Tank Over Rear Axle Surrounded by Body Sheet Metal Barrier & Proposal] II — Tank in Tank Filled With Polyurethene Vs. 1971 Ford Design — 1975 Ford/Mercury. So far as we can determine from the face of the document and on the basis of remarks by counsel for Ford during oral argument, this “Trend Cost Estimate” was prepared in anticipation of a revised National Highway Traffic Safety Administration safety standard of 30 m. p. h. for rear end collisions. It compares the costs of parts and labor associated with two proposed alternate fuel tank designs based on the design of a 1971 full-sized Ford. Without dispute the 1969 Galaxie 500 model was a full-sized Ford car. Apparently, Ford planned to begin using a new fuel tank design sufficient to satisfy the 30 m. p. h. standard in its 1975 full-sized models. In her motion for a new trial pursuant to Rule 60(b)(3), Mrs. Rozier contended that the 1971 Trend Cost Estimate should have been produced in response to the district court’s January 6, 1976, order because it was “applicable to the 1969 Ford Galaxie 500”. In support of her motion, Mrs. Rozier filed affidavits by Frederick E. Arndt, a safety engineering consultant, and Byron Bloch, a product safety consultant, both of whom had testified on her behalf at the trial. Each affidavit stated: [T]he 1969 Ford Galaxie 500 is so similar in size, design,- and fuel tank location to the 1971 Ford that the subject “Confidential Cost Engineering Report” is as applicable and valid for the 1969 Ford Galaxie 500 as it is for the 1971 Ford. S.R. [Supplemental Record] 49, 51. In response to Mrs. Rozier’s motion to vacate, Ford produced an affidavit by Thomas G. Grubba, an attorney on the house legal staff of Ford Motor Company, who “was involved in” the case. Mr. Grubba swore that he was unaware of the Trend Cost Estimate when the answers to plaintiff’s interrogatories were prepared, but that he became aware of the document on February 25, 1976. 6 S.R. 114. We note *1341 that although Ford’s answers were filed on January 22, 1976, the trial did not begin until March 1, 1976, a week after Mr. Grubba discovered the document. By any fair reading, the district court’s January 6, 1976, discovery order called for production of this Trend Cost Estimate. Ford, in response to the motion to vacate and in this appeal, has urged that, as a term of art, a “trend cost estimate” is not a “cost/benefit analysis”. Whether the document in question technically is or is not a cost/benefit analysis — to our non-expert eyes, the terms are synonymous, as alike as “Tweedledum and Tweedledee”- — is largely irrelevant in this case because plaintiff’s interrogatories were not limited to “cost/benefit analyses”. Interrogatory 19 asked whether Ford had “prepared any written reports or analyses of the comparative advantages or disadvantages of alternate locations ... for fuel tanks” (emphasis added); the court’s January 6 order compelled production of all such “written reports or analyses which are applicable to the 1969 Ford Galaxie 500”, specifically noting that written reports also applicable to vehicles other than the 1969 Ford Galaxie 500 were not beyond the scope of the discovery order “and must be disclosed in response to the aforesaid interrogatories”. Undeniably, the Trend Cost Estimate is a report “of the comparative advantages or disadvantages” of alternate fuel tank locations, since the alternative which could satisfy the safety standard for the least cost would, in terms of Ford’s interests, be the most advantageous. 7 Also, in light of the unchallenged assertions made in the Arndt and Bloch affidavits, this estimate based on the design of a 1971 full-sized Ford is applicable to a 1969 full-sized Ford. 8 We recognize that the Trend Cost Estimate does not purport to estimate the cost of installing new fuel tanks on automobiles already manufactured, such as a 1969 Galaxie 500; however, neither the interrogatory nor the court order limited discovery to reports of that description. We conclude that Mrs. Rozier has proved by clear and convincing evidence that Ford engaged in misrepresentation and other misconduct. 9 Plaintiff’s interrogato *1342 ry 19, as limited by the district court’s order, called for production of the Trend Cost Estimate. In its written response, Ford stated that it could find no such report. A month later, an in-house attorney for Ford involved in this case discovered the Trend Cost Estimate but failed to disclose it or to amend the inaccurate response to interrogatory 19. 10 If Ford in good faith believed that the district court’s order was not intended to compel production of this document, the appropriate remedy was to seek a ruling by the district court at that point and not a year after the trial and then only when, by chance, the plaintiff learned of it. The more vexing question is whether nondisclosure of the Trend Cost Estimate prevented Mrs. Rozier from fully and fairly presenting her case. At trial, Mrs. Rozier contended that Ford was negligent in designing a fuel tank that could not withstand an impact such as that involved in the accident which took her husband’s life. Prior to trial, she expressed an intention to rely on 14 theories to explain how Ford deviated from the appropriate standard of care. R. 471-72. Inevitably, information developed in the discovery stages of the case influenced the decision as to which theories would be emphasized at trial. We are left with the firm conviction that disclosure of the Trend Cost Estimate would “have made a difference in the way plaintiff’s counsel approached the case or prepared for trial”, Rock Island Bank & Trust Co. v. Ford Motor Co., 54 Mich.App. 278, 220 N.W.2d 799 (1974), and that Mrs. Rozier was prejudiced by Ford’s nondisclosure. See Seaboldt v. Pennsylvania RR. Co., 290 F.2d 296, 299-300 (3d Cir. 1961). Although Ford does not specifically dispute this conclusion, it has argued, in a related context, that the Trend Cost Estimate, if admissible at trial, would merely have been cumulative because plaintiff’s experts testified at length as to the feasibility of alternative fuel tank designs. Additionally, Ford contends that the document would have been inadmissible by virtue of Rule 407, Fed.R.Evid. We think that these arguments misconstrue the significance of the withheld document. [1L 12] The admissibility of evidence is irrelevant in the discovery process so long as “the information sought appears reasonably calculated to lead to the discovery of admissible evidence”. Fed.R.Civ.P. 26(b)(1). *1343 The Trend Cost Estimate clearly satisfies this test. It was not an isolated document, but rather one in a series; some of the other documents, also not produced by Ford during discovery, are referred to in the estimate itself: “Request from J. M. Chiara”, “Layouts LA-901277 and 901278”, “our report dated February 4, 1971”, “A proposal for a fuel tank installation over the rear axle is illustrated in the attached drawing”. At a minimum, production of the Trend Cost Estimate could have led to discovery of these other documents. Nevertheless, while admissibility is not a sine qua non for granting relief under Rule 60(b)(3) where the wrongful withholding of information during discovery is alleged, it may be a relevant factor in weighing the prejudice suffered by the moving party. Ford argues against admissibility by invoking the protective mantle of Rule 407, Fed.R.Evid.: When, after an event, measures are taken which, if taken previously would have made the event less likely to occur, evidence of the subsequent measures is not admissible to prove negligence or culpable conduct in connection with the event. This rule does not require the exclusion of evidence of subsequent measures when offered for another purpose, such as proving ownership, control, or feasibility of cautionary measures, if controverted, or impeachment. Rule 407 does not preclude admission of the Trend Cost Estimate for several reasons. First, this 1971 document was not written after the “event” in question, a 1973 automobile accident, and was not, in itself, a remedial measure taken. Hence, the threshold requirements for invoking the rule are absent. Secondly, the rule of exclusion is based “on a social policy of encouraging people to take, or at least not discouraging them from taking, steps in furtherance of added safety”. Notes of Advisory Committee on Proposed Rules, Fed.R.Evid. 407. Invoking this policy to justify exclusion here is particularly inappropriate since the estimate was prepared not out of a sense of social responsibility but because the remedial measure was to be required in any event by a superior authority, the National Highway Traffic Safety Administration. Finally, even if we assume that the Trend Cost Estimate qualifies as evidence of a subsequent remedial measure, it would be admissible as proof of subsidiary issues in the case, such as knowledge of the dangerous condition or feasibility of precautionary measures. 11 Our acceptance of Ford’s 407 argument would effectively “turn the blade inward”. We cannot know what use, if any, plaintiff’s counsel would have made of the Trend Cost Estimate had it been produced by Ford prior to trial. However, consideration of one likely use reveals the prejudice that Mrs. Rozier may have suffered as a consequence of Ford’s misconduct. The negligence alleged by Mrs. Rozier may have taken place in one or both of two time frames: (1) up to and including the production of the 1969 Ford Galaxie 500, and (2) between the time of production in 1969 and the time of the fatal collision in 1973. Under the facts of this case, the task of proving negligence in the pre-production period was the more difficult. Given the *1344 industry practices and standards in 1969, the jury was less likely to find that Ford’s conduct fell below that to be expected from an ordinarily prudent manufacturer in designing the fuel system for a full-sized sedan. In the post-production period, however, Ford may have had a duty to warn consumers of a latent danger such as a defectively designed fuel tank. Here the plaintiff need not have been hamstrung by the less sophisticated state of the art in 1969, and may have been able to make a convincing case based on the Trend Cost Estimate. Georgia courts 12 recognize a cause of action against automobilé manufacturers for negligence in failing to warn of latent dangers arising from defective design. Friend v. General Motors Corp., 118 Ga.App. 763, 165 S.E.2d 734, 737 (1968). And, as the Eighth Circuit has explained in discussing this duty to warn: The failure to use reasonable care in design or knowledge of a defective design gives rise to the reasonable duty on the manufacturer to warn of this condition. Larsen v. General Motors Corp., 391 F.2d 495, 505 (8th Cir. 1968) (emphasis added). 13 Here, arguably, the Trend Cost Estimate furnishes evidence that Ford, in 1971, had knowledge that the fuel tanks on those models with designs comparable to the 1969 Galaxie 500 could not withstand rear-end collisions at 30 m.p.h. or greater and that a design to correct this condition was both economically and structurally feasible. If a jury were persuaded that the fuel tank of the 1969 Galaxie was latently but not negligently defective at the time of production, it might still have found that once Ford acquired the knowledge evidenced by the Trend Cost Estimate it negligently failed to warn Galaxie users of the defect. 14 *1345 By a peculiar turn of events, the jury in this case was foreclosed at the last minute from considering Ford’s failure to warn as a possible negligent omission. At least as early as December 10, 1975, when the parties filed a “Consolidated Proposed Pre-Trial Order”, Mrs. Rozier indicated that one “act of negligence” upon which she intended to rely was “[fjailure to warn owners and users of the vehicle of the risk of injury or death due to fire in the event of a collision”. R. 471-72. The trial judge so instructed the jury, T. 5A, noting that although several acts of negligence were alleged, the plaintiff “is merely required to prove one such act. . . .” T. 14A. Counsel for Ford objected to the warning instruction because it did not specify that the duty to warn extended only to unreasonable risks of injury. The exchange which followed this objection suggests that plaintiff’s counsel, in the light of the information then available to him, viewed the duty to warn as an inconsequential part of the case: THE COURT: What do you say about the last exception? MR. DEVINE (For Plaintiff): Your Hon- or, I think it might do well to clarify that. The Jury has to find that there is a danger before the Defendant has a duty to warn of that danger. I believe that there is plenty of evidence to support that there is a danger. THE COURT: Nevertheless, what do you say . . . MR. DEVINE: I think it should be clear. THE COURT: . . . You concede that I should withdraw then that charge from the Jury’s consideration? MR. DEVINE: I have no objection, Your Honor. * * * * * * The Court thereupon further charged the Jury as follows: THE COURT: Members of the Jury, I had charged you that the Plaintiff claimed that the Defendant was negligent in failing to warn owners and passengers of the 1969 Ford Galaxie 500 of the risk of burn injury attendant to occupying that vehicle. I now withdraw that charge, and you will not give that charge any weight whatsoever in your deliberations. The following colloquy ensued out of the presence of the jury: THE COURT: Any exceptions to that? MR. WEINBERG: No, sir. MR. DEVINE: No, sir. T. 32A-33A. It is apparent, then, that the Trend Cost Estimate, far from being a cumulative tidbit of evidence already subsumed in the case presented to the jury, might have been the catalyst for an entirely different approach to the case on a theory that the plaintiff, lacking the document, let die before it reached the jury. Under these circumstances, we hold that Ford’s wrongful withholding of information prevented Mrs. Rozier from fully and fairly presenting her case. 2. Policy Considerations: Our system of civil litigation cannot function if parties, in violation of court orders, suppress information called for upon discovery. “Mutual knowledge of all the relevant facts gathered by both parties is essential to proper litigation. To that end, either party may compel the other to disgorge whatever facts he has in his possession”. Hickman v. Taylor, 329 U.S. 495, 507, 67 S.Ct. 385, 392, 91 L.Ed. 451 (1947). The Federal Rules of Civil Procedure substitute the discovery process for the earlier — and inadequate — reliance on pleadings for notice-giving, issue-formulation, and fact-revelation. As the Supreme Court stated in Hickman v. Taylor, supra, “civil trials in the federal courts no longer *1346 need be carried on in the dark. The way is now clear, consistent with recognized privileges, for the parties to obtain the fullest possible knowledge of the issues and facts before trial”. 329 U.S. at 501, 67 S.Ct. at 389. The aim of these liberal discovery rules is to “make a trial less a game of blind man’s bluff and more a fair contest with the basic issues and facts disclosed to the fullest practicable extent”. United States v. Proctor & Gamble Co., 356 U.S. 677, 683, 78 S.Ct. 983, 986, 2 L.Ed.2d 1077 (1958). It is axiomatic that “[discovery by interrogatory requires candor in responding”. Dollar v. Long Mfg., N.C., Inc., 561 F.2d 613, 616 (5th Cir. 1977). Through its misconduct in this case, Ford completely sabotaged the federal trial machinery, precluding the “fair contest” which the Federal Rules of Civil Procedure are intended to assure. Instead of serving as a vehicle for ascertainment of the truth, the trial in this case accomplished little more than the adjudication of a hypothetical fact situation imposed by Ford’s selective disclosure of information. The policy protecting the finality of judgments is not so broad as to require protection of judgments obtained in this manner. Within a year of their entry, judgments obtained through fraud, misrepresentation or other misconduct may be set aside under Rule 60(b)(3). Although the granting of such relief is within the discretion of the trial court, Hand v. United States, 441 F.2d 529 (5th Cir. 1971), the rule “is remedial and should be liberally construed”. Atchison, Topeka & Santa Fe Ry. Co. v. Barrett, supra, 246 F.2d at 849. In reviewing the instant denial of plaintiff’s Rule 60(b)(3) motion for abuse of discretion, it is not without significance that the trial judge stated no reasons for the denial. S.R. 127. Cf. Dollar v. Long Mfg., N.C., Inc., supra, 561 F.2d at 618. We have searched for reasons justifying denial of the motion and can find none sufficient to sustain this exercise of discretion. Under the unique facts of this case, the policy of deterring discovery abuses which assault the fairness and integrity of litigation must be accorded precedence over the policy of putting an end to litigation. We do not reach the question of whether the district court abused its discretion in denying plaintiff's motion for a new trial based on newly discovered evidence, pursuant to Rule 60(b)(2). Under that Rule, the moving party must show, inter alia, that the newly discovered evidence is “such that a new trial would probably produce a new result”. Ag Pro, Inc. v. Sakraida, 512 F.2d 141, 143 (5th Cir. 1975). To hold the plaintiff in this case to such a showing would be manifestly unfair: [I]t cannot be stated with certainty that all of this would have changed the result of the case. But, as said by the Supreme Court, a litigant who has engaged in misconduct is not entitled to “the benefit of calculation, which can be little better than speculation, as to the extent of the wrong inflicted upon his opponent”. Minneapolis, St. Paul & S.S. Marie Ry. Co. v. Moquin, 1931, 283 U.S. 520, 521-22, 51 S.Ct. 501, 502, 75 L.Ed. 1243. Seaboldt v. Pennsylvania R.R. Co., supra, 290 F.2d at 300. We hold that the district court abused its discretion in denying plaintiff’s Rule 60(b)(3) motion and that a new trial is required. III. THE EVIDENTIARY QUESTION In her original appeal, Mrs. Rozier alleged that the district court erred in admitting into evidence over plaintiff's objection the plea of guilty by Benjamin Wilson, the driver of the impacting vehicle, to charges of involuntary manslaughter in the deaths of Mr. Rozier and Frank Mitchell, the driver of the Ford Galaxie. Although not necessary to our disposition of this appeal, we address this issue because of the near certainty that the question will arise at a new trial. Ford introduced a certified copy of Wilson’s guilty plea for two stated purposes: (1) to corroborate the testimony of the investigating officer, Trooper Swindell, that Wilson’s car was traveling at approxi *1347 mately 68 m.p.h. at the moment of impact, and (2) as evidence that Mr. Rozier’s death was caused by Wilson’s criminal act and not by Ford’s negligence. While we agree with Ford that this evidence is not excluded by the hearsay rule by virtue of Rule 803(22), Fed.R.Evid., we hold that it is inadmissible under the test for legal relevancy set forth in Rule 403. Under the Federal Rules of Evidence, admissibility is predicated on more than mere logical relevance: Although relevant, evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence. Fed.R.Evid. 403. In determining legal relevance under Rule 403, the trial judge has broad discretion, reviewable only for abuse. United States v. Johnson, 558 F.2d 744 (5th Cir. 1977); United States v. Bailey, 537 F.2d 845 (5th Cir. 1976). As relevant to the speed of Wilson’s car, there is little question that the guilty plea amounted to a “needless presentation of cumulative evidence”. Trooper Swindell testified that the impact speed was about 68 m.p.h., and that figure is recorded on his accident report, Defendant’s Exhibit 19, and on the speeding citation that he issued to Wilson, Defendant’s Exhibit 20. 15 Additionally, plaintiff did not dispute that Wilson was traveling at 68 m.p.h. Wilson’s speed was never a contested issue at trial. Whether admission of this cumulative evidence of speed constituted an abuse of discretion under Rule 403 we need not decide in light of the discussion which follows. We recognize that Wilson’s guilty plea was logically relevant to the issue of causation in that, by the plea, Wilson admitted that he “did cause the death of William Burel Rozier”. We conclude, however, that the slight probative value of this evidence was clearly outweighed by the danger that it would confuse and mislead the jury. The legal concepts of cause in fact, proximate cause, and policy cause are confusing in any case, especially one such as this involving the doctrine of “second collision” injuries. Under the second collision principle, [t]here is no rational basis for limiting the manufacturer’s liability to those instances where a structural defect has caused the collision and resulting injury. This is so because even if a collision is not caused by a structural defect, a collision may precipitate the malfunction of a defective part and cause injury. In that circumstance the collision, the defect, and the injury are interdependent and should be viewed as a combined event. Such an event is the foreseeable risk that a manufacturer should assume. Huff v. White Motor Corp., 565 F.2d 104, 109 (7th Cir. 1977). There are thus several dimensions to the cause question in this case. On a purely factual level, Mr. Rozier’s death was caused by a variety of factors, among them the impact of Wilson’s car as well as the design of the Galaxie fuel tank. The legal cause of death, however, must be evaluated in the context of the social policies sought to be advanced by attaching liability to the consequences of specific actions by specific persons. As Professor Prosser explains, “[o]nce it is established that the defendant’s conduct has in fact been one of the causes of the plaintiff’s injury, there remains the question whether the defendant should be legally responsible for what he has caused”. W. Prosser, Handbook of the Law of Torts 244 (4th ed. 1971). In this case, the State of Georgia charged Wilson with causing the death of Mr. Rozier in an effort to enforce *1348 its traffic laws; in terms of this important social interest, it makes sense to hold the driver of a speeding vehicle responsible for damage inflicted by it. Georgia also allows a private action in damages against an automobile manufacturer who negligently produces a car with a defect which causes injury when activated by a foreseeable collision; thus, in the context of encouraging automotive safety, Georgia law has determined that the negligent manufacturer “causes” second collision injuries even though the primary collision was caused by a third party. The jury in this case hardly needed Wilson’s guilty plea to demonstrate what had caused the accident in which Mr. Rozier lost his life. The fact of the accident was conceded by the plaintiff, and numerous photographs put in evidence by both parties graphically depicted the devastating consequences of the impact. Certainly, the jury realized that the fuel tank would not have ruptured had it not been for the crushing blow inflicted by Wilson’s car. The real purpose for introducing the guilty plea would seem to be as evidence of legal cause. It is here that the danger of confusion arises. 16 Because Wilson’s admission of responsibility for Mr. Rozier’s death was made in the context of a criminal action to deter reckless driving, its relevance in a civil action based on a legal doctrine presumably intended to deter the negligent manufacture of automobiles is attenuated at best. The liability which Wilson admitted in his guilty plea is based on a different social policy than that with which Mrs. Rozier charged Ford Motor Company. Furthermore, the second collision doctrine assumes that the primary collision is caused by a party other than the manufacturer. Here, Mrs. Rozier has alleged that the primary collision with Wilson caused the secondary collision with the fuel tank which, in turn, caused the death of her husband by creating the fire that incinerated him. The key question in the case is whether Ford owed a duty to Mr. Rozier to build a fuel tank that could withstand an impact such as that inflicted by Wilson’s car. If Ford owed such a duty then, other considerations aside, Ford caused Mr. Rozier’s death. We think that the interrelationship between cause and duty in the second collision context was difficult enough for the jury to grasp so that it should not have been further complicated by the unnecessary, confusing, and potentially misleading element of the Wilson guilty plea. We hold that the trial court abused its discretion in admitting this evidence. Although we find an abuse of discretion, our holding is not that this error would, in itself, have required reversal. Plaintiff’s counsel had an opportunity to subpoena Mr. Wilson to question him about the circumstances surrounding the plea and could have argued to the jury that by his plea of guilty, Wilson did not absolve Ford of liability for negligent design of the fuel tank. Additionally, we note that the able trial judge may have been led astray by the contradictory and often merely conclusory arguments for and against admissibility made by counsel for both sides. Initially, in arguing a pretrial motion, plaintiff’s counsel took the position that the guilty plea was inadmissible to impeach Wilson’s testimony; the court then ruled in favor of Ford that such evidence was admissible on cross-examination under Rule 609(a), Fed.R. Evid. T. 10-11. During the trial, however, neither party called Wilson as a witness. In cross-examining Trooper Swindell, counsel for Ford elicited testimony concerning *1349 the manslaughter charge against Wilson. In response to an objection by plaintiff’s counsel, Ford assured the court that the information sought was relevant to the issue of speed. T. 68. At the close of the testimony, Ford tendered and the court admitted over plaintiff’s objection the plea and judgment of conviction of Wilson. At this point, Ford explained that the plea was relevant “[t]o illustrate an admission by a person involved in the occurrence of responsibility for the occurrence”. The court sought a contrary argument from plaintiff’s counsel and was assured, seven times, that the plea was either “irrelevant”, “totally irrelevant”, or “absolutely irrelevant” but never told why. Apparently impressed that counsel could generate so much heat with no light, the court overruled the objection. T. 1132-39. IV. CONCLUSION The record in this case establishes that defendant Ford was aware of a document in its files relevant to the plaintiff’s case, sought by the plaintiff through interrogatories, and included within a discovery order, but that it failed to disclose the document or to amend its response to an interrogatory, falsely stating that it was unable to locate such a document. Ford’s misconduct prejudiced the plaintiff by denying her information which might well have reshaped the case she ultimately presented to the jury. Under these circumstances, plaintiff’s timely motion for a new trial pursuant to Rule 60(b)(3) should have been granted. We reverse the district court’s denial of the 60(b)(3) motion and remand this case for a new trial. Additionally, because any probative value that it might have is substantially outweighed by the danger of confusing the issues and misleading the jury, we hold that evidence of Wilson’s guilty plea to manslaughter charges arising out of the fatal collision in this case is inadmissible under Rule 403, Fed.R.Evid. REVERSED and REMANDED, with directions. Footnotes [1] A saving clause in Rule 60(b) provides: “This rule does not limit the power of a court to entertain an independent action ... to set aside a judgment for fraud upon the court.” See Dausuel v. Dausuel, 90 U.S.App.D.C. 275, 195 F.2d 774 (1952). [2] See, e. g., United States v. Standard Oil Co. of Calif., 73 F.R.D. 612 (N.D.Cal.1977); H. K. Porter Co., Inc. v. Goodyear Tire & Rubber Co., 536 F.2d 1115 (6th Cir. 1976); Kupferman v. Consolidated Research & Mfg. Corp., 459 F.2d 1072 (2d Cir. 1972); Keys v. Dunbar, 405 F.2d 955 (9th Cir. 1969); England v. Doyle, 281 F.2d 304 (9th Cir. 1960). [3] Cf. Lockwood v. Bowles, 46 F.R.D. 625, 633 (D.D.C.1969): “Except in extraordinary circumstances such as those truly contemplated by the term ‘fraud upon the court’ in Rule 60(b), the law favors an end to lawsuits rather than a free reopening and retrial of them”. [4] Factually incorrect judgments are the subject of Rule 60(b)(2), which provides for relief from a judgment on grounds of “newly discovered evidence which by due diligence could not have been discovered in time to move for a new trial under Rule 59(b)”. We have held that under Rule 60(b)(2), the newly discovered evidence must be “such that a new trial would probably product a new result”. Ag Pro, Inc. v. Sakraida, 512 F.2d 141, 143 (5th Cir. 1975). [5] Stenning, a Ford engineer, had testified that “Ford has had at least in the last ten years and prior its own in-house research and development work on alternate fuel tank designs, number one, and we have looked at alternate fuel tank locations . . . ” [6] In his affidavit, Mr. Grubba stated that the Trend Cost Estimate “would have no application to the 1969 Ford Galaxie nor to any other vehicle produced by the defendant during that model year”. This conclusory statement does not necessarily contradict the Arndt and Bloch affidavits filed on behalf of Mrs. Rozier. Arndt and Bloch concluded that the document was *1341 applicable to the 1969 model because the 1969 and 1971 models were, in effect, structurally indistinguishable. The brief to which the Grubba affidavit was appended makes clear that Ford’s theory of nonapplicability is based on cost, not structural factors. According to the brief, the document could not apply to a 1969 Ford because “the estimated cost of materials are measured against the 1971 Ford automobile”. S.R. 110. Even assuming, as we do not, that the only possible relevance of the document concerned the cost of alternative fuel tank designs, the difference of two or more years would not eliminate it from the scope of the discovery order because the costs could be adjusted to reflect the time differential. [7] In his deposition, Ford engineer Stenning noted that one alternate location was disfavored because it would not allow different bodies to be installed on the same base, thus cutting down on “interchangeability”. “[W]e are in the business of- making money”, he added. Stenning Deposition, May 1, 1975, Vol. II, at 32. [8] Ironically, Ford objected to the fourth interrogatories as filed not because they were not limited to reports applicable only to the 1969 Galaxie 500, but on the broader ground that “the information sought therein does not relate to vehicles of the same size, chassis and fuel system as the 1969 Ford Galaxie 500”. R. 445. Ford thus appears hardly in a position to argue that the Trend Cost Estimate was not covered by the district court’s order simply because it does not refer specifically to a 1969 Galaxie but rather to a 1971 model which is “of the same size, chassis and fuel system as the 1969 Ford Galaxie 500”. [9] In this respect we are not substituting our judgment for that of the district court. See Atchison, Topeka & Santa Fe Ry. Co. v. Barrett, supra, 246 F.2d at 849. Unfortunately, the district court made no findings of fact, and its order denying the motion for a new trial stated simply: “the Court is of the opinion that said motion should be and is hereby denied upon each and every ground therein stated”. S.R. 127. Our conclusion that Ford engaged in misconduct is not inconsistent with the district court’s holding because, conceivably, the court could have found that Ford engaged in misrepresentation or other misconduct but that a new trial was nevertheless unwarranted. This reading of the court’s holding is buttressed by comments made by the trial judge at the hearing on plaintiff’s motion. When counsel for Ford attempted to argue that plaintiffs interrogatories were intended to discover only that which went into the design of the 1969 Ford Galaxie 500, the trial judge interrupted: THE COURT: Wait a minute. But then if anything was learned by Ford up between *1342 that time and the time of the death of Rozier that would have application to the safety of the ’69 Ford, that would be relevant. That’s why we limited it to the time of his death, didn’t we? 2/26/77 Transcript at 26. Shortly thereafter, the following exchange ensued: MR. WEINBERG (for Ford): In any event, the Court’s order, we contend is clear and that there has been no violation of this Court’s order by Ford Motor Company, that the report simply does not fit into the ambit of the definition, even though so stated by Mr. Arndt and Mr. Block [sic], and that the fair . . . THE COURT: Let me say this, you know. In a nontechnical sense, I would say it does, just by looking at it. I mean, they’re talking about variable costs on the proposed design . . . MR. WEINBERG: It doesn’t have anything to do with the cost of tooling, the cost of production? This, Your Honor . . . THE COURT: Whatever. Whatever. It does have reference to variable costs . MR. WEINBERG: But this is costs and materials only, and that’s different from a cost/benefit analysis, we urge upon the Court. THE COURT: It would seem to me to be a factor that would go into it. Id. at 29-30. [10] Ford’s duty to amend its inaccurate response is based on Rule 26(e)(2), Fed.R.Civ.P.: “A party is under a duty seasonably to amend a prior response if he obtains information upon the basis of which (A) he knows that the response was incorrect when made, or (B) he knows that the response though correct when made is no longer true and the circumstances are such that a failure to amend the response is in substance a knowing concealment”. See Havenfield Corp. v. H & R Block, Inc., 509 F.2d 1263 (8th Cir. 1975), cert. denied, 421 U.S. 999, 95 S.Ct. 2395, 44 L.Ed.2d 665. Although we are unaware of the precise nature of Mr. Grubba’s involvement in this case, he did not participate in the trial itself or in the appeal. The attorneys retained by Ford to represent the company at trial and in the appeal have assured this Court that they were personally unaware of the Trend Cost Estimate until Mrs. Rozier filed her motion to vacate almost a year after the trial. We accept their statements. [11] Vockie v. General Motors Corp., 66 F.R.D. 57, 61 (E.D.Pa.1975), cited by Ford as illustrative of the policy behind Rule 407, is not pertinent to this case. Vockie concerned the admissibility of a recall notice and campaign by General Motors involving a design defect which plaintiffs alleged had existed in their car and had resulted in injuries to them. The district court excluded the recall evidence, reasoning that “[m]anufacturers should not be inhibited in, or prejudiced by, a good faith effort to protect the public safety and comply with their statutory duty”. Id. at 61. Additionally, the court noted that evidence of the recall “has minimal probative value to the existence of a defect in a particular vehicle”. The key to the case, however, is that the plaintiffs argued for admissibility of the recall notice on the grounds that it constituted an admission by an adverse party; Rule 407 was not raised by either party, although its policy was relied upon by the court for analogy to the facts before it. The memorandum opinion in Vockie is dated February 20, 1975; the effective date of the Federal Rules of Evidence was July 1, 1975, 180 days after their enactment, H.R. 5463, P.L. 93-595, approved January 2, 1975. [12] In this diversity case, we are of course “Erie-bound” to apply the law of Georgia, both the forum state and the state where Mr. Rozier’s fatal injury occurred. Erie R.R. Co. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938). [13] Although Georgia courts have not specifically cited Larsen, Friend, supra, and its progeny make clear that Larsen, a seminal case on manufacturer liability for “second collision” injuries, accurately reflects the law in Georgia. Cf. Ford Motor Co. v. Lee, 137 Ga.App. 486, 224 S.E.2d 168 (1976). Georgia has been listed as a follower of Larsen in Huff v. White Motor Corp., 565 F.2d 104, 110 (7th Cir. 1977) and Polk v. Ford Motor Co., 529 F.2d 259, 264 n.5 (8th Cir. 1976). See also Note, Products Liability in Georgia, 12 Ga.L.Rev. 83, 96-98 (1977). In discussing the duty to warn, the court in Larsen quoted from our decision in Blitzstein v. Ford Motor Co., 288 F.2d 738, 744 (5th Cir. 1961), in which we used language relevant to the instant case: We think that a jury could reasonably have found that the American Ford Company was negligent in marketing a product which was inherently dangerous, of which danger it should have been aware from its long experience in the design and manufacture of automobiles, and that American Ford failed to exercise reasonable care to inform the buying public of this dangerous condition. [14] In arguing against the relevance of the Trend Cost Estimate, Ford cites our decision in Barnes v. General Motors Corp., 547 F.2d 275 (5th Cir. 1977), regarding the admissibility of experimental and comparative tests. We held such evidence inadmissible in Barnes because the test vehicle lacked a crucial safety feature, a “roll-stop” mount, which had been installed on the plaintiffs automobile. No similar disparity is present in this case. Ford notes that expert testimony at the trial indicated that the speed of impact — the “differential” speed between the Galaxie and the impacting vehicle, a Chevrolet — was in excess of 60 m.p.h., while the Trend Cost Estimate dealt with a proposed 30 m.p.h. safety standard. However, the evidence at trial allowed the jury to conclude that the closing speed at impact ranged from as low as 20 to as high as about 70 m.p.h. Trooper Gary Swindell, the officer who investigated the accident, testified that the speedometer on the Chevrolet was stuck at 68 m.p.h. and that the Galaxie was traveling at approximately 40 to 45 m.p.h. when hit. T. 58-60. On cross-examination, Swindell conceded that the Galaxie might have been moving at less than 10 m.p.h. and the impacting vehicle in excess of 68 m.p.h. at the moment of impact. T. 64-66. One of plaintiff’s experts opined that the speed of the Galaxie was between 2.3 and 8.6 m.p.h. and that of the Chevrolet was between 65.7 and 72 m.p.h., indicating a maximum differential of 69.7 m.p.h. T. 303. Another expert estimated the differential speed as between 60 and 68 m.p.h. T. 766. Because the jury’s verdict is consistent with almost any differential speed, we cannot determine how the jury evaluated this evidence; it could have concluded that the speed of impact was only 30 m.p.h. but that Ford nevertheless had not breached its duty of *1345 care toward Mr. Rozier. Similarly, although the Trend Cost Estimate is directed at satisfying a 30 m.p.h. safety standard, it is conceivable that the design modifications suggested would also satisfy a higher standard, perhaps up to 60 m.p.h. Hence, the fact that the Trend Cost Estimate refers to a proposed 30 m.p.h. standard does not necessarily diminish its potential materiality in this case. [15] Trooper Swindell also testified that when he arrived at the scene shortly after the accident, he observed that the speedometer on Wilson’s car was stuck at 68 m.p.h. By the time the speedometer was photographed, however, it had slipped to the 57 m.p.h. mark. Photographs of the speedometer in this position were admitted into evidence. Plaintiffs Exhibits 24 and 25. [16] Arguably, this evidence was not offered to prove a factual proposition at all, but rather to support a legal position — that the Wilson guilty plea determined the issue of legal responsibility for the death of William Rozier. Of course, the guilty plea would be entitled to no collateral estoppel effect in this case. But, as Prosser explains, the question of factual cause is often difficult to separate from that of legal cause. The question, writes Prosser, is “whether the policy of law will extend the responsibility for the conduct to the consequences which have in fact occurred. Quite often this has been stated, and properly so, as an issue of whether the defendant is under any duty to the plaintiff. . This is not a question of causation, or even a question of fact . . . and the attempt to deal with it in such terms has led and can lead only to utter confusion”. W. Prosser, supra, at 244.